Birth of a Market: The U.S. Treasury Securities Market from the Great War to the Great Depression 9780262016377, 0262016370

Table of contents :
Contents
Preface
Acknowledgments
1 Introduction
I Before the Great War
2 The Payments System before World War I
3 Treasury Debt Management before World War I
II Financing the Great War
4 Treasury Finance during World War I
5 Designing the Liberty Loan
6 Marketing the Liberty Loans
7 Treasury Cash Management: Certificates of Indebtedness
8 Treasury Cash Management: War Loan Deposit Accounts
9 Federal Reserve Support of the Treasury Market during World War I
10 Coda on Treasury Debt Management during World War I
III Paying Down the War Debt
11 Treasury Finance during the 1920s
12 Paying down the War Debt
13 Revival of the Over-the-Counter Market
14 Evolution of the Primary Market and the Introduction of Treasury Bills
15 Coda on Treasury Debt Management during the 1920s
IV The Great Depression
16 Treasury Finance during the Great Depression
17 Nonmarketable Treasury Debt
18 Treasury Debt Management during the Great Contraction
19 Treasury Debt Management during the New Deal
20 The Primary Market during the Great Depression
21 Statutory Control of Treasury Indebtedness
22 The Brief Revival and Subsequent Extinction of National Bank Notes
23 Coda on Treasury Debt Management during the Great Depression
V L'Envoi
24 Treasury Debt Management since 1939
References
Index

Citation preview

BIRTH OF A MARKET

BIRTH OF A MARKET The U.S. Treasury Securities Market from the Great War to the Great Depression

Kenneth D. Garbade

The MIT Press Cambridge, Massachusetts London, England

© 2012 Massachusetts Institute of Technology All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher. For information about special quantity discounts, please email [email protected] This book was set in Times New Roman by Toppan Best-set Premedia Limited. Printed and bound in the United States of America. Library of Congress Cataloging-in-Publication Data Garbade, Kenneth D. Birth of a market : the U.S. Treasury securities market from the Great War to the Great Depression / Kenneth D. Garbade. p. cm. Includes bibliographical references and index. ISBN 978-0-262-01637-7 (hardcover : alk. paper) 1. Government securities—United States—History. 2. Investments—United States—History. I. Title. HG4715.G37 2012 332.63′2420973—dc22 2011011785 10

9

8

7

6

5

4

3

2

1

for Bernice, Larissa, Edward, and Rachel

Contents

Preface Acknowledgments

ix xi

1

Introduction

1

I

BEFORE THE GREAT WAR

2

The Payments System before World War I

13

3

Treasury Debt Management before World War I

29

II

FINANCING THE GREAT WAR

4

Treasury Finance during World War I

49

5

Designing the Liberty Loans

69

6

Marketing the Liberty Loans

91

7

Treasury Cash Management: Certificates of Indebtedness

109

8

Treasury Cash Management: War Loan Deposit Accounts

123

9

Federal Reserve Support of the Treasury Market during World War I

131

10

Coda on Treasury Debt Management during World War I

143

III

PAYING DOWN THE WAR DEBT

11

Treasury Finance during the 1920s

147

12

Paying down the War Debt

161

13

Revival of the Over-the-Counter Market

185

14

Evolution of the Primary Market and the Introduction of Treasury Bills

199

15

Coda on Treasury Debt Management during the 1920s

217

IV

THE GREAT DEPRESSION

16

Treasury Finance during the Great Depression

221

17

Nonmarketable Treasury Debt

247

18

Treasury Debt Management during the Great Contraction

261

19

Treasury Debt Management during the New Deal

279

20

The Primary Market during the Great Depression

303

viii

Contents

21

Statutory Control of Treasury Indebtedness

313

22

The Brief Revival and Subsequent Extinction of National Bank Notes

319

Coda on Treasury Debt Management during the Great Depression

333

23

V

L’ENVOI

24

Treasury Debt Management since 1939

337

References Index

371 381

Preface

The seed that grew into this book was planted in the summer of 2003, when Christine Cumming, then the director of research at the Federal Reserve Bank of New York and now the Bank’s first vice president, asked me to write a set of papers on innovations in the market for U.S. Treasury securities. Her request led to research on the Treasury’s decision to auction bills in 1929 (Garbade 2008), the subsequent extension of the auction process to notes and bonds in the early 1970s (Garbade 2004a), the emergence of “regular and predictable” as a Treasury debt management strategy (Garbade 2007), the origins of the Federal Reserve book-entry system (Garbade 2004b), and the evolution of repo contracting conventions in the 1980s (Garbade 2006). More or less concurrently I also undertook collaborative work on more recent innovations in Treasury cash management (Garbade, Partlan, and Santoro 2004), the Treasury auction process (Garbade and Ingber 2005), and the settlement of secondary market transactions in Treasury Securities (Garbade et al. 2010). All of the foregoing research started by identifying an innovation and then asking why Treasury officials, Federal Reserve officials, or private market participants had come to conceive and implement the innovation—in particular, what problem were they trying to solve, or what inefficiency were they trying to mitigate? Several years ago I decided to reverse the process and begin by identifying the financing problems that Treasury officials faced from time to time and then asking how they responded to those problems and how their responses contributed to the development of the modern Treasury market. I decided to start with World War I for two reasons. First, that war led to the first large Treasury financings since the Civil War and the Federal Reserve System had just been created. As a result both the Treasury’s financing problems and the institutional organization of the money and capital markets would be reasonably familiar to a current observer. Second, I wanted to assess whether innovations introduced almost a century ago persisted to the present day. Surprisingly, three of the four pillars of the modern Treasury market—regular and predictable issuance, auction offerings aimed at a wide spectrum of market participants, and the Treasury Tax and Loan system—can be traced back to the period from 1917 to 1939, although other innovations introduced in that period have since vanished and some further innovations have been introduced since 1939. (Most important, the fourth pillar of the modern Treasury market, the book-entry system, was introduced in 1968.)

Acknowledgments

My first and greatest debt is to the researchers, policy makers, and analysts who wrote the books, papers, memos, and articles that provide the foundations for this book. The depth and breadth of my obligations are evident in the footnotes and citations that follow. I am similarly indebted to the persons who provided access to those works: the staff of the Research Library at the Federal Reserve Bank of New York, including Pat Buckley, Megan Cohen, Amy Farber, Edita Karaciejute, Kara Masciangelo, Kathleen McKiernan, Mary Tao, and Marilyn Warbach; Maryalice Cassidy and Hugh Tidwell of the Bank’s Law Library; and the staff of the Records Management and Archives functions at the New York Fed, including Joseph Komljenovich, Rosalie Kurtz, Rosemary Lazenby, and Gregory Rappa. I have benefited from numerous conversations with colleagues in the academic community, in the Treasury market, at the U.S. Treasury, and in the Federal Reserve System. These include Ken Abbott, Paul Agueci, Yakov Amihud, Chuck Andreatta, Ken Baron, Steve Berardi, Menachem Brenner, Dave Buckmaster, Chris Burke, Francis Cavanaugh, Al Clark, Jim Clouse, Lou Crandall, Mary Edwards, Kurt Eidemiller, Bob Elsasser, Ned Elton, Amy Farber, Michael Fleming, Steve Friedman, Joshua Frost, Oliver Giannotti, Lee Grandy, Harry Heiss, Warren Hrung, Jeff Huther, Jennifer Imler, Jeff Ingber, Frank Keane, Colin Kim, Arlen Klinger, Clinton Lively, Lorie Logan, Jim Mahoney, Paul Malvey, Antoine Martin, John McGowan, Mary Miller, Dave Monroe, Frank Noll, John Partlan, Debby Perelmuter, Fred Pietrangeli, Karthik Ramanathan, Marcy Recktenwald, Amar Reganti, Tony Rodrigues, Allan Rogers, Matt Rutherford, Lori Santamorena, Paul Santoro, Doug and Rita Skolnick, Mark Stalnecker, Charles Steindel, Amanda Stokes, Marti Subrahmanyam, Richard Sylla, Paul Volcker, Tom Wipf, and Jennifer Wolgemuth. Finally, I thank William Silber, my colleague, collaborator, and chief critic for nearly four decades, for the pleasure of innumerable conversations touching on a century of American economic history. This book would never have come to completion without the first-class research facilities and collegial work environment of the Federal Reserve Bank of New York. I am indebted to the Directors of Research during the time I wrote the book: Christine Cumming, Joe Tracy, Jamie McAndrews, and Simon Potter, for their leadership in providing those facilities and that environment. Bernice and our children, Larissa, Edward, and Rachel, were remarkably tolerant of a husband and father who disappeared into the basement on many nights and weekends. That lost time is neither replaceable nor compensable. Although many deserve thanks, I alone am responsible for the errors that follow. The views expressed are those of the author and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System.

1

Introduction

The market for U.S. Treasury securities is a marvel of modern finance. In calendar year 2009 the Treasury auctioned $8.2 trillion of new securities, ranging from 4-day bills to 30-year bonds, in 283 offerings on 171 different days. (See box 1.1 for a description of Treasury securities.) By the end of the year, total Treasury indebtedness stood at $12.3 trillion, $7.2 trillion of which was publicly held marketable debt spread out over 256 different issues.1 That immense indebtedness, coupled with the creditworthiness of the U.S. government, provided the raw material for the most liquid securities market in the world, where market participants buy and sell large amounts quickly, with minimal transaction costs and minimal price impacts. However, it was not always so. In the decade before World War I there was only about $1 billion of interestbearing Treasury debt outstanding, spread out over just six issues. New offerings were rare—six years passed between the last pre-war offering (sold to help finance the Panama Canal in 1911) and the first wartime sale—and the debt was narrowly held—80 percent was owned by national banks. The limited amount of debt, the infrequency of new offerings, and the narrow ownership base made for an illiquid market. This book traces the development of the Treasury market from a financial backwater prior to World War I to something reasonably familiar to a current observer on the eve of World War II. Central to this description is how successive Secretaries of the Treasury (identified in table 1.1) responded to the needs of their time and how their responses contributed to the development of the modern market. Treasury Indebtedness from World War I to the Eve of World War II Figure 1.1 shows the course of aggregate U.S. production from 1910 to 1940. The broad outlines are well known: a world war that began (for the United States) in April 1917 and ended in November 1918, a sharp contraction that began in January 1920 and bottomed out in July 1921, followed by a robust (albeit bumpy) expansion through August 1929 and then by the catastrophe of the Great Depression. Figure 1.2 shows the contemporaneous behavior of Treasury indebtedness: rapid expansion to finance the war, significant repayment during the 1920s, followed by the need to finance the chronic deficits of the 1930s. Figure 1.3 shows the resulting behavior of the ratio of debt to gross 1. In addition to the $7.2 trillion of publicly held marketable debt, total Treasury indebtedness included $4.5 trillion of nonmarketable debt held in intergovernmental accounts (more than half of which was held in the Federal Old-Age and Survivors Insurance Trust Fund), $561 billion of publicly held nonmarketable debt (more a third of which was savings bonds), and $23 billion of marketable securities held in intergovernmental accounts.

2

Chapter 1

Box 1.1 Treasury Securities The U.S. Treasury presently issues four types of marketable securities: bills, notes, bonds, and Treasury Inflation-Protected Securities (“TIPS”). Treasury bills are single payment securities maturing in not more than a year from the date of issue. A Treasury bill promises to pay its face amount at maturity but does not make any explicit interest or coupon payments. Treasury bills are consequently issued at a discount from face amount. Notes and bonds are coupon-bearing securities issued at or near their stated principal value that pay semiannual coupons prior to and at maturity and repay principal at maturity. Notes have a maximum maturity of ten years and are currently issued with two-, three-, five-, seven-, and ten-year maturities. Bonds can have any maturity and are currently issued with a thirty-year maturity. TIPS (first issued in 1997) are coupon-bearing securities whose interest and principal payments are linked to the consumer price index. During the period from 1917 to 1939 the Treasury issued bills (beginning in 1929), notes (beginning in 1919, with a maximum maturity of five years), and bonds, but it did not issue TIPS. It also issued certificates of indebtedness: coupon-bearing securities that matured in not more than a year from the date of issue.

Table 1.1 Secretaries of the Treasury from World War I to the Great Depression Administration of Woodrow Wilson (March 4, 1913–March 4, 1921): William McAdooa March 6, 1913–December 15, 1918 Carter Glass December 16, 1918–February 1, 1920 David Houston February 2, 1920–March 3, 1921 Administrations of Warren Harding (March 4, 1921–August 2, 1923), Calvin Coolidge (August 2, 1923–March 4, 1929), and Herbert Hoover (March 4, 1929–March 4, 1933): Andrew Mellonb March 4, 1921–February 12, 1932 Ogden Mills February 13, 1932–March 3, 1933 Administration of Franklin Roosevelt (March 4, 1933–April 12, 1945): William Woodin March 4, 1933–December 31, 1933 Henry Morgenthau January 1, 1934–July 22, 1945 a. William McAdoo was an unusually colorful person, before, during, and following his term as Secretary of the Treasury. See McAdoo (1931), Broesamle (1973), and Silber (2007). b. While Secretary of the Treasury in the mid-1920s, Andrew Mellon was ranked with Alexander Hamilton in terms of his contributions to American government, but his reputation plummeted with the onset of the Great Depression. See Cannadine (2006).

Introduction

125

Billions of dollars

100

75

50

25

1925

1930

1935

1940

1925

1930

1935

1940

1920

1915

1910

0

Figure 1.1 Gross national product at current prices

50

40

Billions of dollars

30

20

10

1920

1915

0 1910

3

Figure 1.2 Treasury debt, mid-1910 to mid-1940

4

Chapter 1

50

Percent

40

30

20

10

1940

1935

1930

1925

1920

1915

1910

0

Figure 1.3 Treasury indebtedness as a percentage of gross national product

national product—a graph surprising primarily because of the modest increase in the ratio during the New Deal. Linkages from Past to Present The period from 1917 to 1939 witnessed four institutional developments with lasting significance for the Treasury market: (1) the revival of auction sales of Treasury securities, (2) the close integration of Treasury cash and debt management, (3) a change from statutory limits on issuance of various categories of debt to a single limit on total debt outstanding, and (4) the introduction of “regular and predictable” offerings. Auction Sales Treasury officials had had extensive pre-war experience with auction offerings—seven of the ten bond offerings between 1894 and 1917 were auctions— but nevertheless elected to sell the debt that financed World War I at fixed prices in order to attract the interest of less sophisticated individual and institutional investors, many of whom were believed to be reluctant to bid in auctions for fear of overpaying. (New York banks accounted for the bulk of the pre-war auction offerings.) After the war, the Treasury’s borrowing needs were smaller but still large by pre-war standards (because of the need to refinance maturing debt) and Treasury officials continued to rely on fixed-price sales (to retain the direct participation of less sophisticated investors). They

5

Introduction

did not begin to revive auction offerings until they introduced Treasury bills in 1929. This book explains why, after relying on fixed-price subscription sales for twelve years, the Treasury began to return to auction offerings in 1929. It also explains why the Treasury failed when it attempted to auction bonds in 1935. Another thirty-five years would pass before Treasury officials were finally able to extend the auction process to notes and bonds.2 Integration of Cash Management and Debt Management All of the Secretaries of the Treasury between 1917 and 1939 devoted substantial attention to integrating cash management with debt management. Their initiatives included, during World War I, sales of short-term bond and tax anticipation debt and the development of an extensive system of Treasury deposit accounts, called War Loan Deposit Accounts, at commercial banks. The introduction of Treasury bills in 1929 was the principle postwar cash management initiative. The War Loan Deposit Account system was the forerunner of the modern Treasury Tax and Loan system;3 Treasury bills, in the form of cash management bills, remain in active use today as an instrument of Treasury cash management. Evolution of Statutory Limits on Indebtedness The interval from 1917 to 1939 also witnessed an important evolution in how government officials thought about Treasury debt. In the decades prior to the Great War the Treasury issued bonds to finance narrowly defined projects, including the Spanish–American War in 1898 and the Panama Canal between 1906 and 1911. World War I imposed extraordinary financing burdens on the United States but it did not materially alter the “project finance” construction of Treasury debt: Treasury officials financed most of the costs of the war with five large Liberty Loans—four bond issues and one note issue, each of which was the subject of specific congressional approval. After the war the federal government ran persistent budget surpluses but the Treasury continued to sell large quantities of debt (to refinance maturing debt). The view of Treasury indebtedness began to evolve in this new environment. In 1921, Congress changed from regulating the issuance of Treasury notes to regulating the maximum amount of notes that could be outstanding. The change reflected the understanding that note issuance in an era of budget surpluses was more a matter of refinancing maturing debt than financing new projects. Once the Great Depression got underway the prospect of chronic deficits and the decision to refinance maturing debt at longer maturities made 2. Garbade (2004a) describes the introduction of auction sales of notes and bonds in the first half of the 1970s. 3. Garbade, Partlan, and Santoro (2004) describe the Treasury Tax and Loan system.

6

Chapter 1

it clear that it was no more efficient to regulate bond issuance than it had been to regulate note issuance. In 1935 Congress changed to regulating the amount of bonds that could be outstanding, rather than the amount of bonds that could be issued. The last step in the evolution of congressional control of Treasury debt came in 1939, when Congress limited itself to specifying only the maximum total amount of debt that could be outstanding (whether in the form of bills, notes, or bonds) and left decisions regarding issuance and the composition of the debt to the Secretary of the Treasury. Introduction of “Regular and Predictable” Offerings This book also explains the genesis of “regular and predictable” offerings as a related manifestation of the postwar erosion of the project finance view of Treasury debt. Recognizing that the indebtedness represented by a maturing Treasury security was more likely to be refinanced than extinguished, Treasury officials began to think, first, in terms of regular issuance schedules and, later, in terms of regular offerings of predictable amounts of debt with predictable maturities. More generally, officials began to think in terms of issuance programs, designed to support a more or less persistent indebtedness, rather than in terms of individual securities sold to finance particular projects. This ultimately led to the adoption of “regular and predictable” as a fundamental strategy of Treasury debt management in the 1970s.4 Organization of the Book The book is divided into five parts. The first part sets the stage by describing the payments system and Treasury debt management in the decades before World War I. The following three parts examine, respectively, how the Treasury financed the Great War, how it managed the postwar refinancings and paydowns, and how it financed the chronic deficits of the Great Depression. The first chapter in each part sketches the fiscal environment within which Treasury officials operated, including the tax regimes and expenditure programs authorized by Congress as well the general economic environment. Subsequent chapters examine, first, the broad scope of Treasury debt management operations and, subsequently, more fine-grained topics, including Treasury cash management policies and primary and secondary market developments. A coda at the end of each part summarizes the significant events of the period. The last chapter of the book examines several aspects of modern Treasury debt management that grew out of developments from 1917 to 1939. 4. Garbade (2007).

7

Introduction

Treasury Operations during World War I Treasury officials financed American participation in World War I with five large Liberty Loans: four long-term bonds and, following the cessation of hostilities, one intermediate-term note. (Taxes and short-term debt also played important, albeit smaller, roles in financing the war.) In each loan, Treasury officials offered a single security for sale at a fixed price of par and aggressively marketed the security to middle- and upper-income investors. Officials switched from the auction offerings that had been common before the war to fixed-price offerings because they wanted to sell securities directly to less sophisticated individual and institutional investors. The decision to finance the war with large, discrete loans raised the problem of how to avoid disrupting the short-term credit markets when money came in episodically, at the end of each Liberty Loan campaign, while expenditures were more or less continuous. Treasury officials solved the problem with an elaborate system of sales of short-term debt prior to each loan (with the short-term debt to be repaid with the proceeds of the following loan) and by allowing banks to pay for their Liberty Loan subscriptions by crediting War Loan Deposit Accounts maintained by the Treasury at the respective banks. Paying down the War Debt World War I ended on November 11, 1918, and military expenditures receded rapidly thereafter as wartime draftees returned from Europe and were demobilized. The Treasury, however, was left owing $20 billion on the Liberty Loans and $5 billion on other indebtedness. In spite of a robust economic expansion during the 1920s, the war debt was too large, and too concentrated (in just five issues), to extinguish as it came due. Treasury officials had to develop a program for integrating debt refinance with debt repayment. That program, worked out gradually between 1921 and 1928, included an early version of “regular and predictable” issuance. In one phase of the program, officials offered holders of outstanding Liberty Loans opportunities to exchange their soon-to-mature loans for new issues of short- and intermediate-term debt scheduled to mature on future quarterly tax dates. (Taxes were due in four quarterly installments during the year following the year in which income was earned, there being no provision for income tax withholding until World War II.) The exchange offers reduced the amount of a Loan that had to be redeemed for cash and spread out the dates at which the Treasury would ultimately extinguish the debt. When the Treasury needed to raise cash, it sold new securities—primarily certificates of indebtedness (see box 1.1) – on quarterly tax dates and

8

Chapter 1

scheduled the securities to mature on future tax dates. The amount offered on a particular date was set at the amount of debt maturing, plus what would be needed to fund disbursements in the coming quarter, less anticipated tax receipts. The Treasury hardly ever came to market with a cash offering on other than a quarterly tax date. Treasury officials continued to rely on fixed-price sales (in lieu of auctions) during the 1920s in order to retain the direct participation of less sophisticated investors. Two episodes of failed offerings early in the decade left officials reluctant to trim offering yields close to contemporaneous market yields and new issues were chronically underpriced. Underpricing gave rise to large oversubscriptions and created an allocation problem: if not all subscribers were to get all the securities they asked for, who was to be disappointed? The Treasury initialed opted for a scheme that provided proportionally larger awards to small subscribers (to promote a broad distribution of Treasury debt) and to subscribers tendering maturing securities in lieu of cash (to minimize churning of Treasury cash balances). However, while the scheme advanced two nontrivial goals of Treasury debt management, it did not address the fundamental problem that fixed-price offerings led the Treasury to underprice its offerings. The practice of selling securities only at quarterly tax dates meant that the Treasury sometimes had to borrow well in advance of its actual needs and inventory cash balances in low-yielding War Loan Deposit Accounts. To mitigate this inefficiency, and to begin to return to auction offerings, Treasury officials, working with staff of the Federal Reserve Bank of New York, introduced auction sales of a new short-term security, Treasury bills, in 1929. They planned to use the new securities as a cash management tool, auctioning bills to raise funds “as needed” between tax dates and scheduling redemptions on future tax dates so maturing bills could either be paid down with tax receipts or refinanced with longer term debt. Treasury Operations during the Great Depression The economic depression that started in 1929 forced the Treasury to turn its attention from debt reduction to deficit finance. The principle debt management decision of the Great Depression was the decision to use Treasury bills for debt management as well as for cash management. When the Treasury started using bills to finance a veterans’ loan program in the spring of 1931, it standardized bill maturities at 13 weeks and initiated a practice of regularly refinancing maturing bills with new bills. This marked the first instance of “cycle” securities: securities issued on a regular schedule, at a fixed maturity, and in amounts that did not vary appreciably from offering to offering (so that maturing securities were regularly refinanced with new issues).

9

Introduction

Treasury officials also initiated (in 1931) a program of refinancing couponbearing securities at longer maturities. The program was interrupted or postponed repeatedly, first by the high interest rates and the wave of bank failures that followed Britain’s departure from the gold standard in 1931, then by a steep yield curve in 1932 that created opportunities to issue shorter term debt at more attractive yields, and finally by the collapse of the banking system in early 1933, but officials got the program on track in mid-1933. By the end of the decade they were routinely financing deficits and refinancing maturing debt with weekly bill auctions and quarterly sales of 5-year notes and longer term bonds.

I

BEFORE THE GREAT WAR

2

The Payments System before World War I

This chapter describes the payments system as it existed in the United States between 1900 and the beginning of the First World War. It might seem odd to begin a study of the market for U.S. Treasury securities with a discussion of how Americans paid for their purchases of goods and services, but it is difficult to understand the primary market for Treasury bonds before World War I without understanding national bank notes, it is difficult to understand how America financed its wartime expenditures without understanding why Congress created the Federal Reserve System in 1913, and it is difficult to understand several fiscal and monetary policy initiatives of the early 1930s without understanding how the gold standard operated. The Gold Standard Article I, section 8, of the United States Constitution says that “Congress shall have the power . . . to coin money [and] regulate the value thereof . . ..” By the beginning of the twentieth century, Congress had settled on the principle that each of the various forms of money then in use should be convertible, directly or indirectly, into gold at a fixed price of $18.60 per ounce. This “gold standard” was widely viewed as a device to check the historic propensity of governments to finance expenditures by printing money (in lieu of borrowing money or raising taxes). It also constituted an explicit rejection of earlier efforts to establish a bi-metallic standard of both gold and silver.1 The Gold Standard Act of March 14, 1900, “fixed gold as the standard legal tender monetary metal.”2 Section 1 of the act provided that “the dollar consisting of 25.8 grains of gold nine-tenths fine . . . shall be the standard unit of value, and all forms of money issued or coined by the United States shall be maintained at a parity of value with this standard . . ..”3 There are 480 grains in a troy ounce,4 so 25.8 grains is equal to 0.053750 ounces (0.053750 = 25.8/480). The Gold Standard Act thus defined a dollar as consisting of 1. The bi-metallic standard would have allowed silver bullion to be coined into money at a rate of about $1.16 per ounce, or in a price ratio of about 16-to-1 with gold ($18.60 per ounce of gold = 16 × $1.16 per ounce of silver). Silverites achieved some measure of success with the passage of the Bland–Allison Act of February 28, 1878, and the Sherman Silver Purchase Act of July 14, 1890. However, “hard-money” Democrats and eastern Republicans forced repeal of the Sherman Silver Purchase Act in 1893 and set the country firmly on the path to a monometallic gold standard. Efforts to suppress demands for a bimetallic standard were aided by a boom in world gold production between 1893 and 1897. 2. Quoting Timberlake (1978, p. 173). Emphasis in the original. 3. The monetary standard specified in the Gold Standard Act reaffirmed prior statutory pronouncements, including the Act of January 18, 1837, and the Act of February 12, 1873. 4. The weight of precious metals like gold and silver is commonly expressed in troy ounces, rather than the more familiar avoirdupois ounces. One troy ounce is equal to 1.09714 avoirdupois ounces.

14

Chapter 2

0.053750 ounces of gold, and thereby set the dollar price of gold at $18.60 per ounce (18.60 = 1/0.053750). The qualification that the gold should be “nine-tenths fine” meant that the gold should be at least 90 percent (by weight) pure gold. Thus the price of gold was simultaneously set at $20.67 per fine ounce (20.67 = 18.60/0.90). Treasury Money The most fundamental forms of money were those issued by the U.S. Treasury, including gold coins, gold certificates, U.S. notes (commonly called “greenbacks”), silver dollars, and silver certificates. Maintenance of the gold standard required that each form be convertible into gold (of 0.9 fineness) at the statutory rate of 25.8 grains per dollar. Gold Coins In the early part of the twentieth century, U.S. mints at Philadelphia, Denver, and San Francisco accepted gold bullion for minting into coins.5 Coinage was “free and unlimited,” meaning that the Treasury would mint whatever gold bullion was tendered and did not charge for the service. Newly minted coins came in denominations of $2½, $5, $10, and $20 and the gold content of each coin was exactly as dictated by the Gold Standard Act. A $10 gold coin, for example, had 258 grains of gold of 0.9 fineness. Maintenance of the gold standard with respect to gold coins was never a significant economic issue because the amount of gold prescribed by statute was physically incorporated into each coin. Gold Certificates Gold coins unquestionably embodied the standard of value specified by Congress but were costly to store and costly to transport. Recognizing that transactions costs were lower with paper money, the Treasury also issued gold certificates—a form of currency—against the deposit of gold coin.6 (To save the cost of coinage, in 1911 Congress allowed the Treasury to issue gold certificates against the deposit of gold bullion as well as gold coin.) Gold certificates were nothing more than warehouse receipts for gold: a $10 gold certificate stated on its face that “There have been deposited in the Treasury of the United States ten dollars in gold coin payable to the bearer on 5. Watson (1926, p. 20). There was also a mint in New Orleans until 1909. 6. Gold certificates were first authorized by the Act of March 3, 1863. They were reauthorized in section 12 of the Act of July 12, 1882, and again in section 6 of the Gold Standard Act. See generally, Taus (1943, p. 178, n. 30).

15

The Payments System before World War I

demand.” Gold certificates could be redeemed at the Treasury in Washington and at subtreasury offices in New York, Chicago, St. Louis, and other large cities. To ensure that no certificate holder would be denied his or her right to gold, the Gold Standard Act provided that “The coin . . . deposited [against issuance of gold certificates] shall be retained in the Treasury and held for the payment of such certificates on demand, and used for no other purpose.” In mid-1914, the Treasury held $837 million of gold coin and $244 million of gold bullion as collateral against $1,081 million of gold certificates. The Treasury held $55 million of the certificates in its own vaults as part of its operating cash balance. Other Forms of Treasury Money There were three additional forms of Treasury money outstanding on the eve of World War I: U.S. notes, silver dollars, and silver certificates. All three were denominated in dollars, but the silver dollars contained no gold and the notes and certificates were not warehouse receipts for gold. Maintenance of parity with gold coin and gold certificates was, consequently, somewhat problematic. U.S. Notes Congress authorized the issuance of $450 million of U.S. notes during the Civil War as a means of paying for the war. Following various acts providing for redemption and reissuance, the amount of notes outstanding was frozen at $347 million by the Act of May 31, 1878. In mid-1914 the Treasury held $9 million of the notes as part of its operating cash balance. Section 2 of the Gold Standard Act provided that U.S. notes, “when presented at the United States Treasury for redemption, shall be redeemed in gold coin . . ..” To make the promise of redemption credible, the act directed the Secretary of the Treasury to maintain a $150 million reserve fund of gold coin and gold bullion, to be used solely for the purpose of satisfying requests for the exchange of U.S. notes into gold. (The explicit provision of a reserve fund formalized an earlier Treasury custom that dated to the late 1870s.7) If the reserve fund fell below $100 million, the Secretary was authorized to issue debt against payment in gold coin to restore the fund. In mid-1914 the fund stood at exactly $150 million. Silver Dollars and Silver Certificates Pursuant to the requirements of the Bland–Allison Act of February 28, 1878, and the Sherman Silver Purchase Act of July 14, 1890, the Treasury minted a 7. Noyes (1895, pp. 575–77), Dewey (1934, pp. 440–41), and Timberlake (1978, p. 134).

16

Chapter 2

total of 565 million silver dollars in the 1880s and early 1890s.8 By mid-1914, 491 million of those silver dollars (87 percent) were held in Treasury vaults as collateral for an identical amount of silver certificates.9 The Treasury held $12 million of the certificates as part of its operating cash balance. Nothing in the Gold Standard Act explicitly obligated the Treasury to redeem silver dollars and silver certificates with gold coin, but silver dollars were a legal tender (unless otherwise expressly stipulated) and silver certificates were receivable for customs, taxes, and all public dues. Thus, as a practical matter, it was unlikely that silver dollars and silver certificates would be valued below parity with gold.10 Bank Money Treasury money—gold coin and other money that the Treasury promised to convert to gold or to accept in lieu of gold—was the foundation of the American payments system, but the more significant medium of exchange (in terms of the volume of payments effected) was bank money: bank liabilities that banks promised to convert on demand into “lawful” (i.e., Treasury) money.11 In the early twentieth century, as at present, there were two types of banks: those with national charters and those with state charters.12 Both types accepted deposits that were payable on demand and that could be transferred by check. 8. Committee on Banking and Currency (1960, pp. 25–26). The Bland–Allison Act directed the Treasury to purchase between $2 million and $4 million of silver bullion each month and to coin the bullion into silver dollars. The Sherman Silver Purchase Act repealed the Bland–Allison Act and directed the Treasury instead to purchase 4.5 million ounces of silver bullion each month and to coin a portion of the bullion into dollars. 9. The Bland–Allison Act authorized the Treasury to issue silver certificates against the deposit of silver dollars. Silver certificates were not particularly popular until 1886, when Congress authorized (in the Act of August 4, 1886) the Treasury to issue low denomination silver certificates of $1, $2, and $5. 10. See “Currency Bill Taken Up,” New York Times, January 5, 1900, p. 8 (quoting Senator Nelson Aldrich, Republican of Rhode Island and chairman of the Senate Finance Committee, as saying that “Our own experience and that of other countries, notably France and Germany, clearly proves that it is possible to keep in circulation at a parity of value with gold a large but limited amount of legal tender silver or notes based upon such silver without any provision for compulsory redemption in gold.”). 11. Committee on Banking and Currency (1960, p. 32). Banks were not obliged to pay out gold when redeeming their liabilities, but (from the perspective of a depositor) the right to payment in lawful money amounted to very nearly the same thing because the Gold Standard Act obliged the Treasury to maintain the value of its money at 25.8 grains of gold (of 0.9 fineness) per dollar. 12. The National Banking System was authorized by the National Currency Act of February 25, 1863. The new system was intended to help finance the Civil War by stimulating demand for Treasury bonds. Timberlake (1978, ch. 7), Taus (1943, p. 61), Davis (1910, pp. 36, 74, 80, and 106), and Patterson (1954, pp. 27–29 and 151–52). The National Currency Act was repealed and reenacted in a modified form by the Act of June 3, 1864. The Act of June 3, 1864, was renamed the National Bank Act by the Act of June 20, 1874.

17

The Payments System before World War I

National banks, but not state banks, also issued non–interest-bearing demand notes.13 National Bank Notes A national bank note was a promise of the bank of issue to pay to the bearer, on demand at the bank or at the Treasury in Washington, a stated number of dollars in lawful money. To facilitate redemption at the Treasury, an issuing bank had to maintain at the Treasury a redemption fund equal to at least 5 percent of the value of its outstanding notes.14 A bank could not issue notes in excess of its paid-in capital and had to back its notes with Treasury bonds of equal or greater value. The collateral bonds, held in safekeeping at the Treasury, were valued at the lesser of the market value of the bonds and the principal amount of the bonds.15 The limitation on issuance and the strong backing behind national bank notes left the public confident that the notes could always be exchanged for lawful money.16 A national bank could defease some or all of its note liabilities, and regain possession of the bonds securing the defeased notes, by depositing an equal amount of lawful money with the Treasury.17 As the bank’s notes came into the possession of the Treasury in the normal course of business, for example, in payment of taxes, Treasury officials would withdraw the notes and replace them with the lawful money. In mid-1914, $751 million of national bank notes were outstanding. $715 million circulated in the private economy, $9 million was held in Treasury vaults as part of the Treasury’s operating cash balance, and $26 million was held in Treasury vaults in the process of being redeemed. $736 million of the notes was secured by $741 million of Treasury bonds and $15 million was secured by lawful money. Bank Demand Deposits The various forms of currency—including gold and silver certificates, U.S. notes, and national bank notes—were convenient for effecting payment in hand-to-hand transactions and were cheaper to store and transport than gold 13. State banks issued non–interest-bearing demand notes prior to 1865, but the Act of March 3, 1865, imposed a tax on state bank notes that drove the notes out of existence. The tax on state bank notes was held to be constitutional in Veazie Bank v. Fenno, 8 Wallace 533 (1869). 14. The redemption fund was established by the Act of June 20, 1874. 15. The National Bank Act limited bank note issuance to 90 percent of the lesser of the market value and the principal amount of the collateral bonds, but the Gold Standard Act relaxed the limit to 100 percent. Committee on Banking and Currency (1960, p. 32). 16. Friedman and Schwartz (1963, pp. 21–22) (“national bank notes after [1873] circulated at parity with other currency”). 17. Act of June 20, 1874, and Timberlake (1978, p. 110).

18

Chapter 2

coin. However, as the American economy grew and became increasingly integrated across broad geographic regions after the Civil War, relying on currency for effecting large-value payments and for settling transactions between geographically remote buyers and sellers became increasingly costly; bank demand deposits that allowed depositors to transfer book-entry credit balances by check became increasingly popular. Demand deposits were less susceptible to theft than coin or currency, and it was usually cheaper to make payments at distant locations by check. Check Collection Checks did not, however, eliminate the need for coin and currency in connection with large-value and remote payments, because most check payments required coin or currency for final settlement.18 For example, suppose A made a payment to B by writing a check. If A and B were customers of different banks in the same city, the payment was usually settled by a transfer of coin or currency from A’s bank to B’s bank through a local clearinghouse.19 Following receipt of the check from B, B’s bank would present the check to A’s bank at a session of the clearinghouse and receive payment—in coin or currency—for the check. (Importantly, however, A’s bank would likely be presenting checks drawn on B’s bank at the same clearinghouse session. The great efficiency of the clearinghouse system was that only the net amount had to be settled in coin or currency.) In the last step, A’s bank would debit A’s deposit balance and return the paid item to A. If A and B were customers of banks in different cities but in the same region, the payment was likely to be settled with a transfer of coin or currency at a clearinghouse located in a regional money center.20 Following deposit of the check by B at its bank, B’s bank would forward the item to a correspondent bank in the regional money center for collection. In one simple (and relatively inefficient) scheme, the correspondent would forward the check to A’s bank and receive in payment a draft drawn on another bank in the same regional center at which A’s bank maintained a correspondent balance. The correspondent of B’s bank would collect the draft through the clearinghouse in the regional center and credit the correspondent balance of B’s bank. The correspondent of A’s bank would pay the draft and debit the correspondent balance 18. The exception was when a check was deposited for credit to an account at the same bank that the check was drawn on. In that case payment could be effected by journal entries on the books of the single bank. 19. Cannon (1910) gives an excellent description of clearinghouse operations at the beginning of the twentieth century. 20. West (1977, p. 38).

19

The Payments System before World War I

of A’s bank.21 New York City served the same function for interregional payments.22 The use of New York balances to effect settlements between banks in different regions led to the development of a market for “New York exchange” in most large American cities. If a bank needed to replenish the balance that it maintained with a New York correspondent, it could offer to buy—from a nearby bank, against payment in local funds—funds on deposit in New York. If it had excess balances in its New York account it could offer to sell some of the balances to a nearby bank against payment in local funds.23 Depending on whether New York exchange was scarce or plentiful, $1,000 on deposit in New York could trade at more or less than $1,000 in local funds. Checkable demand deposit accounts and markets for New York exchange did not eliminate the need for shipping coin and currency over long distances. A bank that wanted to replenish its New York balances at a time when New York exchange was scarce (and therefore costly in terms of local funds) sometimes found it cheaper to ship coin or currency to New York than to buy funds already on deposit in New York. Similarly a bank that wanted to repatriate some of its New York balances at a time when New York exchange was plentiful (and therefore cheap) sometimes found it more economical to order coin or currency from its New York correspondent. Shipments of cash from New York banks to banks in the south and west were especially heavy during agricultural harvests in the late summer and fall, when cash was needed by commodity merchants to pay farmers for their crops and by farmers to settle their accounts with local suppliers and to pay field hands for the season’s labor.24 After the crops were purchased and accounts settled, coin and currency would tend to flow back to the New York banks.25 21. There were several more efficient schemes for collecting checks drawn on remote banks. In the “Sedalia, Missouri, plan,” for example, the correspondent of B’s bank would collect the check directly from a bank in the same regional center that acted as a correspondent for A’s bank. Following settlement between the money center banks, the correspondent of B’s bank would credit the correspondent balance of B’s bank and the correspondent of A’s bank would debit the correspondent balance of A’s bank. In the last step, the correspondent of A’s bank would send the check to A’s bank so that bank could debit A’s account and return the paid item to A. See Spahr (1926, pp. 124–25). 22. West (1977, p. 38). 23. Garbade and Silber (1979). 24. Andrew (1906), Kemmerer (1910), and Goodhart (1969, pp. 38–41). Kemmerer writes (p. 29) that “The demands upon banks in [the south and west] for funds to use in the moving of the crops requires them to call upon New York banks for the payment of their deposits . . .. In response to this demand there is a very strong movement of cash from New York City to the West and South from August through October . . ..” 25. Kemmerer (1910, p. 28) (“December and January are the months in which the return flow of cash to New York City principally takes place.”).

20

Chapter 2

Reserve Requirements Bank demand deposits, like national bank notes, were payable in lawful money, but they were not secured with Treasury bonds. This left depositors dependant on other bank assets, primarily commercial and industrial loans. If a bank became insolvent, there was a good chance that those assets would not cover the bank’s deposit liabilities. To limit the prospect that a bank might so far expand its loans and deposits that it would be unable to satisfy depositor demands for cash, banks were required to satisfy statutory reserve requirements. National banks in three “central reserve cities,” New York, Chicago, and St. Louis, were required to maintain lawful money in their vaults equal to at least 25 percent of their deposit liabilities. National banks in about fifty other large “reserve cities” also had to keep 25 percent of their deposit liabilities in reserve, but up to half of the required reserves could be kept as a deposit at a central reserve city bank. All other national banks—sometimes called “country banks”—had to keep 15 percent of their deposit liabilities in reserve, three-fifths of which could be held as a deposit at a reserve city or central reserve city bank. (Balances on deposit with the Treasury in the 5 percent note redemption fund counted toward satisfying the reserve requirements of a national bank.) State banks generally had lower reserve requirements and could meet some or all of their reserve requirements with national bank notes. There were two notable features of national bank reserve requirements. First, the requirements for reserve city and country banks were nicely congruent with the structure of the payments system: banks outside of central reserve cities could satisfy part of their reserve requirements with correspondent balances that they would have maintained in any case to cover checks written by their customers.26 Second, the scheme provided for “pyramiding” reserves: only central reserve city banks had to keep all of their reserves in the form of lawful money. Country banks could keep up to 60 percent of their reserves as deposits with reserve city banks, and reserve city banks could keep as much as half of their reserves as deposits with central reserve city banks. This meant 26. West (1977, p. 38) (“Many country banks kept deposits in city banks for clearing purposes and to redeem notes. City banks outside New York also kept deposits in New York banks for the same purposes. If the National Banking Act had not legitimized this common practice, deposits held for clearings or convenience would not have been legal reserves, and a bank’s earning power would have been lowered by the amount of deposits it found necessary to hold in other banks. . . . Interbank deposits benefited country banks through conveniences such as clearing and city banks because of the increased lending power such deposits allowed.”) and Committee on Banking and Currency (1913, p. 19) (“The original reason for creating this . . . system of reserves was that inasmuch as central banking institutions were absent, and inasmuch as banks outside of centers were obliged to keep exchange funds on deposit with other banks in such centers, it was fair to allow exchange balances with such centrally located banks to count as reserves inasmuch as they were presumably at all times available in cash.”).

21

The Payments System before World War I

that central reserve city banks in New York, Chicago, and St. Louis held a large fraction of the lawful money available to satisfy depositor requests for withdrawals from national banks. Because of the unique role of New York banks in settling interregional payments, New York banks held a particularly large fraction of aggregate reserves of lawful money. The Problem of an Inelastic Currency The principle problem with the American banking system at the beginning of the twentieth century was that there was no “lender of last resort.” When Americans wanted to hold relatively more of their money in the form of cash (and relatively less in the form of bank deposits), such as in the fall when cash was needed to “move the crops” or any time the solvency of the banking system came into question, banks in aggregate had only a limited ability to obtain the needed cash. The stock of currency, especially U.S. notes, silver certificates, and national bank notes, was notoriously inelastic. Edwin Kemmerer, professor of economics and finance at Cornell University, opened his classic study of seasonal variation in the demand for money by noting that: The most common criticism of our American currency system is its alleged inelasticity or irresponsiveness to trade demands, this inelasticity is sometimes considered with particular reference to panic periods that occur at more or less irregular and widely separated times, and sometimes with particular reference to regularly recurring seasonal fluctuations in the demand for money . . ..27

The stock of U.S. notes was fixed by statute and the stock of silver certificates was limited by the requirement that those certificates could only be issued against the deposit of silver dollars and by the fact that the Treasury had stopped minting silver dollars. National banks could increase their note issues by depositing Treasury bonds with the Treasury, but commonly chose not to do so in response to seasonal variation in the demand for currency.28 Figure 2.1 shows the variation in national bank notes after 1906. Prior to the onset of World War I (in mid-1914) there is no visible evidence of any seasonal variation and no more than a limited response to an exceptionally strong shift in the demand for money during the Panic of 1907, when public demand for bank deposits contracted sharply and demand for currency expanded equally sharply. (However, the stock of money was not altogether fixed. Box 2.1 describes two sources of short-run elasticity: foreign gold and Treasury vault cash.) 27. Kemmerer (1910, p. 13). 28. Goodhart (1969, pp. 31–33), Friedman and Schwartz (1963, p. 23), Noyes (1910, p. 9), and Kemmerer (1910, pp. 152–53). See also Wallace and Zhu (2007).

Chapter 2

1,000

Millions of dollars

750

500

250

Secured by other assets

Secured by lawful money

19 18

19 17

16 19

19 15

19 14

19 13

19 12

11 19

10 19

09 19

08 19

07 19

06

0 19

22

Secured by Treasury bonds

Figure 2.1 National bank notes outstanding, end of quarter, March 1906 to June 1917. Notes secured by lawful money are notes defeased by deposits of lawful money pursuant to section 4 of the Act of June 20, 1874. Notes secured by other assets are notes issued pursuant to the emergency currency provisions of the Aldrich–Vreeland Act and secured by commercial loans.

The relatively inelastic stock of currency had particularly severe consequences for the New York banks that were called upon annually to supply the cash needed to move the crops. From time to time those banks became so stressed, so drained of the lawful money needed to satisfy reserve requirements, that depositors with no agricultural interests but concerned about the solvency of the banks would convert their deposits to cash as a precautionary matter. Depositor anxiety blossomed into full-fledged banking panics in 1873, 1884, 1890, 1893, and 1907.29 The panics of 1873, 1893, and 1907 were so severe that banks in New York, and subsequently banks around the country, were forced to suspend convertibility of their deposit liabilities into lawful money. The ensuing disruption of the payments system led to a contraction in trade, inventory backups, and, ultimately, a broad decline in aggregate economic activity. For many Americans, the panic of 1907 was the last straw.30 America needed an “elastic” currency if it was to become a global financial power—some mechanism that would allow depositors to convert their bank deposits to cash, 29. Sprague (1910). 30. Bruner and Carr (2007), West (1977), and Warburg (1910).

23

The Payments System before World War I

Box 2.1 Sources of Short-Run Elasticity in the Stock of Currency There were two major sources of elasticity in the stock of currency available to Americans in the first decade of the twentieth century: foreign gold and Treasury vault cash.a Advances in short-term interest rates in the New York money market, such as those that resulted from seasonal demands for currency to move the crops, were sometimes large enough to attract short-term investment funds from European money centers. Since all of the major European countries were on the gold standard, it didn’t take too much of a rate premium in New York to entice foreign investors to convert their local currencies to gold, ship the gold to the United States, and convert the gold into dollars—by, for example, presenting the gold at the subtreasury in New York and requesting gold certificates—and then using the dollars to fund the investments that had triggered the gold flow in the first place.b Discretionary Treasury repurchases of Treasury debt, prepayments of interest, and deposits of currency in national banks were a second source of elasticity in the stock of currency. When the Treasury ran a budget surplus, it received more revenues than it paid out and cash accumulated in its vaults. During the 1890s and early 1900s, Treasury officials developed increasingly sophisticated procedures for returning cash to the private sector as a way of relieving seasonal credit stringencies.c The most innovative policies were developed by Treasury Secretary Leslie Shaw between 1902 and early 1907,d but Treasury operations to relive strains in the New York money market date to before the Civil War.e a. Kemmerer (1910, p. 30). b. Bloomfield (1959, p. 20) (noting that the major world capital markets, including New York, London, Berlin, and Paris, were “linked together, given the widespread confidence in the stability of exchange rates and in the free convertibility of the currencies concerned, by . . . equilibrating flows of foreign balances that were highly sensitive to interest rate differentials and to exchange rate movements within the narrow limits of the gold points.”). See also Bloomfield (1959, p. 42) (“Changes in interest rates had undeniably an important influence on the international flow of capital (especially short-term funds) in the pre-1914 world, at least so far as the more developed money markets were concerned. Given the fact that exchange rate stability was taken for granted among the gold standard countries, international movements of capital proved highly sensitive in the short run to changes in interest rates in different countries.”). However, Goodhart (1969) points out that the time and expense of shipping gold from Europe to New York introduced nontrivial frictions into such interest rate arbitrage. c. Kinley (1910, pp. 119–46, 187, 256–57, and 278–81), Taus (1943, pp. 85–133), and Timberlake (1978, pp. 146, 151, and 174–83). d. Andrew (1907, 1908a), Patton (1907), and Kinley (1910) were sharply critical of Shaw’s policies, but Timberlake (1978, p. 185) quotes Professor Lloyd Mints of the University of Chicago as saying (in 1952) that Shaw was the “only good secretary of the Treasury we have ever had.” e. Timberlake (1978, pp. 77–83) and Taus (1943, pp. 49–57).

seasonally and in times of stress. To fill the gap until a new monetary system could be put in place, Congress provided a backstop measure, the Aldrich– Vreeland Act. Five and a half years later, and after much debate, it adopted the Federal Reserve Act. The Aldrich–Vreeland Act The Aldrich–Vreeland Act of May 20, 1908, provided that, with the approval of the Secretary of the Treasury, a national bank could issue national bank

24

Chapter 2

notes against a pledge of commercial loans.31 This imparted a crucial elasticity to the currency that had been lacking previously.32 The currency provisions of the Aldrich–Vreeland Act were used only once, in August 1914 when European demands to liquidate American investments at the start of World War I threatened to force national banks to suspend convertibility of their deposits and threatened to force the Treasury to suspend convertibility of its money into gold. Secretary of the Treasury William McAdoo invoked the emergency provisions of the Aldrich–Vreeland Act to effect the dramatic expansion in national bank notes shown in figure 2.1.33 Milton Friedman and Anna Schwartz conclude that “the availability of the emergency [money in 1914] probably prevented . . . the restriction of payments by the banking system.”34 The Federal Reserve Act and Federal Reserve Money The preamble to the Federal Reserve Act of December 13, 1913, states that the purpose of the act was, inter alia, “to furnish an elastic currency [and] to afford means of rediscounting commercial paper.” The term “commercial paper” was commonly understood to mean short-term promissory notes that arose out of commercial transactions (e.g., purchases of raw materials for producing finished goods or purchases of finished goods for resale) and that would be “self-liquidating” (in the sense that the activities financed by the notes generated readily saleable goods).35 The Federal Reserve Act was designed to mitigate the problem that had become manifest during the preceding fifty years: that there was no lender of last resort to whom banks in aggregate could go to satisfy the demands of their customers for currency, resulting in regular seasonal strains that sometimes boiled over into crises and panics. The key institutional innovation provided in the Federal Reserve Act was the establishment of a confederation of regional “reserve” banks. The new 31. More specifically, the act provided for the formation of “national currency associations” by groups of ten or more national banks. Upon pledge of commercial loans maturing in not more than four months and with the approval of the Secretary of the Treasury, a bank that was a member of a national currency association could issue national bank notes equal to 75 percent of the value of the pledged loans. Later, after market stresses had subsided, the bank could defease the notes and reclaim its collateral by depositing lawful money with the Treasury. 32. See Taus (1943, p. 129) (noting that “the clause [of the Aldrich–Vreeland Act] allowing the issuance of emergency currency based on commercial paper showed that the legislators and the public were aware of the lack of elasticity of the national bank notes and its consequences”). 33. Silber (2007) describes how McAdoo used the Aldrich–Vreeland Act, as well as other statutory and extra-legal devices, to keep America on the gold standard. See also Sprague (1914, 1915). 34. Friedman and Schwartz (1963, p. 172). 35. For example, section 1 of the Aldrich–Vreeland Act states that “The term ‘commercial paper’ shall be held to include only notes representing actual commercial transactions . . . and have not exceeding four months to run.”). See also Hackley (1973, pp. 27–33).

25

The Payments System before World War I

system was not a conventional central bank, like the Bank of England, because power and responsibility was diffused among the regional banks and between those banks and a supervisory board in Washington. This was acceptable to agrarian Democrats, who feared a central bank controlled by New York bankers, as well as to northeastern Republicans, who feared a central bank controlled by Washington politicians.36 National banks were required to become members of the new system; state banks were permitted to join. The new Federal Reserve Banks were banks of issue and deposit. They issued Federal Reserve notes that were expected to replace national bank notes as the principle currency of the country, they accepted funds on deposit from member banks, and they provided for the transfer of such deposit balances to the accounts of other member banks.37 Federal Reserve notes were not legal tender but they were obligations of the United States, they were “receivable for all taxes, customs, and other public dues,” and they were payable in gold at the Treasury in Washington and in gold or lawful money at any Federal Reserve Bank.38 Each Reserve Bank was required to maintain a gold reserve equal to not less than 40 percent of its note liabilities, plus not less than 35 percent of its deposit liabilities, and it was required to maintain a gold deposit in a Treasury note redemption fund equal to at least 5 percent of its notes. (Gold on deposit in the note redemption fund counted toward the 40 percent note reserve requirement.) The Reserve Banks were further required to pledge commercial paper of equal value against their note liabilities.39 The Federal Reserve Act did away with the pyramiding of bank reserves by providing that only Federal Reserve deposits and lawful money were eligible reserve assets for member banks.40 As shown in table 2.1, reserves had to be held in the form of either lawful money or Federal Reserve Bank deposits and 36. West (1977) and Timberlake (1978, ch. 13) discuss the origins of the governance structure of the Federal Reserve System. 37. Federal Reserve Act, section 13. 38. Federal Reserve Act, section 16, Willis (1915, p. 253) (observing that Federal Reserve notes are not legal tender), and Reed (1922, p. 215) (observing that Federal Reserve notes are not “legal tender for private payments”). 39. Federal Reserve Act, section 16. Section 16 was amended by section 7 of the Act of June 21, 1917, to allow the pledge of either commercial paper or gold, and to allow gold held as collateral against note issues to count toward the gold reserve requirement. The practical effect of this change was to reduce the backing behind Federal Reserve notes to at least 40 percent gold and the balance in gold or commercial paper. 40. Anomalously Federal Reserve notes were not a reserve asset. See Miller (1921, p. 182) (“Under the terms of the Federal Reserve act, federal reserve notes were not available as legal reserve money to member banks.”), Reed (1922, p. 215) (Federal Reserve notes “may not be counted as a part of the legal reserve of a member bank”), Glass (1927, p. 334) (noting that a Senate provision “permitting Federal reserve notes to be used as reserves in the individual banks” was removed in the course of House-Senate conference committee negotiations prior to final passage of the Federal Reserve Act), and Timberlake (1978, p. 203) (the Federal Reserve Act “prohibited the use of Federal Reserve notes as eligible reserves for member banks”).

26

Chapter 2

Table 2.1 Member bank reserve requirements under the Federal Reserve Act Member banks in central reserve cities (New York, Chicago, and St. Louis) 18% of demand deposits and 5% of time deposits • 6/18ths in lawful money in their vaults • 7/18ths deposited with their Federal Reserve Bank • Balance in vault cash or on deposit, at the discretion of the member bank Member banks in reserve cities 15% of demand deposits and 5% of time deposits • 5/15ths in lawful money in their vaults • 6/15ths deposited with their Federal Reserve Bank • Balance in vault cash or on deposit, at the discretion of the member bank Other member banks 12% of demand deposits and 5% of time deposits • 4/12ths in lawful money in their vaults • 5/12ths deposited with their Federal Reserve Bank • Balance in vault cash or on deposit, at the discretion of the member bank Ignores the provisions of a three-year transition period between the reserve requirements of the National Bank Act and the post-transition reserve requirements of the Federal Reserve Act.

could not be kept as a deposit at another bank. The structure of Reserve Bank and member bank reserve requirements was designed to concentrate U.S. gold reserves in the Reserve Banks, ending the dispersion of those reserves among reserve city and central reserve city banks.41 (However, gold reserves were not as concentrated as they would have been had Congress provided for a single central bank.) The new Federal Reserve Banks were also banks of discount. The Federal Reserve Act authorized the banks to purchase, or discount, short-term commercial loans from member banks at discount rates posted by the banks.42 To be eligible for discount, a loan had to arise out of an actual commercial 41. In an effort to further concentrate gold reserves following the entry of the United States into World War I, the Act of June 21, 1917, excluded vault cash as a reserve asset and limited reserve assets to Federal Reserve bank deposits. See Miller (1921, p. 180) (stating that “the policy . . . was to impound as much of the stock of monetary gold in the country as possible in the federal reserve system”), Warburg (1930, p. 157) (stating that the legislation “enabled the Reserve System to corral and mobilize the country’s gold”) and Meltzer (2003, p. 79) (noting that the legislation increased gold held by the Federal Reserve in excess of requirements by $300 million). The same act also reduced reserve requirements below the levels shown in table 2.1, to 3 percent for time deposits, 7 percent for country bank demand deposits, 10 percent for reserve city bank demand deposits, and 13 percent for central reserve city bank demand deposits. 42. It was expected that the Banks would raise their posted discount rates when they needed to retain or attract gold and lower their discount rates when gold reserves were ample. West (1977, p. 163) (“When the United States returned to central banking in 1914 [the ‘rules of the game’ were] widely accepted [i.e., that the central bank would raise the discount rate to reverse gold outflows and lower the discount rate to shut off gold inflows]. The discount rate was made a chief tool of the [Federal Reserve] system.”). Bloomfield (1959, pp. 23 and 25) points out that central bank discount rate policies were not automatic even under the pre-war gold standard, that central bankers exercised substantial discretion in setting discount rates, and that there was a significant asymmetry between the need to raise discount rates in the face of gold outflows and the ability to lower discount rates in the face of gold inflows. See also Walker (1934) and Warburg (1910).

27

The Payments System before World War I

transaction and could not have been made for the purpose of purchasing or carrying securities (other than U.S. Treasury securities).43 The discount facility was intended to give member banks a reliable way to satisfy customer demands for currency, materially enhancing the elasticity of the nation’s stock of currency. Limiting the discount privilege to short-term commercial loans was intended to allow the stock of currency to expand and contract according to the needs of business—such as when cash was needed to move the crops—but not as a function of, for example, speculative demand for loans to purchase and carry common stock. Importantly, however, the ability of the Reserve Banks to create money was not unlimited, because the Reserve Banks were themselves subject to gold reserve requirements and were obliged to redeem their notes for gold on demand. Speaking of the interval during which Congress worked out the architecture of the new system, between the panic of 1907 and late 1913, Richard Timberlake observed that “No one in this era thought of controverting the gold standard as the secular determinant of the money stock. Gold was to be the reserve, though it was to be centralized and economized.”44 Retiring National Bank Notes The Federal Reserve Act contemplated that, over time, national bank notes would be withdraw from circulation and replaced by Federal Reserve notes. Section 18 of the act provided that, beginning in late 1915, a national bank that wanted to retire some of its outstanding notes could sell at par to a Federal Reserve Bank the Treasury bonds securing the notes and use the proceeds to defease the notes. At the option of the Secretary of the Treasury, the Reserve Bank could, in turn, exchange the bonds for special issues of 1-year Treasury 43. Federal Reserve Act, section 13. The Act of September 7, 1916, amended the Federal Reserve Act to allow a Reserve Bank to lend to a member bank for up to 15 days if the loan was secured by eligible commercial loans or U.S. Treasury securities. The Board recommended adding this authority primarily as a matter of convenience and administrative efficiency. 1915 Federal Reserve Board Annual Report, p. 22 (“In order to enable member banks to obtain prompt and economical accommodations for periods not to exceed fifteen days, the Federal Reserve Banks should be permitted to make advances to member banks against their promissory notes secured by such notes, drafts, bills of exchange, and bankers acceptances as the law at present permits to be discounted or purchased; or against the deposit or pledge of United States Government bonds, the purchase of which is now permitted under the law.”). Following adoption of the provision, the Federal Reserve Bank of New York observed that “Member banks are [now] enabled to borrow for short periods . . . without the accounting labor either to them or to the Reserve Bank of recording and computing interest on large numbers of small notes.” 1916 Federal Reserve Bank of New York Annual Report, p. 11. See also Beckhart (1924, p. 214) (“The passage of this amendment had been recommended by the Board . . . as a convenience to large city banks temporarily in need of funds. With this change they would be enabled to borrow on their own portfolios as collateral without the knowledge of their customers, and at the same time save the Reserve Banks the arduous labor of computing the interest charge on each of a large number of small items.”). 44. Timberlake (1978, p. 191). Section 26 of the Federal Reserve Act explicitly reaffirmed the Gold Standard Act of 1900.

28

Chapter 2

notes and 30-year Treasury bonds.45 In its 1916 annual report, the Federal Reserve Board noted that the purpose of section 18 “was to maintain the market for United States bonds deposited to secure circulation, and to insure a gradual but steady retirement of national-bank notes . . ..”46 Pursuant to the provisions of section 18, in 1916 and the first three months of 1917 the Federal Reserve Banks acquired a total of $56 million of Treasury bonds that national banks had pledged as collateral against outstanding notes. The Reserve Banks converted the bonds to $27 million of special issue 1-year Treasury notes and $29 million of special issue 30-year Treasury bonds.47 As a result the stock of national bank notes fell from $776 million at the end of 1915 to $718 million at the end of March 1917 (see figure 2.1). National banks lost interest in retiring their notes in the spring of 1917, largely as a result of increasing economic activity following the American declaration of war on Germany, and the Treasury and the Federal Reserve Banks suspended their efforts to retire national bank notes.48 45. Alternatively, the Reserve Bank could use the bonds it received from a national bank as the basis for issuing its own Federal Reserve bank notes (which were not the same as Federal Reserve notes and which were not subject to either a gold reserve requirement or a collateral requirement). This option, authorized in section 4 and section 18 of the Federal Reserve Act, allowed for the replacement of national bank notes with Federal Reserve bank notes in the event a Reserve Bank did not have enough gold and commercial loans to issue conventional Federal Reserve notes. The Federal Reserve Bank of Dallas and the Federal Reserve Bank of Kansas City issued a total of $12 million of Federal Reserve bank notes pursuant to this provision. 1916 Federal Reserve Board Annual Report, pp. 8–9 and 52–53. 46. 1916 Federal Reserve Board Annual Report, p. 8. 47. 1918 Treasury Annual Report, pp. 78 and 79. 48. 1918 Treasury Annual Report, p. 78.

3

Treasury Debt Management before World War I

This chapter describes Treasury debt management in the decades prior to the First World War—including the characteristics of Treasury debt and the primary market for Treasury securities. Treasury Debt Six Treasury issues, with an aggregate principal value of $963 million, were outstanding in mid-1914 (table 3.1). Congress had authorized each issue for a specific purpose. For example, the War Revenue Act of June 13, 1898, authorized the bonds sold to finance the Spanish–American War: [T]he Secretary of the Treasury is hereby authorized to borrow on the credit of the United States from time to time as the proceeds may be required to defray expenditures authorized on account of the existing war (such proceeds when received to be used only for the purpose of meeting such war expenditures) the sum of four hundred million dollars, or so much thereof as may be necessary . . ..

This narrow authority illustrates how Congress viewed Treasury debt: as a vehicle for financing specific projects. Congress left little or no room for the exercise of discretion by Treasury officials with respect to issue characteristics. The War Revenue Act of 1898 specified the maturity of the bonds to be issued (20 years), the coupon rate (3 percent), and even the price at which the bonds were to be offered to the public (100 percent of principal value). One historian remarked that “it was considered the duty of Congress to determine the interest rates and maturities of the government’s obligations. It had always been the custom of the legislative body to decide these terms, and not to leave them to the discretion of the Secretary of the Treasury.”1 Bearer Bonds and Registered Bonds Treasury bonds came in two varieties: bearer and registered. A bearer bond (sometimes called a coupon bond) was an engraved certificate that consisted of a “corpus,” or main body, reciting the government’s promise to pay principal and interest, and a series of detachable coupons, each of which was a claim to an interest payment on a specific date. To obtain an interest payment an investor detached the appropriately dated coupon and sent it (through the banking system) to the Treasury for collection. When the bond matured the investor asserted his claim for repayment of principal by sending the corpus for collection. The government’s promises ran to whomever held the bond, so a bondholder could transfer ownership of his claims simply by delivering the bond to a new owner. 1. Patterson (1954, p. 84).

30

Chapter 3

Table 3.1 Treasury bonds outstanding June 30, 1914 4% Bonds redeemable in 1925 • Dated February 1, 1895, redeemable at the option of the Treasury on or after February 1, 1925, interest paid quarterly • Authorized by the act of January 14, 1875, to replenish Treasury holdings of gold coin • $62.3 million issued March 1895 and an additional $100.0 million issued February 1896 3% Spanish–American War bonds of 1918 • Dated August 1, 1898, due August 1, 1918, redeemable at the option of the Treasury on or after August 1, 1908, interest paid quarterly • Authorized by the War Revenue Act of June 13, 1898, to finance the Spanish–American War • $198.8 million issued August 1898 2% Consols redeemable in 1930 • Dated April 1, 1900, redeemable at the option of the Treasury on or after April 1, 1930, interest paid quarterly • Authorized by the Gold Standard Act of March 14, 1900, to refinance higher coupon, shorter maturity debt to a lower coupon rate and a more distant redemption date • Issued in exchange for 5 percent bonds redeemable in 1904, 4 percent bonds redeemable in 1907, and the 3 percent Spanish–American War Bonds of 1918, on the basis of equal principal amounts 2% Panama Canal bonds of 1936 • Dated August 1, 1906, due August 1, 1936, redeemable at the option of the Treasury on or after August 1, 1916, interest paid quarterly • Authorized by the Spooner Act of June 28, 1902, to finance construction of the Panama Canal • $30.0 million issued August 1906 and an additional $24.6 million issued November 1907 2% Panama Canal bonds of 1938 • Dated November 1, 1908, due November 1, 1938, redeemable at the option of the Treasury on or after November 1, 1918, interest paid quarterly • Authorized by the Spooner Act • $30.0 million issued December 1908 3% Panama Canal bonds of 1961 • Dated June 1, 1911, due June 1, 1961, interest paid quarterly • Authorized by the Payne–Aldrich Tariff Act of August 5, 1909, to finance construction of the Panama Canal • $50.0 million issued June 1911

A bond was said to be registered if the government’s promise to pay principal and interest ran to a person whose name and address were recorded with the Treasury. There was an engraved certificate associated with a registered bond, on the face of which the name of the owner appeared, but the certificate served primarily as a device for effecting change in the Treasury’s record of ownership. An investor who wanted to convey a registered bond to a new owner inscribed the re-registration instructions on the back of the bond and sent it to the Treasury. Upon receipt, the Treasury would change its records, destroy the old certificate, and issue a new certificate to the new owner. Because it knew the name and address of the owner of a registered bond, the Treasury could send interest payments directly to the owner on its own initiative. It could have done the same for the principal payment, but chose instead to require tender of the certificate as a way to recover matured certificates.

31

Treasury Debt Management before World War I

Bearer bonds and registered bonds had both advantages and disadvantages. Bearer bonds could be transferred readily, but had to be kept safe from loss, destruction, and theft. Registered bonds could be replaced if they were lost, stolen, or destroyed, so it was less important to provide exceptionally secure safe-keeping. Additionally the Treasury paid interest on registered bonds without requiring tender of any coupons. However, changing the record of ownership of a registered bond was time-consuming, and the Treasury did not allow transfers for several weeks preceding an interest payment. The Secondary Market for Treasury Bonds The secondary market for Treasury bonds was limited by the relatively large quantity of securities held for safekeeping in Treasury vaults. In mid-1914, 81 percent of outstanding Treasury bonds were pledged against national bank notes or Treasury deposits at national banks (table 3.2). (The preceding chapter discussed the requirement that national bank notes had to be secured with Treasury bonds. Box 3.1 discusses collateral requirements for Treasury deposits in national banks.) Trading in Treasury bonds was largely limited to purchases by national banks seeking to expand their note issues (or government deposits) from other national banks seeking to contract their notes (or government deposits). During—and in the first few decades following—the Civil War, most secondary market transactions in Treasury bonds took place on the New York Table 3.2 Treasury bonds outstanding June 30, 1914, and pledged as collateral for national bank notes and Treasury deposits in national banks ($millions) Bonds outstanding on June 30, 1914 4% Bonds redeemable in 1925 3% Spanish–American War bonds of 1918 2% Consols redeemable in 1930 2% Panama Canal bonds of 1936 2% Panama Canal bonds of 1938 3% Panama Canal bonds of 1961 Total

Held as collateral for national bank notes

Held as collateral for Treasury deposits in national banks

Percent held as collateral

118.5

32.9

3.9

31.1

63.9

21.5

4.6

40.8

646.3

604.7

12.7

95.5

54.6

52.9

1.3

99.3

30.0

28.9

0.5

98.0

NA

14.7

29.4

740.9

37.7

80.8

50.0 963.3

Source: 1914 Treasury Annual Report, pp. 211–13 and 302.

32

Chapter 3

Box 3.1 Treasury Deposits in National Banks The U.S. Treasury, like households and businesses, maintains a cash balance to buffer unanticipated and short-run differences between receipts and expenditures. Early in the twentieth century, Treasury balances were relatively large by contemporary standards. In mid-1914 the Treasury held a total cash balance of $242 million—compared to receipts and expenditures of about $700 million per annum.a The Treasury kept its cash in two forms: as currency and coin in vaults at subtreasuries located in major cities around the country and as book-entry deposits at national banks.b On June 30, 1914, the Treasury held a total of $165 million at its subtreasuries and at its main office in Washington and had $77 million on deposit with more than 1,500 national banks around the country.c The Treasury deposited funds in national banks for either of two reasons: to facilitate its own transactions, or to return reserves to the banking system. Banks that held Treasury transaction balances were called “regular” depositories. The Wall Street Journal described regular depositories as “for the convenience of the Government for the holding of internal revenue and customs receipts.”d Banks that received Treasury balances for purposes of reserves management were called “temporary” depositories. The Journal described temporary depositaries as “designated for special deposits of public moneys. When there has accumulated a large surplus in the Treasury the Secretary has the power to place a certain portion of such surplus in circulation through the banks, the object being not to create unnecessary stringency in the money market.”e Special deposits were also maintained in “inactive” accounts at regular depositories. Treasury balances at both regular and temporary depositories had to be secured by a pledge of securities. Up to 30 percent of the required collateral could consist of state and local government bonds; the balance had to be in the form of Treasury securities.f In mid1914 the Treasury held $64 million of bonds pledged to secure deposits at national banks, including $38 million of Treasury bonds.g a. 1914 Treasury Annual Report, pp. 43, 44, and 299. The Treasury had to maintain a large balance because it did not then issue short-term debt—like Treasury bills—for cash management purposes. b. The Independent Treasury Act of August 6, 1846, authorized the Treasury to maintain a network of subtreasuries to make and receive payments and to hold Treasury cash balances. See Kinley (1910, pp. 8, 50, and 53–60). The use of national banks as government depositories was first authorized by the National Currency Act of February 25, 1863, and the National Bank Act of June 3, 1964. c. 1914 Treasury Annual Report, pp. 299, 304, and 306. d. “Money,” Wall Street Journal, April 26, 1912, p. 8. The Federal Reserve Banks assumed responsibility for much of the Treasury’s transactional needs when Treasury Secretary William McAdoo appointed the Banks as depositories and fiscal agents of the United States, effective January 1, 1916. However, the Treasury continued to use its regular accounts at national banks in cities that did not have Reserve Banks or branches. 1916 Treasury Annual Report, p. 275. e. “Money,” Wall Street Journal, April 26, 1912, p. 8. f. 1913 Treasury Annual Report, pp. 211–13, “Must Pay Interest on Nation’s Cash,” New York Times, May 1, 1913, p. 1, and “Banks Must Pay Interest on All Government Deposits,” Wall Street Journal, May 2, 1913, p. 8. National banks were not required to maintain reserves against public deposits. See section 14 of the Aldrich–Vreeland Act of May 30, 1908. g. 1914 Treasury Annual Report, p. 302.

33

Treasury Debt Management before World War I

Box 3.1 Continued Prior to 1908, national banks were not required to pay interest on federal government deposits. That changed when, in 1908, Congress mandated payment of interest on inactive balances at regular depositories and on all balances at temporary depositories at a rate to be prescribed by the Secretary of the Treasury, but not less than one percent per annum.h In June 1908 Treasury Secretary George Cortelyou prescribed the minimum interest required: 1 percent per annum on inactive balances and on all balances at temporary depositories.i The rate was raised to 2 percent in the spring of 1912.j A year later Treasury Secretary William McAdoo extended the requirement to 2 percent on all federal government balances.k h. Aldrich–Vreeland Act of May 30, 1908. The Aldrich–Vreeland Act was silent on the question of interest on transaction balances at regular depositories, neither requiring nor prohibiting payment of such interest. i. “Government Deposits,” Wall Street Journal, June 5, 1908, p. 6, “Effect of New Currency Act for Interest on U.S. Deposits,” New York Times, June 6, 1908, p. 8, and 1908 Treasury Annual Report, p. 165. j. “Banks Must Pay 2 Per Cent,” New York Times, April 24, 1912, p. 15, “Government Special Deposits,” Wall Street Journal, April 25, 1912, p. 8, “Money,” Wall Street Journal, April 26, 1912, p. 8, and 1912 Treasury Annual Report, p. 149. k. “Must Pay Interest on Nation’s Cash,” New York Times, May 1, 1913, p. 1, “Banks Must Pay Interest on All Government Deposits,” Wall Street Journal, May 2, 1913, p. 8, and 1913 Treasury Annual Report, pp. 5 and 211.

Stock Exchange.2 However, by the turn of the century buyers and sellers had shifted to over-the-counter trading with dealers who quoted net prices with no commissions on bid-offer spreads of from 1⁄8 to ¼ percent of principal amount.3 A 1959 study of the Treasury securities market reported that: By 1900 the bulk of the trading in Government securities had . . . shifted . . . to an over-the-counter specialized dealer market. This was primarily the result of the circulation privileges accorded to Government bonds, whereby qualified banks were permitted to issue currency against Government bonds deposited by them. The dealers made it their business to demonstrate to banks the profits that could be derived from such circulation accounts, and to provide all the necessary technical services for putting such accounts into operation. By offering banks such “package” transactions the specialized dealers captured the bulk of the trading in Government bonds from the Exchange . . ..4

2. Childs (1947, p. 360) says of that period, “Sellers relied mostly on the Stock Exchange.” 3. Childs (1947, p. 362). 4. U.S. Treasury and Federal Reserve System (1959, p. 96). Childs (1947, pp. 361–62) states that circulation accounts “created the business motive for private banking houses to perform as dealers” and describes how dealers went about the business of servicing the accounts: “A circulation account was easily put into effect by the dealer, who financed the par value cost of the bonds and received the currency from the Comptroller of the Currency against deposit of the bonds in Washington, assigned to the Treasurer of the United States in trust for a bank’s account. The dealer requested the Treasury to prepare plates and currency for a specified bank, which currency the dealer arranged to complete with facsimile signatures of the bank’s officers. The dealer accepted the currency in payment for the par value of the bonds, and put it into circulation. The bank only paid the dealer the premium price of the bonds. The operation was automatic, as the circulation account was established without disturbing the bank’s balances to pay for the par value of the bonds.”

34

Chapter 3

The dealer market was so firmly established by 1914 that it hardly mattered to market participants that the New York Stock Exchange closed—from July 31 to December 12, 1914 – following the initiation of hostilities in World War I.5 The Primary Market for Treasury Bonds In the decades prior to World War I, Treasury bonds were usually (but not always) sold at auction. The public nature of the offerings: “to all subscribers on terms fixed by public advertisement,” represented an important change from practices during the Civil War and in the decade following.6 Nevertheless, new issues were not widely distributed. Auction offerings usually resulted in the Treasury selling bonds primarily to national banks, because of bank demand for Treasury bonds as collateral for bank notes and public deposits. An unusual private sale of Treasury bonds in 1895—undertaken in circumstances where there was insufficient time for a public offering—produced what many interpreted as inappropriately large profits for the buyers and triggered a groundswell of demand for “popular” offerings conveniently available to individual investors. However, the auction format was not conducive to attracting broad public participation and banks tended to dominate bidding even when Treasury officials made strenuous efforts to attract individual investors. The 1894 Auction Sales of 5 Percent Bonds Redeemable in 1904 In 1893, before the monometallic gold standard was firmly established as the national standard of monetary value, Treasury officials began to encounter persistent problems maintaining a gold reserve sufficient to ensure convertibility of U.S. notes to gold.7 Market commentators viewed a $100 million 5. Childs (1947, pp. 109–10) (“the cessation of trading upon the Stock Exchange was not even a factor with respect to the Government bond market for the reason that virtually all purchases and sales of Government securities were customarily made through dealers specializing in those bonds. . . . For every $1,000 Government bond sales recorded on the Exchange there were $1,000,000 par value unrecorded transactions made by a very few banking houses which were known as Government bond dealers.”). The Stock Exchange closed in support of Treasury Secretary William McAdoo’s efforts to staunch the outflow of gold from the U.S. to Europe at the start of the war. Silber (2007). 6. 1878 Treasury Annual Report, p. xviii. Adams (1887, pp. 235–36) states that the Treasury first began to offer bonds directly to the general public in 1877: “Previous to the summer of 1877, all operations in refunding were carried on by syndicates . . .. But when Secretary Sherman took the Treasury portfolio, the plan of placing bonds by syndicates was abandoned for sale upon public advertisements, or, as it was termed, ‘under circulars.’” 7. Simon (1960, 1968), Dewey (1934, pp. 440–47), and Timberlake (1978, pp. 154–70). See also Timberlake (1978, p. 174) (“Between 1894 and 1899 the government’s budget was constantly in deficit.”).

35

Treasury Debt Management before World War I

gold reserve as the minimum necessary to support a credible promise of convertibility.8 When the reserve fell to $65 million in early 1894, officials replenished it by auctioning, against payment in gold coin, $50 million of 5 percent bonds redeemable after ten years.9 When the reserve fell again to $60 million later in the same year, they auctioned another $50 million of the 5 percent bonds. The two auction offerings are important because they demonstrate the ability of the Treasury to raise funds with auction offerings and because they set the stage for an extraordinarily contentious private placement of Treasury bonds in 1895. The February Auction In the first auction, in February 1894, Treasury officials specified a minimum acceptable price of 117.223 percent of principal—equivalent to a maximum acceptable yield of 3.00 percent per annum to the first call date in ten years.10 Bidding closed at noon on Thursday, February 1, 1894. The Treasury received few bids prior to Monday, January 29.11 A New York bank president remarked that “The general opinion here . . . is that the issue will not be fully subscribed, and that . . . the proposed loan will fail . . .. One objection . . . is that the premium [of more than 17 points] which the Treasury exacts for these bonds is not likely to be favored by the public. There seems to be an impression that it would have been wiser to ask for bids on the bonds without setting a [minimum] price upon them.”12 On January 29, Secretary of the Treasury John Carlisle personally appealed to a group of prominent New York bankers to underwrite the issue.13 The bankers conferred and decided to bid the minimum price (117.223) for $45.9 million of the bonds.14 One banker remarked that “I did not want the bonds; I took them to make the thing go.”15 The Treasury received a total of 238 bids. It accepted in full all bids above the minimum price and filled 94.6 percent of the bids at that price.16 8. Noyes (1895, pp. 575–77), Dewey (1934, pp. 440–41), and Timberlake (1978, p. 134). 9. The bonds had no final maturity date at which the Treasury had to repay principal. The Treasury could, however, redeem the bonds any time after ten years from the date of issue. 10. The offering circular is reprinted in “A Call for Bids for Bonds,” New York Times, January 18, 1894, p. 1. 11. “The New Bond Loan Assured,” New York Times, January 31, 1894, p. 1. 12. “Carlisle Talks to Bankers,” New York Times, January 30, 1894, p. 1. 13. “Carlisle Talks to Bankers,” New York Times, January 30, 1894, p. 1, Barnes (1931, pp. 315–18), and Carosso (1987, p. 316). 14. “Success Crowns the Loan,” New York Times, February 1, 1894, p. 1. 15. Barnes (1931, p. 318). 16. Carosso (1987, p. 316) and “The New Bonds Allotted,” New York Times, February 4, 1894, p. 2.

36

Chapter 3

The November Auction In the second auction, in November 1894, the Treasury reframed from specifying a minimum price.17 Additionally, and unlike the earlier sale, it did not provide for a pro rata allocation of bids at the minimum accepted price. The circular for the February offering had provided that “In case the bids entitled to allotment exceed the bonds to be issued, they will be allotted pro rata.”18 The circular for the November offering stated more simply that “proposals will be received at the Treasury Department …”19 The change in language apparently came at the request of J. P. Morgan. Secretary Carlisle’s biographer quotes a letter from John Stewart, President of the United States Trust Company, to Secretary Carlisle dated November 12, 1895: “This suggestion came from Mr. J. Pierpont Morgan, who thought it possible to make up a bid for the entire amount and wished you to be in a position to decline all other bids, if such a bid would net you the most money.”20 By the close of the bidding on November 24, Treasury officials had received 486 bids for $179 million of bonds.21 Thirty-three New York banks, including J.P. Morgan & Co., formed a syndicate and bid 117.077 on an all-or-none basis for the entire $50 million offering.22 Although there were bids for $10.3 million of bonds at higher prices, the Treasury sold the bonds to the syndicate because that action generated the greatest total proceeds.23 The decision to accept the syndicate’s all-or-none bid was sufficiently controversial that when Treasury officials contemplated a third offering of bonds in early 1895, Assistant Secretary of the Treasury William Curtis recommended against allowing any such bids: “This time have provisions made for allotment, if necessary, and take the highest; no ‘all or none.’ ”24 17. The offering circular for the second auction is reprinted in “Fifty Millions of Bonds,” New York Times, November 14, 1894, p. 1. 18. “A Call for Bids for Bonds,” New York Times, January 18, 1894, p.1. 19. “Fifty Millions of Bonds,” New York Times, November 14, 1894, p. 1. 20. Barnes (1931, p. 356). 21. Barnes (1931, p. 359). 22. Barnes (1931, p. 359) and Carosso (1987, p. 317). The same syndicate also bid 116.8898 on a conventional basis (“Heavy Call for the Bonds,” New York Times, November 25, 1894, p. 4), so it was willing to pay a premium of about $94 thousand for control of the entire block of bonds ($93,600 = 0.1872% of $50 million, where 0.1872 = 117.077 – 116.8898). 23. “Heavy Call for the Bonds,” New York Times, November 25, 1894, p. 4, and “All Go to the Syndicate,” New York Times, November 27, 1894, p. 1. 24. Letter dated January 30, 1895, from Curtis to Carlisle, quoted in Barnes (1931, p. 373–74).

37

Treasury Debt Management before World War I

The 1895 Private Sale of 4 Percent Bonds Redeemable in 1925 In early 1895 the Treasury found its gold reserve dwindling yet again, down to $44 million by the first week of February.25 Market speculation of the need for a third bond offering abounded.26 On February 8, 1895, Secretary Carlisle announced that, with the approval of President Grover Cleveland, the Treasury had entered into a privately negotiated sale of 4 percent bonds redeemable after thirty years to a syndicate headed by J.P. Morgan & Co. and August Belmont & Co. in New York and J.S. Morgan & Co. and N.M. Rothschild & Sons in London. The contract provided for the sale of $62,315,450 principal amount of bonds at a price of 3.5 million ounces of gold.27 Since the price of gold was $18.60465 per ounce, the syndicate effectively paid a price of 104.4946 percent of principal for the bonds.28 The private placement of Treasury bonds with the Morgan–Belmont syndicate remains to this day the most notorious transaction ever undertaken by the U.S. Treasury. Its odious reputation rests on two aspects of the sale: (1) it appeared to be enormously profitable for the members of the syndicate, and (2) the bonds were not offered to the public on the same terms the syndicate received. President Cleveland and Secretary Carlisle opted for a private sale, rather than a conventional public auction offering, for three reasons.29 First, the Treasury needed the gold quickly, within a matter of days, to keep the United States on the gold standard. Second, a public auction offering was likely to fail. The first auction offering of 10-year bonds in 1894 had very nearly failed (albeit more because of Treasury’s minimum price than because of any real shortfall in demand) and the opportunistic, all-or-none, bidding by the Morgan syndicate in the second offering had soured some investors on bidding in Treasury auctions. Third, and probably most important, the contract with the Morgan–Belmont syndicate contained two unusual clauses that could not have been included in a public offering (because the clauses would have been difficult to enforce with respect to public buyers): (1) at least half of the gold 25. 1896 Treasury Annual Report, p. 46. 26. “May Delay a Bond Issue,” New York Times, February 3, 1895, p. 1. 27. “The Terms of Sale,” New York Times, February 9, 1895, p. 5. The contract terms are reported in “The Bill and Contract,” New York Times, February 14, 1895, p. 1. See also Noyes (1895, p. 592). 28. 1.044946 = $65,116,275/$62,315,450, where $65,116,275 is the market value of the gold ($18.60465 per ounce × 3.5 million ounces) and $62,315,450 is the principal amount of the bonds. The contract literally provided for the sale of the bonds at a price of 3.5 million ounces of gold valued at $17.80441 principal value of bonds per ounce. Note that $62,315,450 = 3,500,000 × $17.80441 per ounce. 29. See Carosso (1987, pp. 322, 326, and 337).

38

Chapter 3

was to be obtained in Europe, and (2) the syndicate members agreed, “as far as lies in their power, [to] exert all financial influence and [to] make all legitimate efforts to protect the Treasury of the United States against the withdrawals of gold, pending the complete performance of this contract.” Thus the private placement was not simply a sale of bonds for gold; it also specified the source of some of the gold and it aligned the Morgan–Belmont syndicate with the Treasury in the Treasury’s efforts to keep the United States on the gold standard. On February 20, 1895, the Morgan–Belmont syndicate reoffered the bonds to public investors at a price of 112¼ percent of principal, 7¾ points above what the syndicate paid the Treasury. The bonds were concurrently bid (in open-market trading) on a “when-issued” basis at prices of 116 to 118. (A when-issued transaction is a sale of securities that have not yet been issued. The seller agrees to deliver the securities, and the buyer agrees to pay for the securities, when the securities become available.) Not surprisingly, the offering sold out quickly.30 The apparent profitability of the transaction raised the question of whether Morgan and the other bankers had cleaned the Treasury’s clock. One observer later described the sale as “extremely harsh” and judged that the syndicate had “measured with little mercy the emergency of the Treasury.”31 Secretary Carlisle agreed that “the terms were hard for the Government,” but he also noted that “When you need money you have to get it on the best practical terms. I think we did the best we could.”32 A National Monetary Commission study subsequently concluded that the terms were severe but fair: “the risk which the purchasers ran of failing in their attempt to supply the Treasury with gold was so great that they were justified in marking hard terms.”33 Additionally the transaction as a whole was not as profitable as it appeared in February. The commitment of the syndicate members to “protect the Treasury [against] withdrawals of gold” led to a not entirely successful attempt to control the dollar–sterling exchange market in the spring and summer of 1895.34 The reaction, from politicians and others opposed to the gold standard, was not nearly so measured and has been characterized as “furious,” a “wild clamor,” and a “storm of angry denunciation.”35 The New York World described the sale as “a Wall Street conspiracy” and alluded to both the profitability and 30. “Eager to Get the Bonds,” New York Times, February 21, 1895, p. 1, “Extraordinary Success of the Bond Syndicate’s Operations,” New York Times, February 21, 1895, p. 12, “Allotment of the Bonds,” New York Times, February 24, 1895, p. 1, and Barnes (1931, p. 397). 31. Noyes (1909, p. 234). 32. Quoted in Barnes (1931, p. 388). 33. Kinley (1910, p. 252). 34. See Noyes (1895) and Simon (1960, 1968). 35. Chernow (1990, p. 76) and Noyes (1909, p. 236).

39

Treasury Debt Management before World War I

private nature of the sale: “The history of nations and of finance may be searched in vain for any parallel of this secret bond contract or for any instance where a civilized nation has been robbed by its own citizens on so great a scale in so short a time either in peace or war.”36 Representative William Jennings Bryan (who would make his “cross of gold” speech the following year) remonstrated against “recognizing or ratifying a contract as harsh in its terms and as imperious in its demands.”37 The 1896 Auction Offering of 4 Percent Bonds Redeemable in 1925 In September 1895 the Treasury’s gold reserve declined below $100 million for a fourth time and market participants began to speculate on the need for yet another bond sale.38 When the reserve fell below $75 million in midDecember, The New York Times stated that there was “no doubt” the Treasury would sell additional bonds.39 A week later, after the reserve had declined below $70 million, the paper stated that a sale would come “as soon as it shall be apparent” that the reserve would fall below $65 million.40 And when, a few days later, the reserve fell to $64 million, the paper reported that a sale was “taken for granted.”41 On January 6, 1896, the Treasury announced a public auction offering of an additional $100 million of the 4 percent bonds issued to the Morgan syndicate in February 1895.42 Officials went to unusual lengths to advertise the new issue. Offering circulars were sent to U.S. postmasters with instructions that they should be displayed “in a conspicuous place” and the Comptroller of the Currency sent circulars to all 3,800 national banks, advising them that “It is the desire of the Government that these bonds be distributed as widely as possible . . .. It is suggested that the National banks can materially aid the Government in popularizing this loan by calling attention of their patrons to the desirability of it as an investment, and in stimulating subscriptions thereto.”43 36. Carosso (1987, pp. 335 and 338). In his history of J.P. Morgan & Co., Ron Chernow (1990, p. 76) notes that public reaction to the sale was “laced with anti-Semitism because of the Rothschild participation” and that the New York World described the Morgan syndicate as “a pack of “bloodsucking Jews and aliens.” 37. Quoted in Carosso (1987, p. 330–31). 38. “Mr. Curtis in New York,” New York Times, September 17, 1895, p. 4 (reporting the gold reserve at $95 million). 39. “More Bonds to be Issued,” New York Times, December 16, 1895, p. 1. 40. “The Nation’s Credit Safe,” New York Times, December 24, 1895, p. 1. 41. “Must Issue Bonds Soon,” New York Times, December 28, 1895, p. 1. 42. The offering circular is reprinted in “Call for Bids for Bonds,” New York Times, January 6, 1896, p.1 43. “Gold For the New Bonds,” New York Times, January 12, 1896, p. 8, and “Circular to National Banks,” New York Times, January 17, 1896, p. 2.

40

Chapter 3

The return to an auction format, and the publicity given the new offering, were prompted by the reaction of Congress and the public to the Morgan– Belmont sale. Carlisle’s biographer states that the reopening of the 4 percent bonds was widely advertised “in order that no new charges of dealing with the big moneyed interests might be hurled” at the Secretary.44 Other historians have concluded that “Cleveland didn’t want to incite Populist wrath a second time” and that “neither Cleveland nor Carlisle considered a second contract feasible.”45 The New York Times reported a “personal friend” of President Cleveland as saying: Congress practically served notice on the President to the effect that if he sold bonds to a private syndicate—the Morgan or any other syndicate—a joint resolution would be introduced in the House of Representatives and in the Senate at once condemning the sale of the bonds. He was also notified that such a resolution would without doubt be passed by such a large vote as to make it practically a unanimous vote.46

The desire to publicize the offering as widely as possible led Treasury officials to delay the close of bidding until February 5, a full month after the initial announcement. (The auction offering of the first tranche of 5 percent bonds in February 1894 allowed only 15 days between notice and the close of bidding, and the second tranche in November 1894 allowed only 11 days.) Bankers were unhappy with the delay—they wanted the gold reserve restored promptly. One banker groused that, “Public and political clamor, mostly from people without gold or any other kind of investment funds, carried the day, and for four weeks we are to be haunted by the uncertainties of this loan proposal, with prospects far from reassuring.”47 The auction was a success in spite of the bankers’ concerns: the Treasury received 4,641 bids for $688 million bonds and sold the bonds at an average price of 111.166. (At the time the 4 percent bonds issued in February 1895 were trading at a price of about 115.48) The 1896 bond sale marked the Treasury’s return to public auction offerings, but the bonds were not distributed as widely as the Treasury might have hoped. Carlisle’s biographer commented that “the country in general was well pleased with the sale,” but that “It had little basis . . . on which to claim plaudits, for it was [the New York banks] which had again come to the rescue of the Treasury Department. Of the $111,166,232.65 realized on the sale more than $97,000,000 had come from that city, and most of that amount had been 44. Barnes (1931, p. 416). 45. Chernow (1990, pp. 76–77) and Carosso (1987, p. 344). 46. “Bankers are Satisfied,” New York Times, January 7, 1896, p. 2. 47. “Look to Bond Syndicate,” New York Times, January 9, 1896, p. 5. 48. Barnes (1931, pp. 421–22) and “Harbinger of Prosperity,” New York Times, February 6, 1895, p. 1. See, “Bonds Subscribed Five Times Over,” New York Times, February 6, 1896, p. 1, for a nice description of the noontime opening of bids.

41

Treasury Debt Management before World War I

subscribed by the men who from the beginning had been supplying gold for the Treasury.”49 The 1896 bond sale demonstrated the practical difficulty of effecting a wide distribution of Treasury debt through an auction offering. Even with a four week marketing period, officials were unable to attract very many non–bank bidders. The 1898 Subscription Offering of Spanish–American War Bonds The battleship Maine blew up and sank in Havana harbor on the night of February 15, 1898, with the loss of more than 260 sailors and officers. The catastrophe led to a declaration of war against Spain on April 24 and, two months later, to the first widely distributed Treasury issue since the Civil War. Congress mandated that the Treasury offer the 3 percent, 20-year war bonds as a “popular loan,” conveniently accessible to individual investors. The bonds were to be sold at par in a fixed price subscription offering, rather than auctioned to the highest bidders, so unsophisticated investors would not be deterred by fear of overbidding. Officials were also instructed to give a preference in allocating bonds to individual subscribers and small subscriptions.50 There were two reasons why Congress called for a popular loan, widely distributed to individual investors. First, some legislators remained bitter over the private placement of 4 percent bonds in February 1895 and disappointed with the limited non–bank participation when that issue was reopened in a public auction offering in January 1896. The New York Times reported one congressman as saying that “if this bond issue was not authorized [as a fixedprice popular loan] bonds would be issued under the subterfuge adopted by the Cleveland Administration and Morgan and Drexel & Co. or some other syndicate would reap the profit.”51 Second, and important as a harbinger of what was to come during World War I, Congress clearly intended that subscriptions should express popular support for the war. An early draft of the authorizing legislation provided that the bonds would be sold as well as advertised at post offices throughout the 49. Barnes (1931, p. 422). A group headed by J.P. Morgan & Co. bid for $100 million of the bonds at what proved to be the lowest accepted price and received a pro rata allocation of $33.2 million bonds. “Bond Allotments Made,” New York Times, February 8, 1896, p. 3 and “Bond Allotments Made,” New York Times, February 9, 1896, p. 2. Carosso (1987, pp. 345–46) reports that about 87 percent of the sale proceeds came from New York banks—“mostly from the same banks, trust companies, and other financial institutions that had been included in [the Morgan syndicate].” 50. Section 33 of the War Revenue Act of June 13, 1898, provided that “the bonds authorized by this section shall be first offered at par as a popular loan under such regulations, prescribed by the Secretary of the Treasury, as will give opportunity to the citizens of the United States to participate in the subscriptions to such loan, and in allotting said bonds the several subscriptions of individuals shall be first accepted, and the subscriptions for the lowest amounts shall be first allotted . . ..” (Emphasis added.) 51. “The War Revenue Bill,” New York Times, April 30, 1898, p. 5.

42

Chapter 3

country. That provision, coupled with a minimum denomination of $50, would make the bonds “a great popular loan, to be absorbed by the people.”52 On the eve of passage of the final legislation, Senator Edward Wolcott (Republican, Colorado) expressed confidence that subscriptions would come from “the people of every hamlet of the United States,” and that “the people of the country will take up these bonds as a manifestation of their patriotism and their pride in their country.”53 One market participant later remarked that “it was the opinion of Congress . . . that a wide distribution among the people, at the uniform price of par, was of more value to public interests than the realization of a possible bonus from small groups of professional dealers.”54 Immediately after President William McKinley signed the authorizing legislation, the Treasury offered for sale at par $200 million of 20-year bonds.55 The offering included three provisions intended to foster a wide distribution. First, as with the auction offering of 4 percent redeemable bonds in 1896, the Treasury allowed four weeks before closing the subscription books. Assistant Secretary of the Treasury Frank Vanderlip observed that “This gives ample time even for the most remote sections of the country to receive complete information regarding the details of the issue and to have subscriptions from such points reach the Treasury. Every channel will be used to disseminate in the widest possible manner information regarding this issue, and to make it in every sense a successful popular loan.”56 The Treasury also went to great lengths to allow payment in convenient forms, including postal money orders, checks, and currency sent by registered mail. Finally, pursuant to a statutory requirement, the offering provided that “allotments to all individual 52. “Revenue Plan Adopted,” New York Times, April 13, 1898, p. 2. See also the description of the sales effort reported in “60,000 Bond Sales Agencies,” New York Times, May 13, 1898, p. 5 (“Never before in the history of this country, and probably in the history of any other country, have such comprehensive plans for offering an issue of securities to the whole people been worked out . . .. The Treasury Department has already organized extensive facilities for popular subscriptions throughout the country. Every little town which has a money order Post Office will have an agency of the Treasury. There are 22,000 of them. There are over 16,000 banks . . . who will be asked to act without compensation . . .. The great express companies have generously and patriotically come forward with offers of their extensive facilities.”). 53. War Revenue Bill Passed,” New York Times, June 11, 1898, p. 4. 54. Childs (1947, pp. 83–84). 55. The offering circular is reprinted in “The War Tax Bill Signed,” New York Times, June 14, 1898, p. 3. 56. “The Sale of the Bonds,” New York Times, June 12, 1898, p. 13. The War Revenue Act of 1898 also authorized the sale of interest-bearing certificates of indebtedness with maturities not to exceed one year, subject to the limitation that not more than $100 million of certificates could be outstanding at the same time. The certificates were not required to be offered as a popular loan. Senator Nelson Aldrich (Republican, Rhode Island) remarked that the short-term securities were “intended to meet emergency demands, which rendered it impracticable to make provision for a popular subscription.” “War Revenue Bill Passed,” New York Times, June 11, 1898, p. 4. His comment recognizes the time it took to publicize and process a popular loan.

43

Treasury Debt Management before World War I

subscribers will be made before any bonds will be allotted to other than individuals.”57 The offering was an overwhelming success.58 The day after the announcement of the offering The Wall Street Journal reported “a large crowd in line [at the subtreasury in New York] waiting to secure blank applications for the new 3% loan. . . . The feeling among bond houses is that syndicates will get few of the new 3s.”59 Two days later Secretary of the Treasury Lyman Gage pronounced the offering a success.60 By the time the subscription books closed on July 14, Treasury officials had received more than 450,000 tenders for more than $1.2 billion of the bonds. Secretary Gage announced that sales would be limited to individuals who subscribed for $5,000 or less.61 The fixed-price offering of Spanish–American War bonds accomplished what the 1896 auction offering did not: a broad distribution of bonds to individual investors. However, commentators quickly pointed out the principle disadvantage of the fixed-price offering: a lower sale price. The New York Times editorialized that: The supposed necessity of paying heed to the ignorant clamor about popular loans involves the Government in a direct loss of several million dollars by reason of the sale of the bonds at par instead of at the premium they would have commanded . . . In a few months or a year the bonds will mostly have found their way into the hands of institutions, estates, trustees, and large investors, who will pay a premium they would have been perfectly willing to pay to the Government if the bonds had been sold to the highest bidder.62

A Wall Street banker opined similarly: Uncle Sam is humbugging himself, just giving away $4,000,000 on a sentiment—the popularity of the loan. Without this popular restriction and with the issue of the loan to bids, the bonds should have commanded anywhere from 102 to 103.63

57. “The War Tax Bill Signed,” New York Times, June 14, 1898, p. 3. 58. See generally, Vanderlip (1898). 59. “Government Bonds,” Wall Street Journal, June 15, 1898, p. 1. 60. “The War Loan,” Wall Street Journal, June 17, 1898, p. 2. Consistent with the strong demand for the new securities, the bonds traded at prices in excess of par in when-issued trading. See “The Popular War Loan,” New York Times, June 16, 1898, p. 12, and “War Bond Subscription,” New York Times, June 18, 1898, p. 12. 61. “Over a Billion for Bonds,” New York Times, July 15, 1898, p. 4, and “The Bond Issue Explained,” New York Times, July 16, 1898, p. 3. 62. “The ‘Popular’ Feature of the Loan,” New York Times, June 16, 1898, p. 6. 63. “The Popular War Loan,” New York Times, June 16, 1898, p. 12. See also “The Popular War Plan,” New York Times, July 9, 1898, p. 6 (“The Government loses from five to seven millions on the issue of $200,000,000, since these 3 per cent. bonds would have commanded a premium of from 2½ to 3½ per cent. if they had been offered in the usual manner to the highest bidder.”), and “Delusions About the Loan,” New York Times, July 24, 1898, p. 14 (“These 3 per cent. bonds which the Government sold at par have sold in Wall Street at 1043⁄8, a price which reduces the interest rate to about 2½ per cent. Selling the bonds at this figure the Government would have received $8,750,000 more than it got by selling them at par.”).

44

Chapter 3

Supporters of the subscription method of sale pointed out several purported benefits of a wide distribution to individual investors: “it inspires and deepens patriotism and makes subscribers to the loan better citizens by giving them a direct stake in the stability and wise administration of the Government.”64 Others questioned whether the broad initial distribution was likely to persist and suggested that the bonds would ultimately gravitate to those who valued them most highly.65 One commentator later estimated that “over $77,000,000 of the original holdings of about 116,000 subscribers soon passed into the ownership of 1,001 persons, firms and corporations.”66 The 1900 Exchange Offering of 2 Percent Consols Redeemable in 1930 The largest pre-war Treasury issue was the 2 percent Consols redeemable at the option of the Treasury after April 1, 1930. In mid-1914, almost $650 million of the bonds were outstanding. Section 11 of the Gold Standard Act authorized the 2 percent Consols as part of a program to enhance the profitability, and issuance, of national bank notes.67 Holders of three outstanding bonds: the 5 percent bonds redeemable in 1904, an older issue of 4 percent bonds redeemable in 1907, and the Spanish–American War bonds, were invited to exchange their bonds for an equal principal amount of 2 percent Consols. For purposes of an exchange, the outstanding bonds were valued at a yield of 2.25 percent per annum, calculated from the date of accepted tender. The excess of that amount over par, less accrued interest on the 2 percent Consols (calculated to the date of accepted tender), was paid to a bondholder in cash.68 The 2 percent Consols were attractive to national banks that had been pledging the older, higher coupon, bonds against their issues of bank notes because they could replace those bonds with Consols on a par-for-par basis and capture the cash payment. By mid-1914, 95 percent of the outstanding Consols were pledged against issues of national bank notes or held as collateral against Treasury deposits in national banks (table 3.2). 64. “The Popular War Plan,” New York Times, July 9, 1898, p. 6. 65. “The Popular War Plan,” New York Times, July 9, 1898, p. 6. 66. Childs (1947, p. 84). 67. See “Currency Bill Taken Up,” New York Times, January 5, 1900, p. 8 (statement of Senator Nelson Aldrich (Republican of Rhode Island and chairman of the Senate Finance Committee) that “It is the duty of Congress, in the public interest, to so modify the National Banking act as to give to banking associations an opportunity to issue currency with a reasonable profit. If the provisions of this bill are enacted into law, we may reasonably expect a considerable increase in the National banknote circulation in the near future . . ..”). 68. The offering circular for the 2 percent Consols is reprinted in “Mr. Gage Ready for Business,” New York Times, March 15, 1900, p. 1. Calculation details appear in 1900 Treasury Annual Report, p. LXXIX, note a. See also “Conferees Agree on Currency Bill,” New York Times, February 24, 1900, p. 1.

45

Treasury Debt Management before World War I

The Auction Sales of Panama Canal Bonds The Spooner Act of June 26, 1902, authorized construction of the Panama Canal, the greatest engineering feat of its time.69 Section 8 of the act provided for the sale of up to $130 million of 30-year, 2 percent bonds to finance construction. The offerings had to give “all citizens of the United States an equal opportunity to subscribe,” but Congress did not require that the bonds be sold in fixed-price offerings. The Spooner Act resulted in three bond sales: (1) an auction sale in July 1906 of $30 million of bonds maturing in 1936, (2) an auction sale in November 1907 of an additional $24.6 million of the same issue, and (3) an auction sale in December 1908 of $30 million of bonds maturing in 1938.70 A relatively short, two week, interval between announcement and the close of bidding for each of the three auctions suggests that Treasury officials did not go to any great lengths to attract individual investors or to effect a broad distribution of the bonds. Consistent with Treasury’s experience in the auction offerings of 1894 and 1896, the bonds went mainly to banks.71 On May 18, 1911, the Treasury announced an auction offering of a fourth Canal bond issue, $50 million of 3 percent 50-year bonds.72 Unlike the first three bonds, the fourth issue could not be used to secure national bank notes.73 In light of the reduced demand for the bonds by national banks, Treasury officials made more of an effort to market the bonds to individual investors, providing a four-week interval between announcement and the close of bidding and allowing payment in forms convenient for individuals.74 After the close of bidding on June 17, The Wall Street Journal judged the effort to distribute the 69. The United States and Panama reached agreement on building the canal in November 1903. In May 1904 the United States acquired the property of the French company that had failed in an earlier effort to build a canal; the United States started construction later the same year. The canal was officially opened in August 1914, just days after the beginning of World War I. McCullough (1977, pp. 394, 400, and 609). 70. Offering statements are reprinted in “The Panama Canal Bonds,” Wall Street Journal, July 3, 1906, p. 5 and “Federal Aid up to $150,000,000,” New York Times, November 18, 1907, p. 1. 71. Two-thirds of the 1906 offering went to national banks or investment banks (“Panama Canal Bonds,” Wall Street Journal, July 23, 1906, p. 7), as did two-thirds of the 1908 offering (“Panama Bond Bids Opened,” Wall Street Journal, December 8, 1908, p. 2). The fraction of the 1907 offering that went to buyers other than national banks was characterized as “not large” (“Allots $25,000,000 Bonds,” New York Times, December 7, 1907, p. 1). Childs (1947, p. 472) observes that the 1907 offering was “almost entirely absorbed by the banks.” 72. The offering statement is reprinted in “Treasury Bond Circular,” New York Times, May 19, 1911, p. 13. 73. The 3 percent Canal bonds of 1961 were authorized by section 39 of the Payne–Aldrich Tariff Act of August 5, 1909. Prior to the offering of the bonds, Congress, in anticipation of the currency reforms that ultimately led to passage of the Federal Reserve Act in 1913, adopted the Act of March 2, 1911, stripping bonds issued under the authority of the Payne–Aldrich Act of the currency privilege. 74. “Secretary M’Veagh Offers $50,000,000 Government 3% Bonds,” Wall Street Journal, May 18, 1911, p. 5.

46

Chapter 3

bonds across a wide investor base comparable to the effort to distribute the Spanish–American War loan.75 The Treasury received more than 10,000 tenders for more than $212 million of the bonds. Nevertheless, the Journal concluded that the availability of the bonds as collateral for government deposits “created a demand for over half of the issue.”76 Summary The market for U.S. Treasury bonds prior to the First World War differed from the postwar market in two respects. First, bonds were issued to finance specific projects: to replenish Treasury gold stocks, to finance a war, and to build a canal. The project finance approach is evident in the statutory authorizations for the respective issues. Second, the requirement that national banks had to pledge Treasury bonds against their currency issues and public deposits resulted in a strong demand for Treasury bonds by national banks: 80 percent of all outstanding Treasury bonds were pledged as collateral in 1914. Because Treasury bonds generated ancillary revenues for national banks, they traded at priced in excess of their value as conventional investments. The premium prices limited non–bank demand for the bonds and limited the breadth of both the primary and secondary markets for the bonds. Outside of the Spanish–American War loan (when Congress directed the Treasury to offer bonds on a fixed-price basis and to give allotment preferences to individual subscribers and small subscriptions), Treasury offerings were not widely distributed. 75. “Bids Opened at Washington for New Issue of Panama Bonds,” Wall Street Journal, June 19, 1911, p. 5. 76. “The Bond Market,” Wall Street Journal, June 28, 1911, p. 5. See also “Panama Bond Bids Show 2.50 Premium,” New York Times, June 18, 1911, p. 1 (“[I]t is certain that [the bonds] will be used to replace the $29,000,000 in bonds now out which are available for circulation, but which are temporarily being used to secure Government deposits. The fact that today’s bonds can be used to release this $29,000,000 for use as a basis for circulation admittedly explains in part the good prices paid by banking concerns.”)

II

FINANCING THE GREAT WAR

4

Treasury Finance during World War I

The scale of Treasury financing operations expanded dramatically after the United States entered the First World War and officials had to borrow $25 billion to finance the American participation. A market previously driven primarily by bank demand for collateral (pledged against government deposits and issues of national bank notes) evolved into an investment market, where securities were judged primarily on the basis of their cash flows. One prominent dealer remarked that, “For many years [there were not] enough U.S. bonds suitable for investment purposes to demonstrate by their market value the true credit level of this nation.”1 By war’s end, supply was no longer a problem. This chapter provides an overview of Treasury finance during World War I. The chapter begins with a brief chronology of the war. The chronology is important because, in the final analysis, the course of the war drove everything else. The costs of the war, and how Treasury officials financed the American participation, are then examined. The War World War I began on July 28, 1914, when Austria-Hungary declared war on Serbia following the assassination of the heir to the Austrian throne by a Bosnian Serb. What started as a regional conflict escalated, within a matter of days, to a continental war as Germany declared war on Russia (in response to Russia’s decision to mobilize its army in defense of Serbia) and on Russia’s ally France, and then escalated again when Great Britain declared war on Germany (in response to Germany’s anticipated invasion of Belgium). Germany planned to win a quick victory in western Europe by sending the right wing of its army though Belgium and into northern France while holding the left wing stationary along the French border south of Metz, thereby enveloping and defeating the French army in northeast France. Events followed plan until early September, when French and British forces east of Paris stopped the German advance at the Marne River. Over the course of the following month the opposing armies repeatedly extended their lines to the northwest in an attempt to turn each other’s flank. When the “Race to the Sea” ended, they faced off along a front that ran 475 miles from neutral Switzerland to the English Channel. During the next three and a half years, from the winter of 1914–15 to the summer of 1918, the two armies battled each other in horrific trench warfare that accomplished little at the cost of many hundreds of thousands of dead and 1. “Market Shows Nation’s Credit Level Below 4¼%,” Wall Street Journal, November 23, 1918, p. 10 (quoting a research report from C.F. Childs & Co.).

50

Chapter 4

wounded.2 The German offensive at Verdun, for example, lasted from February to July 1916 and produced 377 thousand French and 337 thousand German casualties. The British offensive at the Somme River lasted from July to November 1916 and produced 420 thousand British, 194 thousand French, and perhaps as many as a half a million German casualties. Although not evident in military operations for another year, the stalemate began to break in January 1917, when Germany, in an attempt to force Britain to sue for peace, resumed unrestricted submarine warfare. Germany had initiated unrestricted submarine warfare earlier, in January 1915, but suspended the effort after eight months following strong protests from the United States. German officials recognized that renewing the effort was likely to draw America into the war on Britain’s side, but reckoned that American troops would not arrive before Britain would be forced to capitulate. They forecast the American reaction correctly but underestimated British resolve. On April 2, 1917, President Woodrow Wilson, addressing a special session of Congress, asked for a declaration of war against Germany to make the world “safe for democracy.” Two days later the Senate voted 82 to 6 in favor of the President’s request. The House of Representatives followed on April 6 with a vote of 373 to 50, and the President signed the declaration later the same day.3 American Participation In the spring of 1917 the United States had a first-class navy but only a small and poorly equipped army; it even lacked plans to train a large force. It would take time to raise, train, and deploy what was expected to be an army of as many as two million men. By November 1917, only 77 thousand American troops had reached France; two months later the number had crept up to 150 thousand and by March 1918—almost a year after America entered the war—to 220 thousand. The slow pace of the American buildup dismayed the French and British but the implications of the buildup—continuation of the war into 1919 and the prospect of gradual exhaustion of Germany’s army and economy—were clear to all. In a daring gamble Germany launched the biggest offensive thus far in the war on March 21, 1918, hoping to win a favorable decision before the American army reached full strength. The French and British armies hung 2. Stevenson (2004, ch. 7) discusses the technological, logistical, and tactical reasons for the stalemate. 3. Although war was probably inevitable once Germany began sinking American ships in midMarch 1917, American public opinion turned decisively against Germany following public disclosure on March 1, 1917, of the “Zimmerman Telegram,” wherein Germany proposed to Mexico that if the United States entered the war and Mexico subsequently agreed to an alliance with Germany, Germany would support Mexican efforts to regain the territory it had lost in the Mexican–American war of 1846–48. Stevenson (2004, pp. 256–61), Tuchman (1958), Keegan (1999, p. 351), and Kennedy (2004, p. 10).

51

Treasury Finance during World War I

on and the offensive stalled. A renewed German effort in late May met a similar fate. The spring offensives cost Germany nearly a million men that it couldn’t replace. In contrast, the rapidly growing American army in France reached one million in July and exceeded 1.8 million in early November. French, British, and American offensives in late September forced Germany to agree to an armistice that went into effect on November 11, 1918.4 The Costs of War The First World War consumed unprecedented quantities of men and materiel. At least in the early days of the conflict, much of the manpower was volunteered; what was not volunteered was requisitioned through drafts. In contrast, materiel was neither volunteered nor requisitioned, but rather purchased in something approximating market transactions. Governments acquired the means of payment by taxing their citizens and by selling bonds. By 1917 France and Britain had exhausted both their human and financial resources. According to Professor David Stevenson, “in 1917 all three main armies on the Western Front [British, French, and German] began to shrink. . . . Although trying to compensate through increased firepower, one army after another converted from an offensive to a defensive posture in response to changed strategic priorities, shortage of men, and ebbing morale.”5 Morale was a particular problem. A futile French offensive at the Chemin des Dames (a highway 60 miles northeast of Paris built for the daughters of Louis XV) in April and May 1917 led to widespread mutinies in the French army that were ultimately suppressed by several thousand courts-martial, over 500 death sentences, and 49 actual executions. Stevenson observes that, “During 1917 units in almost every army refused to attack . . . or even to enter the front lines.”6 French and British financial resources were equally depleted. Stevenson points out the central role played by Britain in financing the war: Not only did Britain place and finance all Russia’s American purchases from 1915, and increasingly those of Italy, but from May 1916 it also financed all France’s American orders, as well as supporting the franc in the foreign exchange markets. The entire Allied war effort would be vulnerable if Britain’s American credit-worthiness deteriorated. . . . 40 percent of all the British government’s war purchases, for itself and for its allies, were made in North America, and the Treasury was expected to come up with more than $200 million a month.7

4. The combatants signed a peace treaty at Versailles on June 28, 1919, but the U.S. Senate rejected the treaty on November 19, 1919. Acting unilaterally for the United States, President Warren Harding proclaimed an end to the war with Germany on July 2, 1921. 5. Stevenson (2004, p. 244). 6. Stevenson (2004, p. 268). 7. Stevenson (2004, p. 184).

Chapter 4

20

16

Billions of dollars

52

12

8

4

0 1910

1911

1912

1913

1914

1915

1916

1917

1918

1919

Fiscal year War and Navy departments

Foreign loans

Other

Figure 4.1 Treasury expenditures

Professor David Kennedy concludes that, by the spring of 1917, Britons “had deeply mortgaged themselves to American creditors . . .. In the London counting houses . . . men knew . . . the day of reckoning was near. It was only a matter of weeks until all the coffers in Britain would be empty.”8 America’s entry into the war opened access for France and Britain to vast quantities of desperately needed resources. There is no question of the significance of American manpower. Professor Stevenson observes that “after the Chemin des Dames disaster [the French generals] made a big American army essential” to their strategy.9 He concludes that “the crucial American contribution was . . . the sheer number of troops, which did much to convince [German generals and soldiers] that they could no longer win.”10 Notwithstanding the importance of American manpower, America’s financial contribution had a greater initial impact and a comparable long-run significance. Figure 4.1 shows the explosive growth in wartime expenditures by the United States. Between mid-1917 and mid-1919, America disbursed $18.3 billion on its own account in connection with the war and lent an additional $8.4 billion to its allies. The 1919 Treasury Annual Report points out the crucial role played by foreign loans: 8. Kennedy (2004, p. 8). 9. Stevenson (2004, p. 300). 10. Stevenson (2004, pp. 359–60).

53

Treasury Finance during World War I

In the most critical stages of the war [the loans] immeasurably assisted America’s gallant associates in obtaining the munitions, supplies, and equipment that were so imperatively needed to meet the enemy’s offensives. . . . [P]robably of equal importance was the fact that they served to hearten the allied armies and peoples by the knowledge that the vast credit resources of the United States were being shared with them for the effective prosecution of a common cause . . . In the beginning, before the creation of our great Army, the principal assistance of America was necessarily through foreign loans, and it was then that these advances proved so very potent in contributing to the final victory. . . . [I]n each grave crisis . . . the loans from the United States gave the Allies the means of replenishing supplies and equipment, and inspired their fighting forces and the peoples behind them with renewed hope and confidence and with strengthened determination in the face of an advancing foe. . . .Without this aid to the allied Governments, the war unquestionably would have been prolonged, if not lost . . ..11

The wartime Secretary of the Treasury, William McAdoo, argued that the loans enabled France and Britain to “gain victories before American troops could be trained and put into action. The dollars that we sent . . . to Europe were, in effect, substitutes for American soldiers . . ..”12 Financing American Participation By the time the United States entered World War I, every major industrialized country was already involved. There was no possibility of borrowing abroad, so American participation had to be financed domestically—by selling bonds to Americans and by taxing Americans. The need to finance the war domestically was widely recognized. Mortimer Schiff, a partner at Kuhn, Loeb and Company, stated that “We are dependent on our own resources, and it is our people who must find the funds for financing our expenditures.”13 Ernest Patterson, an economist at the University of Pennsylvania, observed that “[T]he money we need must be raised within the United States. Our allies may have borrowed and may continue to borrow outside their own borders, but no such resources are open to us.”14 The great debate, among economists and among policy makers, was how much of the costs of the war should be financed with debt and how much with taxes.15 There were proponents all across the spectrum, from 100 percent debt financing to 100 percent tax financing. Senator Furnifold Simmons, Democrat from North Carolina and Chairman of the Senate Finance Committee, took a 11. 1919 Treasury Annual Report, pp. 65–66. 12. McAdoo (1931, p. 376). 13. Schiff (1918, p. 38). 14. Patterson, et al. (1918, p. 36). 15. See, Bullock (1917, p. 359) (“[N]either in theory nor in practice is there universal agreement concerning the proportions in which the two expedients should be employed.”), Durand (1917, p. 888) (“[A] controversy . . . has been raging between the advocates of taxation and the advocates of bond issues . . ..”), and Plehn (1918, p. 574) (“Around the choice of [debt and taxes] there has arisen a pretty controversy.”).

54

Chapter 4

position at the debt end of the spectrum: “It has been the custom of this country to pay war bills by bond issues, and I see no reason for a change in that policy.”16 A bit more moderately, J.P. Morgan thought that no more than 20 percent of war expenses should be paid from taxes.17 Treasury Secretary McAdoo thought that about half should be financed with taxes.18 (McAdoo later revised his thinking to one-third taxes and two-thirds debt.19) There were also proponents for policies further along the spectrum. The New York Times reported in April 1917 that “some members of Congress are advocating the raising of 75 percent of the first year needs by taxation,”20 and leading economists in 43 colleges and universities signed a petition urging taxation as the principle means of finance.21 O. M. W. Sprague, a leading monetary economist, argued that the war should be financed “mainly, if not entirely” from taxes.22 The central issue was whether debt financing could transfer the burden of the war to future generations. Economists agreed that it could—if the debt were sold to foreigners.23 In that case the war effort might require little sacrifice in current living standards. However, since there were no other countries from which to borrow, there was no essential difference—from an aggregate point of view—between debt financing and tax financing. Either way, all of the costs of the war had to born immediately, by Americans, in the form of reduced consumption.24 16. Quoted in Adams (1917, p. 292). 17. Morgan’s view is reported in McAdoo (1931, p. 383). 18. “Senate Will Pass Bond Bill Quickly,” New York Times, April 13, 1917, p. 3, and McAdoo (1931, p. 372). See also the April 14, 1917, letter from McAdoo to Cleveland H. Dodge, a prominent philanthropist and Princeton classmate of Woodrow Wilson, quoted in Synon (1924, pp. 222–23) (“As to taxation, my feeling has been that fifty per cent. of the cost of the war should be financed by it.”). 19. See his letter to the chairman of the House Ways and Means Committee, quoted in “M’Adoo Advises Doubling War Tax,” New York Times, June 7, 1918, p. 1 (“I believe that if we are to preserve the soundness and stability of our financial system, we should raise by taxation not less than one-third of the estimated expenditures for the fiscal year 1919. . ..”). 20. “Big War Loan Bill Ready for Debate,” New York Times, April 12, 1917, p. 2. 21. “Economists United in Favor of War Tax,” New York Times, April 19, 1917, p. 24, “College Men Want Direct Taxes Instead of Bonds,” Boston Daily Globe, April 19, 1917, p. 7, and “Taxation Is Favored to Meet War Expenses,” The Atlanta Constitution, April 19, 1917, p. 6. Economists who signed the petition included E. Dana Durand and Roy Blakey (University of Minnesota), Davis Dewey (MIT), and Irving Fisher (Yale University). 22. Sprague (1917, p. 2). 23. See, for example, Blakey (1918, p. 92). 24. See, for example, Anderson (1917, p. 860) (“[O]ur own citizens must pay now out of current income whatever the government spends now, and, taking the nation as a whole, it is simply impossible for ‘posterity to share the burdens’. . ..”) and Durand (1917, p. 892) (“[F]or the people considered as a whole, domestic borrowing postpones no burden to the future . . .. Borrowing at home, so far as a nation as a whole is concerned, is precisely similar to borrowing by an individual from himself. . . . [The] idea that the burden of war expenditures can be deferred to future generations is the supreme fallacy of finance.”) (Emphasis in the original.)

55

Treasury Finance during World War I

Debt financing and tax financing might be equivalent from an aggregate point of view, but they differed at a more granular level. Debt financing, for example, facilitated intertemporal reallocations of the burdens of war among individuals. Roy Blakey, an economist at the University of Minnesota, observed that “[W]hen a war comes unexpectedly it may find many individuals unprepared to pay their just shares of a new and large burden. It may be best all around to permit some to assume the burden of others temporarily, either wholly or in part.”25 But the crucial difference between debt and taxes was that debt postponed the ultimate allocation of the costs of war among different income classes. Economists recognized that Congress might make very different decisions during the war about who should pay higher taxes than it would make after the war. E. Dana Durand, chairman of the Department of Economics at the University of Minnesota, observed that “[I]t is possible by the bond plan to readjust the burden after the war—possible to put more of it on the poorer classes . . . [P]art of what [the more well-to-do] had advanced to the government would be repaid to them by other social classes.”26 More pointedly, Durand noted that “The patriotism which during the war itself might induce the rich willingly to pay taxes according to the full measure of their ability is bound to wane considerably when the war is over. There is little enthusiasm about paying for a dead horse.”27 In the words of one wit, war bonds were likely to be “a mortgage of the masses to the classes.”28 Taxes The question of how to finance the war was resolved gradually, in a series of incremental actions, including especially new tax legislation.29 25. Blakey (1918, p. 93). Blakey (1917, p. 813) also noted that “Among persons of equal means, some are in a much better position to economize at this time than are others; hence, some borrowing is socially justifiable because it allows accommodation as between individuals.” See similarly, Durand (1917, pp. 906–907). Blakey (1918, pp. 93–94) pointed out that because individuals as well as governments can borrow, individual time preferences can be accommodated either through debt financing or tax financing, but concluded (p. 94) that there are adequate reasons to prefer public, rather than private, finance: “In so far as the government can make easier advantageous credit transactions by itself assuming the borrowing agency instead of leaving the transactions to be arranged through individuals, there is a further net gain. This is the real economic justification of government loans.” 26. Durand (1917, p. 899). 27. Durand (1917, p. 902). Similarly, Blakey (1918, p. 95) noted that “[T]he burden of taxation is apt to be distributed more equitably if the taxes are levied during the war than if after its close. The well-to-do classes have practically always controlled legislation directly or indirectly and they are much more likely to be willing to assume their just burdens during the war while the spirit of patriotism is almost universal and while others are giving their lives to the country.” 28. Quoted in Anderson (1917, p. 864). 29. See, generally, Bank, Stark, and Thorndike (2008, ch. 3).

56

Chapter 4

6

Billions of dollars

5 4 3 2 1 0 1910

1911

1912

1913

1914

1915

1916

1917

1918

1919

Fiscal year Income and profit taxes

Import tarrifs

Other receipts

Figure 4.2 Treasury receipts

Since the end of Reconstruction the United States had relied primarily on import tariffs and excise taxes (on alcohol and tobacco) for revenue.30 Four very different taxes accounted for the bulk of wartime tax financing: • individual income taxes—introduced in 1913, • corporate income taxes—introduced in 1909, and • war profit and excess profit taxes—introduced during the war. Figure 4.2 shows Treasury receipts prior to and during the war. The importance of income and profit taxes during the war is clearly visible. Individual Income Taxes When the United States entered World War I, the federal government had been taxing individual incomes for only four years. Congress had adopted an income tax in 1894, but an 1895 Supreme Court decision31 invalidated the tax, declaring that it was a direct tax within the meaning of article I, section 2 and article I, section 9 of the United States Constitution and that it consequently had to be apportioned among the states in proportion to their respective 30. Dewey (1934, pp. 395, 399, 420, 426, and 474). 31. Pollack v. Farmers’ Loan and Trust Company, 157 U.S. 429 (1895) and 158 U.S. 601 (1895).

57

Treasury Finance during World War I

populations.32 The adoption of the Sixteenth Amendment in early 1913 nullified the Court’s decision.33 Before the year was out, Congress passed the Underwood–Simmons Tariff Act of October 3, 1913, providing a modest individual income tax to offset reduced tariffs on imports. The Underwood–Simmons Tariff Act assessed a “normal tax” on net income and an additional “surtax.” It defined net income as the sum of: • salaries and wages, • interest income (exclusive of interest received on federal, state, and local government bonds), • rent, • royalties, • dividends, and • business income, less: • business expenses, • depreciation on business assets, • interest on indebtedness, • federal, state, and local taxes, and • losses not compensated by insurance. The normal tax was 1 percent of net income in excess of $3,000 for a single person (or $4,000 for a married couple). The surtax started at 1 percent of net income in excess of $20,000 and rose to 6 percent of net income in excess of $500,000.34 The new tax produced less than half the revenue Congress expected and proved to be administratively burdensome.35 It was nevertheless important for 32. Article I, section 2 says, in relevant part: “[D]irect Taxes shall be apportioned among the several States . . . according to their respective Numbers . . ..” Section 9 says, in relevant part: “No capitation, or other direct, Tax shall be laid, unless in Proportion to the Census or Enumeration herein before directed to be taken.” 33. The Sixteenth Amendment provides that “The Congress shall have power to lay and collect taxes on incomes, from whatever source derived, without apportionment among the several States, and without regard to any census or enumeration.” See Blakey and Blakey (1940, chs. 1 and 2) for a history of the Sixteenth Amendment. See also Amar (2005, pp. 405–409). 34. “Income Tax Levy Explained by Hull,” New York Times, September 30, 1913, p. 11, Seligman (1914), Blakely (1914), Seligman (March 1918, p. 37), and Dewey (1934, pp. 489–90). The high personal exemptions were intended, in part, to limit the cost of administering the new tax by limiting the number of taxpayers. Seligman (1914, p. 12), Blakey (1914, p. 38), and Bullock (1917, p. 373). 35. The tax produced $28 million in fiscal year 1914, $41 million in 1915, and $68 million in 1916 (Blakey and Blakey, 1940, p. 103). See Dewey (1934, p. 491) for comments on the administrative burdens of the tax.

58

Chapter 4

sparking the development of the administrative infrastructure needed to collect the much larger income taxes imposed during the war. Frank Taussig, professor of economics at Harvard University, noted (in late 1917) that, “We have on hand, ready for immediate application, the machinery of the income tax. It is far from perfect and yet it is immensely serviceable. If it had now to be set up de novo, we should have to wait at least a year, probably several years, before anything like the revenue now within our grasp could be obtained from it. It makes possible a resort at once to heavy levies without waiting for the slow process of getting the machinery of taxation in working order.”36 The Revenue Act of September 8, 1916, a pre-war measure designed to finance an expanding “preparedness” program, illustrates the ready expandability of the individual income tax. The 1916 act was expected to produce more than $100 million in additional tax receipts simply by raising the normal tax rate from 1 percent to 2 percent and raising surtax rates to a maximum of 13 percent.37 The War Revenue Act of 1917 The War Revenue Act of October 3, 1917, was one of two major wartime tax acts. The House Ways and Means Committee began working on the measure almost as soon as the United States entered the war, but successive increases in projected war costs delayed final passage until the fall of 1917.38 Rather than simply bumping up the normal tax and surtax rates, the War Revenue Act levied an “additional” tax on net income. The resulting structure—characterized by one observer as “indicative of a pious hope for an early and easy repeal”39—was a kluge because taxpayers had to calculate two taxes, the basic tax required by the Revenue Act of 1916 and the additional tax required by the War Revenue Act. The War Revenue Act provided for an additional normal tax of 2 percent of net income in excess of $1,000 for a single person (or $2,000 for a married 36. Taussig (1918, p. 5). See Dewey (1934, p. 504), Blakey (1917, p. 810), Patterson et al. (1918, p. 40), and the 1918 Treasury Annual Report, pp. 935–949, for further commentary on tax infrastructure. 37. “Adds $210,000,000 to Nation’s Taxes,” New York Times, July 2, 1916, p. 1, and Blakey (1916). 38. See, Blakey (1917, p. 791), “Plan Three Billion Credit,” New York Times, April 4, 1917, p. 1, “Bond Issue Bill Will be Rushed,” New York Times, April 10, 1917, p. 1, and “Big War Loan Bill Ready for Debate,” New York Times, April 12, 1917, p. 2, for indications that the Ways and Means Committee geared up quickly for new tax legislation. The progress of the legislation is reported in “Income Rate is Increased,” New York Times, May 9, 1917, p. 1, “Tax Increased on Large Incomes,” New York Times, May 18, 1917, p. 1, “M’Adoo Asks 5 Billions More for War,” New York Times, July 25, 1917, p. 1 (McAdoo’s request “threw the committee’s plans into chaos”), “Year’s War Cost to U.S. $11,651,194,000,” New York Times, July 28, 1917, p. 1, and “Conferees Agree on War Tax Bill,” New York Times, September 28, 1917, p. 1. See also Blakey and Blakey (1940, ch. 6). 39. Haig (1919, p. 378).

59

Treasury Finance during World War I

Table 4.1 Individual income tax assessments Revenue Act of 1916

War Revenue Act of 1917a

Normal tax

2% of net income in excess of credits and exemptions

2% of net income in excess of credits and exemptions

Personal exemptions

$3,000 individual, $4,000 married couple

$1,000 individual, $2,000 married couple, $200 per dependent

$20,000–40,000

$5,000–7,500

$5,000–6,000

1%

1%

1%

Over $2,000,000

Over $1,000,000

Over $1,000,000

13%

50%

65%

Surtaxes Lowest bracket, net income Rate in lowest bracket Highest bracket, net income Rate in highest bracket

Revenue Act of 1918b 6% of first $4,000 of net income in excess of credits and exemptions, 12% of balance $1,000 individual, $2,000 married couple, $200 per dependent

Source: Revenue Act of 1916, War Revenue Act of 1917, and Revenue Act of 1918. a. In addition to taxes imposed by the Revenue Act of 1916. b. Effective for 1918 tax year.

couple) and $200 per child (table 4.1). The act also provided for additional surtaxes, starting at 1 percent of net income in excess of $5,000 and reaching a maximum of 50 percent of net income in excess of $1 million. Figure 4.3 shows the surtax rates specified in the Revenue Act of 1916 and the total surtax rates for the 1916 act and the War Revenue Act. The War Revenue Act also made an important modification to the definition of net income. Interest income from federal, as well as state and local, bonds had previously been excluded from net income. The War Revenue Act removed the exclusion of interest income from Treasury bonds issued after September 1, 1917, except to the extent provided in the legislation authorizing the debt. This meant that Congress had to reconsider the excludability of interest income from Treasury securities every time it authorized a new bond. The sharp advance in tax rates attributable to the War Revenue Act made tax exemption an important consideration in the design of wartime Treasury debt. The Revenue Act of 1918 The Revenue Act of February 24, 1919, commonly known as the Revenue Act of 1918, was the second major piece of wartime tax legislation.40 Following 40. In addition to the works cited below, see Blakey and Blakey (1940) and Haig (1919).

Chapter 4

80

60

Percent

60

40

20

0 1

10

100

1,000

Net income, thousands of dollars Revenue Act of 1916 Sum of Revenue Act of 1916 and War Revenue Act of 1917 Revenue Act of 1918

Figure 4.3 Marginal surtax rates

passage of the War Revenue Act in October 1917, observers did not expect any further tax initiatives until after June 1918, when the first income tax payments required by the War Revenue Act would come due.41 However, in May 1918, at the time of the renewed German offensive in western Europe, it became clear that taxes would have to be raised to finance the expanding costs of the war.42 On May 27, 1918, President Wilson addressed a joint session of Congress to urge prompt action on new tax legislation. The President instructed Congress that its “plain duty was to place the nation upon a strong financial basis,” that the army and navy needed money, and that it was “unsound” to rely too heavily on debt.43 Treasury Secretary McAdoo followed up with a request to double wartime tax collections.44 Prolonged wrangling—mainly over the question of a war profits tax (see below)—delayed action in the House of 41. “Congress Will Defer Tax Changes Till June,” New York Times, December 24, 1917, p. 6. 42. Kennedy (2004, p. 110) and Synon (1924, pp. 230–42). 43. “Wilson Urges New Tax Bill on Congress,” New York Times, May 28, 1918, p. 1. Wilson’s speech is reprinted in “President Calls for Revenue Legislation,” Wall Street Journal, May 28, 1918, p. 1. 44. “M’Adoo Advises Doubling War Tax,” New York Times, June 7, 1918, p. 1.

61

Treasury Finance during World War I

Representatives until late September.45 The Senate was still debating the legislation when the armistice was signed in November. Following the cessation of hostilities, McAdoo reduced his earlier requested tax increase, further delaying Congressional action.46 The Revenue Act of 1918 eliminated the “additional” tax scheme of the War Revenue Act of 1917 and prescribed a consolidated rate structure. The normal tax was set at 6 percent of net income in excess of $1,000 for a single person (or $2,000 for a married couple) and $200 per child, plus an additional 6 percent of net income in excess of the sum of $4,000 and the same personal exclusions (table 4.1). The surtax started at 1 percent of net income in excess of $5,000 and rose to 65 percent of net income in excess of $1,000,000 (figure 4.3). Corporate Income Taxes The 1895 Supreme Court decision that declared income taxes unconstitutional applied to corporate as well as individual income taxes. However, Congress overcame the constitutional impediment to a corporate income tax (without amending the Constitution) by labeling a corporate tax included in the Payne– Aldrich Tariff Act of August 5, 1909, an “excise” tax.47 That act levied a 1 percent tax on net corporate income in excess of $5,000.48 The corporate income tax rate was raised shortly before, and again following, American entry into World War I. The Revenue Act of 1916 doubled the corporate tax rate to 2 percent, the War Revenue Act of 1917 provided for an “additional” tax of 4 percent of corporate net income (thereby raising the total tax rate to 6 percent), and the Revenue Act of 1918 provided for a consolidated tax of 12 percent of 1918 net income and 10 percent of 1919 net income. Excess Profit and War Profit Taxes The first “war profits” tax of World War I—literally, a tax on profits attributable to the war—appeared in Denmark in 1915. The new tax spread to other countries and evolved into a tax on “excess” corporate profits, regardless of the reason for, or source of, the profits.49 45. “House Adopts Kitchin’s Excess Tax Rates,” Wall Street Journal, September 19, 1918, p. 8, and “$8,000,000 Tax Limit This Year,” New York Times, September 20, 1918, p. 1. 46. “M’Adoo Suggests $6,000,000 Cut in Expenditures,” New York Times, November 15, 1918, p. 1. 47. The tax was held to be constitutional in Flint v. Stone Tracy Company, 220 U.S. 107 (1911). 48. Dewey (1934, p. 486) and Blakey and Blakey (1940, ch. 2). 49. Plehn (1920, pp. 285–87).

62

Chapter 4

The Revenue Act of 1917 An excess profits tax first appeared in the United States in the Revenue Act of March 3, 1917, an act adopted to expand funding for the army and navy prior to the declaration of war by the United States. The act levied an 8 percent tax on the profits of a corporation in excess of what was asserted to be a “normal” 8 percent return on invested capital. The idea of an excess profits tax provoked sharp debate in Congress. Nicholas Longworth, Republican representative from Ohio, complained that “The tax . . . is a direct tax upon success, just plain, ordinary, modest success, the success that distinguishes efficiency from shiftlessness and thrift from wastefulness.”50 The bill barely passed in the House, avoiding defeat only because the chairman of the Ways and Means Committee made a last minute plea to his fellow Democrats to stand behind the measure.51 The War Revenue Act of 1917 Congress began to reconsider the matter of an excess profits tax almost as soon as President Wilson signed the Revenue Act of 1917. The tax rate was an issue, but so too was the question of whether the tax should be limited to profits attributable to the war. The House of Representatives wanted to continue to tax excess profits. In its initial draft of the War Revenue Act, the House Ways and Means Committee added an additional 8 percent tax on excess profits, raising the aggregate tax rate to 16 percent.52 The Senate, however, preferred something closer to a tax on profits arising from the war. The Senate Finance Committee proposed to tax war profits, defined as net income in excess of average income earned during the pre-war years of 1911, 1912, and 1913.53 Resolving the differences between the House and Senate versions took several weeks, mainly because of disagreement over whether to tax excess profits (as in the House version) or war profits (as in the Senate version). The final conference bill provided for what was essentially an excess profits tax.54 50. “Savagely Assail the Revenue Bill,” New York Times, January 31, 1917, p. 10. 51. “Pass Revenue Bill; Vote is 211 to 196,” New York Times, February 2, 1917, p. 10. The vote was not as close in the Senate—47 to 33. “Filibuster for Extra Session,” New York Times, March 1, 1917, p. 1. 52. “Income Rate is Increased,” New York Times, May 9, 1917, p. 1. 53. “War Revenue Bill Ready for Debate,” New York Times, July 3, 1917, p. 12, and “G. N. Nelson Explains Excess Profits Tax,” New York Times, July 22, 1917, p. 19. 54. Bank, Stark, and Thorndike (2008, pp. 64–65). The conference bill defined baseline profits as average income in 1911, 1912, and 1913, subject to a minimum of 7 percent of invested capital and a maximum of 9 percent. (This kept baseline profits quite close to 8 percent of invested capital.) The tax was assessed at 20 percent of net income in excess of baseline profits and less than 15 percent of invested capital. Net income in excess of 15 percent of invested capital was taxed at progressively higher rates, with a top rate of 60 percent for net income in excess of 33 percent of invested capital.

63

Treasury Finance during World War I

The Revenue Act of 1918 Interest in taxing war profits did not disappear with the adoption of the War Revenue Act. The New York Times characterized President Wilson’s May 1918 speech urging prompt action on new tax legislation as emphasizing “what has been contended for some time—that the war profits have greatly exceeded popular understanding of them. One reason for enacting another revenue law at this time is to commandeer for the Government some profits which have come indirectly from Government war contracts.”55 Following Wilson’s speech, Treasury Secretary McAdoo was reported to favor a tax on war profits of between 60 and 80 percent.56 His advocacy sparked prolonged wrangling with the House Ways and Means Committee during the summer of 1918.57 McAdoo argued that The theory of a war profits tax is to tax profits due to the war . . .. The excess profits tax must rest upon the wholly indefensible notion that it is a function of taxation to bring all profits down to one level with relation to the amount of capital invested, and to deprive industry, foresight, and sagacity of their fruits. The excess profits tax exempts capital and burdens brains, ability, and energy.58

Members of the Ways and Means Committee, however, were concerned that business would lack “stimulation” if Congress levied anything like an 80 percent tax on war profits.59 The revenue bill ultimately approved by the Ways and Means Committee, and later by the full House, provided for a dual tax scheme, whereby a corporation would be liable for the greater of a war profits tax and an excess profits tax.60 55. “Wilson Urges New Tax Bill on Congress,” New York Times, May 28, 1918, p. 1. 56. “Excess Profits Tax as McAdoo Solves It,” Wall Street Journal, June 10, 1918, p. 1. 57. See, in addition to the articles cited below, “Limits of Levy on War Profits 30 to 80 Per Cent,” New York Times, July 27, 1918, p. 1, “Turn to McAdoo’s Taxation Scheme,” New York Times, August 13, 1918, p. 8, “Tax Bill Delayed; New Levies Sought,” New York Times, August 16, 1918, p. 12, “Revenue Bill Faces Unexpected Delay,” Wall Street Journal, August 16, 1918, p. 5, “Excess Profits Tax Not Yet Approved,” New York Times, August 18, 1918, p. 7, “M’Adoo Rejects Tax Revision,” New York Times, August 20, 1918, p. 6, “Says Bill Will Net Full $8,000,000,000,” New York Times, August 21, 1918, p. 18, and “Exemption of 10% in War Profits Tax,” New York Times, August 22, 1918, p. 7. 58. “M’Adoo Demands Flat 80% Tax on War Profits,” New York Times, August 15, 1918, p. 1. 59. “Tax Bill Framers Reject Big Levy on War Profits,” New York Times, July 26, 1918, p. 1. 60. “Tax Bill Complete; M’Adoo Overruled,” New York Times, August 31, 1918, p. 9, and “House Adopts Kitchin’s Excess Tax Rates,” Wall Street Journal, September 19, 1918, p. 8. The war profits tax defined baseline profits as average income in 1911, 1912, and 1913, plus 10 percent of capital added since 1913. (Anderson, 1917, p. 879, discusses the importance of recognizing capital additions since the end of the pre-war base period.) The tax was assessed at 80 percent of net income in excess of baseline profits. The excess profits tax, on the other hand, defined baseline profits as 8 percent of invested capital. That tax was assessed at 35 percent of net income in excess of baseline profits and less than 15 percent of invested capital. Net income in excess of 15 percent of invested capital was taxed at higher rates.

64

Chapter 4

Following House approval, the bill went to the Senate, where the tax rates on excess profits were reduced after the armistice was signed.61 Summary Assuming peacetime tax receipts and expenditures of $750 million per year (based on the pre-war data in figures 4.1 and 4.2), it appears that about one quarter of the costs of the war were financed with war taxes. Annual receipts and expenditures in excess of peacetime levels were: Excess receipts

Excess expenditures

1917

$0.4 billion

$1.3 billion

1918

3.4

13.0

1919

3.9

18.2

Total

$7.7 billion

$32.5 billion

The excess (war-related) tax receipts of $7.7 billion were 24 percent of the $32.6 billion in excess expenditures. Thus the war was financed about as suggested by J.P. Morgan. The Liberty Loans The $25 billion of war-related expenditures that was not financed with taxes was financed with debt. Five large Liberty Loans, shown in table 4.2, accounted for the bulk of the debt. The most striking feature of the Liberty Loans is their size: more than an order of magnitude larger than the loans the Treasury had floated during the preceding quarter century. They were so large that, unlike earlier loans, they could not be absorbed by banks alone and had to be marketed and sold to individual investors—many of whom had little or no prior experience investing in bonds.62 One of Secretary McAdoo’s most important wartime activities was making Liberty Loans attractive to middle- and upper-income investors. The next chapter describes how McAdoo designed the loans; chapter 6 explains how the loans were marketed. 61. The final tax rates on excess profits were set at 30 percent of net income in excess of baseline profits and less than 20 percent of invested capital and 65 percent of net income in excess of 20 percent of invested capital. The final conference bill also added a cap on baseline profits for purposes of the war profits tax of 10 percent of invested capital (to limit the benefit to firms that had enjoyed unusually high pre-war profits). 62. The Treasury conjectured in 1917 that “there were only about 350,000 bond investors in the United States; the people generally were, therefore, unacquainted with Government bonds.” 1917 Treasury Annual Report, p. 6.

Treasury Finance during World War I

Table 4.2 The Liberty Loans

First Liberty Loan Second Liberty Loan Third Liberty Loan Fourth Liberty Loan Victory Liberty Loan

Form of loan

Issue date

Amount

30-yr bonds 25-yr bonds 10-yr bonds 20-yr bonds 4-yr notes

June 1917 November 1917 May 1918 October 1918 May 1919

$2.0 billion $3.8 billion $4.2 billion $7.0 billion $4.5 billion $21.4 billion

100

80

Millions of dollars

65

60

40

20

0 Jul 1917

Jan 1918

Jul 1918

Jan 1919

Jul 1919

Figure 4.4 Weekly sales of Liberty bonds on the New York Stock Exchange

The immense increase in Treasury debt led to a revival of bond trading on the New York Stock Exchange.63 Figure 4.4 shows that trading in Liberty bonds on the Stock Exchange increased from about $10 million per week in the second half of 1917 to about $45 million per week in the first half of 1919. Trading was so heavy by the spring of 1918 that the Stock Exchange began planning a separate ticker service to speed dissemination of bond prices around 63. However, bonds issued before the war continued to trade in the over-the-counter market. See, for example, “United States Government Bonds,” New York Times, October 27, 1917, p. 19 (stating that “Inasmuch as the bulk of the business in United States Government bonds is done over the counter and not on the Stock Exchange, The Times gives the following quotations obtained from bond dealers,” but then reporting bid and offer prices only for Treasury bonds other than Liberty bonds. Liberty bond prices were reported in a separate column titled “New York Stock Exchange.”)

66

Chapter 4

the country.64 Prices for Liberty bonds were reported in decimal terms, in increments (or “ticks”) of 1/50th of a percent of principal—the smallest tick size of any listed security.65 The unprecedented size of the Liberty Loans, as well as the expanding volume of tax collections, created novel cash management problems for Treasury officials. In particular, how was the Treasury to avoid disrupting shortterm credit markets when receipts came in episodically—from bond sales and tax collections—and expenditures flowed out more or less continuously? Chapters 7 and 8 examine the devices that Treasury officials introduced to smooth fluctuations in cash flow. War-Savings Stamps and Certificates In addition to Liberty bonds targeted to middle- and upper-income investors, the Treasury offered war-savings stamps and certificates to investors of more modest means.66 The first series of stamps and certificates was offered in December 1917 and during 1918. Investors could purchase, at post offices and other convenient locations around the country, war-savings stamps that would pay $5 on January 1, 1923, at a price that started at $4.12 in December 1917 and January 1918 and rose by $.01 per month to a price of $4.23 in December 1918.67 War-savings certificates were cards with space for twenty stamps. A completed certificate would therefore pay $100 on January 1, 1923. A second series of stamps and certificates, offered in 1919, had the same price structure and returned $5 and $100, respectively, on January 1, 1924.68 The war-savings program did not receive nearly as much attention from Treasury officials as the Liberty Loan program and raised only $900 million by the end of the war.69 As shown in figure 4.5, sales dropped off sharply in 1919. 64. “Bond Notes,” Wall Street Journal, April 11, 1918, p. 5, and “Bond Ticker Authorized,” Wall Street Journal, October 5, 1918, p. 5. 65. “Juggle the Bonds, Profit in Stocks,” New York Times, June 16, 1917, p. 1. For the first several months prices were reported in 50ths, so that 99.45 meant 99 and 45–50ths of a percent of principal. However, this proved confusing for many market participants and in early August the Exchange began reporting prices in decimal terms, so that a price of 99 and 45–50ths was reported as 99.90. “Liberty Bond Quotations,” Wall Street Journal, August 4, 1917, p. 1, and “Liberty Bond Quotation,” New York Times, August 4, 1917, p. 10. 66. War-savings stamps and certificates were authorized by section 6 of the Second Liberty Bond Act. 67. See, Treasury Circular no. 94, November 15, 1917, reprinted in 1917 Treasury Annual Report, p. 106. 68. See, Treasury Circular no. 128, December 18, 1918, reprinted in 1919 Treasury Annual Report, p. 291. 69. 1919 Treasury Annual Report, p. 61.

67

Treasury Finance during World War I

250

Millions of dollars

200

150

100

50

Apr 1919

Mar 1919

Feb 1919

Jan 1919

Dec 1918

Nov 1918

Oct 1918

Sep 1918

Aug 1918

Jul 1918

Jun 1918

May 1918

Apr 1918

Mar 1918

Feb 1918

Jan 1918

0

Figure 4.5 Monthly sales of war-savings stamps and certificates

War-savings stamps and certificates were nontransferable but could be redeemed at the current offering price during the year they were sold and on a schedule of gradually rising prices thereafter. Investors who purchased stamps and certificates in 1918 redeemed about 20 percent of their purchases the following year. The Treasury stated that the redemptions “represent chiefly [stamps] purchased as a result of the war-time appeal by persons who were not prepared to hold them permanently as an investment.”70 The Federal Reserve Banks as Fiscal Agents of the United States The burden on the Treasury of selling bonds and managing cash balances was substantially eased by the designation of the twelve new Federal Reserve Banks as “fiscal agents” of the United States. Prior to 1914 the Treasury had received, carried, and dispersed funds through a network of subtreasuries and national bank depositories. Section 15 of the Federal Reserve Act provided a third alternative: that government cash balances “may, upon the direction of the Secretary of the Treasury, be deposited in Federal reserve banks, which banks, when required by the Secretary of the Treasury, shall act as fiscal agents 70. 1919 Treasury Annual Report, p. 62.

68

Chapter 4

of the United States; and the revenues of the Government or any part thereof may be deposited in such banks, and disbursements may be made by checks drawn against such deposits.” On November 23, 1915, Secretary McAdoo appointed the Reserve Banks as government depositaries and fiscal agents and authorized them to accept deposits of public funds and to pay checks written against those deposits, “as well as [to perform] any other services incident to or growing out of the duties and responsibilities of fiscal agents.”71 During 1916 and the first three months of 1917, the Banks’ activities as fiscal agents were limited to receiving funds from tax collectors, paying Treasury checks, and paying coupons clipped from bearer Treasury bonds.72 However, their duties expanded dramatically following the U.S. entry into the war, to include: • processing subscriptions to new issues of Liberty bonds, delivering new issues, and receiving payment for new issues,73 and • administering Treasury deposits in banks and other financial institutions resulting from tax collections and securities sales, and maintaining custody of securities pledged to secure those deposits. Additionally, as discussed in chapter 6, Federal Reserve officials provided critical leadership in the five Liberty Loan campaigns.74 71. “Federal Reserve Banks as Fiscal Agents,” Federal Reserve Bulletin, December 1915, p. 395, and 1916 Treasury Annual Report, p. 6. 72. 1916 Federal Reserve Bank of New York Annual Report, p. 21. See also Hollander (1919, p. 127). 73. The role of the Reserve Banks as fiscal agents in connection with an offering of Treasury securities was first noted in a Federal Reserve Bank of New York circular dated May 2, 1917 describing an offering of certificates of indebtedness. On May 14, 1917, the Secretary of the Treasury designated the Reserve Banks as “fiscal agents of the United States to collate applications and to give notices of the allotments which the Secretary of the Treasury will eventually make to subscribers and to issue interim certificates for payments made on loan subscriptions.” Federal Reserve Bulletin, June 1917, p. 423. See also the offering circular for the first Liberty Loan, Treasury Circular no. 78, May 14, 1917, reprinted in 1917 Treasury Annual Report, p. 87. 74. Federal Reserve officials actively sought expanded fiscal agency responsibilities after the United States entered the war. The biographer of Benjamin Strong, the first Governor of the Federal Reserve Bank of New York, concluded that “With the United States in the war, Strong increased his efforts to acquire fiscal agency functions for the Reserve Banks . . .. He believed [a transfer of functions from the subtreasuries] would help the Federal Reserve and the country’s financial system in several ways: as ‘banker to the Government’ the Federal Reserve would gain prestige and a chance to demonstrate its usefulness; the transfer of deposits would augment the gold reserves of the Reserve Banks; moneys could be handled with less disturbance to money markets; and the system for redeeming Treasury currency would be improved.” Chandler, (1958, p. 105).

5

Designing the Liberty Loans

The five Liberty Loans that financed two-thirds of the cost of American participation in World War I were—like the Spanish–American War bonds of 1898—sold in fixed-price subscription offerings. Treasury officials specified the terms of an issue and offered the issue at par; subscribers indicated how much they wanted to buy. Following a three- or four-week subscription period, officials totaled the orders and announced what each subscriber would receive. Table 5.1 and figure 5.1 show the amount offered, the amount subscribed, and the amount allotted for each of the five loans.1 None of the loans failed; to the contrary, every loan was materially oversubscribed. The success of the Treasury in financing the war was especially notable because the Liberty Loan program evolved over time. Expanding financing requirements and wartime changes in federal tax legislation necessitated unprecedented experimentation with the terms of each successive offering. The crux of the problem was to design securities that, at least cost to taxpayers, would attract the immense sums required by the war. It wouldn’t do to have a failed offering and for the Treasury to have to resort to short-term loans or fiat money—as it did during the Civil War,2 nor would it do to generate a flood of subscriptions for bonds priced too cheaply—as had happened with the Spanish–American War bonds. Treasury officials needed to offer securities that were attractive, but not overly attractive, to a broad range of middle- and upper-income investors. The First Liberty Loan The First Liberty Bond Act of April 24, 1917, authorized the sale of up to $5 billion of bonds. The pressing need to fund loans to France and Britain propelled the act through Congress at an astonishing pace. The largest loan authorization in American history secured House approval on April 14, Senate approval on April 17, and was signed into law by President Wilson on April 1. The respective offering circulars stated whether the Treasury would fill all subscriptions, fill a specified fraction of subscriptions in excess of the amount offered, or sell only what was offered. See Treasury Circular no. 78, May 14, 1917, reprinted in 1917 Treasury Annual Report, p. 87; Treasury Circular no. 90, October 1, 1917, reprinted in 1917 Treasury Annual Report, p. 101; Treasury Circular no. 111, April 6, 1918, reprinted in 1918 Treasury Annual Report, p. 165; Treasury Circular no. 121, September 28, 1918, reprinted in 1918 Treasury Annual Report, p. 179; and Treasury Circular no. 138, April 21, 1919, reprinted in 1919 Treasury Annual Report, p. 242. There was no overallotment on the first issue because Treasury Secretary McAdoo wanted to minimize investor uncertainty in light of the unprecedented size of the offering. Committee on Ways and Means (1917, p. 41). McAdoo changed to a 50 percent overallotment on the second loan, and filled all subscriptions on the third and fourth loans, as financing requirements became more pressing. 2. Patterson (1954, ch. 3) describes Treasury debt management during the Civil War. See also Robinson (1955).

70

Chapter 5

Table 5.1 Terms of the five Liberty Loans First Liberty Loana

Second Liberty Loanb

Third Liberty Loanc

Fourth Liberty Loand

Victory Liberty Loane

Coupon rate

3½%

4%

4¼%

4¼%

Dated date

Jun 15, 1917 Jun 15, 1932 Jun 15, 1947 Full

Nov 15, 1917

May 9, 1918

Oct 24, 1918

Nov 15, 1927

Not callable

Oct 15, 1933

3¾% if non-taxable, 4¾% if taxable May 20, 1919 Jun 15, 1922

Nov 15, 1942

Sep 15, 1928

Oct 15, 1938

Normal and corporate

Normal and corporate

Partial

Yes

None

None

$2.0 billion

One time only $3.0 billion

May 20, 1923 Full (3¾% issue) or partial (4¾% issue) None

$3.0 billion

$6.0 billion

$4.5 billion

$3.0 billion

$4.6 billion

$4.2 billion

$7.0 billion

$5.2 billion

$2.0 billion

$3.8 billion

$4.2 billion

$7.0 billion

$4.5 billion

May 15–Jun 15, 1917

Oct 1–Oct 27, 1917

Apr 6–May 4, 1918

Sep 28–Oct 19, 1918

Apr 21–May 10, 1919

First call date Maturity date Income tax exemption

Conversion option Amount offered Amount subscribed Amount allotted Subscription period a. b. c. d. e.

30-yr bond, callable in 15 yrs. 25-yr bond, callable in 10 yrs. 10-yr bond. 20-yr bond, callable in 15 yrs. 4-yr note, callable in 3 yrs.

24—a mere eighteen days after the declaration of war. The act specified a maximum coupon rate (3½ percent) and a minimum issue price (par) for the bonds, but left maturity up to the Secretary of the Treasury. The act further provided that the bonds were to be offered “as a popular loan, under such regulations prescribed by the Secretary of the Treasury as will give all citizens of the United States an equal opportunity to participate therein.” The key design issues for the First Liberties were maturity, tax exemption, and mitigation of the risk of higher interest rates on subsequent issues. Maturity was easy. Treasury Secretary McAdoo was well aware of the financial lessons of the Civil War.3 Thus the First Liberty bonds did not mature for thirty years. The other two issues were not so easily resolved. 3. McAdoo (1931, pp. 372–74).

71

Designing the Liberty Loans

8 7

Billions of dollars

6 5 4 3 2 1 0 First

Second Offered

Third Subscribed

Fourth

Victory

Allotted

Figure 5.1 Liberty Loan offerings, subscriptions, and allotments

Tax Exemption Recent past practice in the United States had been to exempt interest on Treasury bonds from federal income taxes as a matter of tax law and issuance authority. The income tax provisions of both the Underwood–Simmons Tariff Act of 1913 and the Revenue Act of 1916 exempted interest on Treasury securities. In addition, interest on all six outstanding Treasury bonds (the 4 percent bonds redeemable in 1925, the Spanish–American War bonds, the 2 percent Consols, and the three Panama Canal bonds) was, pursuant to the statutes authorizing their issuance, “exempt from all taxes and duties of the United States.”4 Congress opted in April 1917 to continue in the same vein. The First Liberty Bond Act specifically exempted interest on bonds issued pursuant to the act from all federal taxes other than estate taxes. Secretary McAdoo later testified that he was “unwilling, in connection with the first war-loan act and before the first great issue of bonds, to propose anything involving so radical a departure from the recent practice of the United States as the taxation by the United States of its own bonds . . ..”5 However, the relatively low coupon rate made 4. Patterson (1954, ch. 9) traces the contractual exemption of Treasury bonds from federal taxes to the Refunding Act of July 14, 1870. 5. Committee on Ways and Means (1917, p. 20).

72

Chapter 5

possible by the bonds’ tax exemption reduced the attractiveness of the bonds to middle-income investors. The New York Times reported that, “Lawyers familiar with such matters were of the opinion that the conditions . . . of the Government war bonds . . . were not such as to attract investments from men of small means. They were termed rich men’s bonds . . ..”6 The Risk of Higher Interest Rates on Later Issues France, Britain, and Germany had been fighting for almost three years when the United States entered World War I. There was no way of knowing how much longer the war might continue, nor how high interest rates might rise, but it was reasonable to assume that investors would accept a lower rate of interest on the first series of Liberty bonds if they did not have to bear the risk of future losses due to rising interest rates. The Treasury could assume the risk of rising interest rates by issuing shortor intermediate-term debt and refinancing the debt from time to time at current market rates. However, that approach would leave it exposed to the risk of a financing crisis if the war went badly or lasted longer than expected. Alternatively, and more innovatively, it could sell long-term debt and agree to step up the coupon on the debt if it later sold additional debt at a higher yield. The First Liberty Bond Act left McAdoo to decide whether to include what was commonly known as a “conversion option” in the First Liberties. Section 5 of the act provided that “any series of bonds issued under authority of . . . this Act may, under such terms and conditions as the Secretary of the Treasury may prescribe, be convertible into bonds bearing a higher rate of interest than the rate at which the same were issued if any subsequent series of bonds shall be issued at a higher rate of interest before the termination of the war . . ..” McAdoo chose to include a conversion option in the First Liberties, but in an unusual way. The bonds provided that: If any subsequent series of bonds . . . shall be issued by the United States at a higher rate of interest than 3½ per cent per annum before the termination of the war . . ., the holders of any of the bonds of the present issue shall have the privilege of converting the same . . . into an equal par amount of bonds bearing such higher rate of interest and substantially identical with the bonds of such new series, except that the bonds issued upon such conversion are to be identical with the bonds of the present series as to maturity of principal and interest and terms of redemption.7

The conversion option for the First Liberties was not a simple option to convert the bonds into a subsequent series of bonds with a higher interest rate. If an investor chose to convert her bonds, she would receive new bonds that matured on the same date as the First Liberties and that had the same call provisions as the First Liberties. Keeping the redemption provisions unchanged precluded 6. “Bankers Welcome Plan for War Loan,” New York Times, April 12, 1917, p. 2. 7. Treasury Department Circular no. 78, May 14, 1917, reprinted in 1917 Treasury Annual Report, p. 87. Emphasis added.

73

Designing the Liberty Loans

an unwieldy amount of debt maturing on a single date.8 However, the conversion option was not a simple coupon step-up; a converting bondholder would receive bonds that had all the provisions of the higher coupon series other than its redemption provisions. Thus, for example, an investor might convert into a security that had less liberal tax exemptions as well as a higher coupon rate. The Second Liberty Loan The sale of the First Liberty bonds was hardly complete when, in August 1917, McAdoo was back before Congress requesting additional authority to issue bonds. The Treasury had $3 billion of bond issuance authority remaining under the First Liberty Bond Act (it had offered and sold only $2 billion of the $5 billion of bonds authorized by the act), but McAdoo wanted to borrow more than $3 billion in the second offering and he wanted to limit the tax exemptions on the new bonds—an action that would require a coupon rate in excess of the 3½ percent maximum specified in the First Liberty Bond Act. Congress responded to McAdoo’s request with the Second Liberty Bond Act of September 24, 1917, authorizing the sale of not more than $7.5 billion of bonds with a coupon rate not in excess of 4 percent priced at not less than par. The key features of the Second Liberties were, again, the tax and convertibility provisions. Tax Exemption The Second Liberty Bond Act substantially narrowed the tax exemption of interest on bonds issued pursuant to the act. The act removed the exemption with respect to individual income surtaxes and excess profit and war profit taxes. However, it retained the exemption with respect to normal income taxes. It also provided an exemption for interest on up to $5,000 principal amount of bonds with respect to surtaxes and profit taxes. As a result the interest on the Second Liberties would remain exempt for middle-income investors but would be taxed when paid to wealthy investors. Treasury officials chose to limit tax exemptions for the second series of Liberty bonds—and to pay a higher interest rate—because they wanted to attract more middle-income investors. McAdoo testified before the House Ways and Means Committee that: The indications are that the field of the great middle class of . . . people of moderate means was barely [benefited by the tax exemptions on the First Liberties], and the belief is that a 4 per cent bond, exempt from the normal tax, will appeal much more strongly to such people. . . . Bonds can not be sold in billions on the basis of what they may prove to be worth to the very rich.9

8. See, Committee on Ways and Means (1917, p. 23) (McAdoo testimony). 9. Committee on Ways and Means (1917, p. 21).

74

Chapter 5

McAdoo acknowledged that substantially eliminating the exemption from surtaxes and excess and war profit taxes, even while retaining the exemption from normal income taxes, was a judgmental matter: “My judgment, after the experience we have had with [the first Liberty bond], . . . is that a 4 per cent bond, . . . exempt from United States normal income taxes . . ., will appeal to a greater number of our people and to a larger amount of the investment wealth of this country than any other bond we can issue.”10 McAdoo also wanted to avoid a concentration of fully tax-exempt bonds in the hands of a limited number of wealthy investors. He argued that “If the war continues for a long period and the bond issues increase greatly, there will, if the policy of total exemption is carried out, come to be a class of people of great wealth who are in the enjoyment of great incomes wholly free from tax burden.”11 Convertibility The desirability of a conversion option was scrutinized more closely during congressional hearings on the Second Liberty Bond Act than had been possible in the rush to pass the first act. Cordell Hull, a Democratic representative from Tennessee, wanted to eliminate the option out of concern that “at some critical period the Government would find it necessary to sell bonds at a very high rate, and that all the existing billions of bonds would have their interest rates increased to parallel the new issue.”12 McAdoo agreed that he “would much prefer not to have the privilege of conversion in the bonds,” but felt it “very essential for the present.”13 The Second Liberty Bond Act left the decision to include a conversion option to Treasury officials, but substantially limited the scope of the option: “If the privilege of conversion so conferred under this Act shall once arise, and shall not be exercised . . ., then such privilege shall terminate as to such bonds and shall not arise again though again thereafter bonds be issued bearing interest at a higher rate or rates.” The limitation corrected what The Wall Street Journal described as an “oversight” in the earlier legislation.14 10. Committee on Ways and Means (1917, p. 42). 11. Committee on Ways and Means (1917, p. 20). Thomas Adams, professor of political economy at Yale University, made a similar point more pungently: “We want no class of taxless creditors unless it is demonstrated beyond all doubts that it pays those who are bearing the taxes to support such a group of apparently burdenless drones.” Adams (1917, p. 293). 12. “Bond Bill Ready for House Debate; Changes Defeated,” New York Times, September 1, 1917, p. 1. 13. Committee on Ways and Means (1917, p. 48). 14. “Holders of 3½% Bonds Seek Conversion Data,” Wall Street Journal, October 3, 1917, p. 8. As a result of the “oversight,” the First Liberty bonds would be convertible following issuance of the Third and Fourth Liberty bonds as well as following issuance of the Second Liberty bonds. In contrast, the Second Liberties would be convertible only once, following issuance of the Third Liberties.

75

Designing the Liberty Loans

Consistent with his reply to Hull, McAdoo included a conversion option in the Second Liberties. The option was identical to that in the First Liberties, except that a bondholder would have only one opportunity to convert. The Third Liberty Loan The sale of $3.8 billion of Second Liberty bonds in October 1917 left McAdoo with authority to issue an additional $3.7 billion of bonds. However, by the end of March 1918 McAdoo had concluded that the rate on the third offering would have to be greater than the 4 percent maximum specified in the Second Liberty Bond Act, so he returned to Congress to ask for authority to issue bonds with a higher coupon rate.15 The Third Liberty Bond Act of April 4, 1918, authorized the sale of not more than $12 billion of bonds with a coupon rate not in excess of 4¼ percent and priced at not less than par. The $3.8 billion of Second Liberties counted against the $12 billion ceiling, so McAdoo was left with authority to issue up to $8.2 billion of new bonds. Interest on the new bonds was subject to the same taxes as interest on the Second Liberties. The Problem of Falling Bond Prices In designing the Third Liberties, McAdoo had to address a new problem: how to make the bonds attractive to investors who had suffered losses during the winter of 1917 to 1918 on earlier purchases of First and Second Liberty bonds. The First Liberties were issued at par and did not drop below 99 before the end of the subscription period for the Second Liberties on October 27 (figure 5.2). However, by mid-February 1918 the First Liberties were below 98 and the Second Liberties were below 95. Observers offered several explanations for the decline in bond prices. In December 1917 the Federal Reserve Banks had increased their discount rates on loans to member banks secured by Treasury bonds from 3½ percent to 4 percent (see chapter 9), leading the banks to increase what they charged their customers on similarly secured loans. The increased cost of financing the First and Second Liberties prompted some investors who had been relying on bank credit to liquidate their holdings.16 Opportunistic behavior may also have contributed to the weak bond market. The Wall Street Journal reported that an unnamed Treasury official attributed some bond sales to “patriots for publication”—investors who bought large 15. “New Liberty Loan to be 3 Billions; Interest Rate 4 1-4,” New York Times, March 26, 1918, p. 1, and Committee on Ways and Means (March 1918, p. 12). 16. “Why the Liberty 4s Declined,” New York Times, February 16, 1918, p. 14.

Chapter 5

104

102

Percent of principal

76

100

98

96

94

92 Jul 1917

Jan 1918 First Liberties

Jul 1918 Second Liberties

Jan 1919

Jul 1919

Third Liberties

Figure 5.2 Liberty bond prices

blocks of Second Liberties “in order that gossip or the newspapers might make the facts known,” but who subsequently “unloaded in order to reinvest for larger profits. That their continuous, heavy sales have eventually depressed the Liberty issues . . . did not much concern these public spirited gentlemen: accustomed to taking traders’ risks, they were quite willing to accept a small loss as the price of their reputation for generously ‘doing their bit.’ ”17 Falling prices were hardly conducive to future offerings. Mortimer Schiff of Kuhn, Loeb observed that, “One of the greatest handicaps to the successful floatation of government issues is a weak and rapidly declining market for current securities. People do not feel encouraged to subscribe even to national loans when they see their other investments shrink alarmingly in value and when they find it is impossible to liquidate except at great loss.”18 McAdoo underscored Schiff ’s observation: “I think it is very disturbing to . . . purchasers when the fluctuations of Governments is as wide as [it has been].”19 17. “Liberty Bonds Below Par Laid at Door of Wealthy,” Wall Street Journal, January 21, 1918, p. 8. 18. Schiff (1918, p. 45). 19. Committee on Ways and Means (March 1918, p. 30). McAdoo nevertheless railed at what he interpreted as an unreasonable sense of entitlement on the part of investors: “There is a curious feeling in the breast of the average man that, if he buys a Government bond, even though he contracts to lend his money to the Government, nevertheless if he gets tired of his investment and wants to get his money back, that he ought to be able to sell the bond at par. . . . It is extraordinary the extent to which that feeling exists.” Committee on Ways and Means (March 1918, p. 24).

77

Designing the Liberty Loans

Falling prices focused attention on a problem that Treasury officials thought they had avoided when they gave investors an option to convert their bonds in the event of a subsequent issue of higher coupon bonds. The conversion options on the First and Second Liberties failed to keep those bonds near par because, in the spring of 1918, McAdoo campaigned for an interest rate on the Third Liberties that was below what most observers thought was a market-clearing level. Treasury officials understood that “It was the general banking opinion that the rate should be 4½ per cent,” but McAdoo nevertheless believed that 4¼ percent was a “reasonable” and “sufficient” coupon rate.20 An option to increase a coupon rate to a below-market yield was, of course, less valuable than an option to increase the rate to a market yield. Additionally during the winter of 1917 to 1918 market participants were concerned that the Treasury might issue a short-term security that would not trigger the conversion options.21 The conversion option for the First Liberties could be triggered only by the sale of “bonds (not including Treasury certificates of indebtedness and other short-term obligations)”; the conversion option for the Second Liberties could be triggered only by the sale of “bonds (not including United States certificates of indebtedness, war savings certificates, and other obligations maturing not more than five years from the issue of such obligations, respectively).” Some observers thought that the Treasury should continue to rely on conversion options to stabilize bond prices. The New York Times reported that “Bankers who have given the subject consideration are of the opinion that the only way to maintain Liberty Loan bonds at par, or thereabouts, is for the Government to issue the next loan at an attractive rate, holders of the First and Second loans having the privilege of converting their bonds into the new issue.”22 McAdoo appreciated the value of the conversion options on the first two Liberty Loans23 but concluded that the options created a constituency for higher interest rates.24 At his request the Third Liberty Bond Act limited the discretion of the Secretary of the Treasury to include a conversion option to bonds with coupon rates not in excess of 4 percent. 20. 1918 Treasury Annual Report, p. 5, and Committee on Ways and Means (March 1918, pp. 12 and 25). 21. “Question of Short-Term Notes,” New York Times, January 25, 1918, p. 14, “Bankers Approve M’Adoo Loan Plan,” New York Times, February 8, 1917, p. 17, and “Bankers Look for 4½% Rate for Liberty Loan,” Wall Street Journal, February 20, 1918, p. 10. 22. “Oppose Boistering Up Prices,” New York Times, February 9, 1918, p. 18. 23. McAdoo testified that he had “felt that the convertible feature was essential until we could measure the market better and measure the temper of the American people.” Committee on Ways and Means (March 1918, p. 25). 24. Committee on Ways and Means (March 1918, p. 24) (“We have at present a certain expectation which is manifest each time before a new bond issue comes out, that the rate of interest either can be increased or ought to be increased . . .”) and 1918 Treasury Annual Report, p. 6 (the conversion privilege “had a tendency to create a demand for a constantly rising rate of interest on successive issues”).

78

Chapter 5

Box 5.1 The Estate Tax Privilege The Revenue Act of 1916 put in place what was intended to be a permanent federal tax on decedent estates.a The tax ranged from 1 percent of estates valued at less than $50,000 to 10 percent of that part of an estate valued in excess of $5 million. The Revenue Act of 1917 increased tax rates by 50 percent and the War Revenue Act of 1917 added an additional estate tax that brought total tax rates to between 2 percent and 25 percent. Commentators theorized that allowing a bond to be valued at par for purposes of paying estate taxes could enhance the value of the bond and stabilize its price. Mortimer Schiff of Kuhn, Loeb suggested that: [A]ll government bonds issued under any of the war acts be permitted to be used to pay the . . . tax on the estate of a decedent . . . As the law now is, the tax must be paid in cash within one year after the death of the testator, which means liquidation, possibly forced, of a considerable portion of an estate, while if the option were given to pay the tax in government bonds . . . there would be a great inducement to wealthy men and women to keep a considerable portion of their capital invested in such securities, which after their death could be used for this purpose.b The Third Liberty Bond Act provided for just such an option. In proposing the measure, McAdoo testified that the option would be “a very substantial attraction to induce the purchase of these bonds. It will give assurance to every bondholder that in case of his death his estate will contain assets which the United States will accept at par, whatever may be the market conditions at the time, in payment of the taxes upon his estate. That would also have a very beneficial effect upon the general market for Governments because it would keep in the market a constant demand for Government bonds . . ..”c a. Blakey and Blakey (1940, p. 113). Federal estate taxes had been adopted during earlier wartime emergencies, including the Civil War and the Spanish–American War, but did not last long after the end of an emergency. Dewey (1934, p. 501) and Blakey (1916, pp. 843–844). b. Schiff (1918, p. 44). c. Committee on Ways and Means (March 1918, p. 15). Love (1925) argued that it was unlikely that the estate tax privilege either materially enhanced the demand for Liberty bonds or stabilized the value of the bonds around par.

The failure of conversion options to stabilize the prices of the First and Second Liberties, coupled with McAdoo’s growing disenchantment with the options, led Treasury officials to seek other ways to stabilize the price of the Third Liberties. They opted for a shorter, ten-year, maturity and a novel “bond purchase fund.” (Additionally, as discussed in box 5.1, the Treasury included an “estate tax privilege” in the Third Liberties.) The contribution of a shorter maturity to limiting fluctuations in the price of a bond is self-evident. The bond purchase fund merits some discussion. The Bond Purchase Fund The Third Liberty Bond Act authorized, but did not require, the Secretary of the Treasury to repurchase outstanding Liberty bonds, subject to several restrictions on volume and price. (The Treasury could not repurchase annually more than 5 percent of the amount of a bond outstanding at the beginning of the year

79

Designing the Liberty Loans

and it could not pay an average price in excess of par.) What came to be known as the “bond purchase fund” was not intended to redeem war debt, but rather to stabilize the secondary market price of the debt near its “true value.”25 McAdoo recognized that using the proceeds of bond sales to repurchase outstanding bonds was anomalous, but argued that the exigencies of the war demanded aggressive measures: “It is a most difficult thing, but a thing which we must face, to try to keep Government bonds measurably around par. We have got to continue Government borrowings at a reasonable rate of interest. I think that some such thing as this sinking fund will be much cheaper for the Government than to increase the interest rate; at least, we ought to try it.”26 (Contrary to McAdoo’s hopes, the bond purchase fund proved to be ineffective in limiting further price declines. In early September 1918 he admitted that the fund had been unable to “sustain the price [of the bonds] against adverse developments.”27) The Fourth Liberty Loan The sale of $4.1 billion of Third Liberties in the spring of 1918 left McAdoo with authority to issue an additional $4.1 billion of bonds. However, he wanted to offer more than that amount in the fall of 1918, and he wanted room to accept all subscriptions, as he had in the Third Liberty bond sale. Congress agreed and, in passing the Fourth Liberty Bond Act of July 9, 1918, gave McAdoo authority to issue up to $12.1 billion of new bonds with a coupon rate not in excess of 4¼ percent.28 In designing the Fourth Liberty Loan, Treasury officials had to cope with the concurrent Congressional debate over what ultimately became the Revenue Act of 1918. On May 27, 1918, President Wilson had urged Congress to increase taxes substantially. He made it clear that he expected legislation before the November elections—and thereby quashed expectations that Congress would not consider any new tax measures before the fourth tranche of 25. Representative Cordell Hull (Democrat, Tennessee) observed that “If the true value of Liberty 4s . . . were less than the issue price in the light of conditions present and prospective, the proposed legislation to prevent depreciation would not be so fully justified.” “Seeks Protection for Liberty Bonds,” New York Times, February 9, 1918, p. 2. The bond purchase fund was modeled after a similar UK facility that was said to keep the price of British war bonds within a fraction of a percent of par value. “Seeks Protection for Liberty Bonds,” New York Times, February 9, 1918, p. 2, and “Believe $3,000,000,000 of Bonds at 4¼ Will Go,” Wall Street Journal, March 27, 1918, p. 7. 26. Committee on Ways and Means (March 1918, p. 24). 27. “M’Adoo Asks Liberty Loan Tax Legislation,” Wall Street Journal, September 12, 1918, p. 10. See also 1918 Treasury Annual Report, p. 14. Treasury officials did not subsequently make any substantial use of the bond repurchase fund before the war ended. Prior to November 1918 they repurchased only about $250 million of bonds. 1918 Treasury Annual Report, p. 71. 28. The Fourth Liberty Bond Act authorized the issuance of up to $20 billion of bonds. The $3.8 billion of Second Liberties and the $4.1 billion of Third Liberties counted against the $20 billion ceiling, so McAdoo was left with authority to issue up to $12.1 billion of new bonds.

80

Chapter 5

Liberty bonds came to market. Immediately following Wilson’s speech, the tax-exempt First Liberties rose almost a point in price and the taxable Second and Third Liberties fell a point. By the end of August the Third Liberties were trading below 95, and it was clear that the Treasury was not going to be able to float a fourth Liberty Loan except at an interest rate in excess of the statutory maximum (4¼ percent) or with greater tax exemptions than those on the Third Liberties. Expanding the Tax Exemptions on the Fourth Liberties In a September 5, 1918, letter to the chairman of the House Ways and Means Committee, McAdoo requested expanded tax exemptions for the Fourth Liberties, explaining that he wanted to avoid going above the 4¼ percent interest rate that he had previously characterized as reasonable and sufficient: “I have been anxious to stabilize the interest rate upon Government bonds, believing that by so doing we should be reducing the cost of the war.”29 He observed that: [W]e can not keep the interest rate on Government bonds stationary, or substantially so, and continue indefinitely to increase the surtaxes, to which the income from those bonds is subject, without at the same time limiting the market for Liberty bonds to those who have little or no surtaxes to pay. . . . In order to give the numerous small holders of Liberty bonds the advantage of a market upon which they may sell their bonds in case of necessity, and also to attract subscriptions from the great number of investors of ample means, but not of great wealth, it will be necessary immediately either to increase the interest rate or to neutralize the increased surtaxes by freeing the bonds to a limited extent from such taxes.30

McAdoo chose to request a change in the tax treatment of the next series of Liberty bonds: I recommend that a portion of the income of these bonds should be free from surtaxes for the period of the war and for a brief interval thereafter. This course would make it possible to meet the exigencies of the present situation and to counterbalance the adverse effect on the market value of Liberty bonds of the increased surtax rates, and at the same time would not be open to the very grave objection which exists against any unlimited or permanent exemption, which would deprive the Government of the United States of the power to meet its necessities in the future by [surtaxes] on incomes derived from Liberty bonds.31

Congress accepted McAdoo’s idea and included it in the Supplement (of September 24, 1918) to the Second Liberty Bond Act. The Supplement provided for the same basic tax exemptions as the Second and Third Liberty Bond Acts, but also provided: That until the expiration of two years after the date of the termination of the war between the United States and the Imperial German Government . . . (1) The interest on an amount of bonds of the Fourth Liberty Loan the principal of which does not exceed $30,000, owned by any individual, partnership, association, or corporation, shall be

29. 1918 Treasury Annual Report, p. 14. 30. 1918 Treasury Annual Report, p. 15. 31. 1918 Treasury Annual Report, p. 15.

81

Designing the Liberty Loans

exempt from graduated additional income taxes, commonly known as surtaxes, and excess profits and war-profits taxes, now or hereafter imposed by the United States . . ..

McAdoo further recommended that those who subscribed to the Fourth Liberty Loan should be given some additional measure of tax exemption for interest on outstanding Liberty bonds that were not already tax exempt.32 This innovative feature would enhance demand for the Fourth Liberties—because buying Fourth Liberties would provide additional partial tax exemption for interest on older bonds—as well as enhance demand for previously issued Liberty bonds (because some of the interest on those bonds would be exempted from some taxes). Congress accepted this recommendation and, as described in more detail in box 5.2, included it in the Supplement to the Second Liberty Bond Act. Box 5.2 Tax Exemptions on Interest Income from Outstanding Liberty Bonds Provided by the Fourth Liberty Bonds Pursuant to Secretary McAdoo’s recommendation that those who subscribed to the Fourth Liberty Loan should be given some measure of tax exemption for interest on outstanding Liberty bonds,a the Supplement (of September 24, 1918) to the Second Liberty Bond Act provided that, until two years after the end of the war, The interest received after January 1, 1918, on an amount of the First Liberty Loan Converted, dated either November 15, 1917, or May 9, 1918, the Second Liberty Loan, converted and unconverted, and the Third Liberty Loan, the principal of which does not exceed $45,000 in the aggregate, owned by any individual, partnership, association, or corporation, shall be exempt from [surtaxes, excess profit taxes, and war-profit taxes]: Provided, however, That no owner of such bonds shall be entitled to such exemption in respect to the interest on an aggregate principal amount of such bonds exceeding one and one-half times the principal amount of bonds of the Fourth Liberty Loan originally subscribed for by such owner and still owned by him at the date of his tax return . . ..b and that, The interest on an amount of bonds, the principal of which does not exceed $30,000, owned by any individual, partnership, association, or corporation, issued upon conversion of 3½ per centum bonds of the First Liberty Loan in the exercise of any privilege arising as a consequence of the issue of bonds of the Fourth Liberty Loan, shall be exempt from such taxes. a. 1918 Treasury Annual Report, p. 16. b. The bonds described as “the First Liberty Loan Converted, dated either November 15, 1917, or May 9, 1918,” refers to the 4 percent bonds obtained from conversion of the First Liberties following the sale of the Second Liberties and the 4¼ percent bonds obtained from conversion of the First Liberties following the sale of the Third Liberties, respectively. The bonds described as “the Second Liberty Loan, converted and unconverted,” refers to the 4¼ percent bonds obtained from conversion of the Second Liberties following the sale of the Third Liberties and the Second Liberties as originally issued, respectively. The $45,000 was in addition to the $5,000 exemption previously provided in the Second and Third Liberties. Committee on Ways and Means (September 1918, p. 26).

32. 1918 Treasury Annual Report, p. 16.

82

Chapter 5

Assistant Secretary of the Treasury Russell Leffingwell commented that the limited (in time and amount) tax exemptions were “calculated to attract the dollars of people of moderate means and to hold them, at least, during the period of the war. It is the belief of the Treasury that after the war there will be an ample market on which these bonds may be sold, and that those investors who are concerned about their taxes at that time will have no trouble in selling their bonds . . .. The problem is a war problem . . .. When the war is over and the Government’s expenditures begin to shrink, there will be ample time for a secondary redistribution of these bonds.”33 The tax provisions on the Fourth Liberties were well-received. The New York Times reported that bankers believed the provisions “would supply an important stimulus for the forthcoming Liberty Loan by attracting subscriptions from those prospective buyers who could find relief therein from heavy inroads on their income . . ..”34 The Victory Liberty Loan After the armistice went into effect on November 11, 1918, policy makers could see a clear end to wartime expenditures. The prospect of declining deficits reduced the incentive to issue long-term debt, and McAdoo announced that future Liberty Loans would have short maturities.35 Carter Glass replaced McAdoo as Secretary of the Treasury on December 16, 1918, but retained his predecessor’s preference for short-term financing.36 Glass had $5 billion of unused bond issuance authority, but he needed new authority to issue shorter term notes. He received that authority when Congress passed the Victory Liberty Loan Act of March 3, 1919, and authorized issuance of up to $7 billion of notes with a maturity between one and five years. The key design issue for the Victory notes was, once again, taxes. In a highly unusual development, Congress gave Glass authority to issue notes in any of four categories: 33. Committee on Ways and Means (September 1918, p. 24). 34. “Welcomed by Bankers,” New York Times, September 11, 1918, p. 8. 35. “Loan Drive’s Total was $6,989,047,000,” New York Times, November 20, 1918, p. 7. 36. Glass stated in testimony before the House Ways and Means Committee in mid-February 1919 that “Secretary McAdoo announced before he retired, and I have confirmed the announcement, that the Victory Liberty loan will be of short maturities.” Committee on Ways and Means (1919, p. 14). Short-term debt was also favored by market participants. See, “Bankers Consider Next Liberty Loan,” New York Times, January 28, 1919, p. 4 (“it has been quite generally agreed [among bankers and bond men] that the new Liberty Bonds should be of comparatively early maturity”) and “May Hold Up Loan Until End of Lent,” New York Times, January 30, 1919, p. 4 (“On one thing Wall Street is quite generally agreed. That is a five-year issue. The majority of the men who are paving the way for the new loan want a short-term offering [and] many of the out-of-town banks have indicated that they favor a short-term issue.”).

83

Designing the Liberty Loans

1. notes exempt from all federal taxes except estate taxes; 2. notes exempt from all federal taxes except estate taxes, income surtaxes, excess profit taxes, and war profit taxes, subject to an exemption from income surtaxes, excess profit taxes, and war profit taxes for the interest on notes not in excess of $30 thousand principal amount; 3. notes exempt from all federal taxes except estate taxes, income surtaxes, excess profit taxes, and war profit taxes; and 4. notes exempt from all federal taxes except estate taxes, income taxes, excess profit taxes, and war profit taxes. The unusual grant of discretion arose because Glass had to obtain authority to issue the notes before the 65th Congress expired on March 4, 1919, but did not want to begin offering the notes until mid-April. In view of the long interval between authorization and offering, he was reluctant to have Congress specify the tax provisions for the notes.37 Congress was unwilling to give him blanket authority to determine the tax provisions,38 but it did agree to let him chose from a menu with four choices. Congress additionally provided—see box 5.3—that those who subscribed to the Victory Liberty Loan would receive a measure of further tax exemption for interest on outstanding Liberty bonds. On April 13, 1919, Glass announced that the Treasury would offer a total of $4.5 billion of Victory Liberty Notes in two series, both callable in three years and both maturing in four years: • 3¾ percent notes exempt, like the First Liberties, from all federal taxes except estate taxes, and • 4¾ percent notes exempt, like the Second and Third Liberties, from all federal taxes except estate taxes, income surtaxes, excess profit taxes, and war profit taxes. Investors could specify whether they wanted the 3¾ percent notes or the 4¾ percent notes, and they could interconvert the notes at par at any time prior to 37. In an appearance before the House Ways and Means Committee, Glass stated “[I]n view of the early expiration of the life of the present Congress and the apparent impossibility of convening and organizing the new Congress in time to enact further bond legislation before the Victory Liberty loan campaign begins, I reluctantly ask greater latitude in the exercise of a sound discretion as to the terms of the Victory Liberty loan than has been conferred by the Congress in respect to previous loans.” Committee on Ways and Means (1919, p. 7). 38. Allen Treadway, a Republican Representative from Massachusetts, remarked that “I think there is no secret at all that the sentiment of the membership of the House . . . is absolutely against this authorization asked for.” Committee on Ways and Means (1919, p. 67). See also the comments of J. Hampton Moore, Republican Representative from Pennsylvania (Committee on Ways and Means, 1919, p. 21) (“The big proposition in this bill . . . is the removal of limitations heretofore placed upon the issue, in regard to . . . taxation . . .”) and “Short-Term Notes for Victory Loan Alone Find Favor,” New York Times, February 19, 1919, p. 1. (“The conclusion [of the Ways and Means Committee] is that it would be unwise to . . . permit [the Secretary of the Treasury] unrestricted latitude in fixing tax exemption features . . .”).

84

Chapter 5

Box 5.3 Tax Exemptions on Interest Income from Outstanding Liberty Bonds Provided by the Victory Liberty Loan As a means of increasing the attractiveness of the notes of the Victory Liberty Loan to holders of previously issued Liberty bonds, and as a means of enhancing demand for those older bonds, Congress provided that the notes would give a measure of tax exemption for interest on previously issued Liberty bonds. The offering circular for the notes stated that: [T]he interest received on and after January 1, 1919, on any amount of bonds of the first Liberty loan converted, dated November 15, 1917, May 9, 1918, or October 24, 1918, the second Liberty loan, converted and unconverted, the third Liberty loan, and the fourth Liberty loan, the principal of which does not exceed $20,000 in the aggregate, owned by any individual, partnership, association, or corporation, shall be exempt from graduated additional income taxes, commonly known as surtaxes, and excessprofits and war-profits taxes, now or hereafter imposed by the United States, upon the income or profits of individuals, partnerships, associations, or corporations; Provided, That no owner of such bonds shall be entitled to such additional exemption in respect to the interest on an aggregate principal amount of such bonds exceeding three times the principal amount of notes of the Victory Liberty loan originally subscribed for by such owner and still owned by him at the date of his tax return.a Secretary Glass defended the provision before the House Ways and Means Committee: “I believe it will be wise and proper to confer upon those holders of the old bonds who subscribe to the new loan additional exemptions from taxation. . . . Such a course would not only be a great aid to the sale of the obligations of the new loan, but should be effective to improve the market price of existing issues which has suffered from heavy liquidation due, I believe, in large measure, to the changed conditions following the cessation of hostilities.”b a. Treasury Circular no. 138, April 21, 1919, reprinted in 1919 Treasury Annual Report, p. 242. b. Committee on Ways and Means (1919, p. 13).

redemption. Glass stated that, “In fixing the terms of the issue the Treasury has been guided largely by the desire to devise a security which will not only prove attractive to the people of the country in the first instance, but the terms of which should insure a good market for the notes after the campaign is over and identical prices for the two series . . ..”39 Following the end of the subscription period on May 10, 1919, the Treasury announced that it had received tenders for a total of $5.2 billion of notes and that it had sold $0.7 billion of the 3¾ percent notes and $3.8 billion of the 4¾ percent notes.40 39. “New Loan Fixed at $4,500,000,000; Interest 4¾%,” New York Times, April 14, 1919, p. 1. 40. The notes first became interconvertible on July 16, 1919. Treasury Circular no. 139, May 20, 1919, reprinted in 1919 Treasury Annual Report, p. 362. By June 30, 1920, investors had tendered $98.4 million of the 3¾ percent notes (15 percent of the amount originally issued) for conversion into the 4¾ percent notes and had tendered $333.4 million of the 4¾ percent notes (9 percent of the amount originally issued) for conversion into the 3¾ percent notes. 1920 Treasury Annual Report, p. 427.

85

Designing the Liberty Loans

Summary The contract terms for the five Liberty Loans exhibited a level of ingenuity and a willingness to experiment not typically associated with a sovereign issuer. Interest income varied from fully tax exempt on the First Liberties to varying degrees of partial exemption on subsequent issues. Additionally the Treasury offered debt convertible into debt with higher coupon rates (the First and Second Liberty bonds) and interconvertible debt with different tax provisions and coupon rates (the Victory Liberty notes). Three factors led to the variation in contract terms. First, Treasury officials had no recent experience with large-scale financings in the presence of steeply progressive individual income taxes and were unlikely to have had any good appreciation, ex ante, for how they might finance the war at least cost to taxpayers. Second, financing requirements depended on the course of the war. A war that, in May of 1918, looked like it might continue for several years was over in six months. And finally, wartime tax legislation (which was itself related to the course of the war) had important consequences for the value of tax exemption. The most innovative contract terms were the conversion options on the First and Second Liberties. Whether those options helped to stabilize the prices of the first two Liberty bonds is difficult to assess, but the appendix to this chapter suggests that McAdoo may have been overly hasty in abandoning the options in the spring of 1918. Appendix: Conversions of the First and Second Liberty Bonds The first two Liberty bonds carried contingent conversion options providing that, if the Treasury issued a bond with a higher coupon rate before the end of the war, an investor could convert into a new bond with the higher rate. The new bond would have the same redemption provisions as the original bond but would have all of the other features of the higher coupon bond. The First Liberties were convertible every time the Treasury issued a higher coupon bond; the Second Liberties were convertible only the first time the Treasury issued a higher coupon bond. This appendix examines the conversion experiences of the two bonds. Discussing the conversion experience of the First and Second Liberties is complicated by the practical difficulty of keeping track of a bewildering variety of bonds. (The two original bonds gave rise to four additional bonds.) The wartime practice, as is followed in this book, was to describe a bond by the name of its original series and by its coupon rate. For example, the “First Liberty 4s” denotes the 4 percent bonds created by conversion of the originally issued “First Liberty 3½s.”

86

Chapter 5

Conversions of the First Liberty Bond Pursuant to the contingent conversion option on the First Liberty bonds, the First Liberty 3½s were convertible on three occasions: • following the sale of the Second Liberties in November 1917, they were convertible (between November 15, 1917, and May 15, 1918) into the First Liberty 4s; • following the sale of the Third Liberties in May 1918, they were convertible (between May 9 and November 9, 1918) into the First Liberty 4¼s; and • following the sale of the Fourth Liberties in October 1918, they were convertible (between October 24, 1918, and April 24, 1919) into the First Liberty second 4¼s. The Treasury issued $1.989 billion of the First Liberty 3½s in June 1917. Investors converted 29 percent of that amount, $568 million, into First Liberty 4s.41 Of the remaining $1.421 billion of First Liberty 3½s, investors subsequently converted $8 million into First Liberty 4¼s and $3 million into First Liberty second 4¼s.42 Conversion of the Second Liberty Bond The $3.808 billion of Second Liberty 4s issued in November 1917 became convertible into Second Liberty 4¼s following the sale of the Third Liberties in May 1918. Investors converted 76 percent, $2.884 billion, of the Second Liberty 4s between May 9 and November 9, 1918.43 Conversion of the First Liberty 4s Because the First Liberty 4s (obtained by conversion of the First Liberty 3½s following the sale of the Second Liberties) had all of the features of the Second Liberties other than redemption, they too became convertible (into the First Liberty 4¼s) following issuance of the Third Liberties. Investors had converted $568 million of the First Liberty 3½s into First Liberty 4s. They converted 65 percent of that amount, $368 million, into First Liberty 4¼s between May 9 and November 9, 1918.44 41. 1918 Annual Treasury Report, p. 68. 42. 1919 Treasury Annual Report, p. 80, and 1920 Treasury Annual Report, p. 421. 43. Committee on Ways and Means (1919, p. 33, showing $2.752 billion of the Second Liberty 4¼s outstanding on January 31, 1919, and p. 34, showing $132 million of the Second Liberty 4¼s repurchased before January 31, 1919). 44. Committee on Ways and Means (1919, p. 33, showing $376 million of the First Liberty 4¼s outstanding on January 31, 1919, $8 million of which were obtained by conversion of First Liberty 3½s), 1919 Treasury Annual Report, p. 80. It should be noted that the First Liberty 4¼s could be obtained by conversion of either the First Liberty 3½s or the First Liberty 4s.

87

Designing the Liberty Loans

Did the Conversion Options Stabilize the Prices of Convertible Liberty Bonds? The most illuminating data bearing on the question of whether conversion options stabilized the prices of convertible Liberty bonds comes from conversions following the sale of the Third Liberties, when the First Liberty 4s became convertible into the First Liberty 4¼s and the Second Liberty 4s became convertible into the Second Liberty 4¼s. The bonds tendered for conversion and the bonds received in conversion had similar tax provisions, so neither conversion involved a change in tax treatment. Since conversion of both the First Liberty 4s and the Second Liberty 4s involved a coupon step-up (to 4¼ percent) with no change in redemption provisions (conversion never changed redemption provisions) or tax treatment, the conversion options should have had positive value and the market prices of the First Liberty 4s and Second Liberty 4s should have declined following expiration of their respective conversion options on November 9, 1918. Additionally, substantially all of the First Liberty 4s and substantially all of the Second Liberty 4s should have been tendered for conversion. Figure 5.3 shows the prices of the First Liberty 4s and the First Liberty 4¼s over the two-month interval bracketing the expiration of the conversion option on the First Liberty 4s. It is clear that market participants believed that the option had value: the market price of the First Liberty 4s fell following 100

Percent of principal

98

96

94

92 Oct 13

Oct 20

Oct 27

Nov 3

Nov 10

First Liberty 4s

Nov 17

Nov 24

Dec 1

Dec 8

Dec 15

First Liberty 4¼s

Figure 5.3 Price of the First Liberty 4s and First Liberty 4¼s around the expiration of the option to convert the 4s into the 4¼s on November 9, 1918

Chapter 5

100

98

Percent of principal

88

96

94

92 Oct 13

Oct 20

Oct 27

Nov 3

Nov 10

Second Liberty 4s

Nov 17

Nov 24

Dec 1

Dec 8

Dec 15

Second Liberty 4¼s

Figure 5.4 Price of the Second Liberty 4s and Second Liberty 4¼s around the expiration of the option to convert the 4s into the 4¼s on November 9, 1918

expiration of the option. The average difference between the prices of the First Liberty 4¼s and the First Liberty 4s in the two weeks after the expiration of the option was 2.07 percent of principal value, suggesting that the option was worth about 2 points prior to expiration. We noted above that investors converted 65 percent of the First Liberty 4s. The 35 percent shortfall may be attributable to a lack of understanding of the value of conversion on the part of small investors.45 Figure 5.4 shows the prices of the Second Liberty 4s and the Second Liberty 4¼s over the same interval. It is clear that market participants also believed 45. In January 1919, Secretary Glass proposed to renew the options to convert First Liberty 4s to First Liberty 4¼s and Second Liberty 4s to Second Liberty 4¼s. In a letter to the Chairman of the House Ways and Means Committee, Glass stated his belief that “those who did not avail themselves of the conversion privilege within the period fixed by the terms of the contract . . . fall among the class of small holders who are unaccustomed to bond investments and who, on account of the very wide distribution of Liberty loan bonds were not reached by general publicity, and could not, except in the case of registered bonds, be reached by department circular. Insistence upon the letter of the contract will result in loss to a group of patriotic bondholders towards whom a special duty of consideration exists.” Committee on Ways and Means (1919, p. 14). Congress agreed and, in section 5 of the Victory Liberty Loan Act of 1919, authorized the Secretary of the Treasury to renew the conversion options. The Treasury announced the extensions in Treasury Circular no. 137, March 7, 1919, reprinted in 1919 Treasury Annual Report, p. 348. See also Supplement to Treasury Circular no. 137, June 10, 1919, reprinted in 1919 Treasury Annual Report, p. 352; Second Supplement to Treasury Circular no. 137, November 1, 1919, reprinted in 1919 Treasury Annual Report, p. 354; and Treasury Circular no. 158, September 8, 1919, reprinted in 1919 Treasury Annual Report, p. 358.

89

Designing the Liberty Loans

that the option to convert the Second Liberty 4s had value. The average difference between the prices of the Second Liberty 4¼s and the Second Liberty 4s in the first two weeks after the expiration of the option was 1.78 percent of principal value. We noted above that investors converted 76 percent of the Second Liberty 4s. These results are consistent with the proposition that conversion options buoyed the prices of the First Liberty 4s and the Second Liberty 4s by about 1¾ to 2 points. There were three other opportunities to convert Liberty bonds: (1) the opportunity to convert the First Liberty 3½s following sale of the Second Liberties in November 1917, (2) the opportunity to convert the First Liberty 3½s following the sale of the Third Liberties in May 1918, and (3) the opportunity to convert the First Liberty 3½s following the sale of the Fourth Liberties in October 1918. All three opportunities involved conversion of a tax-exempt bond into a bond subject to income surtaxes, war profit taxes, and excess profit taxes. The first two conversion options were worthless when they expired.46 The First Liberty 3½s traded at 99 and the First Liberty 4s traded at a lower price of 95.15 on May 15, 1918; the First Liberty 3½s traded at 100 and the First Liberty 4¼s traded at a lower price of 98.60 on November 9, 1918. The options were worthless because increases in surtax rates made tax exemption more valuable than a 50 or 75 basis point increase in coupon rate. 46. We can not assess the value of the option to convert the First Liberty 3½s into the First Liberty second 4¼s. Only a trivial amount of the First Liberty second 4¼s was created (by conversion of First Liberty 3½s) and there is no market price data for that bond.

6

Marketing the Liberty Loans

More than a decade after the end of the First World War, the wartime Secretary of the Treasury proudly recalled the Liberty Loan campaigns: “We went direct to the people; and that means to everybody to business men, workmen, farmers, bankers, millionaires, school teachers, laborers.”1 Secretary McAdoo’s recollection focused on what was perhaps the most striking feature of the Liberty Loans: the immense effort devoted to attracting individual investors. The Treasury had sought out individual investors before, when it financed the Spanish–American War, but the Liberty Loan campaigns went far beyond that earlier episode. During World War I the Treasury, the Federal Reserve Banks, and an army of volunteers collaborated in an unprecedented marketing effort. Making the Bonds Accessible Congress specified that Liberty bonds were to be offered “as a popular loan, under such regulations . . . as will give all citizens of the United States an equal opportunity to participate . . ..”2 The first step in marketing the bonds as popular loans was making them accessible to individual investors. Three features addressed the issue of accessibility: fixed-priced offerings, installment purchase plans, and prompt delivery sales. Fixed-Price Offerings All five Liberty Loans were offered on a subscription basis at a fixed price of par. The decision to rely on fixed-price offerings in lieu of auctions is striking because auctions might have resolved the difficult problem of identifying market-clearing yields for the enormous loans. The decision not to auction Liberty bonds can not be attributed to a lack of experience: seven of the ten Treasury offerings in the preceding quarter century had been auction offerings. However, Treasury officials had learned—in the course of financing the Spanish–American War—that fixed-price offerings had an advantage over auctions when they wanted to appeal directly to a large number of relatively unsophisticated investors: fixed-price offerings were simple, transparent, and did not put individual investors at a disadvantage to large institutional investors by exposing them to the risk of bidding too high and paying more than necessary. Fixed-price offerings also offered the prospect that large oversubscriptions—orders for substantially more bonds than what the Treasury was offering—would signal popular support for the war. In early 1918 The New York Times pointed out that the war loans: 1. McAdoo (1931, pp. 378). 2. Section 1 of the First Liberty Bond Act. The other Liberty Bond acts used similar language.

92

Chapter 6

Table 6.1 Installment payments on Liberty Loans First Liberty Loan: 2% on application (on or before June 15, 1917): 18% on June 28, 1917 20% on July 30, 1917 30% on August 15, 1917 30% on August 30, 1917 Second Liberty Loan: 2% on application (on or before October 27, 1917): 18% on November 15, 1917 40% on December 15, 1917 40% on January 15, 1918 Third Liberty Loan: 5% on application (on or before May 4, 1918): 20% on May 28, 1918 35% on July 18, 1918 40% on August 15, 1918 Fourth Liberty Loan: 10% on application (on or before October 19, 1918): 20% on November 21, 1918 20% on December 19, 1918 20% on January 16, 1919 30% on January 30, 1919 Victory Liberty Loan: 10% with application (on or before May 10, 1919): 10% on July 15, 1919 20% on August 12, 1919 20% on September 9, 1919 20% on October 7, 1919 20% on November 11, 1919 Source: Treasury Annual Reports.

. . . afford opportunities for gauging public sentiment in respect to the war, for it is reasonable to assume that if the public’s approval of the war is lacking a Government cannot sell its bonds. The element of patriotism is an important one, and the large oversubscriptions of the First and Second Liberty Loans gave clear indications of the fact that the people of the United States were strongly back of the Administration and were willing to lend their surplus funds and future savings to the United States Government to be used by it in the prosecution of the war. The fact that there were such a large number of individual subscribers . . . gives testimony to the popularity of United States bonds and to the hearty support that the people of the country are giving their Government.3

Installment Purchase Plans Each of the Liberty Loans provided that an investor could pay for his or her bonds on an installment basis instead of all at once. Table 6.1 shows the payment schedules for the five loans.4 The Treasury program of installment sales was supplemented by a variety of private plans. The Federal Reserve Bank of New York reported that even in the first campaign “there was a very general demand throughout the country 3. “Federal Reserve System’s Place in Government Financing,” New York Times, January 6, 1918, p. 92. 4. The Treasury permitted, but did not require, investors to pay for their bonds in installments. In fact, more than 70 percent of the proceeds of each of the first four Liberty Loans was paid in the first installment. Hollander (1919, p. 67).

93

Marketing the Liberty Loans

for installment purchases on an easier basis than that provided in the Government plan.”5 The sponsor of a private plan undertook to buy Liberty bonds with its own funds and to then resell the bonds to small investors on installment contracts that required smaller, more frequent, payments than what the Treasury allowed. The New York Fed reported that “Employers of labor in industrial centers and elsewhere [provided] partial payment plans by which small denomination bonds could be purchased and paid for gradually out of salaries or wages.”6 During the Third and Fourth Liberty Loan campaigns, more than one hundred New York banks formed the Liberty Loan Association and agreed to subscribe for as much as $100 million of bonds for resale to small investors. The Association sold subscription booklets for $2 each, with the understanding that a buyer would make additional payments of $1 per week for 48 weeks. When all 48 payments were completed, the buyer could exchange his booklet for a $50 bond.7 Prompt Delivery Sales Beginning with the Second Liberty Loan, the Treasury offered prompt delivery on retail orders received during the subscription period, where a buyer tendered his money and was immediately handed a bond. Prompt delivery sales were limited to $1,000 in the Second Liberty Loan campaign, $10,000 in the third campaign, and were unlimited in the fourth. The sales provided instant gratification for investors who wanted “something to show.” Preferential Allotments In addition to fashioning devices intended to make Liberty bonds more accessible to retail investors, the Treasury actively allocated relatively more bonds to small subscribers. As shown in table 6.2, it filled larger percentages of the subscriptions of smaller investors in the first, second, and Victory loans. (The Treasury had done something similar when it limited sales of Spanish– American War bonds to individual investors who subscribed for $5,000 of bonds or less.) The Marketing Campaigns The Liberty Loan campaigns were characterized by long subscription periods, intense marketing efforts, and numerous subscribers. Table 6.3 shows the duration and number of subscribers for each of the campaigns. 5. 1918 Federal Reserve Bank of New York Annual Report, p. 46. 6. 1917 Federal Reserve Bank of New York Annual Report, p. 36. 7. 1918 Federal Reserve Bank of New York Annual Report, p. 47.

94

Chapter 6

Table 6.2 Partial allotments on Liberty Loans Subscription

Allotment

First Liberty Loan: Up to and including $10,000 $10,000–$100,000 $100,000–$250,000 $250,000–$2,000,000 $2,000,000–$6,000,000 $6,000,000–$10,000,000 Over $10,000,000 Second Liberty Loan: Up to and including $50,000 $50,000–$100,000 $100,000–$200,000 $200,000–$1,000,000 $1,000,000–$8,000,000 $8,000,000–$30,000,000 Over $30,000,000 Third Liberty Loan: All subscriptions filled in full Fourth Liberty Loan: All subscriptions filled in full Victory Liberty Loan: Up to and including $50,000 $50,000–$200,000 $200,000–$500,000 $500,000–$2,500,000 $2,500,000–$15,000,000 $15,000,000–$30,000,000 Over $30,000,000

100.0% 60.0%, but 45.0%, but 30.0%, but 25.0%, but 21.0% 20.2%

not not not not

less less less less

than than than than

$10,000 $60,000 $112,500 $600,000

100.0% 90.0%, but 75.0%, but 60.0%, but 50.0%, but 41.2%, but 40.8%

not not not not not

less less less less less

than than than than than

$50,000 $90,000 $150,000 $600,000 $4,000,000

100.0% 80.0%, but 70.0%, but 60.0%, but 50.0%, but 45.0%, but 42.4%

not not not not not

less less less less less

than than than than than

$50,000 $160,000 $350,000 $1,500,000 $7,500,000

Source: Treasury Annual Reports.

Table 6.3 Characteristics of the Liberty Loan campaigns

First Liberty Loan Second Liberty Loan Third Liberty Loan Fourth Liberty Loan Victory Liberty Loan

Duration

Number of subscribers

32 days (May 15–Jun 15, 1917) 27 days (Oct 1–Oct 27, 1917) 29 days (Apr 6–May 4, 1918) 22 days (Sep 28–Oct 19, 1918) 20 days (Apr 21–May 10, 1919)

4 million 9½ million 17 million 21 million 15 million

Source: Treasury Annual Reports.

95

Marketing the Liberty Loans

The Federal Reserve Banks were the lynchpins of the campaigns. Professor Lester Chandler described the basic structure: A Liberty Loan committee was formed in every Federal Reserve district, with the governor of the Reserve Bank as chairman. These committees developed comprehensive organizations involving almost every group from the Girl Scouts to Chambers of Commerce and used practically every known technique of propaganda and salesmanship to sell bonds and promote saving.8

The New York Times described the role of the Reserve Banks in similar terms: “In each district the officers of the Reserve Bank were requested to appoint a central committee and any number of subcommittees for carrying on the campaign calculated to arose the people of their respective communities . . ..”9 Chandler’s description of the efforts of Benjamin Strong, the wartime Governor of the Federal Reserve Bank of New York, supports the proposition that Reserve Bank leadership was more than nominal: [Strong] appointed the other members of the [district Liberty Loan] committee, directed the organization of the district, and threw himself into its work without regard to effects on his health. He drove his associates and drove himself, organizing bond-selling groups of many types and spurring them on to greater efforts, setting up payroll savings plans, bringing pressure on financial institutions whenever subscriptions lagged, conferring with Treasury officials, and making many public speeches.10

The Liberty Loan committees had two principal missions: to inform the public of the terms of a Treasury offering and to solicit subscriptions. The New York Times observed that neither task was simple: The American people, prior to the war, had not been accustomed to investments, and as far as Government securities were concerned they were totally ignorant. It was therefore necessary for the committees to conduct an educational campaign designed to impress upon the people of the country not only their duty of subscribing to the loan, but to advise them of the nature and character of obligations of the United States. The terms of the loan had to be explained again and again, and after the loan became fairly well understood it was necessary to arouse the public, to appeal to their sense of patriotism, and to get signatures to the subscription blanks.11

The actual leg work, the door-to-door solicitations, the speeches, and not a small part of the paperwork, was done by volunteers. The so-called Liberty Loan army totaled about one million men and women in the spring of 1918 and ultimately reached a total of about two million.12 Secretary McAdoo effusively praised their efforts in the first campaign: “Bankers, businessmen, bond houses, newspapers, press associations, and citizens generally cooperated in a 8. Chandler (1958, pp. 118–19). 9. “Federal Reserve System’s Place in Government Financing,” New York Times, January 6, 1918, p. 92. 10. Chandler (1958, p. 119). 11. “Federal Reserve System’s Place in Government Financing,” New York Times, January 6, 1918, p. 92. 12. St. Clair (1919, p. 28).

96

Chapter 6

great movement that vibrated with energy and patriotism and swept the country from coast to coast in the greatest bond-selling campaign ever launched by any nation.”13 Prior to the launch of the fourth campaign he again took note of their efforts: “The magnificent patriotism of our people and the fervor and efficiency of the Liberty Loan organization have made it possible to place the Liberty bonds in the hands of many millions of persons who had never before been investors in securities of any kind.”14 The First Liberty Loan Campaign The First Liberty Loan campaign was hastily arranged on short notice.15 President Wilson signed the First Liberty Bond Act on April 24, 1917, and the campaign began three weeks later, on May 15. Because it was the first campaign, the most pressing needs were to elevate public awareness of the financing and to enhance public understanding of what the Treasury was offering. Labert St. Clair, a chronicler of the campaigns, described a few of the efforts at elevating awareness: Pittsfield, Mass., suspended all business for one hour in order that the entire town might turn its attention to purchasing Liberty Loan Bonds. A Liberty Loan Sunday on which thousands of clergymen urged the support of the campaign opened the Liberty Loan Week and proved a great aid to the cause; a special women’s day was set aside; the Boy Scouts of America made a special campaign, and various other organizations lent their assistance . . ..16

Enhancing understanding was equally important. St. Clair noted that “[I]t was a difficult task to make the buyers understand [the bonds] . . .. It was not unusual for the [Treasury] Department to receive letters from buyers of bonds asking when they have to pay their interest on them.”17 Perhaps most striking was the early emergence of an emphasis on oversubscriptions. McAdoo announced at the beginning of the first campaign that the Treasury would sell no more than the $2 billion of bonds that it was offering. Nevertheless, within a few days The New York Times was reporting that Treasury and Federal Reserve officials “were unanimous . . . that a tremendous campaign must be carried on . . . if the Liberty Loan is to meet with the oversubscriptions which officials desire . . ..”18 Officials emphasized that a large 13. 1917 Treasury Annual Report, p. 6. 14. “M’Adoo Asks Law to Help Sales of Liberty Bonds,” New York Times, September 11, 1918, p. 1. 15. The Federal Reserve Bank of New York later noted that “The progress of the organization at first was hampered by lack of opportunity for preliminary preparation and to unavoidable delay in settling important details of the issue until the campaign had been in progress for some time.” 1917 Federal Reserve Bank of New York Annual Report, p. 35. 16. St. Clair (1919, p. 42). 17. St. Clair (1919, pp. 42–45). 18. “Must Stir Nation to Buy the Bonds,” New York Times, May 18, 1917, p. 1. Emphasis added.

97

Marketing the Liberty Loans

oversubscription would show, “in terms which could not be misconstrued by Germany that the United States is in the war to stay until Germany is defeated, and that there will be no stinting of national resources of any kind to bring about the end of autocracy.”19 The Second Liberty Loan Campaign The Second Liberty Loan campaign was more carefully planned, better organized, and far larger than the First. In New York City, for example, the first campaign had two dozen trade committees and 400 house-to-house canvassers; the second campaign had ten dozen trade committees and 35,000 canvassers.20 The New York Times reported that: From the moment when its doors open tomorrow morning until the close of business on Oct. 27, the last day of the campaign, the Treasury Department will put on full steam to make the Liberty Loan bonds the topic of discussion in schoolroom, workshop, office and home. Many cities have made preparations to open the campaign with distinctive demonstrations. Factory whistles will blow, aerial bombs will be sent up, flags will fly, and various other features will be employed to arouse enthusiasm.21

The Second Liberty Loan campaign saw the introduction of “spectacles” designed to attract public attention. Workers in Cleveland, for example, “erected a huge striking machine in the public square and permitted every purchaser of a $50 bond to strike a trigger with a maul in such a manner as to ring a bell concealed in a papier-mâché head of the Kaiser at the top of the machine.”22 The emphasis on oversubscriptions continued in the second campaign. On the eve of the campaign, McAdoo expressed his hope that “when the campaign is over it will be found that the total number of subscribers is at least ten million and the total subscriptions in excess of five billion dollars. Such a response would be notice to our enemies that the American people as a whole intend to support with all their power their Government in the vigorous prosecution of this war and the achievement of an early and lasting peace.”23 During the campaign McAdoo argued that “It is essential to the success of the war and to the support of our gallant troops that these loans shall not only be subscribed, but oversubscribed.”24 And at the end of the campaign, after 19. “Must Stir Nation to Buy the Bonds,” New York Times, May 18, 1917, p. 1. 20. 1917 Federal Reserve Bank of New York Annual Report, pp. 33–34 and 37–39. 21. “Bond Campaign On; M’Adoo Expects 10 Million to Buy,” New York Times, October 1, 1917, p. 1. 22. St. Clair (1919, p. 46). 23. “M’Adoo to Offer 3 Billion or More in Bonds on Oct. 1,” New York Times, September 28, 1917, p. 1. 24. “Bond Campaign On; M’Adoo Expects 10 Million to Buy,” New York Times, October 1, 1917, p. 1. See also “M’Adoo Warns Nation against Failure of Loan,” New York Times, October 13, 1917, p. 1 (quoting McAdoo as saying that the “efficiency of the United States in the war would be impaired and the prestige of the United States before the world would be jeopardized if the Second Liberty Loan was not oversubscribed”).

98

Chapter 6

collecting subscriptions for $4.6 billion of bonds from 9½ million subscribers, he concluded that “The fact that such a vast number subscribed for bonds is significant of the widespread interest of the people in the purposes of the war, and of their determined support of the Government in all measures required for its vigorous prosecution.”25 The Second Liberty Loan campaign also saw the emergence of “borrow and buy,” where investors were encouraged to take out bank loans to buy bonds. Bank loans were not unknown in the first campaign but they became more widely and openly encouraged in the fall of 1917.26 The Federal Reserve Bank of New York actively supported the practice: It soon became evident that in order to secure the [maximum amount of subscriptions assigned to the second Federal Reserve district] it would be necessary for a considerable volume of the subscriptions to be carried by credit. Accordingly, the phrase “borrow and buy” was used freely in all parts of the district, and the banks in New York City and many other places cooperated generously by offering to make loans on the bonds at the coupon rate of interest.27

The Third Liberty Loan Campaign The tone and conduct of the Third Liberty Loan campaign was not materially different from the second. The opening day of the campaign, April 6, was declared a holiday in almost every state. St. Clair observes that “special days were set aside in behalf of the loan . . .. On the night of April 12 Liberty Loan rallies were held in more than a hundred thousand school-houses throughout the country. On April 21, 114,000 preachers delivered Liberty Loan sermons.”28 Spectacles continued to attract public attention: “[W]ar exhibit trains were used for the first time. Exhibits of war materiel, including cannon and other implements necessary to trench warfare, were . . . routed on six special trains through the St. Louis, Dallas, and Atlanta [Federal Reserve] districts. Speakers of national reputation and soldiers who had seen active service in France accompanied the trains. The exhibits proved an enormous success, carrying the war into the homes and the hearts of the people as it never had been brought to them before.”29 The Fourth Liberty Loan Campaign By the time McAdoo launched the Fourth Liberty Loan campaign on September 28, 1918, U.S. participation in combat operations in western Europe 25. 26. 27. 28. 29.

“$4,617,532,300 Total of Second Liberty Loan,” New York Times, November 8, 1917, p. 1. 1917 Federal Reserve Bank of New York Annual Report, pp. 36 and 41. 1917 Federal Reserve Bank of New York Annual Report, p. 41. St. Clair (1919, p. 61). St. Clair (1919, p. 61).

99

Marketing the Liberty Loans

had increased sharply. There was a greater emphasis during the campaign on supporting the troops and less interest in entertaining spectacles.30 In a kickoff speech to campaign workers in the Second Federal Reserve District, McAdoo observed that: There is hard work before you before success can be secured, but it is glorious work. Every dollar you provide brings us nearer to victory and to a righteous and enduring peace. Success means life to thousands of our brave men, who, if we fail, must pay the price of that failure with their blood and find their final resting places far from their native soil in foreign lands across the seas. I was asked not long ago what argument could be made for the Fourth Liberty Loan. I replied immediately, “Pershing and our heroes in France!” No other argument is necessary. The American people know and feel the need of those men and are prepared to go the limit to supply that need.31

The Victory Liberty Loan Campaign By the spring of 1919 the war had been over for five months and Americans wanted to get on with their lives. It fell to McAdoo’s successor, Carter Glass, to preside over the last, and arguably most difficult, campaign. Shortly after the beginning of the year, a prominent government securities dealer had suggested that the last Liberty Loan should be floated “as a cold business proposition.”32 Glass responded heatedly: It is my deliberate judgment that it would be a profanation of the spirit that has already been exhibited to say that we should consider the matter cold blood and on a strict business basis. There are yet two millions of American boys on the soil of France and of Germany who must be brought home.33

He reiterated his position a month later: The requirements of the Treasury are imperative and cannot be financed without . . . a campaign. . . . [I]t is absolutely essential that the widest possible measure of distribution be realized. This will necessitate a campaign of the same character as that conducted in the past through the existing Liberty Loan organizations. They have pledged again their united support in order that the Victory Liberty Loan may be distributed as widely as possible among the American people.34

Despite Glass’s desire to continue the momentum of the wartime financings, the Victory Liberty Loan was floated in a short and unemotional campaign. 30. St. Clair (1919, p. 66) (“On the opening day of the loan it was noticed that a great change had taken place in the attitude of the entire country toward popular financial campaigns. The flamboyant holiday spirit which had been so noticeable in some of the previous loans, if indeed, frequently it had not been the dominant note, had almost entirely disappeared.”) 31. “New Liberty Loan of $6,000,000,000 Offered at 4 1-4%,” New York Times, September 25, 1918, p. 1. 32. “New Liberty Loan Must Sell on a Business Basis,” Wall Street Journal, January 13, 1919, p. 8. 33. “Glass Bases Fifth Loan on Patriotism,” New York Times, January 14, 1919, p. 4. 34. “Short-Term Notes for Victory Loan Alone Find Favor,” New York Times, February 19, 1919, p. 1.

100

Chapter 6

Spectacles were gone entirely.35 The Federal Reserve Bank of New York observed that, “The central thought was presented in a poster . . . which showed a mechanic putting his hand into his pocket and saying, ‘Sure! We’ll finish the job.’ ”36 Were the Liberty Loan Campaigns Coercive? War on the scale of World War I necessarily involves an immense transfer of control over goods and services to the national government. Frank Anderson, an economist at Columbia University, observed that the methods of war finance “may be . . . classified as voluntary or coercive, according to whether those who in the last analysis surrender the goods do so freely consenting or not”37 Most commentators considered debt financing to be voluntary and taxes to be coercive. The New York Times noted that “In respect to taxes, the citizen has no choice. He must pay what the government decrees he should pay, while when a loan is floated the citizen is appealed to for subscriptions and he may invest or refrain from investing his money in Government bonds as he wishes.”38 But Anderson was not so sure, noting that borrowing could be coercive “if carried beyond certain limits.”39 The Liberty Loan campaigns brought intense pressure on financial institutions to generate subscriptions. The Federal Reserve Bank of New York reported that, during the first campaign, Governor Strong “wrote a letter to the banks in [the second Federal Reserve] district outside of New York City, asking their officers, in consultation with other bankers in the same place, to undertake the formation of local committees to secure subscriptions for the bonds. Each locality was advised of the amount of subscriptions which it was expected to obtain . . ..”40 In the second campaign, “quotas were assigned to each locality, and local publicity given to the progress in filling these quotas . . ..”41 Also during the second campaign, Governor Strong wrote to New York City banks that: 35. St. Clair (1919, p. 89) (“It was obvious from the outset that in most districts there was no longer great interest in the so-called ‘circus-stunts.’”) 36. 1919 Federal Reserve Bank of New York Annual Report, p. 46. 37. Anderson (1917, p. 861). 38. “Federal Reserve System’s Place in Government Financing,” New York Times, January 6, 1918, p. 92. See also Seligman (January 1918, p. 76) (“taxes involve a compulsory, and loans only a voluntary, appeal to saving”). 39. Anderson (1917, p. 861). 40. 1917 Federal Reserve Bank of New York Annual Report, p. 34. Emphasis added. 41. 1917 Federal Reserve Bank of New York, p. 40. Emphasis added.

101

Marketing the Liberty Loans

Subscriptions to the Second Liberty Loan have not yet been received in sufficient volume to demonstrate that the issue will be the unqualified success which it must be both for the moral effect and the financial needs of our Government. The amount of bonds which the [Liberty Loan] Committee has felt must be subscribed in this district, in order to make the loan an unqualified success, has been prorated over the banking communities in this district; and the amount allotted to New York City has in turn been prorated among all the banking institutions of this City upon an impartial and equitable basis . . .. Your share in the amount thus to be raised by New York City is $_____. It is the belief of the Committee that the seriousness of the present situation has not yet been thoroughly appreciated by banking institutions; and it is felt that it is vitally essential that the banks and trust companies in their city should consider it their unquestioned duty to produce oversubscriptions, including their own and their customers, to the amount which has been allotted to them.42

Secretary McAdoo took an even more aggressive tone with respect to individual subscribers: “If the thousands of patriotic citizens who want to win this war . . . don’t subscribe for this loan and lend their money to the Government, they are helping the enemy.” He was further quoted as saying that, “Every person who refused to subscribe or who takes the attitude of ‘let the other fellow do it,’ is a friend of Germany and I would like nothing better than to tell it to him to his face . . .. A man who can’t lend his government $1.25 per week at the rate of 4% interest is not entitled to be a citizen.”43 As the war progressed, reports of coercion began to surface. During a debate in the U.S. Senate, Robert La Follette, Republican from Wisconsin, referred to “numerous instances of alleged coercion [in the sale of Liberty bonds] by local Councils of Defense,” and Thomas Gore, Democrat from Oklahoma, declared that “soldiers and sailors had been ‘compelled’ to subscribe for Liberty bonds.”44 A postwar report of the American Economics Association concluded that “The conditions under which loan ‘drives’ have been made in some cases brought a moral pressure to bear upon prospective buyers to purchase . . . which has amounted almost to compulsion.”45 Other researchers 42. Letter dated October 18, 1917, 1917 Federal Reserve Bank of New York circulars volume in Federal Reserve Bank of New York Research Library. Emphasis added. 43. “M’Adoo Warns Nation against Failure of Loan,” New York Times, October 13, 1917, p. 1. Emphasis added. 44. “Filibuster Fails; Loan Bill Passed,” New York Times, March 3, 1919, p. 1. 45. Bogart et al. (1919, pp. 80–81). Almost ninety years after the end of the war, the New York Times, in an article on sedition and free speech in Montana during the war, reported that “Local groups called third-degree committees were formed to ferret out people not supportive of the war, especially those who did not buy Liberty Bonds.” Clemens Work of the Montana School of Journalism was quoted as saying that the committees “leaned on people to ante up and buy bonds, and if they didn’t, they were disloyal and considered pro-German.” “Silence Broken, Pardons Granted 88 Years After Crimes of Sedition,” New York Times, May 3, 2006, p. 1.

102

Chapter 6

concluded similarly that the Liberty Loan campaigns were either actually or very nearly coercive.46 Did the Treasury Sell Liberty Bonds at Below-Market Interest Rates? From the narrow perspective of Treasury finance, the principle question raised by the Liberty Loan campaigns is whether the solicitations and coercive pressures of the campaigns led to primary market sales at below-market interest rates.47 McAdoo certainly implied as much when he later claimed that the campaigns “capitalized” patriotism.48 The postwar report of the American Economics Association concluded that there was “no doubt” that the Treasury had secured below-market rates by “an appeal to patriotism” and “some measure of compulsion.”49 Historian David Kennedy is more emphatic: The super-heated patriotism . . . in the Liberty Loan campaigns was deliberately cultivated by the Wilson administration. It was a calculated consequence of the administration’s reluctance to make the true material costs of the war visible and to lay them explicitly on the people. McAdoo’s unwillingness to rely either on heavy taxation or on market-rate borrowing led directly to the effort to mobilize emotions instead, to substitute aroused patriotic fervor for the real economic price he would not ask the country to pay.50

Sung Won Kang and Hugh Rockoff of the Department of Economics at Rutgers University come to a different conclusion, that “McAdoo did not expect that investors would buy bonds that yielded far less than comparable assets. His faith in his ability to capitalize patriotism was qualified from the start by a realization that the market would be interested first and foremost in financial returns.”51 Whether the Treasury sold Liberty bonds at below-market rates can be addressed by examining the reactions of market participants at the time of and following the offerings: Did investors complain that they were being taken 46. See, Stevenson (2004, p. 302) (“[T]o sell the bonds . . . [McAdoo] needed to mobilize public patriotism, and he did so through saturation publicity. The authorities hoped to rely on patriotic ‘self-mobilization,’ which the American public did eventually deliver, albeit more slowly than in Europe. But to achieve it the government used a surprising degree of coercion and manipulation . . ..”). Rockoff (2004) (Liberty Loan campaigns “undoubtedly created enormous social pressures to buy bonds”) and Kang and Rockoff (2006) (the campaigns “attempted to create strong social pressures to buy bonds”) come to more moderate conclusions. 47. Rates on the Victory Liberty notes are less interesting because of the notes’ relatively short maturity. The evidence suggests that Treasury Secretary Glass put commercially attractive rates on the notes. See, for example, “Bankers Highly Pleased with Victory Loan Terms,” Wall Street Journal, April 15, 1919, p. 1. 48. McAdoo (1931, pp. 378–79). 49. Bogart et al. (1919, pp. 80 and 124). 50. Kennedy (2004, p. 106). Emphasis added. 51. Kang and Rockoff (2006).

103

Marketing the Liberty Loans

advantage of, and did they value at less than par, in secondary market trading shortly after an offering, bonds that the Treasury sold at par? Reactions of Market Participants to Interest Rates on Liberty Bonds There is no evidence that investors thought they were being asked to buy bonds at below-market rates in the first two Liberty bond offerings. The third offering was a different matter. Early in 1918, a consensus formed among market participants that the third Liberty bond offering should carry a 4½ percent coupon. The Wall Street Journal speculated in January that “the next Liberty loan issue will probably . . . be at a higher rate that the previous issue, probably 4½%.”52 In late February the Journal reported that “bankers who have been closely identified with the floatation of the 3½% and 4% Liberty Loans look for a 4½% rate on the next issue.”53 The Treasury itself subsequently acknowledged that “It was the general banking opinion that the rate should be 4½ per cent . . ..”54 McAdoo, however, had come to view 4¼ percent as a “reasonable” and “sufficient” rate of interest for the third series of Liberty bonds.55 In late March he testified to that effect before the House Ways and Means Committee: [W]hen we started out [with the First and Second Liberties] we embarked upon unknown seas . . . because we had never undertaken such bond sales in this country before . . . I think we have reached the time now when . . . we ought to make a stand on a reasonable rate for Government bonds, and see if we can not educate the people to finance the Government, because it is to their interest to do it, upon that reasonable basis.56

McAdoo’s advocacy of a 4¼ percent coupon was an unwelcome surprise to financial market participants.57 The Wall Street Journal reported that “Some bankers are of the opinion that making the Loan only 4¼%, reflected a cheeseparing policy on the part of the Government and amounted to profiteering at the expense of patriotism, involving a sacrifice on the part of both bankers and the investing public. Some bankers thought that this was not the time to quibble over a ¼ of 1% interest on bonds.”58 52. “Expect Government to Raise Loan Rate to 4½%,” Wall Street Journal, January 15, 1918, p. 10. 53. “Bankers Look for 4½% Rate for Liberty Loan,” Wall Street Journal, February 20, 1918, p. 10. 54. 1918 Treasury Annual Report, p. 5. 55. Committee on Ways and Means (March 1918, pp. 12 and 25). 56. Committee on Ways and Means (March 1918, p. 25). 57. See, “New Liberty Loan to be 3 Billions; Interest Rate 4 1-4,” New York Times, March 26, 1918, p. 1 (“The interest rate, 4¼ per cent., is a surprise, 4½ per cent having been urged by many of the Secretary’s advisers in banking circles.”) and “Believe $3,000,000,000 of Bonds at 4¼% Will Go,” Wall Street Journal, March 27, 1918, p. 7 (“Nearly everyone was at first surprised at the smallness of . . . the rate of interest, from what the market had been led to expect.”) 58. “Believe $3,000,000,000 of Bonds at 4¼% Will Go,” Wall Street Journal, March 27, 1918, p. 7.

104

Chapter 6

When the Treasury came to market in the fall of 1918 to sell the Fourth Liberties it was clear that 4¼ percent was a below-market rate. In a campaign kickoff address, Benjamin Strong took note of the fact that the outstanding Third Liberties were trading below par; but he nevertheless argued that “with over a million and a half of our American boys in the fighting line in France, whose victory depends upon the success of these loans, the American people will not subject their patriotism; their resolution to support that army, to be measured by a rate of interest or by a premium or discount on the bonds of their Government.”59 Secondary Market Prices of Liberty Bonds Another way to assess whether the Treasury sold Liberty bonds at belowmarket rates is to examine whether the bonds fell to a discount from the original offering price of par in secondary market trading following the close of an offering. The postwar report of the American Economics Association summarized the basis for examining secondary market prices: “Patriotism might induce large subscriptions to bonds bearing an unduly low rate of interest, but would scarcely be able to support the price when the process of redistributing on the exchange took place.”60 (It important to bear in mind that we are here examining only whether the Treasury was able to sell Liberty bonds at a premium relative to secondary market prices. We are not asking whether other forces, including Federal Reserve monetary policies (discussed in chapter 9), might have bolstered the prices of Treasury securities in both the primary and secondary markets.61) Figure 6.1 shows secondary market prices of the First Liberty bonds after the bonds began trading on the New York Stock Exchange on June 15, 1917. The data give no reason to believe that the secondary market price of the bonds immediately after the close of the offering was materially different from par. Figure 6.2 shows secondary market prices of the First and Second Liberties during the two weeks prior to the end of the Second Liberty Loan campaign on Saturday, October 27, 1917, and continuing for four weeks after the Thursday, November 15, dated date of the Second Liberties.62 The data show that 59. “New Liberty Loan of $6,000,000,000 Offered at 4 1-4%,” New York Times, September 25, 1918, p. 1. 60. Bogart et al. (1919, p. 81). 61. Kang and Rockoff (2006) examine whether secondary market yields on Liberty bonds rose, relative to secondary market yields on other debt securities, following the end of hostilities. They find that any increase was small at best and conclude that wartime secondary market yields on Treasury securities were not materially depressed. 62. The New York Stock Exchange did not allow trading in Second Liberties on a when-issued basis (“To Trade in Bonds Monday,” New York Times, October 27, 1917, p. 2). There was a modest amount of trading between October 27 and November 17 in Second Liberties that had been bought from the Treasury for cash in prompt delivery transactions (“Morgan Subscription to Loan is $50,000,000,” New York Times, October 29, 1917, p. 7, and “Loan Subscriptions are Still Pouring

Marketing the Liberty Loans

Percent of principal

101

100

Last day of campaign and dated date, Jun 15

99

98 Jun 10

Jun 17

Jun 24

Jul 1

Jul 8

Jul 15

Figure 6.1 Prices of First Liberty bonds, June 15, 1917, to July 14, 1917

101

Dated date, Nov 15 100 Percent of principal

105

99 Last day of campaign, Oct 27

98

97 Oct 14

Oct 21

Oct 28

Nov 4

Nov 11

First Liberties

Nov 18

Nov 25

Dec 2

Dec 9

Dec 16

Second Liberties

Figure 6.2 Liberty bond prices, October 15, 1917, to December 15, 1917

In,” New York Times, October 30, 1917, p. 16), but significant trading did not develop until after Saturday, November 17, when the Treasury began delivering Second Liberties on fully paid subscriptions (“$678,000,000 Paid on Liberty Bonds,” New York Times, November 16, 1917, p. 3, “Broader Market for Liberty Bonds,” New York Times, November 16, 1917, p. 14, and “New Liberty Bonds Slump on Delivery,” New York Times, November 20, 1917, p. 15).

Chapter 6

the secondary market price of the Second Liberties was about 98 in late November and suggest that in selling the bonds at par in October the Treasury extracted a 2 point premium. However, the figure also shows that First Liberties fell from about 100 at the end of October to about 99 in late November. A parallel change in the value of the Second Liberties would put the market value of the Second Liberties at about 99 in late October. In other words, an investor would have been no worse off buying Second Liberties at 99 at the end of October than if he had bought First Liberties in the secondary market. This suggests that the Treasury extracted only a 1 point premium in the sale of the Second Liberties. Figure 6.3 shows secondary market prices of the Second and Third Liberty bonds during the two weeks prior to the end of the Third Liberty Loan campaign on Saturday, May 4, 1918, and continuing for four weeks after the Thursday, May 9, dated date of the Third Liberties. The data show that the secondary market price of the Third Liberties was about 97½ in mid and late May and that the Second Liberties fell from about 96½ at the end of April to about 94½ in late May. A parallel two-point change in the value of the Third Liberties would put the market value of the Third Liberties at about 99½ at the end of the third campaign. Figure 6.4 shows secondary market prices of the Third and Fourth Liberties during the two weeks prior to the end of the Fourth Liberty Loan campaign on Saturday, October 19, 1918, and continuing for four weeks after the 100

98 Percent of principal

106

96

94

92 Apr 21

Last day of campaign, May 4

Apr 28

Dated date, May 9

May 5

May 12

May 19

Second Liberties

Figure 6.3 Liberty bond prices, April 22, 1918, to June 8, 1918

May 26

Third Liberties

Jun 2

Jun 9

107

Marketing the Liberty Loans

101

Percent of principal

100

99

Last day of campaign, Oct 19

Dated date, Oct 24

98

97

96

95 Oct 6

Oct 13

Oct 20

Oct 27

Third Liberties

Nov 3

Nov 10

Nov 17

Nov 24

Fourth Liberties

Figure 6.4 Liberty bond prices, October 7, 1918, to November 23, 1918

Thursday, October 24, dated date of the Fourth Liberties. The data show that the secondary market price of the Fourth Liberties was about 98 in November. Since the price of the outstanding Third Liberty bonds did not fall between mid October and early November—in fact it rose almost a point—it seems reasonable to conclude that the Treasury was able to issue the Fourth Liberties at least 2 points, and possibly as much as 3 points, above the market-clearing price of the bonds. Conclusions The foregoing evidence is consistent with the proposition that Treasury officials offered the First Liberties at a market rate of interest and that investors did not think otherwise, either at the time the offering was announced or shortly after the end of the offering. The reaction of market participants to the announcement of the terms of the Second Liberties also does not provide any evidence that investors thought McAdoo was seeking to take advantage of their patriotism. However, secondary market prices suggest that the Second Liberties were in fact sold at what turned out to be a below-market interest rate. Conversely, investors complained vociferously about the rate on the Third Liberties (possibly in reaction to the unexpectedly weak postoffering market for the Second Liberties), but the ex post evidence suggests that those bonds were priced at no more than a modest half-point premium to a market-clearing level.

108

Chapter 6

Only the last series of Liberty bonds provides compelling evidence that Treasury officials knowingly “capitalized” the patriotism of American investors by selling bonds at a below-market rate of interest. Summary To finance America’s participation in World War I, Treasury Secretary McAdoo proceeded along two tracks: designing Liberty bonds that would be attractive to investors and aggressively marketing the bonds. Unlike the sale of Spanish–American War bonds in 1898, there is no indication that the Treasury sold Liberty bonds at below-market prices. The first three offerings were, at best, no more than modestly attractive as investments; the Fourth Liberty Loan was priced visibly and significantly above the market. But McAdoo not only sold the bonds at market prices or better, he was also able to generate oversubscriptions that publicly demonstrated the breadth of American support for the war. He understood that successful prosecution of a war requires Americans up and down the economic ladder to buy into the effort. The Liberty Loan campaigns provided one means of generating the emotional commitment that Wilson and McAdoo wanted.63 The downside of going beyond marketplace economics was that people who appeared to be shirking and not doing “their part” were liable to be singled out for harsh criticism as voluntarism slipped over to coercion. 63. As had been the case with the Spanish–American War bonds, the Liberty Loan campaigns stimulated as well as expressed popular support for the war. Chandler (1958, p. 119) points out that the campaigns “were not merely efforts to raise money; they were also campaigns to ‘sell the war’. . ..”

7

Treasury Cash Management: Certificates of Indebtedness

In the absence of any off-setting disbursements, wartime bond sales and tax collections would have produced large, episodic flows of funds into Treasury accounts at Federal Reserve Banks, draining the supply of reserves available to the banking system and putting upward pressure on short-term interest rates. The contraction in reserves would not have been permanent—military expenditures would quickly return the funds to the banking system—but would have been nonetheless disruptive in view of the magnitude of the flows.1 Treasury officials adopted two strategies to dampen fluctuations in government balances at Federal Reserve Banks: (1) anticipatory sales of short-term certificates of indebtedness, and (2) redeposits of tax collections and the proceeds of securities sales into Treasury accounts at commercial banks. This chapter examines the first strategy.2 The following chapter describes War Loan Deposit Accounts. Certificates of indebtedness were short-term securities, similar to Treasury bills except that they were interest-bearing, rather than discount, instruments. (That is, a certificate paid principal plus interest at maturity, rather than simply its face amount.3) All of the wartime certificate offerings were at par (plus accrued interest if a certificate was purchased after its dated date).4 The First Liberty Bond Act authorized the Secretary of the Treasury to issue certificates of indebtedness. Unlike the authorization for bonds, there was no limit on the amount of certificates the Treasury could issue, but the act stipulated that the Treasury could not have more than $2 billion of certificates outstanding at any point in time. The difference meant that the Treasury could refinance maturing certificates with new certificates without depleting its statutory authority. The distinction between limiting certificates outstanding and limiting bond issuance reflected the view of bonds as instruments for financing specific projects—building a canal or waging a war—and short-term debt as greasing the day-to-day operations of government—something likely to expand and contract with the ebb and flow of receipts and expenditures. 1. Asynchronous cash flows were not a novel problem for the Treasury. Taus (1943, pp. 78–79) observes that, during the 1880s, “custom receipts . . . flow into [Treasury] vaults in a steady stream, while interest, amortization and salary payments are paid periodically. . . . [To] meet the interest on bonds money may have to be withdrawn from circulation for three months; and at the end of the quarter it will be released and find its way back into the banks. This irregularity of Treasury disbursements and receipts seriously interferes with the stability of the money market.” 2. Hollander (1919) is the leading study of the Treasury’s use of certificates of indebtedness. See also Hollander (1918a, b). 3. Interest was prorated over a 365-day year. A $1 million certificate paying interest at the rate of 3½ percent per annum that matured 100 days after its dated date would pay $1,009,589.04 at maturity: the $1,000,000 principal amount of the certificate, plus $9,589.04 interest, where 9,589.04 = (100/365) × 3½% of 1,000,000. 4. Accrued interest was also prorated over a 365-day year. A $1 million certificate paying interest at the rate of 3½ percent per annum would have $863.01 of accrued interest 9 days after its dated date, where 863.01 = (9/365) × 3½% of $1,000,000.

110

Chapter 7

Even while it guarded its prerogatives to approve project financings, Congress thought it best to leave decisions regarding short-term debt issuance to the executive branch and to limit only the total amount of short-term debt that could be outstanding at any one time. In subsequent Liberty Bond acts, Congress expanded the Secretary’s authority in line with the growth in Treasury expenditures. The Second Liberty Bond Act raised the ceiling on outstanding certificates to $4 billion, the Third Liberty Bond Act raised it to $8 billion, and the Victory Liberty Loan Act raised it to $10 billion. Certificates of Indebtedness Issued in Anticipation of Bond Sales On Friday, April 20, 1917, Treasury officials announced the first wartime public offering of certificates: a $200 million issue bearing a 3 percent interest rate and maturing June 30, 1917, to be issued promptly upon the signing of the First Liberty Bond Act.5 President Wilson signed that act into law on Tuesday, April 24. The Federal Reserve Banks, acting as fiscal agents for the Treasury, accepted orders for the new certificates the same day and the securities were delivered (against payment) to buyers the following day.6 Treasury officials intended to use certificates of indebtedness as a cash management tool. The basic idea was that banks would buy certificates with their own funds, hold them to maturity, and then use the principal and interest payments to fund depositor withdrawals made in the course of paying for Liberty bond subscriptions.7 The bond payments would not then lead to any unwieldy drain of funds from the banking system.8 Certificates would be 5. “Will Offer Banks $200,000,000 Issue,” New York Times, April 21, 1917, p. 3. The Treasury had earlier sold $50 million of three-month certificates directly to the Federal Reserve Banks pursuant to a provision in the Revenue Act of 1917 that authorized up to $300 million of certificates. 1917 Treasury Annual Report, p. 14. 6. “U.S. Certificates to be Paid for Today,” Wall Street Journal, April 25, 1917, p. 8. 7. When the first issue was announced the New York Times reported that the certificates could be “employed to buy bonds, principal and interest both being available in the purchase. The plan is devised to relieve the money market from any derangement . . ..” “Will Offer Banks $200,000,000 Issue,” New York Times, April 21, 1917, p. 3. See also the Treasury press release reported in the same article: “[I]n the financial operations in which the Government is about to engage [it will be the purpose of the Secretary of the Treasury] to adjust receipts and disbursements in such a way that as far as possible money paid in will be promptly returned to the market.” 8. See, “Secretary McAdoo to Sell Certificates,” Wall Street Journal, April 21, 1917, p. 5 (“Should the banks during the next few weeks absorb several hundred million of these certificates, . . . the banks will possess ready means with which to meet withdrawals made later by depositors in paying for bond subscriptions.”), and Treasury Circular no. 79, May 16, 1917, reprinted in 1917 Treasury Annual Report, p. 131 (“In order to avoid, even temporarily, a derangement of the money situation, the Secretary of the Treasury earnestly requests that all banks and trust companies which have or expect to have payments to make for themselves or others on account of subscriptions to the [First Liberty Loan], acquire, as and when offered from time to time, Treasury certificates of indebtedness . . . to as large an amount as practicable and at least equal to 50 per cent of the payments which they will have to make . . . on account of subscriptions, and that they utilize such certificates of indebtedness in making payment.”).

111

Treasury Cash Management: Certificates of Indebtedness

issued in small, readily digestible amounts from time to time before a bond sale. The proceeds of each offering would either be returned to the banking system in the normal course of the Treasury’s wartime expenditures or, if expenditures were expected to lag for a few days or weeks, be redeposited in War Loan Deposit Accounts. Following the close of the subscription books for the first Liberty Loan in mid-June 1917, Treasury Secretary McAdoo noted the significance of certificate sales: “This method [of anticipatory financing] proved eminently successful . . ., preparing the way for the loan and preventing to a large extent the accumulation of vast payments on one day.”9 He emphasized that certificates were a temporary, not a permanent, means of financing the war.10 Table 7.1 shows wartime certificate sales made in anticipation of subsequent bond sales. The Treasury pursued a variety of allotment policies in offering bond-anticipation certificates. In the initial series, I-A, it offered $200 million of certificates, received subscriptions for $268 million, and filled all of the subscriptions. Treasury officials subsequently concluded that filling all subscriptions was liable to weaken demand for later offerings and switched to a policy of selling, on a first-come, first-served basis, only as many certificates as were offered.11 With two exceptions, this remained the Treasury’s allotment policy until the offering of series III-D in March 1918.12 (The exceptions were two short-term certificates, series II-E and series II-F, in which Treasury filled all subscriptions, evidently to facilitate payments on the Second Liberty bond.) Beginning with series III-D, the Treasury reverted to its original policy of filling all subscriptions.13 (At about the same time—the spring of 1918—Treasury also began filling all subscriptions for Liberty bonds.) Contractual Innovations Facilitating the Use of Certificates as Cash Management Tools Following the close of the first Liberty Loan campaign, Treasury officials introduced two innovations that enhanced the utility of bond-anticipation certificates as cash management tools. 9. Committee on Ways and Means (1917, p. 17). 10. Committee on Ways and Means (1917, pp. 17 and 66) (testimony of Secretary McAdoo that “certificates are used merely as a temporary measure in advance of and [in] preparation for a large bond issue and in anticipation of taxes” and that “[Certificates] do not remain outstanding. They are only a temporary means of facilitating the permanent operations.”). 11. “Treasury Certificates More Popular in Interior,” Wall Street Journal, June 8, 1917, p. 8. 12. “Certificates of Indebtedness,” Wall Street Journal, May 3, 1917, p. 6, “Secretary Raises Rate to 3¼% on $200,000,000 Certificates,” Wall Street Journal, May 19, 1917, p. 8, and 1918 Treasury Annual Report, pp. 195–99 (offering circulars for series III-A, III-B, and III-C). 13. 1918 Treasury Annual Report, p. 200.

112

Chapter 7

Table 7.1 Certificates of indebtedness issued in anticipation of bond sales

Series

Announcement

Subscription books closed

Dated date

Issued in anticipation of First Liberty bonds I-A Apr 20, 1917 Apr 24, Apr 25, 1917 1917 I-B May 2, 1917 May 9, 1917 May 10, 1917 I-C May 18, 1917 May 23, May 25, 1917 1917 I-D Jun 1, 1917 Jun 6, 1917 Jun 8, 1918 Issued in anticipation of Second Liberty bonds II-A Jul 31, 1917 Aug 7, 1917 Aug 9, 1917 II-B Aug 18, 1917 Aug 25, Aug 28, 1917 1917 II-C Sep 4, 1917 Sep 11, Sep 17, 1917 1917 II-D Sep 24, 1917 Oct 2, 1917 Sep 26, 1917 II-E Oct 7, 1917 Oct 15, Oct 18, 1917 1917 II-F Oct 23, 1917 Oct 31, Oct 24, 1917 1917 Issued in anticipation of Third Liberty bonds III-A Jan 17, 1918 Jan 29, 1918 Jan 22, 1918 III-B Feb 6, 1918 Feb 15, Feb 8, 1918 1918 III-C Feb 21, 1918 Mar 5, 1918 Feb 27, 1918 III-D Mar 11, 1918 Mar 22, Mar 20, 1918 1918 III-E Mar 27, 1918 Apr 10, Apr 10, 1918 1918 III-F Apr 12, 1918 Apr 25, Apr 22, 1918 1918

Maturity

Jun 30, 1917 Jul 17, 1917 Jul 30, 1917 Jul 30, 1917 Nov 15, 1917 Nov 30, 1917 Dec 15, 1917 Dec 15, 1917 Nov 22, 1917 Dec 15, 1917 Apr 22, 1918 May 9, 1918 May 28, 1918 Jun 18, 1918 Jul 9, 1918 Jul 18, 1918

Term (days)

Coupon rate (%)

Amount issued ($millions of principal value)

66

3

268

68

3

200

66

3¼

200

52

3¼

200

98

3½

300

94

3½

250

89

3½

300

80

4

400

35

4

385

52

4

685

90

4

400

90

4

500

90

4½

500

90

4½

543

90

4½

551

87

4½

518

113

Treasury Cash Management: Certificates of Indebtedness

Table 7.1 (continued)

Series

Announcement

Subscription books closed

Dated date

Issued in anticipation of Fourth Liberty bonds IV-A Jun 18, 1918 Jul 2, 1918 Jun 25, 1918 IV-B Jul 3, 1918 Jul 16, 1918 Jul 9, 1918 IV-C Jul 19, 1918 Jul 30, 1918 Jul 23, 1918 IV-D Aug 1, 1918 Aug 13, Aug 6, 1918 1918 IV-E Aug 27, 1918 Sep 10, Sep 3, 1918 1918 IV-F Sep 12, 1918 Sep 24, Sep 17, 1918 1918 IV-G Sep 24, 1918 Oct 8, 1918 Oct 1, 1918 Issued in anticipation of Victory Liberty notes V-A Nov 27, 1918 Dec 10, Dec 5, 1918 1918 V-B Dec 12, 1918 Dec 26, Dec 19, 1918 1918 V-C Dec 28, 1918 Jan 7, 1919 Jan 2, 1919 V-D Jan 8, 1919 Jan 21, 1919 Jan 16, 1919 V-E Jan 24, 1919 Feb 6, 1919 Jan 30, 1919 V-F Feb 7, 1919 Feb 20, Feb 13, 1919 1919 V-G Feb 21, 1919 Mar 6, 1919 Feb 27, 1919 V-H Mar 7, 1919 Mar 20, Mar 13, 1919 1919 V-J Apr 7, 1919 Apr 17, Apr 10, 1919 1919 V-K Apr 22, 1919 May 8, 1919 May 1, 1919 Source: Treasury Annual Reports.

Maturity

Oct 24, 1918 Nov 7, 1918 Nov 21, 1918 Dec 5, 1918 Jan 2, 1919 Jan 16, 1919 Jan 30, 1919 May 6, 1919 May 20, 1919 Jun 3, 1919 Jun 17, 1919 Jul 1, 1919 Jul 15, 1919 Jul 29, 1919 Aug 12, 1919 Sep 9, 1919 Oct 7, 1919

Term (days)

Coupon rate (%)

Amount issued ($millions of principal value)

121

4½

840

121

4½

754

121

4½

585

121

4½

576

121

4½

639

121

4½

625

121

4½

641

152

4½

613

152

4½

572

152

4½

752

152

4½

600

152

4½

687

152

4½

621

152

4½

532

152

4½

542

149

4½

646

159

4½

591

114

Chapter 7

The first innovation was a call option that allowed the Secretary of the Treasury to call a certificate for early redemption on ten days’ notice.14 The Treasury exercised its right to call certificates for early redemption several times during the war. For example, on November 23, 1917, it called series II-C for redemption on December 6 and series II-D for redemption on December 11 because investors had made unexpectedly large installment payments for the Second Liberty bonds on November 15 and officials wanted to accelerate the return of the funds to the banking system.15 The two series had been scheduled to mature on December 15.16 The second innovation was a limited-use put option that allowed investors to tender unmatured certificates in payment for Liberty bond subscriptions. For example, series II-A provided that “The certificates of this series, whether or not called for redemption, will be accepted at par, with an adjustment of accrued interest, if tendered . . . as payment on the purchase price of any bonds issued and allotted hereafter and before the maturity of the certificates.”17 Subsequent Use of Certificates of Indebtedness for Debt Management Table 7.1 shows that all of the certificates issued in anticipation of the First Liberty bonds were redeemed before any certificates were issued in anticipation of the Second Liberties. In particular, series I-C and I-D matured on July 30, 1917, and the Treasury did not begin to issue series II-A until August 9. Similarly all of the certificates issued in anticipation of the Second Liberty bonds were redeemed before any certificates was issued in anticipation of the Third Liberties. In contrast, the Treasury issued certificates in anticipation of the Fourth Liberty bonds before all of the certificates issued in anticipation of the Third Liberty bonds had been redeemed. Series IV-A was first issued on June 25, 1918, and series IV-B was first issued on July 9, 1918, prior to or contemporaneous with the redemption of series III-E on July 9, 1918 and series III-F on July 18, 1918. As shown in figure 7.1, from then on through the fall of 1919 the Treasury was continuously indebted on outstanding bond-anticipation 14. For example, Series II-A certificates included a provision that “Upon ten days’ notice . . . the series of . . . certificates now offered may be redeemed as a whole, at par and accrued interest . . ..” 1918 Treasury Annual Report, p. 188. Two short-term bond-anticipation certificates, series II-E (35 days) and series II-F (52 days), did not carry call options. 15. “M’Adoo to Circulate Liberty Loan Money,” New York Times, November 24, 1917, p. 18, “Record Treasury Balance,” New York Times, November 25, 1917, p. RE11, and Hollander (1919, pp. 85–87). 16. Similarly, the Treasury called series IV-D (scheduled to mature on December 5, 1918) for redemption on November 21, 1918, series IV-E (scheduled to mature on January 2, 1919) for redemption on December 19, 1918, and series V-G (scheduled to mature on July 29, 1919) for redemption on July 1, 1919. “Call for Prepayment,” New York Times, November 9, 1918, p. 17, “Indebtedness Certificates Called,” Wall Street Journal, June 21, 1919, p. 5, and 1919 Treasury Annual Report, pp. 288–89. 17. 1918 Treasury Annual Report, p. 188.

115

Treasury Cash Management: Certificates of Indebtedness

7 6

Billions of dollars

5 4 3 2 1 0 Jan 1917

Jan 1918

Jan 1919

Jan 1920

Figure 7.1 Aggregate outstanding quantity of bond-anticipation certificates of indebtedness

certificates. This suggests that, beginning in mid-1918, the Treasury used certificates for debt management as well as for cash management. It should also be noted that, beginning with series IV-B in mid-1918, the Treasury adopted a policy of announcing a new series of bond-anticipation certificates immediately following the close of the subscription books for the previously announced series. For example, series IV-A closed on July 2, 1918, and the Treasury announced series IV-B on July 3. Thus, after mid-June 1918, the Treasury offered bond-anticipation certificates more or less continuously and filled all subscriptions in full. This is further evidence that the Treasury began to use certificates for debt management in mid-1918. Certificates of Indebtedness Issued in Anticipation of Tax Receipts Taxes on income earned in 1917 and taxed at the sharply higher rates specified by the War Revenue Act of 1917 were due to be paid on June 15, 1918. A ten-day grace period pushed the effective due date back to June 25.18 In anticipation of the immense payments to be received, the Treasury announced on November 20, 1917, a new series of tax-anticipation certificates, bearing interest from November 30, 1917, and maturing on June 25, 1918.19 18. “Expect Rush of Tax Payers on June 25,” Wall Street Journal, June 7, 1918, p. 8. 19. “New Certificate Issue,” New York Times, November 21, 1917, p. 18, and 1918 Treasury Annual Report, pp. 193–94. The certificates were authorized by section 1010 of the War Revenue Act of 1917.

116

Chapter 7

A letter from the Federal Reserve Bank of New York to banks in the second Federal Reserve district made clear that the certificates were being offered as a cash management tool: The amount of Federal income tax and excess profits taxes which will be due and payable June 15, 1918, has been estimated to be in excess of two billion dollars. Payment of such a huge sum in currency or by bank checks within a short period would occasion an unprecedented strain on the banks and money markets of the country; but if the Government from time to time borrows in anticipation of the receipt of those revenues and gradually disburses the proceeds of such borrowings between now and June 15th, the strain may be avoided. The purpose of this issue of certificates is, therefore, to produce a medium for the gradual payment of taxes.20

The new tax-anticipation certificates carried the same sort of limited-use put option that had been introduced for the bond-anticipation certificates, the Treasury announcing that “collectors of internal revenue will receive certificates of this issue at par and accrued interest . . . in payment of income and excess-profits taxes . . ..”21 The Treasury initially indicated that it would sell only “a limited amount” of the first series of tax-anticipation certificates,22 but the offering proved immensely popular. Investor response “literally swept . . . Treasury officials . . . off their feet.”23 The Treasury had subscriptions for almost $700 million of certificates when it closed the books on the offering on November 30.24 Beginning on January 2, 1918, the Treasury began to offer tax-anticipation certificates maturing on June 25, 1918, on a continuous basis. To relieve investors “of the necessity of paying so much on account of accrued interest,”25 it opened a new series of certificates in the middle of each month from February to May. By the middle of June 1918, the Treasury had issued $1.6 billion of tax-anticipation certificates (table 7.2). The Treasury pursued a similar program of issuing certificates in anticipation of taxes on income earned in 1918. However, as shown in table 7.2, those certificates did not mature on a common date because the Revenue Act of 1918 provided that taxes on income earned in 1918 would be due in four quarterly installments, on March 15, June 15, September 15, and December 15, 1919.26 20. Letter dated November 24, 1917, from Federal Reserve Bank of New York to banks in the second Federal Reserve district, in bound Bank circulars collection, Research Library, Federal Reserve Bank of New York. 21. 1918 Treasury Annual Report, p. 193. 22. “New Certificate Issue,” New York Times, November 21, 1917, p. 18. 23. “U.S. 4% Certificates Popular,” Wall Street Journal, December 1, 1917, p. 6. 24. “$690,000,000 Total In Certificate Issue,” New York Times, December 1, 1917, p. 1. 25. 1918 Treasury Annual Report, p. 198. 26. One of the drawbacks to the certificates issued in anticipation of the 1918 tax payments was that the Treasury had to issue six different securities. The securities shared a common maturity date and coupon rate, but were distinguished by their dated dates. In mid-August 1918 the Treasury announced a clever innovation for certificates issued in anticipation of 1919 tax payments: a certificate (series T-G) dated August 20, 1918, and maturing July 15, 1919, that paid a bi-monthly coupon (on November 15, 1918, and January 15, March 15, May 15, and July 15, 1919). Until

117

Treasury Cash Management: Certificates of Indebtedness

Table 7.2 Certificates of indebtedness issued in anticipation of tax receipts

Series

Dated date

Maturity

Term (days)

Coupon rate (%)

Amount issued ($million of principal value)

Issued in anticipation of 1918 T-A Nov 30, 1917 T-B Jan 2, 1918 T-C Feb 15, 1918 T-D Mar 15, 1918 T-E Apr 15, 1918 T-F May 15, 1918

tax receipts Jun 25, 1918 Jun 25, 1918 Jun 25, 1918 Jun 25, 1918 Jun 25, 1918 Jun 25, 1918

207 174 130 102 71 41

4 4 4 4 4 4

692 492 74 111 72 184 1,625

Issued in anticipation of 1919 T-G Aug 20, 1918 T1 Nov 7, 1918 T2 Jan 16, 1919 T3 Mar 15, 1919 T4 Jun 3, 1919 T5 Jun 3, 1919 T6 Jul 1, 1919 T7 Jul 1, 1919

tax receipts Jul 15, 1919 Mar 15, 1919 Jun 17, 1919 Jun 16, 1919 Sep 15, 1919 Dec 15, 1919 Sep 15, 1919 Dec 15, 1919

329 128 152 93 104 195 76 167

4 4½ 4½ 4½ 4½ 4½ 4½ 4½

158 794 392 408 526 239 326 511 3,354

Source: Treasury Annual Reports.

Interest Rates Certificates of indebtedness were sold primarily to financial institutions and were not the object of intense marketing campaigns, as was the case with the Liberty Loans.27 Interest rates on certificates were largely determined by November 15, 1918, the Treasury offered the certificate at a price of par plus accrued interest from the August 20, 1918, dated date. Thereafter it offered the certificate at a price of par plus accrued interest from the last coupon payment date. Treasury Circular no. 120, August 16, 1918, reprinted in 1918 Treasury Annual Report, p. 215, and “U.S. Tax Certificates Have Five Coupons Attached,” Wall Street Journal, October 12, 1918, p. 10. This structure allowed the Treasury to issue a single certificate for payments due on or before July 15. Unfortunately, the 4 percent rate of interest on the certificates proved unattractive to investors and the Treasury sold only $158 million of the certificates before closing the subscription books prematurely in early November 1918. “Tax Certificates at 4½% Meet Investment Conditions,” Wall Street Journal, November 7, 1918, p. 10, and 1918 Treasury Annual Report, p. 216. The Treasury then reverted to its earlier practice of selling tax-anticipation certificates with a single coupon payable at maturity. 27. “Look for Two More War Bond Issues,” New York Times, February 16, 1919, p. 21 (“certificates of indebtedness have been [sold] almost exclusively to banks”). See also Hollander (1918a, p. 847) (“of the certificate issues prior to January 1, 1918, the banks took for their own account slightly less than seven-eights, and . . . of the issues since emitted the banks have taken something more than three-fifths”) and Hollander (1919, p. 59) (“The certificates have been taken and held in the main by the financial institutions of the country—national banks, state banks and trust companies.”).

118

Chapter 7

market conditions, McAdoo himself acknowledging that “we have to meet the conditions of the market.”28 In the first public offering of certificates in April 1917, McAdoo initially announced that the new securities would carry a 2½ percent interest rate.29 Contemporaneous rates on short-term commercial paper were 3½ to 4 percent and the Treasury’s offer quickly encountered resistance. One banker grumped that “We took our share, but we didn’t like it.”30 Others saw McAdoo as taking advantage of their patriotism.31 Facing a bankers’ rebellion not even a month after the U.S. entry into the war, McAdoo backpedaled and raised the offer rate on the first series of certificates to 3 percent.32 There were numerous explanations for the rate increase—The New York Times variously reporting that, The Secretary was assured that reasonable immediate requirements could be met by a sale of certificates bearing as low a rate as 2½ per cent. interest, but that there would be no doubt about the sale of the largest amount of such debt certificates and that a wide market for the same could be created if they were offered at 3 per cent. interest.33

and that, [W]hile the responses from the larger cities had been entirely satisfactory, the Secretary of the Treasury had decided to place the certificates of indebtedness at 3 per cent. interest instead of 2½ per cent. in order to create a wider market for them, and to afford opportunities to country banks to subscribe to the issue.34

But it was clear that, in mid-April 1917, the certificate market was not going to clear at 2½ percent. 28. Committee on Ways and Means (1917, p. 44). McAdoo was particularly reluctant to have Congress specify a maximum rate of interest on certificates of indebtedness: “I feel that with the tremendous financial operations in which the Government must engage over the period of the next year it is a very unsafe thing for the Secretary of the Treasury to be unable to meet the conditions of the money market on these temporary certificates. . . . [I]t would not do to have the hands of the Treasury so tied that we could not finance the actual necessities of the Government.” Committee on Ways and Means (1917, p. 18). 29. “M’Adoo Will Offer $150,000,000 Issue,” New York Times, April 20, 1917, p. 1, and “Federal Reserve Bank Offers Treasury Bills,” Wall Street Journal, April 20, 1917, p. 5. 30. “Rate on Treasury Bills Raised from 2½ to 3%,” Wall Street Journal, April 21, 1917, p. 8. See also “Washington Blunder Mexicanizes Banks,” Wall Street Journal, April 24, 1917, p. 1 (reporting “protests and refusals”). 31. “Rate on Treasury Bills Raised from 2½ to 3%,” Wall Street Journal, April 21, 1917, p. 8. 32. “Will Offer Banks $200,000,000 Issue,” New York Times, April 21, 1917, p. 3, and “Rate on Treasury Bills Raised from 2½ to 3%,” Wall Street Journal, April 21, 1917, p. 8. 33. “Will Offer Banks $200,000,000 Issue,” New York Times, April 21, 1917, p. 3. 34. “Will Offer Banks $200,000,000 Issue,” New York Times, April 21, 1917, p. 3. See also “Banks Rush to Buy Treasury Bills,” New York Times, April 22, 1917, p. 22 (“The raising of the interest rate from 2½ to 3 per cent. had the effect of stimulating subscriptions from country banks. The large New York City institutions already had filled their subscriptions when Secretary McAdoo decided to change the rate.”).

119

Treasury Cash Management: Certificates of Indebtedness

The second series of certificates, series I-B, was also offered at 3 percent but investor response was tepid35 and Treasury officials moved the rate to 3¼ percent on series I-C.36 The rate was increased a third time, to 3½ percent on series II-A, following a spike in short-term interest rates in mid-July 1917.37 The first seven series of certificates were issued under the authority of the First Liberty Bond Act and benefited from the broad tax exemptions of that act. In late September 1917, beginning with series II-D, the Treasury began issuing certificates under the authority of the Second Liberty Bond Act. In view of the narrower tax exemptions provided in that act, the Treasury had to raise the certificate rate to 4 percent. The offering rate for new certificates remained at 4 percent during the winter of 1917 to 1918, despite waning investor interest. In early February 1918, series III-B attracted only “scanty” subscriptions,38 and bankers again complained that they were being asked to buy certificates at a below-market rate.39 The Federal Reserve Bank of New York was forced to take 20 percent of the $500 million offering.40 The weak response signaled that the certificate rate would have to be moved up yet again—as it was, to 4½ percent on series III-C. The rate thereafter remained at 4½ percent through the middle of 1919. Rates on tax-anticipation certificates started out at 4 percent, in line with the contemporaneous level of rates on bond-anticipation certificates. However, the tax-anticipation rate stayed at 4 percent when the bond-anticipation rate moved to 4½ percent in late February 1918. Tax-anticipation series T-C, T-D, T-E, and T-F—all issued during or after mid-February 1918 and all bearing a 35. “Big Excess Demand for Treasury Issue,” New York Times, May 24, 1917, p. 3. The Treasury offered $200 million of series I-B certificates and reported that it had received subscriptions for exactly that amount. Federal Reserve Bulletin, June 1917, p. 423. This suggests that government officials had to cajole some banks into underwriting the offering, much as Treasury Secretary Carlisle had done in January 1894. See “Carlisle Talks to Bankers,” New York Times, January 30, 1894, p. 1, Barnes (1931, pp. 315–18), and Carosso (1987, p. 316). 36. “Secretary Raises Rate to 3¼% on $200,000,000 Certificates,” Wall Street Journal, May 19, 1917, p. 8 (reporting that “the rate of 3¼% has been established with the particular purpose in view of encouraging the sale of these certificates to investors and the smaller banks even in those districts where higher money rates usually prevail and thus to bring about a gradual general preparation throughout the country for a convenient method of anticipating the payment for the Liberty Loan.”). 37. “Call Money at 10% Highest This Year,” New York Times, July 17, 1917, p. 13. 38. “Next Liberty Loan 4½ Per Cent?,” New York Times, February 21, 1918, p. 1. 39. “Next Bond Issue Puzzle Makes Bankers Hesitate,” Wall Street Journal, February 14, 1918, p. 12. (“[T]he investment which [bankers] are asked to make does not appeal to them, from a business point of view. They are asked to invest in a 4% certificate . . . at a time when the market for outstanding [4 percent] Liberty bonds is considerably below par. Should investors, under the circumstances, not invest in the forthcoming bond issue as expected, the banks would then find themselves loaded with a considerable amount of the certificates of indebtedness. This would mean a dangerously over-extended credit situation.”) 40. “U.S. Certificates Are a General Accommodation,” Wall Street Journal, February 19, 1918, p. 10.

120

Chapter 7

4 percent interest rate—consequently attracted limited interest. After an exceptionally poor reception for series T-G during the late summer and fall of 1918, the Treasury raised the rate on tax-anticipation certificates to 4½ percent.41 Were Sales of Certificates of Indebtedness Coercive? Offering rates on certificates of indebtedness were generally reflective of market conditions, but Secretary McAdoo nevertheless made fervent appeals to patriotism to sell the securities. In a telegram sent to banks and trust companies in early February 1918, McAdoo stated that the Treasury planned to offer $500 million of bond-anticipation certificates every two weeks until the launch of the Third Liberty bond in early April and requested (or demanded) bank support: In order . . . to provide for the expenditures for the military operations of the Unites States and the allies, I must have the whole-hearted cooperation of the bankers of the United States, and to that end I request the board of directors or trustees of each bank and trust company to reserve each week out of its loanable funds for the use of the Government of the United States about one per cent of gross resources of their institutions, not to exceed in the aggregate ten per cent. and to invest that amount in Treasury certificates of indebtedness. We are approaching a critical test on the battle fronts in Europe. America’s sons are now actually shedding their blood in the trenches. If the banks, which are the first line of financial defense, fail to support the Government fully in its necessary operations we shall imperil America’s army and America’s safety. I know that I have only to state the case to command the support of every patriotic bank and banker. This is a supreme duty of patriotism. May I count upon you to do your part and to telegraph me immediately at my expense that you will?42

The Wall Street Journal remarked that “McAdoo’s telegram to the banks looks like coercion to float Government loans.”43 McAdoo’s telegram was sent simultaneously with the announcement of series III-B bond-anticipation certificates—an offering that generated little investor interest and that (as noted earlier) had to be bailed out by the Federal Reserve. The fact that the Treasury bumped the offering rate to 4½ percent on the very next series (series III-C) suggests that whatever coercive pressure McAdoo might have sought to apply was largely ineffective. There were several other instances where McAdoo and others appealed to patriotism to sell certificates,44 but the only truly effective remedy for lagging certificate sales was a higher offering rate. 41. Hollander (1919, p. 58) characterized the results of the series T-G offering as “gravely disappointing.” 42. 1918 Treasury Annual Report, pp. 196–97. 43. “Certificates Pave Way for Government Loan,” Wall Street Journal, February 8, 1918, p. 10. 44. Most prominently, in June 1918 McAdoo sent a letter to banks and trust companies to inform them that the Treasury would be offering $750 million in bond-anticipation certificates every two weeks and to seek commitments for 1 1/4 percent of gross bank resources weekly: “America’s sons are dying daily in . . . battles of fire and poison gases that are now raging in France. The

121

Treasury Cash Management: Certificates of Indebtedness

Summary Treasury officials adopted two strategies to dampen fluctuations in government balances at Federal Reserve Banks that would have otherwise resulted from the immense flows of funds attributable to wartime tax payments and bond sales: (1) raise (and quickly disburse) money in smaller amounts, several hundred million dollars at a time, by issuing certificates of indebtedness that could be used later to satisfy tax liabilities or to fund bond subscription payments, and (2) redeposit bond and tax receipts in Treasury accounts at commercial banks. The Treasury made 33 public offerings of bond-anticipation certificates and 14 public offerings of tax-anticipation certificates between April 1917 and mid-1919. In view of the novelty and unprecedented scale of the operation, the certificate program worked quite well. Officials showed remarkable ingenuity in adapting the terms of the certificates to better achieve their cash management objectives by including call and limited-use put options. The relatively small size of individual offerings ensured that the Treasury was not at substantial risk if a particular offering failed to sell well. And the more or less continuous offerings meant that the Treasury received regular feedback from market participants about interest rates. heart of every American must thrill with pride and emotion as he thinks of the sacrifices our sons are making for our safety and liberty. The bankers of the United States can render a particularly useful service to our gallant sons by keeping the Treasury of the United States supplied with the money required by the Government to furnish every American hero with the things he must have to fight victoriously or to die gloriously. I am sure no patriotic banker in the United States will fail to do his full meed of essential service to his country and to her noble defenders.” 1918 Treasury Annual Report, pp. 21–22. McAdoo, and later Carter Glass, continued to make patriotic appeals to promote purchases of certificates of indebtedness after the signing of the armistice but, in the absence of hostilities, the appeals lacked credibility. See McAdoo’s letter of November 27, 1918, (1919 Treasury Annual Report, pp. 56–57), Glass’s telegram to the Governors of the Federal Reserve Banks in January 1919 (“Carter Glass Calls on Bankers for Funds,” New York Times, January 4, 1919, p. 14), and Glass’s statement promoting the benefits of series T2 tax-anticipation certificates (“Secretary Glass Urges Buying Treasury Certificates,” New York Times, January 16, 1919, p. 10).

8

Treasury Cash Management: War Loan Deposit Accounts

Based on earlier experience in connection with bond sales to finance the Spanish–American War and the Panama Canal, most observers in the spring of 1917 expected that the Treasury would redeposit in commercial banks the proceeds of wartime securities sales until the funds were needed.1 Shortly after President Wilson signed the declaration of war against Germany, The New York Times reported that “it is taken for granted that the proceeds of the sale of . . . bonds will not be locked up in the Treasury or deposited exclusively with the Federal Reserve Banks, but will be kept on deposit with banks throughout the country.”2 The Treasury already had a network of national bank depositories— described in box 3.1—but Treasury Secretary McAdoo wanted a broader network that included state-chartered banks.3 Congress agreed, and provided for a broad depository program in the very first Liberty Loan act.4 The program had several important features: • The Secretary of the Treasury was not limited to placing deposits with national banks, but could also place deposits with state-chartered banks, including banks that were not members of the Federal Reserve System.5 • The Secretary of the Treasury was authorized to prescribe the rate of interest to be paid on the deposits, as well as other terms and conditions (e.g., collateral requirements). • Reserve requirements on Treasury balances were waived for national banks and state banks that were members of the Federal Reserve System. 1. The Treasury provided redeposit programs during the sale of Spanish–American War bonds in 1898 (Taus 1943, p. 96), during a sale of Panama Canal bonds in 1906 (Timberlake 1978, p. 178, “The Bond Issue,” New York Times, July 4, 1906, p. 6, “The Panama Canal Bonds,” Wall Street Journal, July 6, 1906, p. 1, “Banks May Keep Funds Paid for Canal Issue,” New York Times, July 13, 1906, p. 10, and “Panama Bonds,” Wall Street Journal, August 28, 1906, p. 8), and during a sale of Panama Canal bonds in 1907 (Taus 1943, p. 124). 2. “Bankers Welcome Plan for War Loan,” New York Times, April 12, 1917, p. 2. 3. McAdoo’s position on depositing funds with state-chartered banks appears in Committee on Ways and Means (1917, p. 73): “I felt it was very desirable to keep these funds distributed as widely as possible so that we should not make a demand all at once that might cause some disturbance in the country . . ..” See also Hendricks (1933, pp. 267 and 273). 4. Section 7 of the First Liberty Bond Act provided that “the Secretary of the Treasury, in his discretion, is hereby authorized to deposit in such banks and trust companies as he may designate the proceeds . . . arising from the sale of the bonds and certificates of indebtedness authorized by this Act . . ..” 5. During Senate debate on the First Liberty Bond Act, Senators James Reed and William Stone, Democrats from Missouri, objected to the deposit of Treasury funds in nonmember banks but Senator Henry Cabot Lodge, Republican from Massachusetts, responded that “The United States is now at war. This bill is drawn for the sole purpose of enabling the United States to prosecute that war with vigor to a successful termination. It is not intended to encourage one system of banks at the expense of another . . ..” “Senate Is a Unit for Bond Issue,” New York Times, April 18, 1917, p. 1.

124

Chapter 8

• Deposits were subject to recall when the Treasury needed to replenish its transaction balances at the Federal Reserve Banks.6 In view of the need for close and continuing contact with a large number of depositories, the Federal Reserve was tapped to administer the program on behalf of the Treasury. The importance of redepositing the proceeds of securities sales became evident following the very first public wartime financing, the sale of $268 million of certificates of indebtedness on April 24, 1917. The certificates were issued on April 25 against payment by several large New York banks—acting in their own behalf and as agents for correspondent banks in other cities— through the New York Clearing House to the Federal Reserve Bank of New York.7 The payment depleted the surplus reserves of the New York banks and, as shown in table 8.1, led to a temporary tightening of overnight call loan rates and a virtual suspension of term lending on stock exchange collateral. The experience put Treasury and Federal Reserve officials on notice that they would have to have a redeposit program in place before the first offering of Liberty bonds settled. Early Experience with Redeposits Following Sales of Certificates of Indebtedness A redeposit mechanism was still wanting when the Treasury announced the second series of certificates on May 2, 1917, and remained wanting when that issue settled on May 10. This time, however, the New York Fed devised a temporary expedient: redepositing funds tendered by New York banks and accounting for the redeposits as “uncollected items.”8 On May 24, 1917, the day before the third certificate offering was due to settle, the New York Fed announced a bare-bones redeposit program. The Bank informed subscribing banks that “a portion of the proceeds of the sale of [series I-C certificates] will not be required for several days; and in order not to disturb the money market we have been instructed to redeposit part of the proceeds in the banks subscribing to [the] certificates. . . . You will be required to pay interest at the rate of 2% per annum on the amount deposited with you and to deposit with this bank 100% security for such deposit. Security may consist of [the certificates] or other Government securities in negotiable form.”9 The 6. Hollander (1919, p. 143) describes the mechanics of a recall. 7. “U.S. Certificates to be Paid for Today,” Wall Street Journal, April 25, 1917, p. 8. 8. “Reserve Banks Get Federal Deposits,” New York Times, May 13, 1917, p. E7. See also “Previous Sales of Certificates,” New York Times, May 18, 1917, p. 16 (“proceeds of the sale of [the last issue of certificates] were deposited with the local banks and trust companies”). 9. Federal Reserve Bank of New York circular dated May 24, 1917, in bound Bank circulars collection, Research Library, Federal Reserve Bank of New York.

125

Treasury Cash Management: War Loan Deposit Accounts

Table 8.1 Call loan rates and comments on time money around April 25, 1917 Call loan rate Monday, April 16 April 17 April 18 April 19 April 20

2½% 2½ 2½ 2½ 2¼

Monday, April 23 April 24

2½ 2¾

April 25

4

April 26

2½

April 27

2¼

Comments on time money

“Time money is hardly obtainable. . . . This condition is primarily due to Government’s offering of certificates of indebtedness which has temporarily absorbed the supply of available funds.” “Trading in time funds is practically suspended.” “Time money continues scarce and firm as a result of the coming payment of subscriptions to the Government Treasury bills.” “Dealings in time money are practically suspended on account of the payment of the subscriptions to the Government certificates of indebtedness. The situation is somewhat unsettled in spite of the fact the money paid into the Federal Reserve Bank is being at the same time disbursed, and institutions are reluctant to enter into new commitments pending the restoration of more normal conditions.” “Bidding [for time funds] is . . . light as borrowers are holding aloof in anticipation of lower rates in the next few days, when the temporary unsettlement . . . has passed away.”

Source: “Call and Time Money,” Wall Street Journal, various issues between April 17 and April 28, 1917.

mechanics were cumbersome but transparent: “[O]n receipt of payment tomorrow for your allotment of . . . certificates of indebtedness and advice from you that you desire [a redeposit of the payment], we will hand you our check for the amount of deposit drawn to your order for account of this bank as fiscal agent of the Treasury Department.” Unlike the work-around used for the second series, the Fed’s redeposit program provided for both collateral and the payment of interest on redeposited funds. (The 2 percent interest rate was identical with the rate on Treasury balances deposited with national banks under the existing depository program; see box 3.1.) The Wall Street Journal reported that payments for the third series of certificates “caused no embarrassment to the money market because the [Federal Reserve Bank of New York] at once redeposited the funds . . ..”10 10. “Money Undisturbed by Certificate Payments,” Wall Street Journal, May 26, 1917, p. 8.

126

Chapter 8

The Redeposit Program for the First Liberty Loan Providing for the redeposit of the proceeds of a Liberty bond sale was more complicated than providing for the redeposit of the proceeds of a certificate sale. Certificates were bought primarily by banks, so banks had ready access to collateral—the certificates themselves—that could be pledged to secure a Treasury deposit. However, a significant fraction of Liberty bonds were bought by individuals and by institutions other than banks. If bonds were bought with funds borrowed from a bank, the bank would likely retain possession of the bonds (as collateral for its loan) and could rehypothecate the bonds. But if bonds were bought with cash, for example, by check drawn against an existing bank balance, the buyer would take possession of the bonds. Thus there was a compelling need to establish a redeposit program that could accept a wide range of bank assets as collateral.11 On May 16, 1917, the Treasury announced that it would allow banks to pay for subscriptions to the First Liberties by book-entry credit to accounts that would pay 2 percent interest.12 Rather than pay for the bonds and then receive back a redeposit of funds (as was done in the settlement of the third series of certificates), a bank could simply write up what it owed the Treasury.13 Two weeks later the Treasury announced that balances in what were called “Liberty Loan Deposit Accounts” could be secured with the collateral shown in table 8.2 as well as with First Liberty bonds.14 (The accounts were renamed “War Loan Deposit Accounts” later in 1917.15) Extension of the Redeposit Program to Tax Collections The first big wave of wartime tax collections came due in June 1918 and posed the same problem as Liberty bond sales: the prospect of large payments draining bank reserves and disrupting the money market. Observing the Treasury’s 11. “Difficulties in Depositing the Big Loan Proceeds,” Wall Street Journal, April 27, 1917, p. 8. 12. Treasury Circular no. 79, May 16, 1917, reprinted in 1917 Treasury Annual Report, p. 131. See also “Treasury Plans to Let the Banks Keep Funds Paid for Liberty Bonds,” New York Times, May 19, 1917, p. 1, “Credit Payment in Banks to Avoid Disturbance,” Wall Street Journal, May 21, 1917, p. 8, and Taus (1943, p. 160). 13. Payment by book-entry credit was subsequently extended to bank payments for certificates of indebtedness. Hollander (1919, pp. 37–39). 14. Treasury Circular no. 81, May 29, 1917, reprinted in 1917 Treasury Annual Report, p. 124. The requirement that at least 25 percent of any bank’s collateral had to be Treasury securities was reduced to 10 percent when depository banks found themselves short of Treasury collateral at the time of the June 28, 1917, installment payment on the First Liberties. “Liberty Loan Ruling,” Wall Street Journal, June 28, 1917, p. 8. The requirement was subsequently eliminated altogether. Treasury Circular no. 92, October 6, 1917, reprinted in 1917 Treasury Annual Report, p. 133. 15. Treasury Circular no. 92, October 6, 1917, reprinted in 1917 Treasury Annual Report, p. 133.

127

Treasury Cash Management: War Loan Deposit Accounts

Table 8.2 Securities acceptable as collateral for government deposits Collateral

Terms

Treasury bonds and certificates of indebtedness Bonds issued under the United States Farm Loan Act and bonds of the Philippine Islands, Porto Rico, and the District of Columbia Bonds of any state of the United States

Par value Par value

3½% bonds of the Territory of Hawaii Other bonds of the Territory of Hawaii Bonds of the Manila Railroad Co. Dollar-denominated bonds and obligations of foreign governments engaged in war against Germany and issued since July 30, 1914 Bonds of any county or city in the United States that are direct obligations of the county or city as a whole Railroad mortgage bonds secured by direct mortgage upon lines of railroads within the United States, but not including any such bonds that yield more than 5½% per annum to maturity on May 29, 1917 Commercial paper eligible for rediscount or purchase by Federal Reserve Banks and approved by the Federal Reserve Bank of the district in which the depositary bank is located

Lesser of market value and par value 90% of par value Lesser of market value and par value 90% of the lesser of market value and par value 90% of the lesser of market value and par value 75% of the lesser of market value and par value Lesser of (a) 75% of market value and (b) par value

75% of par value

Source: Treasury Circular no. 81, May 29, 1917, reprinted in 1917 Treasury Annual Report, p. 124. Note: At least 25 percent in value, as determined above, of the securities deposited by any bank or trust company to secure deposits must consist of Treasury bonds and certificates of indebtedness.

success with redepositing the proceeds of securities sales in War Loan Deposit Accounts, Congress granted the Secretary of the Treasury authority to deposit tax collections as well.16 Market participants expected that the process for redepositing tax payments would be virtually identical to the process for redepositing bond payments.17 In particular, it was expected that “no money [will] be withdrawn immediately from the money market; but, instead, the proceeds of the tax payments will be credited by the banks, through a bookkeeping process, to account of the Government.”18 However, the process was more cumbersome in practice and 16. Section 5 of the Third Liberty Bond Act. 17. The New York Times reported that “It will be exactly the same thing as was done in the case of the Second Liberty Loan and in payments that have been made from time to time by the banks on the purchase of U.S. Certificates of Indebtedness.” “To Leave Income Taxes on Deposit in Banks,” New York Times, May 23, 1918, p. 10. 18. “To Leave Income Taxes on Deposit in Banks, ” New York Times, May 23, 1918, p. 10.

128

Chapter 8

reminiscent of the redeposit process for the third offering of certificates of indebtedness: Under the system adopted the Collectors [of Internal Revenue] deposited with the Federal Reserve Bank [of New York] every day, or twice a day, all checks received. Thousands and tens of thousands of checks were brought in to the Reserve Bank daily last week. Clerks in the Reserve Bank sorted out all the checks according to the institutions on which they were drawn, and then the bank turned over to each of the institutions all of the checks drawn on it. Each bank and trust company in this way received a Government deposit of an amount equal to the aggregate of the checks drawn by its depositors to the order of the Collector of Internal Revenue.19

The process may have been cumbersome, but it worked. Frank Vanderlip, the president of National City Bank, remarked that, “The tax payments were made without a ruffle. . . . The transaction was put through with perfect balance and without the least disturbance to business or banking. The Treasury Department, the Federal Reserve Banks and the people of the country deserve commendation for the foresight that they all have shown in arranging for the payments.”20 Redeposits of Tax Payments in 1919 The Revenue Act of 1918 provided that taxes on income earned in 1918 would be due in four quarterly installments, on March 15, June 15, September 15, and December 15, 1919, rather than in a single lump sum. When the first installment came due, Treasury officials employed the same redeposit scheme they had used in 1918 and the payments were processed without incident.21 Three months later, in June 1919, Treasury officials decided to retain, rather than redeposit, the second installment—apparently because they believed the tax payments, when netted against prospective redemptions of certificates of indebtedness, would produce only a relatively small increase in Treasury balances at the Reserve Banks.22 They soon regretted their decision. On Friday, June 13, the money market tightened dramatically in anticipation of the 19. “Money Market Firm Under Income Taxes,” New York Times, June 16, 1918, p. 17. 20. “$3,000,000,000 Tax Paid with No Strain,” New York Times, June 17, 1918, p. 10. Walter Frew, president of the Corn Exchange Bank, observed similarly that the Treasury “manifested great wisdom in working out a plan for the redistribution of tax receipts among the banks, and in this connection the Federal Reserve Banks again demonstrated their great usefulness as fiscal agents of the Government. The Reserve Banks handled the work very satisfactorily indeed.” “$3,000,000,000 Tax Paid with No Strain,” New York Times, June 17, 1918, p. 10. 21. “Congestion Causes Delay in Clearing Tax Payments,” Wall Street Journal, March 19, 1919, p. 12. 22. On June 9, 1919, the Treasury had offered to redeem series T2, T3, V-D, and V-E certificates on demand at par plus accrued interest to the date of redemption. The Treasury made the offer because of unexpectedly heavy installment payments on sales of Victory Liberty notes. 1919 Treasury Annual Report, p. 289, Exhibit 35. In addition taxpayers could satisfy their June 1919 tax liabilities with any of the outstanding tax-anticipation certificates.

129

Treasury Cash Management: War Loan Deposit Accounts

looming tax date, the overnight call loan rate rising to 12 percent.23 The following Monday the call loan rate hit 15 percent, the highest level in seven years.24 By the end of the week the Treasury was forced to call $530 million of series V-G certificates for early redemption on July 1 in order to get reserves back into the banking system.25 Summary The disruptions to the short-term credit markets and the spikes in short-term interest rates associated with the settlement of the first series of bond-anticipation certificates of indebtedness and the June 1919 tax payments, as well as the absence of disruptions between those two settlements, demonstrate the practical usefulness of the Treasury redeposit program. The conceptual basis for the program was not new; the Treasury had been managing its cash balances with an eye to mitigating strains in the money market for decades.26 However, the scale of the program, involving hundreds of millions of dollars, was new. 23. “Financial Markets,” New York Times, June 14, 1919, p. 16, and “Action of Call Money Disturbs Stock Market,” Wall Street Journal, June 14, 1919, p. 1. 24. “Call Money Reaches Highest in 7 Years,” New York Times, June 17, 1919, p. 29, and “Banks Calling Loans to Restore Their Reserves,” Wall Street Journal, June 17, 1919, p. 1. 25. “Indebtedness Certificates Called,” Wall Street Journal, June 21, 1919, p. 5, and 1919 Treasury Annual Report, p. 289, Exhibit 36. 26. See box 2.1 in chapter 2.

9

Federal Reserve Support of the Treasury Market during World War I

Federal Reserve support of the Treasury market during World War I was founded on two statutory credit facilities: the original power of the Reserve Banks to discount, or purchase, loans made by member banks to their customers, and the subsequently added power of the Reserve Banks to lend to member banks.1 Reserve Bank discounts were limited to (1) short-term loans “arising out of actual commercial transactions” and (2) short-term loans to purchase or carry Treasury securities.2 The authority to lend to member banks was added in 1916,3 when section 13 of the Federal Reserve Act was amended to provide that: Any Federal reserve bank may make advances [i.e., loans] to its member banks on their promissory notes for a period not exceeding fifteen days . . . provided such promissory notes are secured by such notes . . . as are eligible for rediscount or for purchase by Federal reserve banks . . ., or by the deposit or pledge of bonds or notes of the United States.

Interest rates on loans and discounts were set by the individual Reserve Banks, subject to the “review and determination” of the Federal Reserve Board in Washington.4 The First War Financing The first war-related Treasury financing was a sale, on March 28, 1917, of $50 million of certificates of indebtedness directly to the Federal Reserve Banks.5 The transaction provoked the ire of both the Banks and the Board and played 1. The Banks also had the power to purchase Treasury securities in open market operations, but rarely used that power during the war. 2. Section 13 of the Federal Reserve Act provided that, “Upon the endorsement of any of its member banks . . ., any Federal reserve bank may discount notes, drafts, and bills of exchange arising out of actual commercial transactions; . . . the Federal Reserve Board to have the right to determine or define the character of the paper thus eligible for discount, . . . but such definition shall not include notes, drafts or bills covering merely investments or issued or drawn for the purpose of carrying or trading in stocks, bonds, or other investment securities, except bonds and notes of the Government of the United States. Notes, drafts and bills admitted to discount under the terms of this paragraph must have a maturity at the time of discount of not more than ninety days: Provided, That notes, drafts, and bills drawn or issued for agricultural purposes . . . and having a maturity not exceeding six months may be discounted in an amount to be limited to a percentage of the capital of the Federal reserve bank, to be ascertained and fixed by the Federal Reserve Board.” 3. Act of September 7, 1916. 4. Section 14(d) of the Federal Reserve Act empowered each Reserve Bank, “to establish from time to time, subject to review and determination of the Federal Reserve Board, rates of discount to be charged by the Federal reserve bank . . ..” The Act of September 7, 1916, provided similar authority for the Banks to set, and for the Board to review, interest rates on loans to member banks. 5. “Reserve Banks Lend M’Adoo $50,000,000,” New York Times, March 29, 1917, p. 2, “Treasury Notes Allotted,” New York Times, March 30, 1917, p. 11, and “Federal Reserve Banks Used by Government,” Wall Street Journal, March 30, 1917, p. 8. The certificates were part of a total of $300 million of certificates authorized by the Revenue Act of 1917.

132

Chapter 9

an important role in shaping the character of Federal Reserve support of the Treasury market during the war. On March 28, 1917, Treasury Secretary McAdoo advised the Federal Reserve Board that he wanted to borrow $50 million for 90 days at 2 percent interest in anticipation of personal and corporate income taxes coming due in June. He suggested that: This is an excellent opportunity for the Federal Reserve Banks to secure a desirable short-term investment and to demonstrate their usefulness as fiscal agents of the Government. I purpose, therefore, to offer the Federal Reserve Banks, the opportunity to take these certificates. Will you please get in touch with the Federal Reserve Banks and ascertain whether or not they care to take this loan and what amount they respectively desire to take?6

By framing the loan as a fiscal agency activity, McAdoo blithely conflated a principal function—lending money—with the responsibilities of the Banks as fiscal agents of the United States. The country being on the verge of war, Reserve Bank officials agreed to buy the certificates—but they were dismayed with what they regarded as a below-market rate of interest and with the idea that McAdoo should so misconstrue the role of a fiscal agent.7 The Governor of the Federal Reserve Bank of Boston replied to the Board’s inquiry that: We feel that it is of the utmost importance that at the present time the resources of the Reserve Bank should be available so far as is possible for the benefit of their member banks, and we should have preferred, had the taking of this issue by member banks been burdensome to them, to have rediscounted for them to meet their requirements, rather than to have taken the issue direct, and our committee are in hopes that should future financing of this sort become necessary that it will be dealt with in this way, and the loans placed at a rate that will induce the commercial banks to absorb the issue.8

The directors of the Federal Reserve Bank of New York similarly expressed their displeasure at McAdoo’s presumptuousness: Whenever in the future circumstances may seem to indicate the advisability of any action being taken by the Bank as Fiscal Agent [the directors of the Federal Reserve Bank of New York] look forward to such consultation with the Government relative thereto as is customary between

6. Quoted in Harding (1925, pp. 87–88). 7. Contemporaneous as well as later observers remarked on the 2 percent interest rate. See “Reserve Banks Lend M’Adoo $50,000,000,” New York Times, March 29, 1917, p. 2 (bankers characterized as believing that “other institutions . . . would not care to invest their funds . . . at the very unattractive rate”), “Federal Reserve Banks Used by Government,” Wall Street Journal, March 30, 1917, p. 8 (remarking that McAdoo “could not hope to place [the certificates] in the market” at 2 percent), Harding (1925, p. 88) (“The opinion of a majority of the members of the Federal Reserve Board, and of all of the Federal Reserve Bank Governors, was that the rate proposed . . . was too much below the market, and that it should have been at least two and one half per cent.”), and Chandler (1958, p. 113) (“The 2 percent rate was well below market and unattractive to private buyers.”). Meltzer (2003, p. 86n. 45) states that the Board acquiesced to the below-market rate when McAdoo threatened to invoke the Overman Act. However, the Overman Act did not become law until May 20, 1918. 8. Quoted in Chandler (1958, pp. 113–14).

133

Federal Reserve Support of the Treasury Market during World War I

fiscal agents and their principals, in order to be in a position to render it the most effective service. . . . [The directors] are strongly of the opinion that the normal services of the Bank as fiscal agent will best be rendered by assisting in distributing Government securities rather than by acting as a purchaser of them. . . . The Directors desire, therefore, that their present action in subscribing for 90-day certificates of the Government should not be considered as establishing a precedent or a policy which will necessarily be followed in the future.9

Except for a limited number of short-term “cash management” loans, almost all of which matured in seven days or less, the Treasury did not subsequently obtain credit directly from the Federal Reserve Banks.10 Instead, the Banks lent to member banks to fund bank investments in Treasury securities and to fund bank loans to customers who invested in Treasury securities. At the end of the war the Reserve Banks had a substantial portfolio of loans secured with Treasury debt, but they did not themselves own any substantial quantity of Treasury securities.11 Financing the First Liberty Loan The broad characteristics of the Fed’s indirect support of the Treasury market during World War I were established in the very first Liberty Loan financing. On Wednesday, May 2, 1917, Secretary McAdoo announced that the Treasury would soon offer an unprecedented $2 billion of Liberty bonds.12 It was clear that, unlike the sale of Spanish–American War bonds in 1898, the offering would not be a blowout. Congress had set a maximum rate of interest on the bonds—3½ percent—that was realistic but hardly generous. Treasury and Federal Reserve officials believed that the bonds would have to be sold in large part to investors who financed their purchases with comparably priced bank credit. Their banks would, in turn, need access to similarly priced central bank credit. Table 9.1 shows the rates posted by Federal Reserve Banks at the end of April 1917 for short-term loans to member banks and for discounts of member bank commercial paper. All of the discount rates were at or above 4 percent and, except for New York, all of the loan rates were at or above 3½ percent. The Banks would have to post lower rates if member banks and their customers 9. Quoted in Chandler (1958, p. 114). 10. The cash management loans are noted in Hollander (1919, p. 25), Hendricks (1933, p. 272), 1918 Treasury Annual Report, pp. 24 and 27, 1919 Treasury Annual Report, pp. 55 and 260, 1917 Federal Reserve Board Annual Report, p. 265, 1917 Federal Reserve Bank of New York Annual Report, p. 60, 1918 Federal Reserve Bank of New York Annual Report, p. 13, and 1919 Federal Reserve Bank of New York Annual Report, pp. 16 and 65. 11. Meltzer (2003, p. 88) suggests that the relatively small amount of Treasury debt owned by the Reserve Banks facilitated the postwar contraction of Reserve Bank credit and restoration of Federal Reserve control of monetary policy. 12. “McAdoo Offers $2,000,000,000 of Liberty Loan,” New York Times, May 3, 1917, p. 1.

134

Chapter 9

Table 9.1 Loan rates and discount rates posted by Federal Reserve Banks, 1917 15-Day loan rate

Boston New York Philadelphia Cleveland Richmond Atlanta Chicago St. Louis Minneapolis Kansas City Dallas San Francisco

61- to 90-Day discount rate

April 30

May 31

June 28

April 30

May 31

June 28

3½ 3 3½ 3½ 3½ 3½ 3½ 3½ 4 4 3½ 3½

3½ 3 3 3* 3½ 3½ 3** 3½ 3** 3** 3½ 3½

3½ 3 3½ 3* 3½ 3½ 3** 3½ 3** 3** 3½ 3½

4 4 4 4½ 4 4½ 4½ 4 4½ 4½ 4½ 4½

4 3½ ** 4 4½ 3½ ** 3½ ** 3½ ** 3½ ** 4 3½ ** 3½ ** 3½ **

3½ ** 3½ ** 3½ ** 3½ ** 3½ ** 3½ ** 3½ ** 3½ ** 3½ ** 3½ ** 3½ ** 3½ **

Source: Federal Reserve Bulletin. Note: Rates are generally applicable to eligible collateral unless otherwise noted. * if secured with certificates of indebtedness. ** if secured with certificates of indebtedness or Liberty bonds.

were to have an economic incentive to finance purchases of 3½ percent Liberty bonds with Reserve Bank credit. The first step to lower interest rates was the mid-May announcement of the Federal Reserve Board that it would approve any Reserve Bank request to lower its 15-day loan rate to 3 percent.13 The Board took the view that attractively-priced short-term loans to member banks could ease “temporary shortages [of reserves] which may arise here and there” while investors were paying for their bond purchases.14 As shown in table 9.1, five Reserve Banks acted before the end of May to lower their loan rates, with four of the five Banks limiting the 3 percent rate to loans secured with Treasury securities. In late May the Board announced that it would also approve any Reserve Bank request to lower its discount rate for bank loans secured by Liberty bonds or certificates of indebtedness to 3½ percent. The Board envisioned a conventional financing role for the preferential discounts: The Board has been advised that many corporations, including savings banks, have agreed to subscribe to substantial amounts of the Liberty Loan, carrying the bonds for their employees

13. Federal Reserve Bulletin, June 1, 1917, p. 425, and Wicker (1966, p. 14). 14. Federal Reserve Bulletin, June 1, 1917, p. 430 (noting that “The facilities offered by Federal Reserve Banks for [15-day loans] secured by commercial paper or Government securities, should prove of great value in meeting [unusual] withdrawals [of bank deposits]. . . . Banks should realize that, under present circumstances, it will be their patriotic duty freely to avail themselves of these facilities. They will thus be able to keep money rates easy.”).

135

Federal Reserve Support of the Treasury Market during World War I

subject to payment in small installments. It has been suggested that it would be helpful to the banks in placing the Liberty Loan, and particularly in assisting corporations which make subscriptions of this character, if Federal Reserve Banks were authorized to establish a special Liberty Loan rate of 3½ per cent for notes . . . drawn by customers of the banks . . . having a maturity not in excess of 90 days, and secured by Liberty Loan bonds or United States Treasury certificates of indebtedness.15

By the end of June, all twelve Banks had adopted a 3½ percent preferential discount rate (table 9.1). In announcing its decision to act on the Board’s recommendation, the Federal Reserve Bank of New York emphasized the importance of accommodating investors: “It is clearly desirable that the coming issue of bonds should be absorbed by the ultimate investor as rapidly as possible, and it will, therefore, be the policy of this bank to aid its member banks freely, if desired, in order that they in turn may give their customers every facility for purchasing the bonds, permitting them, if necessary, to take a reasonable time to make complete payment therefore.”16 Table 9.2 shows that the first Liberty Loan financings went as planned. The initial payment on the bonds was due on June 15 and the last installment on August 30. Reserve Bank loans and discounts rose sharply, from $98 million Table 9.2 Reserve Bank loans and bills discounted by Federal Reserve Banks, 1917 Loans and discounts ($millions) June 1 June 8 June 15 June 22 June 29 July 6 July 13 July 20 July 27 August 3 August 10 August 17 August 24 August 31

50.8 98.0 202.8 241.0 197.2 129.9 140.2 161.4 138.5 130.9 134.2 143.9 128.4 147.3

Source: Federal Reserve Bulletin.

15. Federal Reserve Bulletin, June 1, 1917, p. 430. 16. Federal Reserve Bank of New York Circular no. 64, May 22, 1917. Several months later, Federal Reserve Board member Adolph Miller defended the preferential rates, saying that they “were established by the Board at the time when the First Liberty Loan was in process of placement and when fear was expressed in banking circles that the floatation of the loan would be hindered unless distinct encouragement was given to bankers (who were expected to co-operate liberally in placing the bonds with their customers) . . ..” Quoted in Wicker (1966, p. 14).

136

Chapter 9

to $241 million, between June 8 and June 22 and thereafter declined to $147 million at the end of August. The relatively brief surge in Federal Reserve credit is broadly consistent with the stated intention of the Federal Reserve System to do no more than facilitate the initial distribution of the Liberty bonds. The “Borrow and Buy” Program The Fed’s credit support program evolved after the first Liberty Loan campaign into a program of longer term, but still indirect, support for the Treasury market. The catchphrase of the expanded program, “borrow and buy,” expressed the idea that investors should finance purchases of Liberty bonds and certificates of indebtedness with Reserve Bank credit channeled through commercial banks and then repay their loans from future savings. One prominent banker argued that: Our people must learn to borrow and our banks to rediscount. . . . Current savings are not sufficient; the people of the country . . . must pledge their future savings as well. They must borrow from their banks and invest in government bonds the funds thus received. The banks in turn must rediscount their customers’ notes at the federal reserve banks . . ..17

At the end of the war the Federal Reserve Bank of New York reflected on the significance of the borrow and buy program: “During the war the Federal Reserve Bank as fiscal agent of the Government had to encourage a programme of ‘borrow and buy,’ among banks in order that they might buy certificates of indebtedness far in excess of their available funds, and among individuals that they might buy bonds far in excess of their current savings.”18 The borrow and buy program accomplished its principle objective of facilitating sales of Treasury securities, but at the cost of a substantial increase in Federal Reserve loans and discounts.19 Figure 9.1 shows three categories of Reserve Bank assets: Bank loans and discounts, bills bought (in open market operations), and Treasury securities owned. The growth in loans and discounts, beginning around the time of the Second Liberty Loan campaign in the fall of 1917, is readily apparent. 17. Schiff (1918, p. 49). 18. Report on Business Conditions, Federal Reserve Bank of New York, November 20, 1919, p. 1. 19. Report on Business Conditions, Federal Reserve Bank of New York, November 20, 1919, p. 1 (“The result [of the borrow and buy program] was an immense increase in the loan accounts of the member banks, and a proportionately great increase in their rediscounts with the Federal Reserve Bank.”). Meltzer (2003) examines the implications of the wartime growth in Reserve Bank credit for the money supply and inflation.

137

Federal Reserve Support of the Treasury Market during World War I

2.5

Billions of dollars

2.0

1.5

1.0

0.5

0 Jan 1917

Jul 1917 Loans and discounts

Jan 1918

Jul 1918

Bills bought

Jan 1919

Jul 1919

Treasury securities owned

Figure 9.1 Federal Reserve loans and discounts, bills bought (in open market operations), and Treasury securities owned by Federal Reserve Banks

Wartime Discount Rate Policy The growth in Reserve Bank credit was achieved by holding Federal Reserve loan and discount rates at levels that made it profitable, or at least costless, to finance a position in Treasury securities with borrowed funds.20 Table 9.3 shows loan and discount rates posted by the Federal Reserve Bank of New York between 1917 and mid-1919. Figure 9.2 shows that during the four Liberty bond campaigns the Bank kept its short-term loan rate below—and its preferential discount rate at or below—the coupon rate on the bonds the Treasury was offering. (Figure 9.3 shows that the New York Fed’s preferential discount rate was substantially below the open market rate on short-term commercial paper throughout the war.) The first wartime increase in preferential discount rates, from 3½ to 4 percent, followed the completion of the second Liberty Loan campaign in the fall of 1917. Secretary McAdoo announced on September 27 that the Second 20. Wicker (1966, p. 15) (“To induce the public to borrow and buy, the lending rate at commercial banks could not be in excess of the rate on Liberty Bonds. To induce the banks to lend to prospective customers, the Federal Reserve thought they would have to supply reserves at a rate equal to or below the coupon rate on the bonds.”).

Chapter 9

Table 9.3 Loan rates and discount rates posted by the Federal Reserve Bank of New York, January 1917 to June 1919 15-Day loan rate

16- to 90-Day discount rate

As of January 1, 1917 May 22, 1917

3% 3

December 21, 1917

3½

April 6, 1918

4

4% 3½ for notes secured by Liberty bonds or certificates of indebtedness 4 all other eligible bills 4 for notes secured by Liberty bonds or certificates of indebtedness 4½ all other eligible bills 4¼ for notes secured by Liberty bonds or certificates of indebtedness 4¾ all other eligible bills

Source: Federal Reserve Bank of New York Annual Reports. Note: Rates effective on April 6, 1918, remained in effect through mid-1919.

5.00

4.50

Percent

138

4.00

3.50

3.00

2.50 Jan 1917

Jul 1917 Liberty Bond coupon rate

Jan 1918

Jul 1918

Preferential discount rate

Jan 1919 Loan rate

Figure 9.2 15-Day loan rates and 16- to 90-day preferential discount rates (for notes secured by Liberty bonds or certificates of indebtedness) posted by the Federal Reserve Bank of New York. The heavy black lines show the subscription periods for the first four Liberty Loan campaigns.

Federal Reserve Support of the Treasury Market during World War I

7.00

6.00

5.00 Percent

139

4.00

3.00

2.00 Jan 1917

Jul 1917

Jan 1918

Jul 1918

Jan 1919

Jul 1919

Commercial paper rate Discount rate, other than Treasury collateral Discount rate, Treasury collateral

Figure 9.3 60- to 90-Day commercial paper rates and 16- to 90-day discount rates posted by the Federal Reserve Bank of New York

Liberties would bear a 4 percent coupon,21 the maximum allowed by the Second Liberty Bond Act. Officials at some Federal Reserve Banks began to contemplate moving their preferential discount rates to 4 percent as well, but the Federal Reserve Board resisted—on the grounds that an increase would impair prospects for the imminent Treasury offering.22 The second Liberty Loan campaign ended on October 27 and the Banks were subsequently allowed to raise their rates. The Reserve Banks raised their preferential discount rates a second time, to 4¼ percent, in early April 1918, at the start of the third Liberty Loan campaign, to match the interest rate on the new bonds. The increase did not come as a surprise to market participants. The Wall Street Journal reported that “A higher Federal Reserve Bank rate has been expected in banking circles, especially in connection with Liberty Loan operations . . .. The rates were out of keeping 21. “McAdoo to Offer 3 Billion or More in Bonds on Oct. 1,” New York Times, September 28, 1917, p. 1. 22. See, Meltzer (2003, p. 91, fn. 52) (stating that “Chairman Perrin [of the Federal Reserve Bank of San Francisco] wrote on October 3, 1917, to ask whether the discount rate should be raised to 4 percent for loans on 4 percent certificates and 4 percent Liberty bonds. [Board member Adolph] Miller replied for the Board, saying that an increase in rates would hurt the Second Liberty Loan”).

140

Chapter 9

with the prevailing rates in the money market and in view of the higher interest rate which the new Liberty bonds bear.”23 The Federal Reserve Board characterized the rate increase as “moderate” and noted that “The action taken makes the rate on 90-day paper secured by Government obligations identical with the rate borne by the third Liberty bonds, so that banks desiring to obtain accommodation for the carrying of such obligations can do so without expense to themselves.”24 Following the rate increase in the spring of 1918, the Reserve Banks kept their discount rates unchanged through the fall of 1919, largely because of actions taken during the fourth Liberty Loan campaign in October 1918, when government and campaign officials encouraged banks to commit to lending to bond buyers at a fixed rate for one year.25 Their actions created a moral obligation on the part of the Reserve Banks to keep their discount rates similarly fixed. Whether the Reserve Banks should post higher loan rates became an issue following the armistice in November 1918. As shown in figure 9.4, short-term (15-day) loans to member banks surged during and after the summer of 1918, from $550 million at the end of June to $1.3 billion at the end of November, and several Board members wanted to raise the (4 percent) loan rate to the level of the (4¼ percent) preferential discount rate. However, the new Secretary of the Treasury, Carter Glass, demurred, arguing that an increase would “gravely prejudice” prospects for the Victory Loan campaign planned for the spring of 1919. The Board sided with Glass and postponed any rate increase until after the completion of that campaign.26 Summary There is no question about the locus of Federal Reserve credit policy during the First World War: it was in the office of the Secretary of the Treasury.27 23. “Federal Reserve Bank Raises Discount Rates,” Wall Street Journal, April 8, 1918, p. 10. 24. Federal Reserve Bulletin, May 1, 1918, p. 362. 25. Wicker (1966, p. 31) (“An important feature of the Fourth Liberty Loan was the commitment made by almost all national banks to lend to their bond purchasing customers at 4¼% for ninety days with renewals at the same rate for one year. The various Liberty Loan organizations had made this request with the approval of the Federal Reserve Board and the Federal Reserve Banks.” Emphasis in the original.) and Chandler (1958, p. 140) (“During the Fourth Liberty Loan campaign . . . many banks, with the approval and even urging by the government, had contracted to lend for a year at an interest rate equal to the coupon rate on the bonds (4¼ per cent).”). 26. See, Wicker (1966, pp. 29–30). 27. There are several explanations for the primacy of the Treasury during the war, including (1) the country was at war and power tends to migrate to the executive branch during war, (2) the Secretary of the Treasury was also the chairman of the Federal Reserve Board, (3) the Federal Reserve risked precipitating a failed offering if it removed or reduced its support of the Treasury

Federal Reserve Support of the Treasury Market during World War I

2.5

2.0

Billions of dollars

141

1.5

1.0

0.5

0 Jan 1917

Jul 1917

Jan 1918 Total

Jul 1918

Jan 1919

Jul 1919

Maturing within 15 days

Figure 9.4 Total Federal Reserve Bank loans, bills bought, and bills discounted and loans and discounts maturing within 15 days

Following the armistice, the Federal Reserve Bank of New York explained that: The policy adopted by the Government during the war of paying low rates of interest on its borrowings . . . necessitated the adoption of a correspondingly low rate policy by the Federal Reserve Banks, in order to facilitate the sales of Government securities. These always exceeded in volume the savings available at the time and, therefore, could only be absorbed through recourse to bank loans for a considerable part of each issue. In fact, ‘borrow and buy’ was one of the slogans common to all of the later loan campaigns.28

market (because the Treasury sold certificates and bonds in fixed-price offerings), and (4) the role of the Reserve Banks as fiscal agents and principal organizers of the Liberty Loan campaigns. See Chandler (1958, pp. 106–107) (“The distinction between fiscal agency functions and the function of formulating and administering monetary policy became blurred, and the subordination of monetary policy to the wishes of the Treasury may have been facilitated during the war period. . . . [T]he distinction between Treasury finance and monetary policy was so tenuous that it was easy for the Treasury financing plans to determine monetary policy within narrow limits. Reserve Bank officials may also have confused their functions. For example, [Benjamin] Strong [wartime Governor of the Federal Reserve Bank of New York] acted in at least three important capacities: as a formulator and administrator of monetary policy, as fiscal agent for the Treasury, and as chairman of the Liberty Loan Committee for the New York Federal Reserve District. At times he probably did not know in which capacity he was acting, and he would have been reluctant to insist, in one capacity, on policies that would jeopardize his success in his other capacities. Specifically, he would have been reluctant to insist on a monetary policy that would make much more difficult his job of raising great sums for the Treasury.”). 28. 1919 Federal Reserve Bank of New York Annual Report, pp. 6–7.

142

Chapter 9

Professor Benjamin Beckhart of Columbia University concluded similarly that: During the war period the Reserve Banks shaped their discount policies to harmonize with the Treasury Department’s plans of war finance. The Treasury Department had determined upon floating government obligations at low rates of interest . . .. The Federal Reserve Banks, acting in accordance with the advice of the Board, fixed their discount rates in order to facilitate McAdoo’s program.29

However, it should not be concluded that the Federal Reserve policy of keeping preferential discount rates matched to interest rates on Liberty bonds meant that Treasury officials effectively set short-term interest rates. The Fed acted to make an investment in Liberty bonds financed with borrowed money costless (at least in the short run), but the Treasury still had to set bond rates at a level that would attract sufficient demand from investors. 29. Beckhart (1924, pp. 275–76). Beckhart additionally suggested (p. 276) that “It is quite possible that the System did not enter upon this voluntarily, but was forced to do so by pressure from the Treasury Department.”

10

Coda on Treasury Debt Management during World War I

At the end of June 1916, interest-bearing Treasury debt amounted to $1.0 billion. Three years later, it was $25.2 billion, including $3.5 billion in certificates of indebtedness, $4.5 billion in notes, and $17.2 billion in bonds.1 The explosion in debt provided the raw material for the emergence of a broad market for Treasury securities and marked the end of an era in which Treasury securities were owned primarily by national banks and used primarily as collateral for bank notes and government deposits.2 Several important changes in Treasury debt management accompanied the wartime growth in debt, including how the Treasury sold securities, the contractual provisions of Treasury securities, and the way in which the Treasury moderated the mismatch between episodic receipts (stemming from bond sales and tax collections) and its more nearly continuous disbursements. During the war Treasury officials relied on fixed-price offerings, rather than auction offerings, to sell a tidal wave of securities. Adhering to the congressional mandate that Liberty bonds should be offered as “popular” loans, officials set out to effect a broad public distribution of the securities. On the basis of the Treasury’s experiences in the preceding quarter century, this appeared to be feasible only with fixed-price offerings. However, unlike the case with the Spanish–American War bonds, Treasury officials did not set interest rates above market-clearing levels. To the contrary, they set rates at or below market levels and they relied on aggressive marketing and supportive Federal Reserve monetary policies to generate the subscriptions that they needed and the oversubscriptions that they wanted. Treasury officials had more leeway with respect to sales of certificates of indebtedness. They were not bound by a congressional mandate to offer certificates as popular loans and the investor base for certificates was largely limited to banks (for bond-anticipation certificates) and corporate and highincome individual taxpayers (for tax-anticipation certificates). Nevertheless, the Treasury sold certificates as well as bonds in fixed-price offerings. The evidence suggests that Treasury officials opted for fixed-price offerings in an attempt to hold interest rates on certificates below market-clearing levels, but that they were largely unsuccessful in the attempt. Lackluster sales and banker 1. The $4.5 billion of notes reflects the amount sold in the Victory Liberty loan campaign. As of June 30, 1919, receipts from sales amounted to $3.5 billion, with the balance due on installment payments. Total interest-bearing debt included $954 million in war savings stamps and certificates. 2. On June 30, 1919, $692 million of Treasury bonds was pledged against issues of national bank notes and $44 million of Treasury securities was pledged by national banks against Treasury deposits (1919 Treasury Annual Report, p. 717). In addition $905 million received from the sale of certificates of indebtedness and Victory Liberty notes was on deposit in War Loan Deposit Accounts and secured by Treasury and other securities (1919 Treasury Annual Report, p. 112). The total amount of collateral was therefore a far smaller proportion of outstanding Treasury debt than the 81 percent pledged at the end of fiscal year 1914 (table 3.2).

144

Chapter 10

complaints led to rate increases in May 1917, February 1918, and November 1918. The war years were a time of remarkable innovation in the contractual provisions of Treasury securities. Liberty bonds had several complex provisions: 1. To mitigate the risk that an investor’s securities would depreciate if rising interest rates forced the Treasury to sell higher coupon debt, the Treasury sold bonds that could be converted into new bonds earning the higher rate. 2. To attract high-income investors during a period of temporarily high tax rates, the Treasury sold bonds with tax exemptions limited to the duration of the war. 3. To attract investors in a variety of tax brackets (in the absence of detailed information on the prospective demand from investors in different brackets), the Treasury sold interconvertible notes with different coupon rates and tax provisions. The willingness of Treasury officials to introduce such novel structures was remarkable in view of the fact that those officials bore the not insubstantial risk of offering a mispriced security. (Alternatively, it could be argued that Treasury officials were audacious in their reliance on fixed-price offerings in view of the novelty and complexity of the securities they were selling.) Finally, the war years witnessed a significant expansion in Treasury cash management techniques. The Treasury sold certificates of indebtedness—in anticipation of both bond sales and tax collections—to moderate what might otherwise have been disruptively large inflows into Treasury accounts at Federal Reserve Banks. Redemption options, including a Treasury option to call a certificate on ten days’ notice and investor options to pay for bond purchases or satisfy tax liabilities with outstanding certificates, enhanced the usefulness of certificates for cash management purposes. Additionally the Treasury developed a vastly expanded network of depository banks to speed the return of reserve balances to the private sector.

III

PAYING DOWN THE WAR DEBT

11

Treasury Finance during the 1920s

At the end of the First World War the American fiscal landscape exhibited two prominent features: the national debt stood at $25 billion—an immense expansion over the $1 billion of pre-war indebtedness—and tax rates were so high that—in the opinion of both Democrats and Republicans—they significantly dampened investment incentives. The wartime expansion in the national debt was widely viewed as a temporary phenomenon, an alternative to the taxes that otherwise would have been required. The costs of war had to be funded and the country chose to fund three-quarters of the costs by issuing promises of future payment to those willing to give up current claims on goods and services. With hostilities at an end, many believed that it was time to begin liquidating those promises. David Houston, the Secretary of the Treasury in the last year of the Wilson administration, believed that “The war debt should be paid, not perpetuated, and the time to pay it is as soon as possible after the end of the war.”1 At the same time, political leaders recognized that the tax system had become badly warped in the haste of responding to wartime demands. In his 1919 State of the Union message, President Wilson suggested that: The Congress might well consider whether the higher rates of income and profits taxes can in peace times be effectively productive of revenue, and whether they may not, on the contrary, be destructive of business activity and productive of waste and inefficiency. There is a point at which in peace time high rates of income and profits taxes discourage energy, remove the incentive to new enterprise, encourage extravagant expenditures, and produce industrial stagnation . . ..2

The excess profits tax and surtaxes on individual incomes attracted particular scrutiny. Wilson’s penultimate Secretary of the Treasury, Carter Glass, argued that the excess profits tax “encourages wasteful expenditure, puts a premium on overcapitalization and a penalty on brains, energy, and enterprise, [and] discourages new ventures.”3 He further argued that the upper bracket surtaxes had “already passed the point of productivity”: [The] only consequence of any further increase would be to drive possessors of these great incomes more and more to place their wealth in the billions of dollars of wholly exempt securities heretofore issued by States and municipalities. . . . This process not only destroys a source of revenue to the Federal Government, but tends to withdraw the capital of very rich men from the development of new enterprises . . ..4

The crucial question was how officials would balance their desire to reduce taxes with their commitment to pay down the debt. 1. 1920 Treasury Annual Report, p. 25. 2. “Text of President Wilson’s Message to Congress,” New York Times, December 3, 1919, p. 6. 3. 1919 Treasury Annual Report, p. 23. 4. 1919 Treasury Annual Report, p. 24. See also Smiley and Keehn (1995, p. 287) (by 1920, “Both Democrats and Republicans believed that tax avoidance had reduced the revenue collected from the wealthiest Americans . . ..”).

Chapter 11

8

6

Billions of dollars

148

4

2

0 1920

1921

1922

1923

1924

1925

1926

1927

1928

1929

1930

Fiscal year Income and profit taxes

Other receipts

Figure 11.1 Treasury receipts

In a series of three revenue measures adopted between 1921 and 1926, Congress eliminated the excess profits tax and dramatically reduced surtaxes on individual incomes.5 Treasury receipts fell from a high water mark of $6.7 billion in fiscal year 1920 to $4.1 billion in fiscal 1922 in the wake of a severe postwar recession,6 but then leveled off—and even rose a bit between fiscal 1925 and fiscal 1927—as a result of rapidly expanding economic activity (figure 11.1). At the same time Congress was able to effect significant expenditure reductions (figure 11.2). Spending for the War and Navy departments plummeted from $11.3 billion in fiscal year 1919 to $1.7 billion in fiscal 1920 and continued to fall (at a more modest rate) until the middle of the decade. Spending in other categories also trended down during the first half of the decade. The result was a surplus in every fiscal year from 1920 to 1930 (figure 11.3). The surpluses underwrote a remarkable 37 percent reduction in Treasury indebtedness—to $16 billion by the end of the decade. 5. See, Blakey (1922, 1924, and 1926), Smiley and Keehn (1995), and Cannadine (2006, pp. 287–88 and 313–18). Murnane (2004) recounts the political history of the surtax reductions and places them in the context of the development of industrial capitalism. 6. 1925 Treasury Annual Report, p. 8. The National Bureau of Economic Research places the peak of the business cycle in January 1920 and the trough in July 1921.

Treasury Finance during the 1920s

8

Billions of dollars

6

4

2

0 1920

1921

1922

1923

1924

1925

1926

1927

1928

1929

1930

Fiscal year War and Navy Departments

Interest

Other

Figure 11.2 Treasury expenditures (other than for debt retirement)

8

6 Billions of dollars

149

4

2

0

1920

1921

1922

1923

1924

1925

1926

1927

1928

1929

Fiscal year Receipts

Expenditures (other than for debt retirement)

Figure 11.3 Treasury receipts and expenditures (other than for debt retirement)

1930

150

Chapter 11

This chapter provides an overview of federal tax policy during the 1920s and describes how Treasury officials funded the debt reduction program. Tax Policy Shortly after being sworn in as President on March 4, 1921, Warren Harding called a special session of Congress to revise the federal tax system. His agenda was well known; among other things, he wanted to eliminate the excess profits tax and he wanted to reduce surtaxes on individual incomes.7 In the two and a half years since the end of the war, businessmen and political leaders had become increasingly exasperated with the excess profits tax. They objected in particular to the substantial expertise required to comply with the tax and to seemingly capricious inequities in application. As well, the Treasury acknowledged that “Owing to the difficulty of determining the capital actually used to carry on any industry, it was impossible to apply the [tax] without very great hardship in many cases.”8 Businessmen, politicians, and economists also understood that high surtax rates were leading to tax avoidance. Thomas Adams, professor of economics at Yale University and a well-known authority on federal tax policy, pointed out that many observers believed that surtax rates were “excessive, that they practically force wealthier taxpayers to avoid or evade the tax, that their yield is shrinking rapidly, and that . . . free investment funds [have] been diverted from railway and industrial bonds to tax-free securities, with consequent repression of private enterprise . . ..”9 In a letter to the chairmen of the House Way and Means Committee and the Senate Finance Committee sent during the 1921 special session of Congress, Harding’s newly appointed Secretary of the Treasury, Andrew Mellon—a wealthy Pittsburgh financier and industrialist—reiterated the administration’s position that the excess profits tax should be eliminated.10 However, he appreciated that debt reduction required maintenance of adequate revenues and 7. “Harding Up Late, Finishing Message for Congress Today,” New York Times, April 12, 1921, p. 1, and “President’s Address to Congress on Domestic and Foreign Policies,” New York Times, April 13, 1921, p. 7. Cannadine (2006, p. 278) describes Harding’s agenda as a restoration of “the prewar climate of low taxes, balanced budgets, manageable national debt, limited government, and a functioning international economy backed by the gold standard.” 8. 1921 Treasury Annual Report, p. 22. Secretary Houston had earlier described the tax as “clogging the administrative machinery [of tax collection] and threatening . . . its possible breakdown.” 1920 Treasury Annual Report, p. 30. See also Adams (1921a, p. 371) (comment that the administration of the excess profits tax was “practically collapsing”). 9. Adams (1921b, p. 529). 10. “Drastic Revision of Federal Taxes Urged by Mellon,” New York Times, May 2, 1921, p. 1.

151

Treasury Finance during the 1920s

suggested raising the corporate income tax rate (from 10 percent to 15 percent) to offset the loss of revenue. Mellon also proposed that surtaxes on individual incomes be reduced to a maximum of 32 percent for the current year and 25 percent thereafter, but he argued that there was no need to offset lower surtax rates because lower rates would stimulate economic activity and produce greater total revenue.11 Congress agreed to eliminate the excess profits tax and to raise the corporate tax rate to 12½ percent when it adopted the Revenue Act of November 23, 1921, but it lowered the maximum surtax rate only to 50 percent (table 11.1). Table 11.1 Individual income tax assessments

Normal tax

Personal exemptions

Surtaxes Lowest bracket, net income Rate in lowest bracket Upper bracket rates: $100,000–150,000 150,000–200,000 200,000–300,000 300,000–500,000 500,000–1,000,000 Over $1,000,000

Revenue Act of 1918a

Revenue Act of 1921b

Revenue Act of 1924

Revenue Act of 1926

4% of first $4,000 of net income in excess of credits and exemptions, 8% of balance $1,000 individual, $2,000 married couple, $200 per dependent

4% of first $4,000 of net income in excess of credits and exemptions, 8% of balance

2% of first $4,000 of net income in excess of credits and exemptions, 4% of next $4,000, 6% of balance $1,000 individual, $2,500 married couple, $400 per dependent

1½% of first $4,000 of net income in excess of credits and exemptions, 3% of next $4,000, 5% of balance $1,500 individual, $3,500 married couple, $400 per dependent

$5,000–6,000

$6,000–10,000

$10,000–14,000

$10,000–14,000

1%

1%

1%

1%

52% 56 60 63 64 65

48% 49 50 50 50 50

37% 37 38 39 40 40

20% 20 20 20 20 20

$1,000 individual, $2,500 married couple ($2,000 if net income in excess of $5,000), $400 per dependent

Source: Revenue Act of 1918, Revenue Act of 1921, Revenue Act of 1924, and Revenue Act of 1926. a. Effective for 1919 tax year. b. Effective for 1922 tax year.

11. Blakely (1922, pp. 80–81).

152

Chapter 11

In late 1921 Treasury officials anticipated a small deficit in fiscal year 1922, to be followed by a larger deficit in fiscal 1923 as the new tax act began to take hold.12 However, an unexpectedly rapid reduction in expenditures produced a $300 million surplus in fiscal 1922. An unexpectedly strong recovery from the recession of 1920 to 1921 led to a similarly strong budget performance in the following fiscal year.13 Mellon was soon back before Congress, requesting another round of tax cuts. He argued that high upper bracket surtax rates had three adverse consequences: they stifled initiative, they diverted investment funds to tax-exempt state and local bonds, and they constricted federal tax revenues, and he restated his belief that more revenue would be forthcoming at lower rates.14 Congress agreed to lower surtax rates, but again declined to reduce the rates to the level proposed by the Secretary. The Revenue Act of June 2, 1924, reduced the maximum surtax rate to 40 percent (table 11.1). In signing the act, President Calvin Coolidge (who had succeeded Harding following Harding’s death on August 2, 1923) made clear that he would have preferred the program recommended by Mellon. He gave notice that he would “bend all [his] energies” to getting a more agreeable tax bill through the next Congress.15 Following an overwhelming Republican victory in the 1924 elections, Mellon went back to Capitol Hill and requested a third round of tax cuts. He finally achieved his goal of reducing surtax rates to no more than 25 percent when Congress adopted the Revenue Act of February 26, 1926 (table 11.1). Figure 11.4 illustrates the dramatic reduction in surtax rates that took place between the Revenue Act of 1918 and the Revenue Act of 1926.16 (The 12. 1921 Treasury Annual Report, p. 153. The principal provisions of the new tax act became effective in 1922 but would not begin to effect Treasury tax receipts until 1923, in the second half of fiscal year 1923 and the first half of fiscal year 1924. (Beginning with the Revenue Act of 1918, the tax on income earned in a given calendar year was due in four quarterly installments during the following year, on March 15, June 15, September 15, and December 15.) 13. 1922 Treasury Annual Report, p. 10, and 1923 Treasury Annual Report, p. 12. 14. Mellon (1924, pp. 13 and 16–18) and Blakey (1924, p. 483). See also Mellon’s letter, dated November 10, 1923, to William Green, acting chairman of the House Ways and Means Committee, reproduced in “Income Tax Cuts of $323,000,000 Urged by Mellon,” New York Times, November 12, 1923, p. 1. Murnane (2004, p. 837) gives a concise expression of the rationale for Mellon’s proposal: “First, surtax-rate reduction was not designed to reduce revenue nor to reduce income tax payments by the rich, but to secure more revenue in the long run, both by making avoidance less profitable and by expanding the economic pie . . .. Second, tax policy should not be a political (partisan) issue but a scientific (economic or business-efficiency) issue . . .. Third, the old income tax system, based on high surtaxes, was a product of war emergency, not intended for permanent use. The country needed an income tax for permanent peacetime use.” 15. “Text of the President’s Statement Criticizing Provisions of Tax Law,” New York Times, June 3, 1924, p. 1, and “Sees Good and Bad in Bill,” New York Times, June 3, 1924, p. 1. 16. Smiley and Keehn (1995) examine the consequences of the reduction in surtax rates for taxexempt interest rates, individual income tax filings, and aggregate tax receipts.

153

Treasury Finance during the 1920s

80

Percent

60

40

20

0 1

10

100

1,000

Net income, thousands of dollars Revenue Act of 1918

Revenue Act of 1926

Figure 11.4 Marginal surtax rates

Revenue Act of 1926 also raised the tax rate on corporate income to 13½ percent, but continued strong economic growth and robust tax receipts led Congress to reduce that rate back to 12 percent in the Revenue Act of May 29, 1928.17) Taxes and Debt Reduction The most striking characteristic of federal tax policy in the 1920s was the near-universal agreement that taxes should not be cut to levels that would impede debt reduction. In an address to the special session of Congress in 1921, President Harding announced a policy of “orderly refunding and gradual liquidation” of the debt.18 Three years later, in the course of arguing for a second round of surtax reductions, Mellon cautioned that “the Government must always be assured that taxes will not be so far reduced as to deprive the Treasury of sufficient revenue with which properly to run its business . . . and to take care of the debt.”19 Debt reduction, Mellon claimed, was “the best method of bringing about tax reduction. Aside from gradual refunding at lower 17. Blakey (1928, p. 428). 18. “President’s Address to Congress on Domestic and Foreign Policies,” New York Times, April 13, 1921, p. 7. Emphasis added. 19. Mellon (1924, pp. 19–20). Emphasis added.

154

Chapter 11

rates of interest, it is the only method of reducing the heavy annual interest charges.”20 Debt reduction required that the Treasury take in more than it disbursed on programs other than debt reduction, so Congress either had to moderate its spending in light of the tax receipts that it could expect or, alternatively, moderate its tax cuts in light of its spending. Figure 11.3 shows that Treasury receipts and Treasury expenditures both fell sharply in the early 1920s but that receipts never fell below expenditures. Mellon thus successfully paid down indebtedness even while reducing tax rates—an achievement that elicited widespread praise.21 Funding the Paydown The $25 billion of interest-bearing Treasury debt outstanding in mid-1919 included $20 billion in Liberty loans, $3.5 billion in short-term certificates of indebtedness, $1 billion in pre-war debt, and $1 billion in other debt—primarily war savings stamps and certificates. By the end of the decade, aggregate indebtedness had fallen to $16 billion, including $8 billion of Liberty loans and $1.4 billion of short-term debt. Table 11.2 and figure 11.5 show the course of the debt retirement program during the 1920s. The reduction in indebtedness that took place between mid-1919 and mid1930 was funded through three major channels:22 • $3.2 billion of sinking fund appropriations, • $1.5 billion of principal and interest payments on wartime loans to foreign governments, and • $3.5 billion in surplus receipts. 20. 1924 Treasury Annual Report, p. 26. Mellon reiterated his view of the linkage between debt reduction and tax reduction in the 1926 Treasury Annual Report (pp. 33–34): “As long as there are enormous fixed debt charges . . ., no large reduction in total expenditures is possible . . .. [T]he more rapidly the debt is retired, the sooner will come the time when these charges can be practically eliminated.” 21. Cannadine (2006, pp. 310 and 322) (by late 1923, “Mellon was increasingly regarded as the administration’s star figure; soon he was being acclaimed as ‘the greatest Secretary of the Treasury since Alexander Hamilton’” and “During these triumphant years [the mid-1920s], Mellon received many indications that his work had made him the most admired member of the Coolidge cabinet.”). See also “Mellon Hailed Here as Second Hamilton,” New York Times, December 5, 1924, p. 3. 22. 1930 Treasury Annual Report, pp. 27 and 56–63. There were several other, relatively minor, channels, including payment of federal estate taxes with Treasury securities in lieu of cash.

Treasury Finance during the 1920s

Table 11.2 Interest-bearing Treasury debt, mid-1919 to mid-1930 ($billions)

1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930

Pre-war debt

Liberty Loans

Short-term debt

Postwar notes and bonds

Other debt

Total debt

0.87 0.87 0.87 0.87 0.87 0.87 0.75 0.75 0.75 0.75 0.75 0.75

19.77 19.58 19.15 17.07 14.89 14.38 14.27 13.86 11.69 9.46 8.22 8.20

3.45 2.49 2.45 1.75 1.03 0.81 0.53 0.45 0.69 1.25 1.64 1.42

0.00 0.00 0.31 2.25 4.87 4.50 4.22 3.74 4.44 5.23 5.39 4.76

1.14 1.12 0.95 0.76 0.35 0.43 0.44 0.58 0.68 0.62 0.64 0.79

25.23 24.06 23.74 22.71 22.01 20.98 20.21 19.38 18.25 17.32 16.64 15.92

Source: Treasury Annual Reports.

25

20

Billions of dollars

155

15

10

5

0 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 Pre-war debt Short-term debt

Other debt

Liberty loans

Postwar notes and bonds

Figure 11.5 Composition of Treasury debt, mid-1919 to mid-1930

156

Chapter 11

The Sinking Fund The sinking fund was the most reliable vehicle for retiring Treasury debt. Section 6 of the Victory Liberty Loan Act established the fund for the retirement of Liberty notes and bonds:23 For the fiscal year beginning July 1, 1920, and for each fiscal year thereafter until all [Liberty notes and bonds] are retired there is hereby appropriated . . ., an amount equal to the sum of (1) 2½ per centum of the aggregate amount of such [notes and bonds] outstanding on July 1, 1920, less an amount equal to the par amount of any obligation of foreign Governments held by the United States on July 1, 1920, and (2) the interest which would have been payable during the fiscal year for which the appropriation is made on the [notes and bonds] purchased, redeemed, or paid out of the sinking fund during such year or in previous years.

The fund was designed to retire the war debt by 1945.24 The initial appropriation to the sinking fund, on July 1, 1920, was $253 million: 2½ percent of the difference between the par value of the Liberty loans outstanding on that date ($19.6 billion) and the par value of loans to foreign governments ($9.4 billion).25 (We explain below why Congress reduced sinking fund appropriations in proportion to the amount of loans to foreign governments.) Appropriations in subsequent years were greater by the amount of interest that would have been paid on securities previously retired; by 1928 this supplemental provision was adding more than $100 million annually to sinking fund appropriations.26 Table 11.3 shows the indebtedness retired during the 1920s by the operation of the sinking fund. Securities were acquired in secondary market purchases, through calls for early redemption, and by redemption at maturity.27 The sinking fund was important primarily because it created a psychological barrier to tax cuts materially deeper than those adopted between 1921 and 1928. Sinking fund appropriations entered the budget before the calculation of the surplus or deficit, so a tax cut that would have forced the financing of sinking fund purchases with sales of new debt would have been stigmatized as creating a deficit. The 1919 Treasury Annual Report argued that “To make the 23. The Act of March 2, 1923, added Treasury notes and bonds issued after July 1, 1920, to the menu of securities that could be acquired for the sinking fund. See also Hendricks (1933, pp. 216–18). 24. 1920 Treasury Annual Report, p. 114. 25. 1920 Treasury Annual Report, p. 113. 26. 1928 Treasury Annual Report, p. 29, and 1929 Treasury Annual Report, p. 16 27. Section 6 of the Victory Liberty Loan Act provided that the average cost of securities purchased for the sinking fund could not exceed par (exclusive of accrued interest). The 1925 Treasury Annual Report explained (p. 23) that “The Treasury is in the market for its securities when they are below par. When they exceed par . . . the fund is applied to the retirement of maturing or called securities.” However, Section 6 did not preclude the Treasury from purchasing securities at a premium—as long as it also purchased securities at a discount. The only restriction was that it could not pay more than par on average. See, “Cost of Liberty Bonds Purchased for Sinking Fund,” Wall Street Journal, October 19, 1922, p. 5.

157

Treasury Finance during the 1920s

Table 11.3 Treasury debt retired pursuant to the sinking fund ($millions) Fiscal year

Par amount of debt retired

1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 Total

261.3 275.9 284.0 296.0 306.3 317.1 333.5 354.7 370.3 388.4 3,187.5

Source: 1930 Treasury Annual Report, p. 593.

[sinking fund] effective, sinking-fund purchases must be met out of revenues received from taxation. Any thought [of] meeting its charges through the sale of securities would be . . . unwise in the extreme from the standpoint of the government’s finances and the ultimate wiping out of the war debt . . ..”28 Payments from Foreign Governments When Congress specified how the sinking fund was to be calculated, it excluded an amount of debt equal to the par value of loans to foreign governments. The exclusion was premised on the belief that the excluded amount would be covered by repayment of the foreign loans. The several Liberty Bond acts directed the Secretary of the Treasury to apply repayments of principal on foreign loans to the purchase and/or redemption of bonds authorized by the acts.29 (There was no comparable requirement that interest payments on foreign loans had to be used to retire Treasury debt.) The first column of table 11.4 shows the indebtedness retired during the 1920s as a result of repayments of foreign loans. Negotiations with foreign sovereign debtors during the mid-1920s led to a series of agreements that extended repayment schedules, reduced interest rates, and gave debtors the option of paying principal and/or interest with Treasury securities valued at par (plus accrued interest) in lieu of cash.30 28. 1919 Treasury Annual Report, p. 85. 29. 1919 Treasury Annual Report, p. 83, citing section 3 of the First Liberty Bond Act and section 3 of the Second Liberty Bond Act. 30. 1925 Treasury Annual Report, p. 23, and Cannadine (2006, pp. 289–91 and 318–20). See also “Tells How England Got Liberty Bonds,” New York Times, July 17, 1923, p. 24, and “British Payment Will Be in Cash,” Wall Street Journal, June 11, 1924, p. 8.

158

Chapter 11

Table 11.4 Treasury debt retired as a result of interest and principal payments on loans to foreign governments ($millions) Par amount of securities received as

Fiscal year

Par amount of securities purchased or redeemed with cash repayments of principal

Principal repayments

Interest payments

1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 Total

7.9 72.7 73.9 64.8 32.1 38.5 0.4 4.4 19.3 19.1 0.6 51.1 384.8

— — — — — 23.0 22.8 29.0 25.0 27.4 37.9 40.3 205.4

— — — — 68.8 87.9 136.0 136.3 135.0 135.3 137.7 69.5 906.4

Source: Treasury Annual Reports.

Box 11.1 Non-cash Treasury Receipts and Expenditures The receipts and expenditures shown in table 11.5 include non-cash as well as cash items. One non-cash item was closely related to debt retirement. As noted in the text, after 1922 foreign debtors had the option of making principal and interest payments with Treasury securities in lieu of cash. A payment made with securities was recorded by the Treasury as a receipt in the amount of the par value of the securities. (The payment would appear on the “Other receipts” line in table 11.5.) Since Congress provided that any securities received in payment of principal or interest on a foreign loan were to be retired, an expenditure was recorded simultaneously for the same amount. (The expenditure would appear on the “Other retirements” line of table 11.5.) Since receipts and expenditures expanded by the same amount, the payment and retirement had no effect on the budget surplus and thus no effect on the Treasury general fund.

Related legislation provided that any securities so received were to be retired by the Treasury. The second and third columns of table 11.4 show the indebtedness retired during the 1920s as a result of payments of principal and interest in the form of securities. Surplus Receipts Table 11.5 shows Treasury receipts and expenditures during the 1920s. With the important exception of certain payments on foreign loans (see box 11.1), most of the receipts were in the form of cash and were credited to the

159

Treasury Finance during the 1920s

Table 11.5 Treasury receipts and expenditures, fiscal year 1920 to fiscal year 1930 ($billions) 1920

1921

1922

1923

1924

Ordinary receipts Customs 0.323 0.309 0.356 0.562 0.546 Income and 3.945 3.206 2.068 1.679 1.842 profit taxes Other 2.427 2.110 1.685 1.766 1.624 receipts Total receipts 6.695 5.625 4.109 4.007 4.012 Expenditures chargeable against ordinary receipts War and navy 1.725 1.213 0.932 0.726 0.681 Interest 1.024 0.997 0.991 1.056 0.941 3.654 2.906 1.449 1.512 1.427 Other nonretirement Total 6.403 5.116 3.372 3.294 3.049 nonretirement Sinking fund — 0.261 0.276 0.284 0.296 retirements Other 0.079 0.161 0.147 0.119 0.162 retirements Total 0.079 0.422 0.423 0.403 0.458 retirements Total 6.482 5.538 3.795 3.697 3.507 expenditures Surplus 0.213 0.087 0.314 0.310 0.505

1925

1926

1927

1928

1929

1930

0.548 1.761

0.579 1.982

0.605 2.225

0.569 2.174

0.602 2.331

0.587 2.411

1.471

1.402

1.299

1.299

1.100

1.180

3.780

3.963

4.129

4.042

4.033

4.178

0.708 0.882 1.473

0.668 0.832 1.598

0.680 0.787 1.507

0.722 0.732 1.650

0.782 0.678 1.838

0.828 0.659 1.953

3.063

3.098

2.974

3.104

3.298

3.440

0.306

0.317

0.334

0.355

0.370

0.388

0.161

0.170

0.186

0.185

0.180

0.166

0.467

0.487

0.520

0.540

0.550

0.554

3.530

3.585

3.494

3.644

3.848

3.994

0.250

0.378

0.635

0.398

0.185

0.184

Source: Treasury Annual Reports.

Treasury’s general fund and most of the expenditures were disbursed from the general fund in the form of cash. The difference, or surplus, remained—in the first instance—in the general fund. As described in more detail in the following chapter, from time to time Treasury officials used accumulated general fund balances to redeem maturing debt. Thus the budget surplus constituted a third channel for funding debt retirement.

12

Paying down the War Debt

In mid-1919 total interest-bearing Treasury debt stood at $25.2 billion dollars. $3.5 billion of the debt was short-term certificates of indebtedness, $4.5 billion was Victory notes scheduled to mature in May 1923, and $4.0 billion was Third Liberty Loan bonds scheduled to mature in 1928. After that, nothing of any significance was due before 1938. By mid-1930 the Treasury had redeemed the Victory notes and the Third Liberty bonds, reduced its short-term indebtedness to $1.4 billion, and retired $4.6 billion of other debt (including—more than a decade early—all $3.6 billion of the Second Liberty bonds that were due in 1942); in total, a reduction of $14.1 billion. Over the same 11-year interval the Treasury issued $4.8 billion of notes and bonds that were still outstanding in mid-1930, making for a net reduction in indebtedness of $9.3 billion. This chapter examines how the Treasury applied the sinking fund appropriations, foreign debt payments, and budget surpluses described in chapter 11 to the redemption of the Victory notes and the Second and Third Liberty bonds and to the reduction in short-term debt. The process of retiring the three Liberty Loans was not simple. The Loans had been sold during the war as large, unitary issues and it was impractical to build Treasury cash balances to a size sufficient to pay off an entire issue in a single go. Treasury officials instead had to identify some combination of early retirements, cash redemptions at maturity, and refinancings that would be compatible with the quarterly flow of income tax receipts. Refinancing the Short-Term Debt: Mid-1921 to Mid-1923 During fiscal year 1920, short-term Treasury debt fell $1 billion, to $2.5 billion. In November 1920, Treasury Secretary David Houston confidently predicted a continuation of the decline in the coming year: “It is the Treasury’s expectation that each quarterly payment of income and profits taxes henceforth will witness an important progressive decline in the floating debt . . ..”1 However, the notion that short-term debt had been paid down with tax receipts was a fiction. Receipts exceeded expenditures (for purposes other than debt reduction) by less than $300 million in fiscal 1920.2 Most of the funding for the paydown had come from a one-time $866 million reduction in the Treasury’s general fund, from $1,226 million in mid-1919 to $360 million in mid1920.3 Houston’s prediction that the country could “look forward confidently 1. 1920 Treasury Annual Report, p. 8. 2. The 1920 budget surplus was $213 million and the Treasury expensed an additional $79 million in debt retirements—primarily purchases of Third Liberty bonds financed with repayments of foreign loans. 1921 Treasury Annual Report, p. 153. 3. 1920 Treasury Annual Report, pp. 258 and 582.

162

Chapter 12

to the retirement of the floating debt out of the taxes provided by existing law”4 proved well off the mark when that debt remained stalled at $2.5 billion through mid-1921. The failure to reduce short-term indebtedness raised the prospect that the Treasury would still be carrying a material amount of that debt in May 1923, when the Victory notes matured. Unless a significant portion of the notes could be redeemed with cash—an unlikely prospect, in mid-1921, in view of a postwar recession that was only then coming to an end—the Treasury was liable to find itself saddled with the task of refinancing more than $6 billion of maturing debt over a relatively short period of time. To head off a prospective crisis, Treasury officials in the newly installed Harding administration decided to refinance much of the short-term debt with intermediate-term notes that would mature during a 5-year interval beginning after the Victory notes had been redeemed and ending before the Third Liberty bonds came due in September 1928. In an April 1921 letter to the chairmen of the House Ways and Means Committee and the Senate Finance Committee, Secretary Mellon announced that: It will . . . be the Treasury’s policy to vary its monthly offerings of Treasury certificates of indebtedness from time to time when market conditions are favorable with issues of short-term notes in moderate amounts with maturities of from three to five years, with a view to gradual distribution of the short-dated debt through successive issues of notes in convenient maturities extending over the period from 1923 to 1928 . . .. This program will make the short-dated debt more manageable and facilitate the refunding operation which will be necessary in connection with the maturity of the Victory Liberty Loan.5

Mellon sold $2.6 billion of notes in seven cash subscription offerings between June 1921 and May 1923 (table 12.1). The sales funded a reduction in shortTable 12.1 Cash subscription sales of Treasury notes, mid-1921 to mid-1923 Issue date

Coupon (%)

Jun 15, 1921 Sep 15, 1921 Feb 1, 1922 Aug 1, 1922 Dec 15, 1922 Jan 15, 1923 May 15, 1923

5¾ 5½ 4¾ 4¼ 4½ 4½ 4¾

Maturity

Amount sold ($millions)

Jun 15, 1924 Sep 15, 1924 Mar 15, 1925 Sep 15, 1926 Jun 15, 1925 Dec 15, 1927 Mar 15, 1927 total:

311 391 402 345 423 362 382 2,616

Source: Treasury Annual Reports.

4. 1920 Treasury Annual Report, p. 12. 5. “Drastic Revision of Federal Taxes Urged by Mellon,” New York Times, May 2, 1921, p. 1.

163

Paying Down the War Debt

term debt to $1.0 billion in mid-1923 and prepared Treasury’s balance sheet for the redemption of the Victory notes. Note sales were not a regular feature of Treasury financing operations during the 1920s. The 1921 to 1923 sales were designed to address a specific problem and the Treasury subsequently sold notes for cash on only one occasion prior to May 1932 (when it began selling notes on a regular basis during the Great Contraction).6 Outside of that single exception, when Mellon needed to raise cash he sold certificates of indebtedness or—on a sporadic basis— bonds. (However, as described below, the Treasury did issue notes in exchange offerings for Victory notes and for Second and Third Liberty bonds.) Retiring the Victory Notes Table 12.2 shows how the Treasury retired the Victory notes. Almost $700 million of notes were retired through open market purchases for the sinking fund, but the bulk of the issue was retired through exchange offerings ($1.8 billion) and redemptions at or shortly before early call dates and final maturity ($1.6 billion). (Table 12.2 also shows that $450 million of notes were retired through purchases for the bond purchase fund. Those purchases are discussed in box 12.1.) The exchanges and redemptions were carried out in a series of operations conducted between January 1922 and May 1923. Four exchange and redemption operations led to the retirement of $1.5 billion of Victory notes in fiscal year 1922: Table 12.2 Victory Note retirements, fiscal year 1920 to fiscal year 1923 ($millions)

Outstanding, beginning of fiscal year Retirements Sinking fund Bond purchase fund Loans to foreign governments Budget surplus Exchangesa Call or maturity redemptions Other retirements Outstanding, end of fiscal year

1920

1921

1922

1923

4,495

4,246

3,914

1,991

261 70

258 131

170

249

689 450

634 1,187

1 3,914

1,143 371 20 1,991

1,777 1,558 21

4,246

Total

0

Source: Treasury Annual Reports. a. Exchanges shown in table 12.3.

6. The single cash offering of notes after mid-1923 came in September 1927, when the Treasury offered and sold $250 million of 5-year notes callable after three years.

164

Chapter 12

Box 12.1 Purchases of Liberty Loans for the Bond Purchase Fund Following the End of World War I Congress established the bond purchase fund in April 1918 at the request of Treasury Secretary McAdoo in an attempt to stabilize the prices of Liberty bonds near their “true” value.a Section 15 of the Third Liberty Bond Act authorized, but did not require, the Secretary of the Treasury to repurchase outstanding Liberty bonds, subject to the restriction that he could not repurchase during any calendar year more than 5 percent of the amount of a bond outstanding at the beginning of the year and could not pay an average price in excess of par. The bond purchase fund was hardly used during the war (only about $250 million of bonds were repurchased before November 1918b), but it was used aggressively for several years thereafter. As shown in the table below, almost $1.8 billion of Liberty bonds and Victory notes were purchased for the bond purchase fund up to the middle of 1920 and another $200 million of notes were purchased in fiscal year 1921 and fiscal 1922.c (The Treasury did not make any purchases for the bond purchase fund after mid-1922 and instead purchased securities for the sinking fund.) The purchases in fiscal year 1918 and fiscal 1919 did not affect aggregate Treasury indebtedness because the Treasury ran large budget deficits in those years. The purchases in fiscal 1920 had only a small effect on aggregate indebtedness because the Treasury recorded only a $213 million surplus in that year. The purchases had a more important effect on the outstanding amounts of individual Liberty bonds and Victory notes.

Purchases for the bond purchase fund, fiscal year 1918 to fiscal year 1922 (in $millions)

First Liberty bonds Second Liberty bonds Third Liberty bonds Fourth Liberty bonds Victory notes Total

1918

1919

1920

1 64 15

4 180 194 165

80

543

32 237 225 403 249 1,146

1921

70 70

1922

Total

131 131

37 481 434 568 450 1,970

a. See “Seeks Protection for Liberty Bonds,” New York Times, February 9, 1918, p. 2, and McAdoo’s testimony before the House Ways and Means Committee in March 1918 (Committee on Ways and Means, March 1918, p. 24). b. 1918 Treasury Annual Report, p. 71. c. 1920 Treasury Annual Report, p. 109, 1921 Treasury Annual Report, p. 257, and 1922 Treasury Annual Report, p. 142.

• In January 1922, the Treasury offered to issue up to $200 million of new 3-year notes in exchange for outstanding Victory notes.7 The offer was fully subscribed.8 • In February 1922, the Treasury announced that it was calling for early redemption on June 15, 1922, all $400 million of the outstanding tax-exempt 7. Treasury Circular no. 276, January 26, 1922, reprinted in 1922 Treasury Annual Report, p. 184. 8. 1922 Treasury Annual Report, p. 48.

165

Paying Down the War Debt

3¾ percent Victory notes.9 For the convenience of investors and to even out its workload, the Treasury also announced that it would redeem any of the called notes tendered prior to June 15 at par plus accrued interest to the tender date. Redemptions prior to June 15 retired $136 million Victory notes; redemptions on and shortly after the call date retired another $235 million notes.10 • Following the redemption of the tax-exempt 3¾ percent Victory notes, the Treasury made two exchange offers to holders of the taxable 4¾ percent notes. An exchange offer in March 1922 of new 4-year notes retired $618 million of the Victory notes; an exchange offer in June 1922 of new 3½-year notes retired an additional $325 million of the Victory notes.11 In fiscal year 1923 the Treasury continued to retire Victory notes with exchange offers and calls for early redemption: • Between July 1922 and May 1923, the Treasury extended five separate exchange offers to holders of Victory notes.12 Subscriptions on the offers— shown in table 12.3—retired $624 million of the notes.13 • On July 26, 1922, Treasury officials announced that they were calling for early redemption on December 15, 1922, about half of the $2 billion of Victory notes that remained outstanding.14 The Treasury also announced that it would redeem any of the called notes tendered prior to December 15 at par plus accrued interest to the tender date. Redemptions prior to December 15 retired 9. Treasury Circular no. 277, February 9, 1922, reprinted in 1922 Treasury Annual Report, p. 175. In the same circular the Treasury announced that, pursuant to a provision in the original offering of Victory notes (Treasury Circular no. 138, April 21, 1919, reprinted in 1919 Treasury Annual Report, p. 242), it was suspending the interconvertibility of tax-exempt 3¾ percent Victory notes and taxable 4¾ percent Victory notes. Thus after June 15, 1922, only taxable 4¾ percent Victory notes remained outstanding. 10. 1922 Treasury Annual Report, p. 142. 11. Treasury Circular no. 280, March 9, 1922, reprinted in 1922 Treasury Annual Report, p. 185, Treasury Circular no. 292, June 8, 1922, reprinted in 1922 Treasury Annual Report, p. 187, and 1922 Treasury Annual Report, pp. 48–49. Another $10 million of the June exchange was effected in fiscal year 1923. 1923 Treasury Annual Report, p. 167. 12. Treasury Circular no. 298, July 26, 1922, reprinted in 1922 Treasury Annual Report, p. 189; Treasury Circular no. 307, October 9, 1922, reprinted in 1922 Treasury Annual Report, p. 168; Treasury Circular no. 315, December 7, 1922, reprinted in 1923 Treasury Annual Report, p. 183; Treasury Circular no. 318, January 9, 1923, reprinted in 1923 Treasury Annual Report, p. 187; and Treasury Circular no. 323, May 7, 1923, reprinted in 1923 Treasury Annual Report, p. 188. 13. The Treasury also allowed subscribers to four series of certificates of indebtedness issued in December 1922 and March 1923 to pay for their subscriptions with Victory notes valued at par plus accrued interest to the dated dates of the new certificates. Treasury Circular no. 314, December 7, 1922, reprinted in 1923 Treasury Annual Report, p. 237, and Treasury Circular no. 321, March 8, 1923, reprinted in 1923 Treasury Annual Report, p. 238. A total of $5.5 million of notes were retired through this device. 1923 Treasury Annual Report, p. 167. 14. Treasury Circular no. 299, July 26, 1922, reprinted in 1922 Treasury Annual Report, p. 179. The Treasury called notes with letters A, B, C, D, E, or F prefixed to their serial numbers. It left uncalled notes with prefix letters G, H, I, J, K, or L.

166

Chapter 12

Table 12.3 Exchange offerings of new Treasury notes and bonds for outstanding Victory notes Issue date

Series

Coupon (%)

Maturity

Amount issued ($millions)

Exchanges in fiscal year 1922 Feb 1, 1922 Note Mar 15, 1922 Note Jun 15, 1922 Note

4¾ 4¾ 43⁄8

Mar 15, 1925 Mar 15, 1926 Dec 15, 1925 Total:

200 618 335a 1,153

Exchanges in fiscal year 1923 Aug 1, 1922 Note Oct 16, 1922 Bond Dec 15, 1922 Note Jan 15, 1923 Note May 15, 1923 Note

4¼ 4¼ 4½ 4½ 4¾

Sep 15, 1926 Oct 15, 1952 Jun 15, 1925 Dec 15, 1927 Mar 15, 1927 Total:

142 145 46 5 286 624

Source: Treasury Annual Reports. a. Includes $10 million of Victory notes received and retired in fiscal year 1923.

$74 million Victory notes; redemptions on the call date retired another $616 million notes.15 • The remaining Victory notes were retired with a $497 million redemption payment in May 1923. Of the $3.9 billion of Victory notes outstanding in mid-1921, the Treasury retired $2.1 billion (55 percent) through purchases in the secondary market, two partial calls for early redemption, and repayment at maturity. It refinanced an additional $1.8 billion through the exchange offerings shown in table 12.3. Statutory Authority By mid-1923 the Treasury had issued $4.2 billion of new notes and about $750 million of new bonds.16 All of the new securities relied, for statutory authority, on the wartime Liberty Loan acts. In the process Treasury officials broke new ground in debt management policy. 15. 1923 Treasury Annual Report, p. 168. 16. The bonds, the 4¼s of October 15, 1952, were sold in a combined cash and exchange offering in October 1922. Treasury Circular No. 307, October 9, 1922, reprinted in 1922 Treasury Annual Report, p. 168. In the cash offering the Treasury offered to sell $500 million of the bonds, received subscriptions for $1.4 billion, and sold $512 million. 1922 Treasury Annual Report, p. 5. The Treasury exchanged $145 million of the new bonds for 4¾ percent Victory notes and $107 million of the bonds for two series of certificates of indebtedness due to mature on December 15, 1922. 1922 Treasury Annual Report, p. 5, and 1923 Treasury Annual Report, p. 167.

167

Paying Down the War Debt

The Second Liberty Bond Act, as amended by the Third and Fourth Liberty Bond Acts, authorized the Secretary of the Treasury “to borrow, from time to time, on the credit of the United States for the purposes of this Act, and to meet expenditures authorized for the national security and defense and other public purposes authorized by law not exceeding in the aggregate $20,000,000,000, and to issue therefore bonds [emphasis added] of the United States . . ..” The Treasury had issued $14.9 billion of bonds in the Second, Third, and Fourth Liberty bond offerings—the $2 billion borrowed in the First Liberty bond offering was not included in the $20 billion limit—and had ample room to issue additional bonds. However, reliance on the broad language of the Second Liberty Bond Act: “to meet expenditures authorized for the national security and defense and other public purposes authorized by law,” was—at least by historical standards—somewhat aggressive. As noted in chapter 3, prior to the First World War Congress authorized bond issues for specific purposes. The broad language of the Liberty Bond acts, and the Treasury’s willingness to rely on that language, substantially insulated the Treasury from congressional oversight of bond sales in the 1920s. The sale of Treasury notes raised a different issue. The Victory Liberty Loan Act authorized the Secretary of the Treasury “to borrow from time to time on the credit of the United States for the purposes of this Act, and to meet expenditures authorized by law, not exceeding in the aggregate $7,000,000,000, and to issue therefore notes [emphasis added] of the United States . . ..” Since the Treasury had already issued $4.5 billion of Victory notes in 1919, the $7 billion limit in the Victory Liberty Loan Act did not leave it with much spare authority when it began (in June 1921) to issue what would grow to be an additional $4.2 billion of notes. The problem was solved with a 1921 amendment that changed the limit to $7.5 billion of notes outstanding at any one time.17 The new language allowed the Treasury to issue notes the same way it issued certificates of indebtedness.18 As a consequence of the broad language in the Liberty Bond Acts and the 1921 amendment limiting outstanding notes, Treasury debt management in the 1920s was subject to three statutory limitations: a limit of $10 billion on outstanding certificates of indebtedness, a limit of $7.5 billion on outstanding notes, and a limit of $20 billion on the amount of bonds that could be issued (of which the wartime Liberty bonds had already absorbed $14.9 billion). None of these limits constrained Treasury debt management actions prior to the onset of the Great Depression. 17. See section 1401 of the Revenue Act of 1921. 18. Statutory authority for certificates of indebtedness was noted in chapter 7.

168

Chapter 12

Paying off the Treasury Notes and Reducing the Short-Term Debt The $4.2 billion of notes issued between mid-1921 and mid-1923 served two distinct purposes: refinancing short-term debt until after the Victory notes had been retired (table 12.1) and refinancing the Victory notes to a variety of later maturity dates (table 12.3). The maturities of the new notes: quarterly dates from mid-June 1924 to mid-December 1927, were deliberately chosen to facilitate the Treasury’s policy of “orderly refunding and gradual liquidation” of the war debt.19 All of the new notes matured on the 15th of the third month of a calendar quarter—when most individuals and corporations made quarterly income tax payments. This was no accident and was part of a larger scheme designed to facilitate redemption of the notes and, more generally, reduction of the war debt.20 As a tax payment date approached, Treasury officials estimated the receipts they were about to receive and the funds they were likely to disburse during the coming quarter. They used balances in the general fund equal to the estimated excess of receipts over disbursements to redeem maturing debt and refinanced the remainder (of the maturing debt) to a subsequent tax date.21 Table 12.4 illustrates how the scheme worked. The first column shows the maturity dates of the notes and certificates of indebtedness that matured in 1926. The second column shows the principal amounts of securities originally 19. 1922 Treasury Annual Report, p. 9. 20. See, for example, 1922 Treasury Annual Report, p. 9 (“All of the outstanding issues of notes and certificates maturing within this period fall due on quarterly tax payment dates, and thus absorb any surplus revenues which may be available. This gives the best assurance of the gradual retirement of the war debt, and is perhaps the greatest advantage of the short-term refunding which the Treasury has been carrying on, for by distributing the debt over early maturities in amounts not too large to be financed each year these refunding operations have given the Treasury control over the debt and its retirement and avoided the tendency to perpetuation of the debt which would have been inherent in long-term refunding upon a comprehensive scale.”) and 1923 Treasury Annual Report, p. 20 (“Except for the issue of about $750,000,000 of 25- to 30-year Treasury bonds in the fall of 1922, the refunding has all been on a short-term basis, and it has been arranged with a view to distributing the early maturities of debt at convenient intervals over the period before the maturity of the Third Liberty loan in 1928 in such manner that surplus revenues may be applied most effectively to the gradual reduction of the debt. With this object in view all of the short-term notes issued in the course of the refunding have been given maturities on quarterly tax-payment dates, and all outstanding issues of Treasury certificates have likewise been reduced to tax maturities.”) 21. See, for example, the 1926 Treasury Annual Report, p. 35 (“A few weeks prior to the 15th of each September, December, March, and June the Treasury determines what income it will need to meet expenditures during the coming quarter, taking into account, on the receipt side, the cash in the general fund and the Government receipts to be expected and, on the expenditure side, the amount of cash required to meet obligations maturing during the quarter, and the probable expenses of the Government during the quarter.”) and p. 39 (“New issues of public debt securities in regular course are made only on tax-payment dates and the amount of the issue is determined by the estimated cash requirements of the Treasury to the next payment date in excess of the cash in hand and the estimated receipts from taxes and other sources of revenue.”).

169

Paying Down the War Debt

Table 12.4 Refinancings and paydowns on tax dates in 1926 ($millions) Date

Maturing

Mar 15, 1926

618

Refinanced

Paid down Note issued Mar 15, 1922, in exchange for Victory notes 3¾% bond to mature Mar 15, 1956

495 123 Jun 15, 1926

124

Certificate issued Jun 15, 1925, for cash Certificate issued Sep 15, 1925, for cash

252 376 Sep 15, 1926

345 142

Note issued Aug 1, 1922, for cash Note issued Aug 1, 1922, in exchange for Victory notes Certificate to mature Jun 15, 1927

379 108 Dec 15, 1926

453

Certificate issued Dec 15, 1925, for cash Certificate to mature Sep 15, 1927

229 224 Source: Treasury Annual Reports.

issued to mature on each of the four dates. (Three of the securities were certificates of indebtedness issued on tax dates in 1925, two were notes issued in 1922 in exchange for Victory notes, and one was a note issued for cash in 1922.) The third column shows the principal amount of securities issued on each date, including a 30-year bond in March and certificates of indebtedness in September and December. (The Treasury did not issue any new securities in June.) The last column shows the amount paid down each quarter.22 Figure 12.1 shows the aggregate amounts issued to mature and refinanced, respectively, on tax dates between March 15, 1924 and December 15, 1927. The Treasury paid off some maturing debt in every quarter but two. After June 1922, Treasury securities were almost always sold on tax dates and almost always matured on tax dates.23 The “regularization” of financing operations on quarterly tax dates was an important innovation in Treasury debt management because it greatly enhanced the predictability of Treasury operations and facilitated the integration of Treasury debt management with Treasury cash management. The 1923 Treasury Annual Report observed that the 22. Some debt was retired before it matured with purchases for the sinking fund. 23. There were only two occasions after 1922 when certificates of indebtedness were sold on other than a tax date: an issue of 213-day certificates sold in November 1927 to finance the redemption of Second Liberty bonds and an issue of 335-day certificates sold in October 1928 to finance the redemption of Third Liberty bonds.

Chapter 12

1,000

750

Millions of dollars

500

250

19

24 Q 24 I 19 QI 24 I 19 QII 24 I Q 19 IV 25 19 QI 25 19 QI 25 I 19 QII 25 I Q 19 IV 26 19 QI 26 19 QI 26 I 19 QII 26 I Q 19 IV 27 19 QI 27 19 QI 27 I 19 QII 27 I Q IV

0 19

170

Issued to mature

Refinanced

Figure 12.1 Treasury refinancings on tax payment dates

structure of maturities was designed “so as to permit . . . refinancing with a minimum disturbance to business and industry . . ..”24 However, the system was not flawless. The relatively infrequent (quarterly) borrowings meant that the Treasury sometimes had to borrow well in advance of actual requirements and inventory funds in commercial bank accounts until they were needed.25 This was expensive, because the Treasury’s commercial bank accounts earned only 2 percent interest—significantly less than the rates paid by the Treasury on its certificates of indebtedness.26 Additionally, as shown in figure 12.2, the maturities of new certificates varied from quarter to quarter and, as shown in figure 12.3, the sizes of new 24. 1923 Treasury Annual Report, p. 23. 25. Committee on Ways and Means (1929, p. 3) (testimony of Under Secretary of the Treasury Ogden Mills that “it is reasonably clear that if you are going to borrow only four times a year, you have got to borrow in advance of requirements”). During the 1920s the Treasury continued to use the War Loan Deposit Accounts that it had established at commercial banks during the war. See, 1920 Treasury Annual Report, p. 171. 26. “Review of Present Methods of Financing the United States Government and Suggested Supplementary Practices,” Federal Reserve Bank of New York memo dated January 25, 1929, Federal Reserve Bank of New York file 413.7 (“Funds have been borrowed by the Government in advance of actual requirements and the interest cost on such borrowing has exceeded the interest received on idle Government deposits. A closer adjustment of borrowings to expenditures would be more economical.”).

Paying Down the War Debt

360

Days

270

180

90

0 1922

1925

1928

1931

Figure 12.2 Term to maturity of new offerings of certificates of indebtedness. Solid balls are certificates issued on a tax payment date and maturing on a tax payment date. Open balls are certificates issued on a date other than a tax payment date and maturing on a tax payment date.

600

450

Millions of dollars

171

300

150

0 1922

1925

1928

1931

Figure 12.3 Issue size of new offerings of certificates of indebtedness. Solid balls are certificates issued on a tax payment date and maturing on a tax payment date. Open balls are certificates issued on a date other than a tax payment date and maturing on a tax payment date.

172

Chapter 12

offerings also varied significantly, depending on how much was maturing, the magnitude of anticipated tax receipts, and the magnitude of expected disbursements. Thus the predictability of two economically important aspects of a financing: amount and tenor, was somewhat limited. Finally, late receipt of tax payments sometimes led to undesirable volatility in the overnight credit markets. The 1925 Treasury Annual Report observed that, “Frequently payments [on maturing issues] exceed [tax] receipts on the tax day, making it necessary to borrow temporarily from the Federal reserve bank on a special securities of indebtedness [sic] in anticipation of the tax receipts which it takes several days to collect. This places reserve bank funds temporarily on the market and results in easier money rates. Rates tighten up again, however, when the loan is repaid, upon the collection of the tax checks.”27 To dampen these episodes of transient ease the Federal Reserve Banks sometimes sold participations in their loan to the Treasury to member banks. At other times member banks used the surplus credits to reduce temporarily their Reserve Bank borrowings.28 On balance, however, the program of regular tax date financings, coupled with the policy of refinancing maturing Liberty loans in manageable amounts to a variety of later tax dates, was an innovative solution to the problem of paying down large unitary issues with budget surpluses that became available gradually and on a quarterly basis. Retiring the Second and Third Liberty Bonds After redeeming the last of the Victory notes in May 1923, Treasury debt managers turned to the next problem: preparing for the maturity of the Third Liberty bonds on September 15, 1928.29 As shown in table 12.5, between mid27. 1925 Treasury Annual Report, p. 44. See also “Treasury Financing Leads Big Turnover,” New York Times, June 16, 1928, p. 30 (“The immediate result of [paying off the securities maturing in June 1928] was to place additional funds in the money market, and call loans ruled easy all day . . .. An easy tone also prevailed in time money. This condition of ease in the money market is expected to continue into next week. The banks’ reserves have been increased to a point that is expected to keep them in comfortable position until next Tuesday, the next settlement day at the Federal Reserve Bank. After that the return of income tax checks to the banks for collection is expected to cause a somewhat firmer range of money rates.”). Treasury borrowings from the Federal Reserve typically lasted for about five days. Committee on Ways and Means (1929, p. 9) (testimony of Under Secretary of the Treasury Ogden Mills). 28. “New Factor Enters Treasury Financing,” New York Times, December 15, 1929, p. N11, and “Big Flood of Funds Due Here Tomorrow,” New York Times, September 14, 1930, p. 41. Meltzer (2003, p. 203n. 106) reports that in July 1924 the Open Market Investment Committee (a forerunner of the Federal Open Market Committee) approved a proposal to sell and repurchase securities to reduce transient dips in money market rates during tax payment periods. Beckhart, Smith, and Brown (1932, p. 357) describe other methods of draining reserves. 29. As early as 1920 Secretary Houston had opined that “after the Victory loan has been provided for, sound fiscal policy would require that the sinking fund and surplus revenues then be applied to the bonds of the third Liberty loan . . ..” 1920 Treasury Annual Report, p. 25.

173

Paying Down the War Debt

Table 12.5 Third Liberty bond retirements, fiscal year 1920 to fiscal year 1929 ($millions)

Outstanding, start of fiscal year Retirements Sinking fund Bond purchase fund Loans to foreign governments Budget surplus Exchangesa Call or maturity redemptions Other retirements Outstanding, end of fiscal year

1920

1921

1922

1923

1924

1925

1926

1927

1928

1929

3,959

3,663

3,612

3,474

3,408

2,997

2,885

2,488

2,148

1,229

17

33

238

18

317

273

65

32

39

365

225

70

44

80

62

93

3,663

56

2

6

3,612

3,474

3,408

2,997

1,262 225

250

128

7

Total

312 607

5 2,885

2,488

2,148

1,229

582 108 752

808 752

4

80

0

Source: Treasury Annual Reports. a. Exchanges: December 1924 offer to exchange new 29-year bonds for Third Liberty bonds; $93 million Third Liberty bonds retired in fiscal year 1925. January 1928 offer to exchange new 5-year notes for Third Liberty bonds; $607 million Third Liberty bonds retired in fiscal year 1929. July 1928 offer to exchange new 14-year bonds for Third Liberty bonds; $108 million Third Liberty bonds retired in fiscal year 1929.

1923 and mid-1927 they allocated $846 million of sinking fund appropriations and $270 million of budget surpluses to repurchasing the bonds.30 In December 1924 they additionally offered new 30-year bonds in exchange for Third Liberty bonds; investors exchanged $93 million of bonds pursuant to the offer.31 By early 1927 the Treasury had reduced the outstanding amount of Third Liberty bonds to $2.2 billion.32 Other things being equal, Treasury officials might have continued to whittle down the Third Liberty bonds during the eighteen months that remained before 30. See also “Bond Retirement by U.S. Is Retarded,” New York Times, September 8, 1924, p. 23 (stating that “Within four years the United States Government will have to complete its plans for retiring the third 4¼ per cent. Liberty bond issue, and already the Treasury Department is understood to have figured out a program to be followed. The third 4¼s are the first of the large war loans to mature and are non-callable until maturity, leaving no opportunity through early redemption to scale down the amount. As a result the Government has concentrated a considerable part of its sinking fund purchases on the third 4¼s and has gone into the open market for these bonds.”). 31. Treasury Circular no. 349, December 3, 1924, reprinted in 1925 Treasury Annual Report, p. 243, and 1925 Treasury Annual Report, p. 203. 32. 1927 Treasury Annual Report, p. 37.

Chapter 12

7.00

6.00

5.00

Percent

174

4.00

3.00

2.00

1.00 1919

1921

1923

1925

Certificates of Indebtedness

1927

1929

1931

Bonds

Figure 12.4 Yields on Treasury securities

the bonds matured, just as they had whittled down the Victory notes prior to maturity. Under Secretary of the Treasury Garrard Winston indicated, in a speech in December 1926, that he intended to do just that.33 However, the Third Liberties had competition for Treasury’s attention. The Second Liberty bonds did not mature until 1942 but the call protection on the bonds was due to expire on November 15, 1927. Although the Second Liberties were originally issued with a 4 percent coupon, virtually the entire issue had been converted to a 4¼ percent coupon.34 The $3 billion of outstanding Second Liberties thus bore the same coupon rate as the Third Liberties but could be redeemed almost a year before the latter issue matured.35 Since Treasury yields were below 4¼ percent in early 1927 (see figure 12.4), the longer term 33. “Sees National Debt Reduced 12 Billion,” New York Times, December 3, 1926, p. 38, and “How War Debt is Being Handled,” Wall Street Journal, December 3, 1926, p. 11. Treasury officials had observed as early as 1924 that “future financing would be simplified . . . to the extent [that the outstanding amount of Third Liberty bonds] could be whittled down.” 1925 Treasury Annual Report, pp. 30–31. 34. The original conversion option expired on November 9, 1918. Section 5 of the Victory Liberty Loan Act of 1919 granted the Secretary of the Treasury the authority to renew the option and Secretary Glass did so in March 1919 (Treasury Circular no. 137, March 7, 1919, reprinted in 1919 Treasury Annual Report, p. 348). Secretary Mellon terminated the option effective June 1925 (Treasury Circular no. 351, December 29, 1924, reprinted in 1925 Treasury Annual Report, p. 250), by which time more than 97 percent of the Second Liberty 4s had been converted to Second Liberty 4¼s. 1925 Treasury Annual Report, pp. 35–36. 35. The $3 billion is from “Mellon to Refund Liberty 4 1-4 Bonds,” New York Times, March 8, 1927, p. 1.

175

Paying Down the War Debt

Second Liberty bonds were a more attractive redemption target than the Third Liberties. However, officials had to take care that exercising the option to redeem the Second Liberties would not jeopardize the obligation to redeem the Third Liberties. Two events in early 1927 caused the Treasury’s attention to shift to the Second Liberty bonds. On March 4, Ogden Mills, a politically astute and personally ambitious former Congressman from New York, was sworn in as Winston’s replacement as Under Secretary of the Treasury.36 At almost the same time, it became clear that the budget surplus for fiscal 1927 was likely to set a record.37 (It ultimately exceeded $600 million.) Mills moved quickly to put his personal stamp on Treasury debt management, throwing over Winston’s step-by-step approach and testing whether, given the unexpectedly strong tax receipts, the Treasury might be able to retire the Second Liberties before the Third Liberties.38 On March 8, 1927, the Treasury offered to exchange 3½ percent, 5-year notes maturing on March 15, 1932, for Second Liberty bonds.39 To enhance the Treasury’s flexibility, the new notes could be called for early redemption as early as March 15, 1930, so they could be retired any time during a 2-year interval that began after the 1928 maturity of the Third Liberty bonds and ended before the first call dates for the First Liberty bonds (June 15, 1932) and the Fourth Liberty bonds (October 15, 1933). The settlement date of the exchange was set for March 15, but—as an inducement to investors who might be reluctant to exchange an issue with a 4¼ percent coupon for a new issue with a 3½ percent coupon—the Treasury agreed to pay interest on the Second Liberties to the next, May 15, coupon payment date; that is, it agreed to pay an extra two months interest.40 The exchange offer attracted a remarkable $1.4 billion of Second Liberties and reduced the outstanding amount of the issue to $1.7 billion.41 36. “Mills Take Oath in Treasury Post,” New York Times, March 5, 1927, p. 2. See, Cannadine (2006, p. 355) (Mills “was every inch a Republican politician” and “clearly had his sight set on higher office”) and “Red Year’s End,” Time, July 13, 1931. 37. “Payments of Income Tax May Reach $600,000,000; Big Surplus Indicated,” New York Times, March 16, 1927, p. 1. 38. “Mills’ Debt Policy One of Expediency,” Wall Street Journal, March 10, 1927, p. 10 (stating that Mills “very promptly made it clear that he has assumed actual control of the public debt policies of the government, the chief activity of his Treasury office” and that the decision to proceed with the refunding of the Second Liberty bonds “furnishes an example of the sharp break from the policies for handling the maturing Libertys which had been outlined by Garrard B. Winston, of Chicago, whom Mr. Mills succeeds. The major financing operation undertaken by Mills at the outset of his career at the Treasury is significant in that it definitely indicates his intention to govern his actions by conditions as they arise rather than under any previously formulated program.”). 39. Treasury Circular no. 379, March 8, 1927, reprinted in 1927 Treasury Annual Report, p. 274. 40. 1927 Treasury Annual Report, p. 38. 41. 1927 Treasury Annual Report, pp. 38 and 276.

176

Chapter 12

Box 12.2 Calling the Second Liberty Bonds for Early Redemption Treasury officials were well aware, in the spring of 1927, that a significant fraction of the Second Liberty Loan was held in small lots by a large number of relatively unsophisticated investors.a When they decided to call the Second Liberties for early redemption on November 15, they knew they would have to make unusual efforts to get the news out. The 1927 Treasury Annual Report recounted their efforts: Because of the intensive nation-wide campaigns conducted when the Liberty loans were issued, at which time every available facility was used to reach the public and secure subscriptions, which resulted in unparalleled widespread distribution of the bonds, the Treasury recognized an obligation to the holders of second Liberty loan bonds to make every effort through the use of every available facility to notify them that their bonds were called for redemption. The press, as usual, responded and carried the announcement widely as a matter of public concern. Banks and trust companies throughout the country were asked to cooperate and generously gave their assistance. The cooperation of the Postal Service was whole-heartedly given. Placards setting forth the call were displayed in practically every banking office and post office throughout the United States. The announcement in the form of an advertisement was placed in all daily, weekly, and semiweekly general newspapers throughout the United States which could be reached. For the first time the radio was used by the Treasury Department as a means of reaching millions of bond holders, the announcement of the call being broadcast through the courtesy of the National Broadcasting Co., its entire facilities being placed at the disposal of the Treasury, covering the country as far as Kansas City. Simultaneously similar broadcasts were made from Denver and from San Francisco by Treasury representatives.b a. Although the response to the March 1927 exchange offer had been unprecedented, an even greater response had been expected. The shortfall was attributed to “a surprising lack of knowledge of the exchange offer on the part of the public. “Bond Market Waits for Treasury Plan,” New York Times, March 27, 1927, p. E14. $59 million of coupon-bearing Second Liberties with a denomination of $50 were outstanding in March 1927; only $1.7 million were offered in exchange for 5-year notes. Nevertheless, $1.026 billion out of $1.366 billion Second Liberties with a denomination of $10,000 were offered in exchange. 1927 Treasury Annual Report, p. 301. b. 1927 Treasury Annual Report, p. 38.

The sharp reduction in Second Liberties gave Treasury officials confidence that they could call the remaining Second Liberties for early redemption in November 1927 and still be able to redeem the Third Liberties in September 1928.42 They announced the call on May 9, 1927.43 In view of the wide distribution of the bonds, officials went to unusual lengths to appraise investors of the early redemption. Box 12.2 describes some of the novel aspects of the call. To further reduce the funds that would be required to finally redeem the Second Liberty bonds, the Treasury undertook two additional exchange offers in 1927 (table 12.6). A June offer to issue new 33⁄8 percent 20-year bonds in 42. 1927 Treasury Annual Report, p. 38 (“The response to [the] exchange offer … made certain a successful refunding of the entire loan.”). 43. Treasury Circular no. 381, May 9, 1927, reprinted in 1927 Treasury Annual Report, p. 278.

177

Paying Down the War Debt

Table 12.6 Exchange offerings of new Treasury notes and bonds for outstanding Second Liberty bonds Issue date

Series

Coupon (%)

Mar 15, 1927 Jun 15, 1927 Sep 15, 1927

Note Bond Note

3½ 33⁄8 3½

Maturity

Amount issued ($millions)

Mar 15, 1932 Jun 15, 1947 Sep 15, 1932 Total:

1,361 245a 369 1,975

Source: Treasury Annual Reports. a. $218 million were issued in fiscal year 1927 and $27 million issued in fiscal year 1928.

exchange for Second Liberties retired $245 million of the latter bonds.44 The Treasury did not offer to pay any extra accrued interest on the Second Liberties but the terms were intended to be attractive: a concurrent cash offering of $200 million of the same 20-year bonds at a premium price of 100½ attracted over $600 million of subscriptions.45 In September 1927 the Treasury offered to exchange new 3½ percent 5-year notes (again callable in three years) for Second Liberty bonds.46 Holders of the Second Liberties had to pay a 1⁄8 percent (of principle) cash premium for the new notes, but were again offered an extra two months of accrued interest on their Second Liberties, from the September 15 settlement date of the exchange through November 15. The June and September 1927 exchange offers left less than $1 billion of Second Liberties to be redeemed on November 15.47 As it had done with the Victory notes, the Treasury announced that, for the convenience of investors and to even out its own workload, it would redeem Second Liberties in the weeks prior to November 15 at a stipulated price plus accrued interest to the date of tender.48 Between mid-October and midNovember it purchased $94 million of bonds.49 The balance of the bonds were 44. Treasury Circular no. 383, May 31, 1927, reprinted in 1927 Treasury Annual Report, p. 281, and 1927 Treasury Annual Report, p. 285. 45. 1927 Treasury Annual Report, pp. 284. 46. Treasury Circular no. 387, September 6, 1927, reprinted in 1927 Treasury Annual Report, p. 290. 47. $1.3 billion of Second Liberty bonds were outstanding in mid-1927 (1927 Treasury Annual Report, pp. 505 and 506) and $369 million of those bonds were tendered in exchange for 5-year notes in the September exchange offering (table 12.6). 48. Between October 17 and October 22, the Treasury offered to pay par and 3/32nds (plus accrued interest), between October 24 and October 29 it offered par and 2/32nds, between October 31 and November 7 it offered par and 1/32nd, and thereafter offered to pay par. Treasury press releases dated October 17, 24, and 31, 1927, reprinted in 1927 Treasury Annual Report, pp. 295–96. 49. 1927 Treasury Annual Report, p. 40.

178

Chapter 12

Table 12.7 Second Liberty bond retirements, fiscal year 1920 to fiscal year 1928 ($millions)

Outstanding, start of fiscal year Retirements Sinking fund Bond purchase fund Loans to foreign governments Budget surplus Exchangesa Call or maturity redemptions Other retirements Outstanding, end of fiscal year

1920

1921

1922

1923

1924

1925

1926

1927

1928

3,567

3,325

3,317

3,311

3,199

3,105

3,105

3,105

1,306

41

334

237 2

69

92

7 3,317

6 3,311

2 3,199

3 3,105

376 237 163

219 1,579

5 3,325

Total

3,105

3,105

1,306

26 396 547

245 1,975 547

2 0

25

Source: Treasury Annual Reports. a. Exchanges: March 1927 offer to exchange new 5-year notes for Second Liberty bonds; $1,361 million Second Liberty bonds retired in fiscal year 1927. June 1927 offer to exchange new 20-year bonds for Second Liberty bonds; $218 million Second Liberty bonds retired in fiscal year 1927; $27 million retired in fiscal year 1928. September 1927 offer to exchange new 5-year notes for Second Liberty bonds; $369 million Second Liberty bonds retired in fiscal year 1928.

paid as they were presented for redemption, funded in part by a cash offering of $400 million of seven-month certificates of indebtedness in mid-November 1927.50 Table 12.7 summarizes how the Second Liberty bonds were paid down and refinanced. About 40 percent of the $3.2 billion of bonds outstanding in the middle of 1923 was repurchased or redeemed for cash and the balance was refinanced in exchange offerings After retiring the Second Liberty bonds, the Treasury returned to the task of paying off the Third Liberties. About $2.1 billion of the bonds remained outstanding at the end of calendar year 1927, “too great an amount to let run to maturity” on September 15, 1928.51 To reduce its cash requirements, the Treasury offered to exchange new securities for the maturing bonds. As shown in table 12.8, in January 1928 it retired $607 million of Third Liberties in exchange for an equal amount of a third issue of 3½ percent 5-year notes (once again callable in three years).52 To overcome investor reluctance to accept a lower coupon, the Treasury continued the practice of paying accrued interest 50. Treasury Circular no. 389, November 7, 1927 (offering series TJ-1928, 31⁄8 percent certificates, dated November 15, 1927, to mature June 15, 1928), reprinted in 1928 Treasury Annual Report, p. 245. 51. 1928 Treasury Annual Report, p. 25. 52. Treasury Circular no. 392, January 9, 1928, reprinted in 1928 Treasury Annual Report, p. 251, and 1928 Treasury Annual Report, p. 254.

179

Paying Down the War Debt

Table 12.8 Exchange offerings of new Treasury notes and bonds for outstanding Third Liberty bonds Issue date

Series

Coupon (%)

Dec 15, 1924 Jan 16, 1928 Jul 16, 1928

Bond Note Bond

4 3½ 33⁄8

Maturity

Amount issued ($millions)

Dec 15, 1954 Dec 15, 1932 Jun 15, 1943 Total:

93 607 108 808

Source: Treasury Annual Reports.

on the Liberty bonds for two months following the settlement date of the exchange. In July 1928 the Treasury retired another $108 million of Third Liberties in exchange for an equal amount of new 33⁄8 percent 15-year bonds, again paying an extra two months of accrued interest on the Second Liberties.53 The remaining Third Liberty bonds were repurchased in secondary market transactions or retired as they were presented for redemption, funded in part by another cash offering of certificates of indebtedness—this time $300 million of eleven-month certificates—in mid-October 1928.54 Summary Paying down the war debt in the 1920s did not present Treasury debt managers with challenges remotely comparable to those faced by McAdoo during the First World War. Nevertheless, Mellon and his staff won deserved praise for the breadth of their success. Tilford Gaines, in his path-breaking study of Treasury debt management, concluded that “the decade of the 1920’s witnessed what is probably the most effective execution of debt management policy, in a technical sense, in the history of the country.”55 Gaines pointed out that, 53. Treasury Circular no. 405, July 5, 1928, reprinted in 1928 Treasury Annual Report, p. 263, and 1928 Treasury Annual Report, pp. 267–68. 54. Treasury Circular no. 410, October 8, 1928 (offering series TS-1929, 4¾ percent certificates, dated October 15, 1928, to mature September 15, 1929), reprinted in 1928 Treasury Annual Report, p. 273. The certificate offering was delayed a month after the maturity of the Third Liberty bonds because Treasury officials had found that “in the case of the long term war issues, which were widely distributed, maturing bonds are not all presented on the maturity date, but a large number are presented for redemption over a considerable period of time.” In the case of the Third Liberties, $955 million of the bonds was outstanding on September 14, 1928, but only $475 million was tendered on the following day and only $733 million was presented prior to September 26. “[T]o avoid borrowing in excess of actual needs and to save unnecessary interest charges,” the Treasury delayed, until mid-October, borrowing some of the funds it knew it would ultimately need to complete the redemption of the Third Liberty bonds. 1928 Treasury Annual Report, p. 27. 55. Gaines (1962, p. 27).

180

Chapter 12

Policy actions of the 1920s were directed toward specific, clearly-stated objectives. Maturing securities were redeemed if funds were available in the sinking fund or from surplus; if not, they were refunded to a carefully selected niche in the debt structure where, when they matured, funds might be expected to be available to redeem them.56

and that, At a purely technical level, the job done by Secretary Mellon during the 1920’s was superb. Each operation, whether intra-year certificate financing or refunding the Liberty and Victory loans, was carefully planned and conducted through a series of steps that at no time overstrained the market’s absorptive capacity. The program was orderly, with ample advance notice to the market before each step, and predictable, in the sense that the Secretary’s program and intentions were clearly understood.57

Mellon and his colleagues addressed first the problem of refinancing the postwar overhang of short-term debt and then put in place the critically important program of regular tax date financings that became the backbone of Treasury debt operations in the 1920s. Between mid-1923 and early 1927 they whittled down the Third Liberty bonds as sinking fund appropriations became available, but when tax receipts exploded in early 1927, they had the mental agility to make a midcourse correction, switching their attention to the Second Liberty bonds and calling those bonds for early redemption. They also made increasingly sophisticated use of intermediate-term securities in the second half of the decade, three times offering 5-year notes callable in three years.58 As discussed in the appendix to this chapter, the only criticism that might be lodged against Treasury debt managers in the 1920s was their failure to make clear the circumstances under which they were prepared to issue longterm bonds. Appendix: The “Mellons” The Treasury did not issue bonds frequently or regularly during the 1920s, but it did issue bonds on an occasional basis throughout the decade. In addition to the four exchange offerings noted in tables 12.3, 12.6, and 12.8, Treasury officials sold bonds in fixed price cash subscription offerings on six occasions between 1922 and 1928. Four of the offerings were joint with exchange offerings of the same bond. Table 12.9 summarizes the cash and exchange offerings. The five bonds issued in the six offerings were commonly and collectively known as “Mellons.” Treasury officials never made clear why they sometimes chose to raise cash by selling bonds (instead of certificates of indebtedness). Available evidence 56. Gaines (1962, p. 29). 57. Gaines (1962, p. 34). 58. The three callable 5-year notes were the 3½s of March 1932 (table 12.6), the 3½s of September 1932 (table 12.6), and the 3½s of December 1932 (table 12.8). All three notes were subsequently called for early redemption.

$200 $1,461 $225 Dec 15, 1924 Third Liberty bonds and other securities maturing in Mar 1925 $532 $757

$500 $1,400 $512 Oct 16, 1922

Victory Liberty notes and other securities maturing in Dec 1922 $252 $764

N/A $495

$500 $647 $495 Mar 15, 1926

$450a $382 $290 Mar 16, 1925

N/A $290

Mar 15, 1926 Mar 15, 1956 Mar 15, 1946

3¾ 30 (callable in 20)

Mar 8, 1926

4 29¾ (callable in 19¾) reopening Dec 15, 1924 Dec 15, 1954 Dec 15, 1944

Mar 5, 1925

a. Joint offering with 9-month certificates of indebtedness, total of $450 million in bonds and certificates offered.

Exchanged (millions) Total amount issued (millions)

Dec 15, 1924 Dec 15, 1954 Dec 15, 1944

Oct 16, 1922 Oct 15, 1952 Oct 15, 1947

Dated date Maturity First call Cash offering Offered (millions) Subscribed (millions) Allotted (millions) Issue date Exchange offering In exchange for

4 30 (callable in 20)

4¼ 30 (callable in 25)

Coupon (%) Term, years

Dec 3, 1924

Oct 9, 1922

Announcement

Table 12.9 The Mellons

$245 $495

Second Liberty bonds

$200 $618 $250 Jun 15, 1927

$108 $360

Third Liberty bonds

$250 $743 $252 Aug 1, 1928

Jul 16, 1928 Jun 15, 1943 Jun 15, 1940

33⁄8 15 (callable in 12)

33⁄8 20 (callable in 16) Jun 15, 1927 Jun 15, 1947 Jun 15, 1943

Jul 5, 1928

May 31, 1927

181 Paying Down the War Debt

182

Chapter 12

suggests that the first two offerings were consistent with contemporaneous Treasury operations but that the four subsequent offerings were not part of any long-term strategic plan; the decisions to sell the bonds were made at the last minute and the offerings came as a surprise to market participants. The first bond offering, in October 1922, was the most conventional. By the fall of 1922 the Treasury was well along on its twin programs of funding the short-term debt with notes and exchanging new notes for Victory notes, and it made sense to further extend the term structure of Treasury debt by offering 30-year bonds for cash and in exchange for Victory notes. In a letter announcing the offerings, Mellon noted that “the time has come for a longer term operation.”59 The Wall Street Journal reported that the bond, the first in four years, was “expected to prove particularly attractive” to those “who require a long-term investment.”60 The second bond offering, the 1924 cash and exchange offering of a 30-year bond in the regular December tax date financing, came as no surprise to market participants.61 The New York Times remarked that Treasury officials decided to offer long-term bonds “because the money market seemed favorable for the successful floatation of a long term issue at a relatively low rate of interest.”62 One dealer later noted that “the time appeared appropriate for the sale of a long-term government bond.”63 Each of the four subsequent bond offerings surprised market participants—a reaction that suggests the market did not fully appreciate the circumstances under which the Treasury was prepared to issue long-term debt. • In March 1925, market participants were described as “confidant that the new financing will take the form of a short-term issue.”64 Instead, the Treasury reopened the 30-year bond first offered three months earlier, largely because of technical conditions in the credit markets. The New York Times reported that the reopening “was due to a combination of influences, which proved an unexpected boon. In the last two weeks . . . there has been a dearth of taxexempt issues with which to meet investment requirements of insurance companies, savings banks, trust companies and individuals. The comparatively 59. Quoted in “$500,000,000 Issue of Long-Term Bonds Offered by Mellon,” New York Times, October 9, 1922, p. 1. 60. “Money Tendency Upward Says Reserve Board,” Wall Street Journal, October 10, 1922, p. 10. 61. “Mellon Silent on Plans,” New York Times, December 2, 1924, p. 4 (“Treasury plans for the Dec. 15 financing are approaching completion and an announcement may be expected within a few days disclosing whether the Treasury will offer a long-term bond issue at that time. There have been reports in the last few days that no short-term issue would be floated in the December operations . . ..”). 62. “Treasury to Offer 30-Year 4% Bonds,” New York Times, December 3, 1924, p. 1. 63. “Expect Long-Term Treasury Issue,” Wall Street Journal, February 27, 1926, p. 8 (quoting a report from C.F. Childs). 64. “Expect Treasury to Issue Short Term 3½s,” Wall Street Journal, March 3, 1925, p. 8.

183

Paying Down the War Debt

small total of municipal issues has been rapidly snapped up. Consequently the United States Government offering which was set before the public yesterday was being rapidly absorbed.”65 • In February 1926, bankers were reported to be expecting an issue of 15- or 18-month notes in the March tax date financing.66 However, a concurrent contraction in stock prices led Treasury officials to conclude that an offering of Treasury bonds would be well received and they put forth a cash offering of new 30-year bonds.67 • The market was surprised again in the June 1927 tax date financing. Participants had expected short-term securities, but the Treasury instead announced a combined cash and exchange offering of 20-year bonds.68 • The market was surprised yet again in July 1928, when the Treasury announced a combined cash and exchange offering of 15-year bonds in the midst of rising interest rates. The New York Times reported that “the offering of long-term bonds by the Treasury at this time came as a considerable surprise in Wall Street . . ..”69 Unlike Treasury certificate and note issuance policies, Treasury policies for issuing long-term debt in the 1920s remain unclear. The first two bond issues were consistent with contemporaneous debt management operations, the second two issues appear to have more opportunistic, and the last two came with virtually no explanation. 65. “Treasury Bonds Sell at Premium,” New York Times, March 6, 1925, p. 29. 66. “Government to Issue $500,000,000 Million Bonds,” New York Times, February 21, 1926, p. E12 (“Bankers expect the new financing to take the form of notes falling due in fifteen or eighteen months, as the Treasury has no short-term obligations maturing in June or September of 1927, and new financing usually is arranged with this factor in mind.”). See also “More Liberty Bonds Near Retirement,” New York Times, February 24, 1926, p. 26 (“The general belief is that the securities to be offered will mature in from fifteen to eighteen months.”). 67. “Bond Issue at 3 3-4 Offered by Mellon,” New York Times, March 8, 1926, p. 1 (“The situation on the Stock Exchange is said to have played no small part in the decision reached. It is felt that the check given to the speculative boom and the losses sustained by some operators and their followers, will slow up activities on the exchanges, release large amounts of money for other forms of investment and make the purchase of treasury bonds . . . appear much more attractive than when the stock boom was at its height.”). See also “New Federal Bonds in Active Demand,” New York Times, March 9, 1926, p. 36 (“It is confidently believed that Secretary Mellon’s judgment in making the offering of long-term bonds at the relatively low interest rate coincident with the severe shrinking of values on the stock market will be proven sound.”). 68. “Mellon Calls in Second Liberties,” New York Times, May 9, 1927, p. 9 (“It is believed here that relatively short-term Treasury notes finally will be decided on . . ..”), “Wall St. Surprised by Mellon’s Plan,” New York Times, June 1, 1927, p. 36 (“The terms of the Treasury’s financing . . . were received with surprise in Wall Street yesterday. . . . The general expectations had been that the financing would take the form of the sale of short-term securities, leaving the next sale of long-term bonds until next year, when the Third Liberty 4½ per cent. issue must be retired.”), and “Mellon Offers Second Liberty Bond Exchange,” Wall Street Journal, June 1, 1927, p. 12 (“Decision of the Treasury to continue the refunding of Second Liberty bonds by means of a long-term issue instead of the Treasury notes which have been used for that purpose was something of a surprise.”). 69. “New Treasury Bond Sells at Premium as Soon as Offered,” New York Times, July 6, 1928, p. 1.

13

Revival of the Over-the-Counter Market

At the end of World War I, Liberty bonds traded in an active secondary market on the New York Stock Exchange; the Exchange reported aggregate Liberty bond sales of almost $3 billion during 1919.1 In contrast, there was only a limited secondary market for certificates of indebtedness. An investor who wanted to liquidate an investment in certificates typically went back to the bank where he bought them and asked the bank to repurchase the securities. Banks were not unhappy to undertake such repurchases because they could readily finance certificates with Federal Reserve Bank credit.2 Federal Reserve credit, rather than a bona fide secondary market, provided liquidity to certificate holders. By the second half of the 1920s, Liberty bond sales on the New York Stock Exchange had contracted to a tenth of the wartime volume (table 13.1) and trading on the Exchange had been displaced by a far more liquid over-thecounter market that included certificates of indebtedness and notes as well as bonds.3 This chapter describes the evolution of the secondary market for Treasury securities during the 1920s. The first part examines the development of an over-the-counter market for certificates of indebtedness. The second part explains the return of trading in Treasury bonds to the over-the-counter market. Development of an Open Market for Certificates of Indebtedness The Treasury began selling certificates of indebtedness in April 1917 to raise funds in advance of the forthcoming sale of First Liberty bonds. The premise of the early offerings was that banks would buy certificates for their own 1. Harris (1948, p. 11). 2. 1920 Federal Reserve Bank of New York Annual Report, p. 17 (stating that “no open market existed for . . . certificates, and a purchaser wishing to realize upon them before maturity was obliged to re-sell them to the banks from which he purchased them. The banks in turn would borrow upon them from the Federal Reserve Bank.”) and Beckhart, Smith, and Brown (1932, p. 334) (“Up to the close of 1919 . . . any holder [of certificates] wishing to convert his certificates into cash had to resell them to the bank from which they were purchased. These banks in turn then borrowed from the Federal Reserve banks, usually from the Federal Reserve Bank of New York, since most certificates were sold in or gravitated to New York.”). 3. U.S. Treasury and Federal Reserve System (1959, p. 97) (“By the mid-1920’s . . . the volume of trading in Government securities on the over-the-counter dealer market far exceeded that on the Exchange.”). See also “Liberty Bonds Active and Higher,” Wall Street Journal, June 2, 1923, p. 5. (“Large part of the dealings in Liberty bonds is done through houses that buy and sell on their own account in the outside market. These sales are not recorded in Stock Exchange transactions. . . . Actual buying is much larger than the bond ticker indicates.”), Meeker (1930, p. 260) (The over-the-counter market “has become the great wholesale or jobbers’ market for bonds, and is particularly notable for its large individual transactions. Compared with it, the bond market in the Stock Exchange is normally thin and concerned primarily with retail transactions.”), and Meeker (1930, p. 266) (“So greatly have Stock Exchange dealings in U.S. Government bonds dwindled since the years immediately following the war, that these are now handled in a rather inconspicuous corner of the [bond] room.”)

186

Chapter 13

Table 13.1 Trading in government securities on the New York Stock Exchange ($millions) Year

Volume

1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930

286 1,437 2,901 2,861 1,957 1,873 796 876 390 262 289 188 142 115

Source: Harris (1948, p. 11).

account and then use the proceeds from maturing certificates to fund withdrawals by depositors paying for bond purchases. Banks were also encouraged to sell certificates directly to investors, to be used by the investors in lieu of bank balances to pay for their bond purchases. One of the Treasury’s primary wartime objectives was financing American participation at low nominal interest rates. To that end, Treasury officials organized bond drives, pressed banks to allocate a portion of their loanable funds to certificate purchases, and discouraged the Federal Reserve Banks from raising their loan and discount rates. However, after hostilities ended in November 1918, after the Treasury completed its offering of Victory notes in the spring of 1919, and after the one-year moratorium on interest rate increases expired in the fall of 1919, Federal Reserve officials began to reassert themselves.4 The Federal Reserve Bank of New York raised its loan and discount rates in early November 1919, in mid-December, and again on the last day of the year (table 13.2). The Bank remarked that “rates which during the war were necessarily related to the rates on Government securities rather than to the rates paid by industry and commerce, began gradually to approach their normal relation to the latter class of borrowing.”5 Reserve Bank officials continued to 4. Chandler (1958, pp. 135–87), Wicker (1966, pp. 25–56), and Meltzer (2003, pp. 90–109) describe the efforts of Federal Reserve officials to regain control of interest rates. 5. 1919 Federal Reserve Bank of New York Annual Report, p. 7. The Bank additionally noted that “the rapidity with which in the late summer and early autumn [of 1919] fresh credit was

187

Revival of the Over-the-Counter Market

Table 13.2 Loan rates and discount rates posted by the Federal Reserve Bank of New York during fiscal year 1920 15-Day loan rates

Apr 6, 1918 Nov 5, 1919 Dec 12, 1919 Dec 31, 1919 Jan 23, 1920 Feb 26, 1920 Jun 1, 1920

16- to 90-Day discount rates

Secured by certificates of indebtedness

Secured by Liberty bonds

Secured by commercial paper

Notes secured by certificates of indebtedness

Notes secured by Liberty bonds

Commercial paper

4 4½a 4½ 4¾ 4¾ 5 5½

4 4½ 4¾ 4¾ 5½ 5½ 6

4 4¾ 4¾ 4¾ 6 6 7

4¼ 4½ 4¾ 4¾ 4¾ 5 5½

4¼ 4½ 4¾ 4¾ 5½ 5½ 6

4¾ 4¾ 4¾ 4¾ 6 6 7

Source: 1918 Federal Reserve Bank of New York Annual Report, p. 75; 1919 Federal Reserve Bank of New York Annual Report, p. 56; and 1920 Federal Reserve Bank of New York Annual Report, p. 70. Note: Rates effective on June 1, 1920 remained in effect beyond the end of the fiscal year. a. 4¼ percent for certificates of indebtedness with 4¼ percent coupon rate.

raise rates in 1920; the loan and discount rates for commercial paper ultimately reached 7 percent on June 1, 1920. (Table 13.2 shows that by the end of 1919 the Fed had eliminated preferential rates for loans and discounts secured by Treasury securities, but then reintroduced preferential rates in 1920. Box 13.1 discusses preferential rates on Treasury debt.) Higher money market rates soon forced the Treasury to begin raising the rates on its certificate offerings. Table 13.3 shows that the Treasury was able to sell certificates at interest rates of 4½ percent or lower through the middle of December 1919 but that it had to offer 4¾ percent on a late December offering of an 11½-month certificate that settled on January 2, 1920. The Treasury successfully offered a short-term, 42-day certificate at 4½ percent in late January 1920, but encountered substantial resistance when it offered a 1-year certificate at 4¾ percent in March. Officials wanted to raise between $300 and $350 million in the March offering, but investors subscribed for only $200 million, even though the subscription books remained open for almost two weeks after the issue date.6 The New York Times described the response as “disappointingly small.”7 In a meeting with Assistant Secretary of demanded and absorbed by the feverish industrial and speculative activity . . . made it clear that natural forces in credit must . . . reassert themselves . . . if credit were to exert any restraint upon further price expansion.” 1919 Federal Reserve Bank of New York Annual Report, p. 7. 6. “The Certificate Sale,” New York Times, March 23, 1920, p. 18, and 1920 Treasury Annual Report, p. 15. 7. “The Certificate Sale,” New York Times, March 23, 1920, p. 18.

188

Chapter 13

Box 13.1 Preferential Loan and Discount Rates for Treasury Securities During World War I the Federal Reserve Bank of New York posted a preferential discount rate for notes secured by Treasury securities, but it did not make any collateral distinctions in rates on short-term loans to banks. During the last six months of the war, the Bank charged 4 percent on all 15-day loans, posted a 4¾ percent discount rate for eligible collateral, and posted a 4¼ percent preferential discount rate for notes secured by Liberty bonds and certificates of indebtedness.a When the Fed began to reassert its control over interest rates in November 1919, the New York Bank raised its base rate for short-term loans to 4¾ percent, but—in the face of Treasury objections to higher rates on loans secured with Treasury securities—raised the rate for loans secured by certificates of indebtedness and Liberty bonds to only 4½ percent (table 13.2).b At the same time, the Bank raised the discount rate for notes secured by certificates and Liberty bonds to 4½ percent. The resulting structure eliminated the difference between loan rates and discount rates but expanded the preference for Treasury collateral from discount rates to loan rates.c Subsequently, in early December 1919, the Treasury dropped its insistence on a preference rate for credit secured with Liberty bonds,d whereupon the Board “promptly wired the Reserve Banks that it would be receptive to proposals to eliminate the preferential rate on paper collateraled by Liberty Bonds and Victory Notes.”e The New York Fed quickly eliminated all of its preference rates other than the preference rate on short-term loans secured by certificates of indebtedness (table 13.2). Despite Treasury objections,f the Bank eliminated that remaining preference at the end of the year. a. 1918 Federal Reserve Bank of New York Annual Report, p. 75. b. 1919 Federal Reserve Bank of New York Annual Report, p. 56. In a meeting on September 4, 1919, Assistant Secretary of the Treasury Russell Leffingwell asked the Federal Reserve Board to refrain from increasing rates “on paper secured by Government obligations.” Meltzer (2003, p. 95). See also Wicker (1966, pp. 35–37) and Chandler (1958, pp. 159–60) (noting the Treasury’s policy preference that “preferential rates on paper collateraled by Treasury obligations should be left alone. A rise of the rate on paper backed by Liberty Bonds might further weaken the government bond market, and an increase of the rate applicable to paper collateraled by certificates of indebtedness would jeopardize current refunding operations.”) c. The November 1919 rate changes included a special, 4¼ percent rate for short-term loans secured by certificates of indebtedness with a 4¼ percent coupon rate. In late November 1919 several Reserve Banks, including the New York Bank, requested Board permission to increase their discount and loan rates. Leffingwell argued against elimination of the special preference: “It would be absolutely impossible for the Treasury to sell 4¼ per cent certificates were the 4¼ percent discount rates eliminated.” Quoted in Chandler (1958, p. 161). See also Wicker (1966, p. 41) (quoting Leffingwell as saying “The Treasury looks upon the action of the [Federal Reserve Bank of New York] in eliminating the 4¼% rate on paper secured by 4¼% certificates as a direct attempt to punish the Treasury of the U.S. for not submitting to dictation on the part of the Governor of the F. R. Bank of New York, even though it be at the cost of a shortage of funds of the Treasury to meet its outstanding obligations.”). The Board disapproved the proposed rate increases and the Treasury sold $422 million of 77- and 105-day 4¼ percent certificates a few days later. d. Meltzer (2003, p. 103) (stating that, on December 10, Leffingwell indicated to the Federal Reserve Board that he “no longer objected to an increase in rates or the elimination of the preferential rate for debt secured by Liberty and Victory bonds. The preference for certificates should remain.”). e. Chandler (1958, p. 162). The Board also said that it would favor an increase in the rate on paper secured by certificates of indebtedness with a 4¼ percent coupon from 4¼ to 4½ percent. f. Chandler (1958, p. 162) (Leffingwell advised the Federal Reserve Board that “he did not consider the increase to be a wise move”) and Meltzer (2003, p. 103) (Leffingwell criticized the increase in the rate on short-term loans secured by certificates as “unwise.”).

189

Revival of the Over-the-Counter Market

Box 13.1 Continued In January 1920 the Board, acting on the recommendation of Assistant Secretary of the Treasury Russell Leffingwell, directed the Federal Reserve Bank of New York and two other Reserve Banks to increase the basic short-term loan rate and discount rate 125 basis points, to 6 percent, but to keep the rate on loans secured by certificates of indebtedness and the discount rate for similarly secured notes unchanged at 4¾ percent. This action came at a time of increasing inflation, when higher interest rates were clearly necessary but when the Treasury was also reluctant to agree to anything that would make its own financings more costly.g As shown in the tables below, preferential rates on notes and loans secured by Treasury securities remained in place until mid-1921.h

15-Day loan rates

Dec 31, 1919 Jan 23, 1920 Feb 26, 1920 Jun 1, 1920 Feb 5, 1921 May 5, 1921 Jun 16, 1921

Secured by certificates of indebtedness

Secured by Liberty bonds

Secured by commercial paper

4¾ 4¾ 5 5½ 6 6 6

4¾ 5½ 5½ 6 6 6 6

4¾ 6 6 7 7 6½ 6

g. Meltzer (2003, p. 104) (“Leffingwell . . . became a proponent of higher rates on commercial loans but continued to demand a preferential rate for borrowing on Treasury certificates.”) and Chandler (1958, p. 166) (“On January 14, 1920, [the directors of the New York Bank] voted to raise [the bank’s] discount rate from 4¾ to 5½ percent. . . . Leffingwell objected strongly, however, insisting that the 4¾ per cent rate on paper backed by certificates should be retained, but adding his opinion that the rate on commercial paper should be raised to 6 rather than 5½ per cent. . . . On January 22 Governor Harding notified the Bank that the Board had determined that its discount rate on commercial paper should be 6 per cent with lower rates for paper backed by Treasury obligations.”). See also Wicker (1966, p. 45) (“What is indeed remarkable about the abrupt increase to 6% at the January 21 meeting of the Board was the fact that it done at the initiative of Leffingwell and not the Federal Reserve Board.”). The Board’s action followed a controversial opinion of the Acting Attorney General of the United States that the Board had the right “to determine what rates of discount should be charged, from time to time by a Federal reserve bank, and, under their powers of review and supervision, to require such rates to be put into effect by such Bank.” Opinions of the Attorney General, vol. 32, p. 81, December 9, 1919. See also Hackley (1973, pp. 171–72) and Joint Committee on the Economic Report (1952, p. 276–77). h. Meltzer (2003, p. 112) (“The outgoing Wilson Treasury at last agreed to end preferential rates on certificates of indebtedness.”) and Wicker (1966, p. 55) (“At the end of March [1921] the new Secretary of the Treasury, Andrew Mellon, expressed the view to Governor Harding that . . . the differential rates in favor of government securities should be eliminated.”).

190

Chapter 13

Box 13.1 Continued

16- to 90-Day discount rates

Dec 31, 1919 Jan 23, 1920 Feb 26, 1920 Jun 1, 1920 Feb 5, 1921 May 5, 1921 Jun 16, 1921

Notes secured by certificates of indebtedness

Notes secured by Liberty bonds

Commercial paper

4¾ 4¾ 5 5½ 6 6 6

4¾ 5½ 5½ 6 6 6 6

4¾ 6 6 7 7 6½ 6

Table 13.3 Certificates of indebtedness issued in fiscal year 1920 Issue date

Maturity date

Term (days)

Coupon rate (%)

Amount ($millions)

Jul 1, 1919 Jul 1, 1919 Jul 15, 1919 Aug 1, 1919 Aug 15, 1919 Sep 2, 1919 Sep 15, 1919 Sep 15, 1919 Dec 1, 1919 Dec 1, 1919 Dec 15, 1919 Jan 2, 1920 Feb 2, 1920 Mar 15, 1920 Apr 1, 1920 Apr 15, 1920 Apr 15, 1920 May 17, 1920 Jun 15, 1920 Jun 15, 1920

Sep 15, 1919 Dec 15, 1919 Mar 15, 1920 Jan 2, 1920 Jan 15, 1920 Feb 2, 1920 Mar 15, 1920 Sep 15, 1920 Feb 16, 1920 Mar 15, 1920 Jun 15, 1920 Dec 15, 1920 Mar 15, 1920 Mar 15, 1921 Jul 1, 1920 Jul 15, 1920 Oct 15, 1920 Nov 15, 1920 Jan 3, 1921 Jun 15, 1921

76 167 244 154 153 153 182 366 77 105 183 348 42 365 91 91 183 182 202 365

4½ 4½ 4½ 4½ 4½ 4½ 4¼ 4½ 4¼ 4¼ 4½ 4¾ 4½ 4¾ 4¾ 5 5¼ 5½ 5¾ 6

326 511 323 534 532 574 101 657 162 260 728 703 305 201 201 84 171 103 177 243

Source: 1920 Treasury Annual Report, p. 284.

191

Revival of the Over-the-Counter Market

the Treasury Russell Leffingwell on March 8, New York bankers had expressed intense dissatisfaction with the 4¾ percent rate and suggested that the new certificates should bear “at least” 5 percent—in line with the Fed’s contemporaneous 15-day loan rate on certificate collateral (table 13.2).8 The Wall Street Journal labeled the 4¾ percent rate a “mistake.”9 The Treasury subsequently offered certificates with coupon rates more in line with other money market rates. By the middle of 1920 the Treasury was offering to pay 6 percent on 1-year certificates, 50 basis points above the 5½ percent 15-day loan rate then posted by the New York Fed. Rising interest rates on new certificate offerings forced the development of an “open,” or unrestricted, secondary market for outstanding certificates. Prior to the spring of 1920 the Treasury, anxious to keep interest rates low, discouraged banks from buying or selling certificates at less than par (plus accrued interest).10 This was acceptable to the banks as long as coupon rates on outstanding certificates were in line with other money market rates and as long as they could borrow freely at comparable rates from their Federal Reserve Banks. However, when the Reserve Banks began to raise their discount rates during the winter of 1919 to 1920, and when other money market rates rose in parallel, bankers became increasingly reluctant to continue to pay par for certificates with coupon rates increasingly below current market rates.11 Had the Treasury maintained its opposition to bank purchases below par, the primary source of liquidity for certificate holders would have evaporated and investors might have begun to demand a yield premium for being asked to hold an essentially illiquid instrument. Faced with that unappealing prospect, Treasury officials began to relax their opposition to transactions below par.12 Following a sale of 6-month certificates at 5¼ percent in mid-April 1920, 8. “Treasury Offers New 4¾% Tax Certificates,” Wall Street Journal, March 10, 1920, p. 12 (characterizing the bankers as “much exercised”). 9. “Treasury Made Mistake in Certificate Rate at 4¾%,” Wall Street Journal, March 30, 1920, p. 12. 10. “Favor Open Market for Federal Paper,” New York Times, March 3, 1920, p. 20 (“the Treasury has on several occasions announced its opposition to any attempt on the part of the banks to purchase . . . certificates from private investors at less than par”) and Harris (1948, p. 7) (“The Treasury Department during the war . . . opposed trading in certificates of indebtedness below a par-plus-accrued-interest basis. The feeling was that trading in certificates of indebtedness below par would be disturbing to the success of new issues.”). 11. “Favor Open Market for Federal Paper,” New York Times, March 3, 1920, p. 20. 12. “Favor Open Market for Federal Paper,” New York Times, March 3, 1920, p. 20 (Treasury seen as “likely to announce its willingness to pursue a more lenient attitude than in the past.”), “See Open Market for Certificates,” New York Times, May 13, 1920, p. 23 (“[I]t is understood that the Treasury officials have receded somewhat from their attitude regarding the ‘pegging’ of certificates at par.”), and Federal Reserve Bank of New York (1940, p. 5) (“As the first move toward a free market the Treasury indicated in April and May, 1920, that it would not continue its policy of discountenancing open market trading by dealers in outstanding certificates of indebtedness at prices below par.”).

192

Chapter 13

The New York Times reported that “sales began to be made at considerable recessions below par in the case of certificates bearing only 4½ per cent . . .. The rate of yield on the sales began to bear a close relation to the rate for money in the open market, and some bankers say that dealings in the certificates quickened somewhat.”13 The Times further reported that, “bankers say that the attempts to hold all certificates at par will be utterly unavailing and that a really open market will be forced.” The Wall Street Journal reported similarly: “Quite an extensive open market for United States Government Treasury certificates of indebtedness is being established here . . .. It is now said that while the Government is not as yet convinced that an open market rate should be created for future Government issues, it is at least not directly opposed to these issues changing hands in the open market.”14 The Federal Reserve Bank of New York noted the developing secondary market for certificates in its 1920 Annual Report: In the early spring [of 1920] it seemed advisable to attempt to create an open market in certificates. This was undertaken with the two-fold object of facilitating the sale of future issues of certificates, and of relieving the banks from the burden of re-absorbing certificates they had sold. Certain dealers in short time investments undertook to effect a secondary distribution of certificates, purchasing them from the banks and re-selling them to private investors and corporations having funds available for temporary investment.”15

The early dealers included C.F. Childs, Charles E. Quincey and Company, Salomon Brothers & Hutzler, and the bond department of Bankers Trust Company.16 They were joined within a year by First National Corporation (the securities affiliate of First National Bank of Boston), Discount Corporation, and Shawmut Corporation (the securities affiliate of National Shawmut Bank of Boston).17 Dealer Financing The willingness of the Federal Reserve Bank of New York to finance the certificate inventories of non–bank dealers with repurchase agreements (“repos”) materially aided the development of an open market for certificates of indebtedness. (As described in box 13.2, the Fed first used repurchase agreements during World War I to support the Treasury’s wartime borrowing program.) 13. “See Open Market for Certificates,” New York Times, May 13, 1920, p. 23. 14. “Open Market for U.S. Bills Growing,” Wall Street Journal, May 13, 1920, p. 11. 15. 1920 Federal Reserve Bank of New York Annual Report, p. 17. 16. Beckhart, Smith, and Brown (1932, p. 335). C.F. Childs, founded in 1911, specialized in making markets in Treasury securities. Quincey, founded in 1887, and Salomon Brothers & Hutzler, founded in 1910, did a general securities business. 17. Beckhart, Smith, and Brown (1932, p. 335). For information on securities affiliates, see Peach (1941).

193

Revival of the Over-the-Counter Market

Box 13.2 Use of Repurchase Agreements by Federal Reserve Banks during World War I Federal Reserve Banks extended credit on repurchase agreements during World War I in circumstances where it would have been costly, or impossible, to make conventional loans.a Effective December 1, 1917, Title VIII of the War Revenue Act of 1917 levied taxes on a variety of financial instruments, including promissory notes. A stamp tax of 2 cents per $100 was held to apply to promissory notes made by banks in the course of borrowing from Federal Reserve Banks. The Federal Reserve Board determined that the tax was so onerous as to practically prohibit short term borrowings. In view of the exigent need to support the Treasury’s wartime borrowing program, the Board encouraged banks to obtain short-term credit by selling Treasury securities to their Reserve Banks under agreement to repurchase the securities.b In April 1918 Congress exempted promissory notes secured by U.S. government securities issued since the start of the warc and the Board requested that the Reserve Banks resume their former practice of extending credit on Treasury securities to member banks with loans.d Federal Reserve Banks also used repurchase agreements to extend credit to nonmember banks to finance bank and bank customer purchases of certificates of indebtedness. The 1918 Annual Report of the Federal Reserve Bank of New York explained that “While the Federal Reserve Bank could not [as a matter of statutory authority] discount directly for nonmember banks, it has freely purchased from them whenever necessary, at the same rate at which it was discounting for its member banks, certificates of indebtedness with an agreement on their part to repurchase within 15 days . . ..”e a. However, the Fed did not invent the repurchase agreement. See, for example, Lockhart (1921, p. 143) (reference to national banks borrowing on repurchase agreements in 1913) and Timberlake (1978, p. 78) (reference to Secretary of the Treasury Robert Walker entering into zero interest rate repurchase agreements in 1847 to relieve pressure on the New York money market). b. Mimeographed Letters and Statements of the Federal Reserve Board, X-545, December 1, 1917, and Harris (1933, p. 289). See also “Federal Reserve Bank Raises Discount Rates,” Wall Street Journal, April 8, 1918, p. 10 (“To get around [application of the stamp tax] the banks have been following the procedure of turning United States Certificates of Indebtedness over to the Federal Reserve bank with the option of repurchase within 15 days.”), and Hollander (1919, p. 60) (stamp tax “was avoided by the use of ‘resale’ or ‘repurchase agreements,’ whereby the Federal Reserve Banks acquired and held temporarily Liberty bonds and certificates of indebtedness until taken over by subscribing banks”). c. Section 301 of the War Finance Corporation Act of April 5, 1918. d. Mimeographed Letters and Statements of the Federal Reserve Board, X-817A, April 6, 1918 (“It is suggested . . . that the practice of purchasing Liberty Bonds and Certificates of Indebtedness under so-called repurchase agreements be discontinued and that such borrowing by member banks be made on their own promissory notes secured by such bonds and certificates.”), Federal Reserve Bulletin, May 1, 1918, p. 360 (“Federal Reserve Banks instead of purchasing [United States war obligations] under ‘repurchase agreements’ may accept from their members United States war obligations again as collateral for promissory notes, as was their practice prior to December 1, 1917.”), and Harris (1933, p. 289). e. 1918 Federal Reserve Bank of New York Annual Report, p. 24. See also 1918 Federal Reserve Board Annual Report, p. 323 (report of Pierre Jay, Chairman of the Board of Directors of the Federal Reserve Bank of New York and Federal Reserve Agent: “In order to facilitate the purchase of certificates of indebtedness by nonmember banks, the [Federal Reserve Bank of New York] has been ready to purchase such certificates, with an agreement on the part of the nonmember Bank to repurchase within 15 days.”) and 1919 Federal Reserve Bank of New York Annual Report, p. 16 (“The custom of purchasing certificates of indebtedness from nonmember banks with an agreement on their part to repurchase within fifteen days, was continued in 1919 as a means of encouraging their subscriptions to the issues.”).

194

Chapter 13

The New York Reserve Bank sometimes described its postwar dealer financing activities in rather modest terms: “For a time this bank, by means of purchase and resale agreements, facilitated [dealer operations in certificates of indebtedness] by making moderate advances against certificates.”18 Nevertheless, Reserve Bank financing was clearly quite important for the new market, a “material assistance” in the words of one later commentator.19 Governor Strong argued that dealer access to Reserve Bank credit was an “essential requirement” for an open market in certificates: [W]hen it is impossible for the dealers to procure [inventory financing] in the market either at all or at rates economically possible for them, assistance must be given to them by the Federal reserve banks by means of spot purchases of a portion of their supply of . . . Government securities. But as they are retailers of these goods and must have them available for sale in the future, the Federal reserve banks have made arrangements with them so that they may repurchase such . . . securities at some time in the future.20

The New York Fed’s practice of lending to dealers on repurchase agreements led to a sharp difference of opinion between the New York Bank and some Federal Reserve Board officials.21 In 1923 the Board’s General Counsel, Walter Wyatt, responding to questions concerning the propriety of Reserve Bank repurchase agreements, opined that repos were loans and that the Reserve Banks lacked authority to enter into such agreements.22 Governor Strong argued that, to the contrary, the Bank’s repos were purchases and sales rather than loans and were authorized by Section 14 of the Federal Reserve Act, which provided, inter alia, that a Reserve Bank could “buy and sell . . . bonds and notes of the United States . . ..”23 18. 1920 Federal Reserve Bank of New York Annual Report, p. 17. In testimony before the House Committee on Banking and Currency in 1926, Benjamin Strong, Governor of the Federal Reserve Bank of New York, testified that dealer repos were “a comparatively unimportant part of our operations.” Committee on Banking and Currency (1926, p. 431). 19. Simmons (1954, p. 26). See also Beckhart, Smith, and Brown (1932, p. 335) (“In order to help the new market, the Federal Reserve Bank of New York agreed to purchase [certificates] from time to time . . . under repurchase agreements.”). 20. Committee on Banking and Currency (1926, pp. 432). In a 1922 letter to Under Secretary of the Treasury S. Parker Gilbert, Strong described dealer repos as “one of the factors which makes the [certificate] market stable . . .. It is a carrying arrangement for the dealers upon which they may rely with certainty and at reasonable rates, and is something upon which the security of this kind of a market, to a certain extent, depends.” Letter dated July 18, 1922, from Strong to Gilbert, Federal Reserve Bank of New York file 410.5. Some observers found such arguments unconvincing. See, for example, Hardy (1932, p. 253). 21. Harris (1933, p. 290). 22. Mimeographed Letters and Statements of the Federal Reserve Board, X-3817, August 18, 1923 and Harris (1933, p. 290). The Reserve Banks could lend only to member banks (they could not lend to either nonmember banks or non–bank dealers) and (pursuant to the Act of September 7, 1916) only on promissory notes for a period not in excess of 15 days. Repurchase agreements were not represented by such notes. 23. Committee on Banking and Currency (1926, pp. 431–36) (written statement submitted by Governor Strong). See also Committee on Banking and Currency (1926 and 1927, pp. 930–35 and 981–1001) (extended discussion among W. R. Burgess, Assistant Federal Reserve Agent of

195

Revival of the Over-the-Counter Market

Whether repos are purchases and sales or loans was not resolved during the 1920s, but Strong’s assessment prevailed as a practical matter: the New York Bank and other Reserve Banks continued to enter into repurchase agreements on Treasury securities throughout the decade.24 Wire Transfers A second, quite different, Federal Reserve initiative further advanced the development of the secondary market for certificates of indebtedness. On July 15, 1921, the Federal Reserve Banks inaugurated a service for the “wire transfer” of certificates from one Reserve Bank to another. A simple example illustrates how the service operated. Suppose that a bank in San Francisco sold certificates to a dealer in New York. To get paid, the San Francisco bank had to deliver the securities to the dealer’s office. Prior to July 15, 1921, this required that the bank send bearer securities by U.S. mail or private express to the New York dealer. After that date the bank could deliver the securities to the Federal Reserve Bank of San Francisco and request that similar securities be made available to the dealer at the Federal Reserve Bank of New York. The San Francisco Fed would take in the securities, add them to the stock of unissued certificates that it held as fiscal agent for the United States, and advise the Federal Reserve Bank of New York (by telegram sent over a private “leased wire” network connecting the Reserve Banks25) to withdraw an equal quantity of the same certificates from its fiscal agency stock and prepare the securities for delivery. The New York Fed would then advise the dealer that it could send a messenger to the Bank to pick up the certificates. The San Francisco bank could either send the securities for delivery against payment—in which case the dealer would have to pay for the securities when it picked them up (and the New York Fed would remit the funds to the San Francisco Fed for the account of the California bank)—or it could send them “free” and arrange for payment in some other way. The Fed’s wire transfer service was established “to assist in the development and maintenance of a uniform countrywide market for Treasury certificates of indebtedness . . ..”26 The obvious advantage of the service was that it eliminated the time and expense of moving bearer securities around the country. the Federal Reserve Bank of New York, Otis Wingo, Democratic Congressman from Arkansas, and T. Alan Goldsborough, Democratic Congressman from Maryland, of whether repos were loans or purchases and sales). 24. See, Committee on Banking and Currency (1931, pp. 818–24). 25. See, Smith (1956) and “Government Holds 12,000 Miles Leased Wire,” Wall Street Journal, March 29, 1920, p. 11. 26. Smith (1956, p. 87). Federal Reserve officials were interested in this objective because banks invested in certificates as a secondary reserve and transacted in the securities in the course of adjusting their reserve positions. Smith (1956, p. 87).

196

Chapter 13

Summary Efforts to develop an open market for Treasury certificates of indebtedness soon bore fruit. As early as February 1921, the New York Times was reporting that certificate dealers were “able to take commitments of large amounts with confidence of being able to place them.”27 The 1921 Treasury Annual Report noted that “Quotations [for certificates] appear regularly in the leading daily newspapers and financial periodicals, and Treasury certificates enjoy a broad and active investment market.”28 By the middle of the decade the Treasury was able to report that “Due to their high degree of security and wide market, certificates of indebtedness yield a lower rate than commercial paper of the same maturity . . ..”29 And in a similar fashion over the same period, Treasury and Federal Reserve officials successfully encouraged the development of an open market for Treasury notes.30 Revival of the Over-the-Counter Market for Treasury Bonds The migration of trading in Treasury bonds from the New York Stock Exchange to the over-the-counter market during the first half of the 1920s was a consequence of two factors: (1) the postwar concentration of new Treasury issues in certificates and notes that were not listed on the Exchange31 and (2) the growing concentration of Treasury bonds in institutional hands.32 27. “Investors Buying U.S. Certificates,” New York Times, February 19, 1921, p. 20. 28. 1921 Treasury Annual Report, p. 5. See also “Investors Buying U.S. Certificates,” New York Times, February 19, 1921, p. 20 (“Until now there has been objection to regular day-to-day quotation of Treasury certificate prices in the open market, on the ground that the certificates were a new form of security not enjoying an established market. Because of the recent growth of the market it is possible to give the quotations and hereafter a table showing the bid, asked, and approximate yield of all issues of certificates will be a feature of the financial pages of The Times.”) 29. 1925 Treasury Annual Report, p. 40. The comparison with commercial paper was particularly apt because the market for commercial paper was also a “one-way” market in the early 1920s but did not subsequently undergone a transformation like that experienced by the market for certificates. 1920 Federal Reserve Bank of New York Annual Report, p. 14 (“The market in commercial paper . . . is only a one-way market. Commercial paper once bought by a bank cannot be resold in the paper market except by special arrangements with the brokers from whom it was bought.”). 30. Federal Reserve Bank of New York (1940, pp. 6–7) (“When in June, 1921, a program of issuing Treasury notes ranging in maturity from three to five years was inaugurated by the Treasury, the Reserve Banks were asked by the Secretary [of the Treasury] to cooperate in establishing a market for Treasury notes in the same way that the certificate of indebtedness market had been developed.”). 31. US Treasury and Federal Reserve System (1959, p. 97) (“only marketable Treasury bonds are listed on the New York Stock Exchange”) and Harris (1948, p. 9) (“the Treasury in June 1921 began to issue Treasury notes ranging in maturity from three to five years, thereby enlarging the volume of Government securities outstanding for which only an over-the-counter market prevailed.”). 32. Biais and Green (2005) report that trading in municipal bonds and corporate bonds also moved from the New York Stock Exchange to the over-the-counter market as a result of the increasing importance of institutional investors in those securities in the late 1920s and late 1940s, respectively.

197

Revival of the Over-the-Counter Market

Many market participants observed that an over-the-counter dealer market for Treasury bonds seemed to better serve the interests of institutional investors than trading on the Exchange. A dealer could step in to buy, with its own capital, a block of bonds from an impatient seller, it could supply bonds from its own inventory to an impatient buyer, and it could keep in virtually continuous contact with potential counterparties around the country throughout the day.33 In addition the failure of the Exchange to list certificates and notes placed the Exchange at a disadvantage in some types of transactions. For example, investors who wanted to swap out of bonds into shorter term securities usually found it easier and cheaper to execute both sides of the swap with a single dealer than to execute the sell side on the floor of the Exchange and the buy side with a dealer. It was hardly coincidental that a 1926 list of dealers in Treasury bonds included many of the same firms that made markets in certificates, including C.F. Childs, Discount Corporation, Charles E. Quincey and Company, and Salomon Brothers & Hutzler.34 In a particularly illuminating episode in June 1923, the British Government acquired almost $70 million of Second Liberty bonds (to make an interest payment on its wartime borrowings from the U.S. Treasury) entirely through purchases in the over-the-counter market. One dealer remarked that “an attempt to acquire a like amount of bonds solely on the Stock Exchange . . . would have materially advanced the market level for all Liberty bonds.”35 The decision of the Treasury to switch the locus of its own trading activities underscored the postwar revival of the over-the-counter market for Treasury securities. During and after World War I the Treasury executed purchases of Liberty bonds for the bond purchase fund and the sinking fund on the New 33. See, for example, the discussions in Federal Reserve Bank of New York (1940, pp. 7–10) and Harris (1948, p. 10–13). Beckhart, Smith, and Brown (1932, p. 361) state that “The transfer of a large proportion of dealings in long-term government bonds from the Stock Exchange to [the] over-the-counter market, took place gradually but persistently, beginning in 1920. It was due primarily to the inability of brokers on the New York Stock Exchange to handle the tremendous volume of business necessary in this market. Quotations [on the Exchange] were as a rule for 100-bond lots or less and transactions involving a million dollars were rare. It was therefore difficult for brokers to match purchases and sales and to carry out large orders without unduly influencing the price.” See also “Liberty Bonds Active and Higher,” Wall Street Journal, June 2, 1923, p. 5. (“Heavy buyers [of Liberty bonds] are said to be [bidding] from three to six thirtyseconds higher for large lots of bonds than exchange prices. They say any buying through exchange would immediately advance official quotations while buying on the outside is not reflected in market quotations. It is partly on this account that volume of official trading has declined to relatively small proportions.”). Meeker (1930, p. 260) states that “the growth of [the over-the-counter bond market] was facilitated by the development of the telephone . . ..” 34. Committee on Banking and Currency (1926 and 1927, p. 928) (testimony of W. Randolph Burgess, Assistant Reserve Agent of the Federal Reserve Bank of New York). Other dealers in Treasury bonds included First National Corporation, Shawmut Corporation, Barr Brothers & Co., First National Company of Detroit, Mabon & Co., and Scholle Brothers. Beckhart, Smith, and Brown (1932, p. 361) noted that “some of the most important dealers in government bonds make a market for both long- and short-term issues.” 35. “Tells How England Got Liberty Bonds,” New York Times, July 17, 1923, p. 24.

198

Chapter 13

York Stock Exchange.36 Officials believed that “The execution of orders on the stock exchange leaves the Treasury undisclosed in the operation, gives a public record of the transaction, and insures reflection of the order in the controlling [i.e., dominant] market for the bonds.”37 Nevertheless, the migration of trading in Treasury securities to the over-the-counter market prompted the Treasury to shift its operations to that market in 1925.38 Summary The first half of the 1920s witnessed the birth of an over-the-counter market for Treasury notes and certificates of indebtedness and the revival of the pre-war over-the-counter market for Treasury bonds. However, the postwar markets were not at all similar to the pre-war market. Treasury indebtedness was immensely larger and only a small fraction of the debt was held as collateral for national bank notes and government deposits. In addition, the Treasury issued short-term certificates of indebtedness and intermediate-term notes as well as long-term bonds. Many investors, including especially banks, held the certificates and notes as reservoirs of liquidity. Postwar trading was consequently far more active, and the secondary markets far more liquid, than had been the case before the war. 36. U.S. Treasury and Federal Reserve System (1959, p. 97) (“During the First World War there had been a substantial increase in Government securities trading on the New York Stock Exchange. The Treasury in fact supported this development by channeling its outright transactions (handled by the Federal Reserve Bank of New York) through the Exchange.”). See also Federal Reserve Bank of New York (1940, p. 5) and Harris (1948, p. 6). 37. Letter dated October 17, 1922, from S. Parker Gilbert, Under Secretary of the Treasury, to Benjamin Strong, Governor of the Federal Reserve Bank of New York, Federal Reserve Bank of New York file 410.5. 38. U.S. Treasury and Federal Reserve System (1959, p. 97) (“By the mid-1920’s . . . the volume of trading in Government securities on the over-the-counter dealer market far exceeded that on the Exchange. Since the bulk of the trading was outside the Exchange, there were obvious advantages in shifting Treasury orders from the Exchange to the over-the-counter market, and at the suggestion of the Federal Reserve Bank of New York its operations for Treasury account were transferred to that market in 1925.”).

14

Evolution of the Primary Market and the Introduction of Treasury Bills

Following the end of World War I, the Treasury continued to sell large quantities of securities on a regular basis. Between mid-1919 and mid-1930, the Treasury sold a total of $28 billion of securities in 100 cash offerings.1 In contrast, between 1894 and 1916, it had sold a mere $600 million of securities in nine cash offerings. Figure 14.1 shows the volume of cash sales by fiscal year. 1920 was a transition year between the war years (when the Treasury had an almost insatiable appetite for money and borrowed on the strength of marketing campaigns that emphasized patriotic sacrifice) and the balance of the decade (when the Treasury had more limited needs and sold securities strictly on an investment basis). The offering volume dropped between 1921 and 1924 primarily because Treasury officials began selling longer term certificates of indebtedness and intermediate-term notes that did not need to be refinanced as frequently as shorter term certificates. The scale of primary market activity raised the question of how best to bring new issues to market. Not surprisingly, Treasury officials initially continued their wartime practices. In fiscal 1920 the Treasury sold certificates of indebtedness on twenty occasions. Officials set the coupon rate on a new issue, offered the securities at par, and filled every order received during a subscription period that typically lasted about a week. In most cases they closed the subscription books on the dated date of the new security, but they sometimes kept the books open substantially longer if sales proved slow. (An offering of 1-year certificates in March 1920 remained open for more than two weeks.) Certificates were sold primarily to banks that either retained them as investments or resold them to other investors. In either case a bank almost always paid for its purchases by crediting the War Loan Deposit Account that the Treasury maintained at the bank.2 (The Treasury also accepted payment in the form of maturing securities.) The period from mid-1920 to late-1922 saw a significant evolution in the structure of primary market offerings, ending with the institutionalization of fixed-price subscription sales of specified (rather than unlimited) quantities of securities and a system of scaled allotments favoring small investors. However, subsequent experience revealed several flaws in that structure. Most prominent, new issues were commonly underpriced. Additionally the ability of banks to pay for new issues with War Loan Deposit Account credits allowed the banks to substitute War Loan deposits for more expensive borrowings. In 1929 the Treasury introduced Treasury bills to address these, and other, issues. 1. The Treasury also issued a total of $5.1 billion of securities in 14 exchange offerings (primarily for Liberty Loans) between mid-1919 and mid-1930. 2. During the 1920s the Treasury continued to use the War Loan Deposit Accounts that it had established during the war. See 1920 Treasury Annual Report, pp. 171–77.

200

Chapter 14

8 7

Billions of dollars

6 5 4 3 2 1 0 1920

1921

1922

1923

1924

1925

1926

1927

1928

1929

1930

Fiscal year Certificates

Notes

Bonds

Figure 14.1 Cash sales of Treasury certificates of indebtedness, notes, and bonds

This chapter describes (1) the emergence of, and the solutions to, the allocation problem created by fixed-price offerings of underpriced securities, (2) the substitution of War Loan deposits for more expensive borrowings, and (3) the introduction of Treasury bills in 1929. Emergence of, and Solutions to, the Allocation Problem In a continuation of wartime practices, the Treasury sold certificates of indebtedness in fixed-price subscription offerings once or twice a month through the middle of 1921. Prior to mid-1920 it filled all subscriptions in full, but ended that practice in May 1920 when, due to a declining need for funds, it began to limit sales to only slightly more than an announced amount. For example, in February 1921 it announced an offering of $100 million of 5-month certificates, received subscriptions for $219 million of the certificates, and issued $133 million.3 The new practice of not filling all subscriptions in full gave rise to an allocation problem: if not all subscribers were to get all the securities they asked 3. Treasury Circular no. 227, February 10, 1921, reprinted in 1921 Treasury Annual Report, p. 179, 1921 Treasury Annual Report, p. 65, and “Certificate Allotments Total $132,886,500,” Wall Street Journal, February 18, 1921, p. 4.

201

Evolution of the Primary Market and the Introduction of Treasury Bills

Box 14.1 Failed Treasury Offerings in the Early 1920s The risk of a failed offering was more than conjectural when Treasury securities were sold in fixed-price subscription offerings. In March 1920, Treasury officials tried to raise between $300 and $350 million in an offering of 1-year, 4¾ percent certificates.a Bankers protested that the rate should have been “at least” 5 percent and subscribed for only $200 million of the certificates.b In December 1922 the Treasury offered $400 million of 3-month and 1-year certificates, but garnered only $310 million in subscriptions, including $45 million of last-minute subscriptions from three Federal Reserve Banks.c The Governor of the Federal Reserve Bank of Boston complained about the last minute subscriptions, saying that “this was the first time that anything of this sort had been done since the days of Secretary McAdoo.”d a. Treasury Circular no. 185, March 10, 1920, reprinted in 1920 Treasury Annual Report, p. 301, and “The Certificate Sale,” New York Times, March 23, 1920, p. 18. b. “Treasury Offers New 4¾% Tax Certificates,” Wall Street Journal, March 10, 1920, p. 12, “Treasury Made Mistake in Certificate Rate at 4¾%,” Wall Street Journal, March 30, 1920, p. 12, and 1920 Treasury Annual Report, p. 15. c. Treasury Circular no. 314, December 7, 1922, reprinted in 1923 Treasury Annual Report, p. 237, 1923 Treasury Annual Report, p. 51, and letter, December 16, 1922, from Under Secretary of the Treasury S. Parker Gilbert to Charles Morss, governor, Federal Reserve Bank of Boston, Federal Reserve Bank of New York file 410.5. d. Letter, December 15, 1922, from Under Secretary of the Treasury S. Parker Gilbert to J. Herbert Case, deputy governor, Federal Reserve Bank of New York, Federal Reserve Bank of New York file 410.5. For an example of a Federal Reserve rescue of a floundering Treasury offering during World War I, see “U.S. Certificates are a General Accommodation,” Wall Street Journal, February 19, 1918, p. 10.

for, who should be disappointed? Importantly, the problem was chronic, and not merely sporadic, because Treasury officials had a continuing incentive to avoid the risk of a failed offering by setting the coupon rate on a new issue above the level that they expected would attract the desired quantity of funds. (The risk of a failed offering was more than conjectural. Box 14.1 describes two failed offerings from the early 1920s.) The Initial Scheme When the Treasury first began to limit certificate sales to not much more than a specified offering amount, it allocated securities by Federal Reserve district. Prior to the announcement of a new offering, Treasury and Federal Reserve officials set a sales target for each of the twelve districts that reflected the volume of banking activity in the respective districts. The sum of the twelve targets was equal to the amount of the offering. A district that failed to generate subscriptions in excess of its target was allotted securities equal to its aggregate subscriptions and every subscription from that district was filled in full. A district that more than met its target was allotted its target plus a percentage of its oversubscription. The percentage, which was the same for all oversubscribing districts, was set following the close of the subscription books at the level that produced the desired volume of total sales.

202

Chapter 14

For example, in November 1920 the Treasury offered $200 million of 6-month certificates.4 It received subscriptions for $293 million of the certificates, including $248 million from banks in seven districts that subscribed for more than their targets and $45 million from banks in five districts that did not exceed their targets. The Treasury sold a total of $232 million of the certificates. It filled all subscriptions from banks in the five undersubscribing districts and made a 40 percent allotment on the oversubscriptions from the seven oversubscribing districts, selling the banks in those districts a total of $187 million of certificates, including their aggregate target of $147 million and $40 million on their oversubscriptions.5 The securities allotted to each of the seven oversubscribing districts were then divided among the district’s subscribers. Banks that undersubscribed relative to their targets got what they asked for; an oversubscribing bank got its target plus a fraction of its oversubscription.6 The target-based scheme was administratively simple but failed to make use of the Treasury’s ability to allocate underpriced securities. Officials abandoned the scheme in the fall of 1922 and introduced a new system of scaled allotments in an effort to broaden the primary market for Treasury securities. Introduction of Scaled Allotments In October 1922 the Treasury brought its first bond offering—$500 million of 30-year bonds—since the Fourth Liberty Loan in 1918. Bond offerings had historically attracted close congressional scrutiny, particularly with respect to the goal of achieving a broad distribution among individual investors.7 Thus it was not surprising that Treasury Secretary Mellon asked U.S. banking institutions to “make a special effort to bring the offering to the attention of your customers, large and small, for it is the Treasury’s desire to secure the widest possible distribution of the bonds among investors.”8 More substantively, the offering circular for the bonds set forth a significant change in how the Treasury would allocate the bonds: all subscriptions for $10,000 or 4. Treasury Circular no. 211, November 8, 1920, reprinted in 1920 Treasury Annual Report, p. 312, and “New Treasury Note Issue,” New York Times, November 8, 1920, p. 29. 5. “New Certificates Taken,” New York Times, November 17, 1920, p. 28, and 1920 Treasury Annual Report, p. 21. Note that $40 million = 40 percent of the $101 million oversubscription, where $101 million = $248 total subscriptions from oversubscribing districts, less $147 million target for the oversubscribing districts. 6. Memo, “Allotment of certificates,” April 18, 1922, from J. Herbert Case, Deputy Governor, Federal Reserve Bank of New York, to George Harrison, Deputy Governor, Federal Reserve Bank of New York, Federal Reserve Bank of New York file 410.5. 7. See the discussion in chapter 3 on the 1896 auction offering of 4 percent bonds redeemable in 1925 and the 1898 subscription sale of Spanish–American War bonds. 8. Letter dated October 9, 1922, reprinted in 1922 Treasury Annual Report, p. 171.

203

Evolution of the Primary Market and the Introduction of Treasury Bills

less would be filled in full, regardless of the Federal Reserve district from which they originated; subscriptions for larger amounts would be subject to allotment.9 Prior to closing the subscription books for the new bonds on October 14, Treasury officials received subscriptions for $1.4 billion of the bonds. They decided to issue $512 million, or 37 percent of the amount subscribed. On October 15, officials announced that subscriptions for $10,000 or less had absorbed $325 million of the issue and that allotments on larger subscriptions would be scaled:10 $10,000–50,000

40 percent of amount subscribed, but not less than $10,000

$50,000–100,000

30 percent of amount subscribed, but not less than $20,000

$100,000–500,000

20 percent of amount subscribed, but not less than $30,000

$500,000–1,000,000

15 percent of amount subscribed, but not less than $100,000

over $1,000,000

10 percent of amount subscribed, but not less than $150,000

Officials said the system of scaled allotments was intended “to secure the widest possible distribution among investors throughout the country.”11 The Treasury quickly extended its new allotment scheme to certificates and notes. While the subscription books for a cash offering were open, the Federal Reserve Banks reported to the Treasury their daily receipts of subscriptions. After enough subscriptions had been received to cover the offering, Treasury closed the books and the Reserve Banks reported their aggregate subscriptions in seven size categories: 1. $1,000 or less 2. $1,000 to $10,000 3. $10,000 to $50,000 4. $50,000 to $100,000 5. $100,000 to $500,000 6. $500,000 to $1,000,000 7. over $1,000,000 9. Treasury Circular no. 307, October 9, 1922, reprinted in 1922 Treasury Annual Report, p. 168. 10. 1922 Treasury Annual Report, p. 5, and “$500,000,000 Issue Gets $1,500,000,000,” New York Times, October 16, 1922, p. 1. 11. 1922 Treasury Annual Report, p. 5.

204

Chapter 14

Table 14.1 Representative allotments for subscription offerings of certificates of indebtedness Example of a modest scaling In March 1929, the Treasury offered to sell $475 million of 9-month certificates of indebtedness. It received subscriptions for $524 million of the certificates and sold $476 million, 91 percent of the amount subscribed. The Treasury filled all subscriptions for less than $100,000 in full. Larger subscriptions were allocated as: $100,000–$1,000,000 90% of amount subscribed, but not less than $100,000 Over $1,000,000 85% of amount subscribed, but not less than $900,000 Example of an intermediate scaling In September 1925, the Treasury offered to sell $250 million of 9-month certificates of indebtedness. It received subscriptions for $568 million of the certificates and sold $252 million, 44 percent of the amount subscribed. Subscriptions were allocated as: Less than $1,000 100% of amount subscribed $1,000–$10,000 60% of the amount subscribed, but not less than $1,000 $10,000–$50,000 50% of the amount subscribed, but not less than $6,000 $50,000–$500,000 40% of the amount subscribed, but not less than $25,000 Over $500,000 30% of the amount subscribed, but not less than $200,000 Example of an aggressive scaling In December 1926, the Treasury offered to sell $200 million of 9-month certificates of indebtedness. It received subscriptions for $1,096 million of the certificates and sold $229 million, 21 percent of the amount subscribed. Subscriptions were allocated as: Less than $1,000 50% of amount subscribed, but not less than $500 Over $1,000 10% of amount subscribed, but not less than $500 Source: Treasury Circular no. 360, September 8, 1925, reprinted in 1925 Treasury Annual Report, p. 254; “Treasury Gets $568,000,000 Subscriptions for its New Loan Offer of $250,000,000,” New York Times, September 12, 1925, p. 6; Treasury Circular no. 373, December 8, 1926, reprinted in 1927 Treasury Annual Report, p. 269; 1927 Treasury Annual Report, p. 270; Treasury Circular no. 413, March 7, 1929, reprinted in 1929 Treasury Annual Report, pp. 265; and 1929 Treasury Annual Report, pp. 266.

On the basis of the Reserve Bank numbers, the Treasury calculated the percentage of securities to be allocated to subscribers in each of the categories. When subscriptions exceeded the amount offered by only a modest amount, scaling was comparably modest and, as illustrated in the first panel of table 14.1, most investors got almost all they asked for. However, if subscriptions were heavy, scaling on large orders was likely to be quite severe. The third panel of table 14.1 gives an example. In December 1922 Treasury officials added an important wrinkle to their system of scaled allotments. To reduce churning of funds through War Loan Deposit Accounts, subscribers tendering maturing securities that the Treasury had announced would be accepted for payment were given a preferred allotment. For example, on September 10, 1923, the Treasury offered to sell $200 million of 6-month certificates of indebtedness dated September 15.12 Investors subscribed for $554 million of the new certificates. $64 million of the subscriptions tendered maturing certificates in payment. The Treasury filled all of the 12. Treasury Circular no. 328, September 10, 1934, reprinted in 1923 Treasury Annual Report, p. 241.

205

Evolution of the Primary Market and the Introduction of Treasury Bills

latter subscriptions, as well as all cash subscriptions for less than $10,000, but allocated only 50 percent of cash subscriptions for between $10,000 and $100,000 (subject to a minimum allocation of $10,000) and 20 percent of cash subscriptions in excess of $100,000 (subject to a minimum allocation of $50,000).13 Scaled allotments, coupled with a preference for subscriptions tendering maturing securities,14 advanced two goals of Treasury debt management: effecting a broad distribution of new issues and limiting churning of Treasury cash balances on tax dates. However, it did not address the fundamental economic problem that the Treasury was selling securities at below-market prices in its fixed-price offerings. Pricing New Issues The October 1922 bond offering illustrates the inefficiencies of fixed-price offerings. In that offering, the Treasury offered to sell $500 million of 30-year, 4¼ percent bonds at par. Immediately following the announcement, the bonds began to trade for when-issued settlement at prices between 100¼ and 100½,15 providing a quick profit to successful subscribers. Not surprisingly, the Treasury received subscriptions for far more than it was offering—for $1.4 billion of the bonds, making for a “cover ratio” (or ratio of amount subscribed to amount offered) of 2.8—and issued $512 million of bonds.16 One banker remarked that “A 4 per cent. bond could probably have been sold by the Government.”17 Figure 14.2 shows that large oversubscriptions—the principle indicia of underpricing—was a persistent characteristic of Treasury offerings throughout the 1920s. The cover ratio averaged 2.7 over the interval from mid-1920 to mid-1930. As shown in figure 14.3, it ranged as high as 7.3 in December 1924, 13. “Mellon Allots New Notes,” New York Times, September 14, 1923, p. 26, and 1923 Treasury Annual Report, p. 54. 14. The Treasury sometimes limited or withheld the preference for subscribers tendering maturing securities. In a March 1925 offering of 29¾-year bonds, officials announced that they would treat all subscriptions the same and would not give a preference to subscriptions tendering maturing securities. “Treasury Issues Draw $747,000,000,” New York Times, March 12, 1925, p. 30, and 1925 Treasury Annual Report, p. 32. In offering certificates of indebtedness in December 1926 and March 1927, officials announced that subscribers tendering maturing securities would be allotted only 50 percent of the amount tendered. “Treasury to Need Only $200,000,000,” New York Times, December 8, 1926, p. 37, “$229,264,500 Allotted on Treasury’s Call,” New York Times, December 12, 1926, p. 21, “Treasury Offers $450,000,000 Loan,” New York Times, March 7, 1927, p. 21, “New Treasury Loan Thrice Oversubscribed; Books Are Closed and Allotments Cut Down,” New York Times, March 10, 1927, p. 27, and 1927 Treasury Annual Report, pp. 270 and 273. 15. “Ten Banks Here Ask Half of New $500,000,00 Loan,” New York Times, October 10, 1922, p. 1. 16. 1922 Treasury Annual Report, p. 5. 17. “Orders Pouring In for New U.S. Bonds,” New York Times, October 11, 1922, p. 9.

Chapter 14

2.5

Billions of dollars

2.0

1.5

1.0

0.5

0 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 Unfilled subscriptions

Additional issued

Offered

Figure 14.2 Amount offered, issued, and subscribed in quarterly, tax payment date, offerings of Treasury securities

10

8

6

Ratio

206

4

2

0 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931

Figure 14.3 Cover ratio (ratio of amount subscribed to amount offered) for fixed-price cash subscription offerings of Treasury securities

207

Evolution of the Primary Market and the Introduction of Treasury Bills

when the Treasury offered $200 million of 30-year, 4 percent bonds at par and received subscriptions for $1.46 billion of the bonds.18 Substitution of War Loan Deposits for Other Borrowings In the 1920s, as during the war, banks were important underwriters of certificates of indebtedness. Banks bought new certificates at par, paying for them by crediting Treasury War Loan Deposit Accounts, and then sold them to other investors, usually at par plus accrued interest. The effective underwriting commission was the difference between the accrued interest and the 2 percent per annum interest paid to the Treasury on its War Loan Deposit Account balances. Banks quickly realized that they could use primary market purchases of Treasury securities to generate War Loan deposits and that they could substitute those deposits for other, more expensive, sources of funds. A bank would first subscribe to a new offering, promising to pay by crediting the War Loan Deposit Account that the Treasury maintained at the bank. After receiving notice of its allotment, the bank would sell the new securities for settlement on the issue date and use the proceeds to reduce other borrowings. The net result was a reduction in the cost of funds to 2 percent per annum over the interval between the issue date of the securities and the date the Treasury finally drew down its War Loan deposit balance.19 War Loan deposits became increasingly attractive when the Reserve Banks started raising their discount rates in 1928 (figure 14.4). Generating War Loan deposits by subscribing to new issues made economic sense even if a bank sold its allotted securities at a discount from par, as long as interest rates were high enough and the interval the bank could expect to retain the War Loan deposits was long enough. In June 1928, The New York Times reported whenissued sales of new certificates at prices of 9931 (99 and 31/32nds percent of principal value) and 9930, even though the certificates had been heavily oversubscribed: 18. A heavy oversubscription for the December 1924 offering was predictable when the bonds began trading for when-issued settlement at prices between 1005⁄8 and 100¾ immediately after the offering was announced. “Big Oversubscription for Federal Loan; One New York Bank Asks $75,000,000,” New York Times, December 5, 1924, p. 1. The New York Times reported that the Treasury was “swamped” with subscriptions and that officials had had to scale the allocation of the bond unusually aggressively: subscriptions for less than $1,000 were filled in full, subscriptions for between $1,000 and $10,000 were allotted 65 percent of what was asked (subject to a minimum of $1,000), and subscriptions for over $10,000 were rejected. “Billion for Bonds Swamps Treasury,” New York Times, December 7, 1924, p. 1. 19. An early version of this opportunistic behavior appeared as soon as the Federal Reserve began raising short-term interest rates in late 1919 and the first half of 1920. See “Banks Profit by Deposit of Certificate Proceeds,” Wall Street Journal, April 19, 1920, p. 12.

Chapter 14

7.00

6.00

Percent

208

5.00

4.00

3.00

2.00 1925

1926

1927

1928

1929

1930

Figure 14.4 Federal Reserve Bank of New York discount rate

This paradox of an issue being apparently heavily oversubscribed [at par] at the same time that sales are being made below par was explained by conditions in the money market and the opportunity for profit arising out of the methods by which the Government securities are sold to the banks, their largest purchasers. When Government securities are awarded on subscription the banks do not pay for them at once, but credit the Treasury’s account with the sum involved. This money is left on deposit until the Treasury calls for it, a period which is usually two or three weeks and sometimes stretches into months. The banks pay the Treasury 2 per cent. interest on these deposits . . ..20

In April 1929 the Times reported that the wide spread between open market interest rates and War Loan Deposit Account rates incentivized banks “to bid for larger amounts of Treasury securities than they ordinarily would take, and they often sell the securities as soon as they are allotted. At times, this produces the spectacle of a Treasury issue being heavily oversubscribed and simultaneously selling below par . . ., a situation that makes for confusion and artificial values in Treasury financing.”21 The incentive for banks to oversubscribe to Treasury offerings and then dump their allotments in the secondary market did not go unnoticed by non– bank investors. Those investors began to abstain from subscribing for new issues, electing instead to acquire the securities in subsequent secondary market transactions. This undermined Treasury efforts to broaden the primary market for its securities. Under Secretary of the Treasury Ogden Mills summed up the situation in testimony before the House Ways and Means Committee in May 1929: 20. “Sales Reported in Treasury Issue,” New York Times, June 9, 1928, p. 25. 21. “New Treasury Plan Similar to English,” New York Times, April 28, 1929, p. 39.

209

Evolution of the Primary Market and the Introduction of Treasury Bills

[T]here has grown up recently a practice on the part of the banks which was somewhat detrimental to the credit of the Government. Banks unquestionably subscribe for Government certificates because of the deposit privilege. During the last year or so, which has been a period of tight money, a practice has developed on the part of the banks of selling these certificates sometimes even before they are issued. In other words a bank can afford to subscribe for these certificates and sell them at a loss of say two-thirty-seconds or four-thirty-seconds and even six-thirty-seconds and still show a profit on the transaction, providing it can keep the Government deposit for 30 or 40 days. What the bank is really doing is borrowing from the Government of the United States at 2 per cent for 30 or 40 days, and it can use that money at once to pay off its indebtedness to the Federal reserve bank, on which it is paying 5 per cent. It can of course afford under these circumstances to sell these certificates at a discount. So that Government certificates during the course of the last year have sold at less than par almost immediately when they were issued. And, of course, those corporations and individuals who want to invest in Government securities, having observed that they show a tendency to go below par almost immediately after issue, refrain from putting in their subscriptions at the time the certificates are offered and rely on their ability to buy them in the market afterwards.22

Bank substitution of War Loan deposits for other sources of funds was a breakeven proposition for nearly everyone. Banks gained from accessing cheap Treasury balances, but lost on the sale of certificates to non–bank investors at prices below par. The Treasury benefited by being able to issue securities to banks at a higher price than the general public would pay, but was limited to earning only 2 percent on the proceeds of issuance. And non–bank investors bought securities at fair market prices—as long as they waited to buy in the secondary market. The principle consequence was an increasingly opaque primary market with declining non–bank participation. Introduction of Treasury Bills By the late 1920s Treasury and Federal Reserve officials had become aware of several defects in the structure of the primary market for Treasury securities and had begun to contemplate corrective measures. First and foremost was the difficulty of setting coupon rates on new issues and the prospective benefit of auctioning securities. As early as April 1921, The Wall Street Journal pointed out the advantage of auctions: “The fact that there has been a big over-subscription to recent offerings of certificates of indebtedness, suggests that the rate fixed for the certificates has lately been slightly higher than the money market warranted,” and noted that the British had already returned to auctioning their Treasury bills.23 Treasury officials were hardly ignorant of auction processes. The Treasury had commonly auctioned securities in the decades before World War I (see 22. Committee on Ways and Means (1929, pp. 4–5). 23. “Government Borrowing by Tender Suggested Here,” Wall Street Journal, April 14, 1921, p. 4. See also “Government Borrowing,” Wall Street Journal, May 19, 1921, p. 4 (“Bankers point out that [the $275 million oversubscription on an offering of $200 million of 9-month certificates] is another illustration of the advantage that might have been afforded by adopting the system of offering the certificates at tender. Had they been offered by tender a considerable saving might have been effected.”).

210

Chapter 14

chapter 3) and (as explained in box 14.2) had recently used reverse auctions to repurchase outstanding securities. Nevertheless, Treasury officials appreciated that small- and mid-size banks, not being in close touch with the Treasury market, might be reluctant to participate in an auction, leaving the primary market to larger banks and securities firms.24 Treasury officials also realized that the practice of offering securities only on quarterly tax dates meant that the Treasury sometimes had to borrow well in advance of its needs, issuing securities with coupon rates between 3 and 5 percent and leaving the proceeds in War Loan Deposit Accounts earning only 2 percent.25 It would clearly be advantageous to borrow between tax dates as funds were needed. Officials further appreciated that the practice of allowing banks to pay for new issues by crediting War Loan Deposit Accounts led to the substitution of War Loan deposits for other sources of bank funds. Nevertheless, officials believed that the “book credit system” was a “very potent sales factor” and that banks “have been and still are our best distributors of Government securities.”26 They were not anxious to tamper with a system that, while perhaps not perfect, worked quite well.27 In late April 1929 the Treasury unveiled a proposal for correcting the defects in the structure of the primary market. Senator Reed Smoot, Republican of 24. Memo, “Discussion of method of handling short term debt by United States Treasury, including comparison with British Treasury method,” January 4, 1928, from J. Herbert Case, Deputy Governor, Federal Reserve Bank of New York, to Benjamin Strong, Governor, Federal Reserve Bank of New York, Federal Reserve Bank of New York file 413.7 (“[M]any banks would feel that they are not sufficiently in touch with the money market to make bids for Government obligations. This would probably tend to restrict somewhat the distribution of Government securities.”). See also memo, August 31, 1928, from William S. Broughton, Commissioner of the Public Debt, to Ogden Mills, Under Secretary of the Treasury, Federal Reserve Bank of New York file 413.7A (“The Treasury would probably be criticized [for introducing competitive bidding] by those not closely in touch with the market.”). 25. Memo, “Review of Present Methods of Financing the United States Government and Suggested Supplementary Practices,” January 25, 1929, Federal Reserve Bank of New York, Federal Reserve Bank of New York file 413.7 (“Funds have been borrowed by the Government in advance of actual requirements and the interest cost on such borrowing has exceeded the interest received on idle Government deposits. A closer adjustment of borrowings to expenditures would be more economical.”) and Committee on Ways and Means (1929, p. 3) (testimony of Under Secretary of the Treasury Ogden Mills that “it is reasonably clear that if you are going to borrow only four times a year, you have got to borrow in advance of requirements.”). 26. Memo, “Changes in investments and other matters,” November 6, 1922, from J. Herbert Case, Deputy Governor, Federal Reserve Bank of New York, to Benjamin Strong, Governor, Federal Reserve Bank of New York, Federal Reserve Bank of New York file 510.5. 27. See, for example, memo, “Proposal that we request bids on our certificate issues, and objections to adoption of British Treasury bills,” July 27, 1928, from John F. Ebersole, Section of Financial and Economic Research, Office of the Secretary of the Treasury, to Ogden Mills, Under Secretary of the Treasury, Federal Reserve Bank of New York file 413.7A (“We have built up an effective ‘underwriting system’ … This underwriting system, with its subscribing and depositary banks, has been built up at a great expense both in effort and in money. It is thoroughly understood by practically all of our bankers and may be said to have become traditional.”)

211

Evolution of the Primary Market and the Introduction of Treasury Bills

Box 14.2 Reverse Auctions of Treasury Securities in the 1920s In the course of redeeming the Second and Third Liberty bonds, the Treasury conducted reverse auctions on three occasions. On November 27, 1925, the Treasury announced a reverse auction for $50 million, “or thereabouts,” of Third Liberty bonds maturing September 28, 1928.a The auction closed on December 10 and settled on December 29. The Treasury received tenders for $176 million of bonds and accepted $66.3 million of the tenders.b The novel transaction (it was characterized as a “departure” in government finance,c an “experiment”d) was undertaken “to determine the feasibility of making . . . purchases . . . directly from the holders, and thus giving all holders of third Liberty loan bonds the opportunity to sell their bonds to the Government . . ..”e Market participants conjectured that the Treasury also wanted to avoid “running up the price” on Third Liberties.f On March 1, 1926, the Treasury announced a second reverse auction, for $100 million, “or thereabouts,” of Third Liberty bonds.g The auction closed on March 10 and settled on March 23. The Treasury received tenders for more than $148 million of bonds and accepted $121.6 million of the tenders.h And on June 16, 1927, the Treasury announced a reverse auction for “a limited amount” of Second Liberty bonds that had been called for early redemption on November 15, 1927.i The auction closed on June 22 and settled on June 28. The Treasury received $72 million of tenders and accepted $63.0 million.j Auction participants other than banks and trust companies were required to submit their proposals through a bank or trust company to a Federal Reserve Bank. Accepted propositions were settled by delivering bonds through a bank or trust company to a Federal Reserve Bank. The Treasury did not itself accept either auction proposals or bonds, and investors other than banks and trust companies could not participate directly in an auction. a. Treasury Department Circular no. 363, November 27, 1925, reprinted in 1926 Treasury Annual Report, p. 199. b. 1926 Treasury Annual Report, p. 43. c. “Treasury Will Buy $50,000,000 3rd Libertys,” Wall Street Journal, November 28, 1925, p. 8. d. “Liberty Bonds Rise on Treasury Offer,” New York Times, November 28, 1925, p. 19. e. 1926 Treasury Annual Report, p. 42. f. “Liberty Bonds Rise on Treasury Offer,” New York Times, November 28, 1925, p. 19, “Treasury Will Buy $50,000,000 3rd Libertys,” Wall Street Journal, November 28, 1925, p. 8, “Dealers are Well Supplied with Bonds,” Wall Street Journal, December 2, 1925, p. 4, and “Liberty Bonds Up on Mellon’s Offer,” New York Times, March 2, 1926, p. 8 (“Before the November innovation Liberty bonds desired for sinking fund purposes were purchased in the open market, and every time the Treasury started to buy the prices advanced. It was to overcome this, as well as to save commission charges . . ., that the new plan was decided upon. The transactions bear out frequent reports in the financial district that Liberty bonds have become concentrated in a comparatively few strong hands, and [that] it is difficult to dislodge them in the amounts desired by the Government for sinking fund purposes.”). g. Treasury Department Circular no. 366, March 1, 1926, reprinted in 1926 Treasury Annual Report, p. 201. h. 1926 Treasury Annual Report, p. 43. i. Treasury Department Circular no. 384, June 16, 1927, reprinted in 1927 Treasury Annual Report, p. 286. j. “Treasury Offer Drew Few,” New York Times, June 23, 1927, p. 23, and 1927 Treasury Annual Report, p. 39.

212

Chapter 14

Utah and chairman of the Senate Finance Committee, and Representative Willis Hawley, Republican of Oregon and chairman of the House Ways and Means Committee, introduced legislation to allow the Treasury to auction a new type of security: zero-coupon bills with a maturity of up to one year, at a discount from face value.28 Secretary Mellon observed that the new securities could be sold as funds were needed and set to mature when taxes were actually received. To allay the fears of small bankers, he stated that the proposed program was intended to supplement, rather than replace, the existing system: “It is not the purpose of the Treasury Department . . . to discontinue the present depositary method, or system of short-term financing, but rather to supplement it with the new system, using both as may prove to be most advantageous to the interests of the government.”29 Congress acted quickly to approve the proposed legislation and President Herbert Hoover signed it into law in midJune, 1929.30 The First Bill Auction The Treasury announced the first bill auction on Tuesday, December 10, 1929, offering $100 million of 90-day bills for settlement on Tuesday, December 17, to mature Monday, March 17, 1930.31 Auction tenders were due by 2 pm on Friday, December 13.32 Each tender was to state the amount bid for and a bid price, specified as a percent of face amount with three digits of precision to the right of the decimal point33 (e.g., 99.172 percent). Participants could submit 28. “Treasury for Sale of Notes Below Par,” New York Times, April 23, 1929, p. 27. The Treasury needed new statutory authority to issue bills because existing statutes did not allow the sale of Treasury securities at a price less than par. Section 5 of the Second Liberty Bond Act authorized the Secretary of the Treasury to issue certificates of indebtedness “at not less than par.” Section 1 of the same act authorized the issue of bonds and required that they “first be offered at not less than par as a popular loan.” The Victory Liberty Loan Act authorized the issue of notes “at not less than par.” 29. 1929 Treasury Annual Report, pp. 273–75. See also the April 24, 1929, speech of Under Secretary of the Treasury Ogden Mills before the American Institute of Banking (1929 Treasury Annual Report, p. 275–80, and “Mills Explains Aim of Treasury Bills,” New York Times, April 25, 1929, p. 42) and Mills’s testimony before the House Ways and Means Committee (Committee on Ways and Means, 1929). 30. Act of June 17, 1929. The act authorized the Secretary of the Treasury to issue bills, subject to a $10 billion limit on total outstanding bills and certificates of indebtedness. 31. Federal Reserve Bank of New York Circular no. 949, December 10, 1929. See also Treasury Circular no. 418, November 22, 1923, reprinted in 1930 Treasury Annual Report, p. 303, and Federal Reserve Bank of New York Circular no. 944, November 23, 1929. 32. The short three-day interval between announcement of the auction and the close of bidding indicates that Treasury officials did not intend to offer the bills as a “popular loan.” This is consistent with their expectation that bidding would be limited to large banks and securities firms. 33. Bidding on a price basis saved Treasury officials from specifying how bids in terms of interest rates would be converted to prices. Market participants used a variety of conventions. For example, the price of a bill with n days to maturity quoted at a discount rate of D is P = 100 – (n/360) × D. The price of the same bill quoted at a simple interest yield of R is P = 100/[1 + 0.01 × (n/365) × R]. In the case of a 90-day bill quoted at 4.50 percent, P = 98.875 if the quoted rate is a discount

213

Evolution of the Primary Market and the Introduction of Treasury Bills

multiple tenders with different bid prices. Tenders would be accepted in order of decreasing price and tenders at the lowest accepted price (the “stop-out” price) would be allotted securities on a pro-rata basis. Payment in “immediately available,” or “Federal,” funds, namely collected balances on deposit at a Federal Reserve Bank, would be due on the settlement date; payment could not be made by crediting a War Loan Deposit Account. On Saturday, December 14, the Treasury announced that it had received tenders for $224 million of the new bills and that it had accepted tenders starting at the highest bid (99.310 percent) and stopping at 99.152 percent, or at yields ranging from 2.82 percent to 3.47 percent.34 Tenders bidding at the stop were allotted 80 percent of the amount bid for.35 The average accepted price was 99.181. Two dealers, Salomon Brothers & Hutzler and the International Manhattan Company, won the bulk of the offering and promptly reoffered $70 million of the bills at a discount rate of 31⁄8 percent, or at a when-issued price of 99.219.36 Under Secretary Mills stated that he was “entirely satisfied” with the results of the auction and that the auction had resulted in “considerably cheaper money than we could get through the medium of certificates of indebtedness . . ..”37 He further stated that, henceforth, the government would be able to tailor its borrowings more closely to its needs: “We will not be in a position where we have to borrow a lot of money and hold it in anticipation of needs for which we have only estimates . . .. We can sell government bills to fit the immediate needs and make the maturities fit into a known time of income, that is, the tax-paying dates.”38 Subsequent Bill Auctions The Treasury auctioned bills on three additional occasions before the end of fiscal year 1930 (table 14.2). Consistent with the statements of officials that rate, namely if D = 4.50 percent, and P = 98.903 if the quoted rate is a simple interest yield, namely if R = 4.50 percent. See letter, October 5, 1929, from W. S. Broughton, Commissioner of the Public Debt, Treasury Department, to Leslie Rounds, Deputy Governor, Federal Reserve Bank of New York, and reply dated October 8, 1929, Federal Reserve Bank of New York file 413.7. 34. “Plan Further Use of Treasury Bills,” New York Times, December 15, 1929, p. 16. The yields are expressed in terms of simple interest using a 365-day year, so that 99.310 = 100/[1 + (90/365) × 0.0282] and 99.152 = 100/[1 + (90/365) × 0.0347]. 35. “Treasury Bills Well Received,” Wall Street Journal, December 17, 1929, p. 12. 36. “3 1-8% Discount Rate on Treasury Bills,” New York Times, December 16, 1929, p. 47. The International Manhattan Company was the securities division formed following the March 1929 merger of the International Acceptance Bank, Inc. and the Bank of Manhattan. The when-issued price is computed as 99.219 = 100 – (90/360) × 3.125. 37. “Issue Satisfies Mills,” Wall Street Journal, December 16, 1929, p. 11, and “Plan Further Use of Treasury Bills,” New York Times, December 15, 1929, p. 16. 38. “Plan Further Use of Treasury Bills,” New York Times, December 15, 1929, p. 16.

214

Chapter 14

Table 14.2 Treasury bill auctions in fiscal year 1930

Auction close

Issue

Maturity

Term (days)

Dec 13, 1929 Feb 14, 1930 Apr 11, 1930 May 15, 1930

Dec 17 Feb 18 Apr 15 May 19

Mar 17 May 19 Jul 14 Aug 18

90 90 90 91

Amount ($millions)

Bid prices (% of face amount)

Offered

Bid

Issued

Highest

Stop

Average

100 50 50 100

224 186 132 276

100 56 51 105

99.310 99.250 99.315 99.400

99.152 99.125 99.250 99.331

99.181 99.174 99.267 99.356

Source: 1930 Treasury Annual Report, pp. 285, 287–88, and 291–93.

they would issue bills to fund disbursements between quarterly tax dates, none of the three bills was issued on or near a tax date. Additionally, however, none of the three bills matured shortly after a tax payment date. This was inconsistent with earlier statements by Mellon and Mills that bills would be issued to mature on dates on which tax payments actually came in. Further the decision to refinance the bill maturing on May 19, 1930, with a new 91-day bill maturing on August 18, 1930, presented the prospect that the Treasury might establish self-sustaining bill cycles. Thus, although bills appeared to be headed toward a permanent place in Treasury debt management, it was unclear exactly how they would be used. An Unanticipated Defect in Design The first several bill issues turned up an unanticipated defect in the design of the new securities. The Act of June 17, 1929, that authorized the Treasury to issue bills provided that the original issue discount on a bill was interest income and would be exempt from federal, state, and local income taxes. However, the Commissioner of Internal Revenue subsequently announced that any gain upon sale prior to maturity, computed net of accrued interest, was taxable.39 To facilitate compliance with this interpretation, each Treasury bill stated on its face the discount at which it had been issued. The decision to print the original issue discount on the face of a Treasury bill meant that bills maturing on the same date were not fungible if they were originally issued at different auction prices. Thus a buyer had to keep track of the original issue discount on each bill that it bought. This imposed a substantial record-keeping burden on market participants, and investors soon grew reluctant to buy the new securities. One dealer complained that “whereas the market anticipated that the United States Treasury bill would become the 39. Treasury Decision 4276, November 22, 1929, reprinted in 1930 Treasury Annual Report, p. 306.

215

Evolution of the Primary Market and the Introduction of Treasury Bills

premier security of the world and the most easily traded in, it is in fact today the least popular of all United States issues.”40 Treasury and Federal Reserve officials understood the problem: a security that was intended to be valued for its liquidity was in fact distressingly illiquid. The chairman of the Board of Directors of the Federal Reserve Bank of New York observed that: The sale of Treasury short-term obligations on favorable terms is dependent upon a group of traders or dealers in short-term investments who always stand ready to buy or sell obligations of this sort. The desirability of the Treasury bill depends on the holders’ being able to liquidate it at any time at a fair price, and these dealers constitute the market where this is always possible. … We have been hoping that the United States Treasury bill would take its proper place in this market and become a desired investment for short-term investments by banks, corporations and individuals. But the fact is that the dealers now find the market almost closed to these new Treasury bills, solely on the ground of the bookkeeping complications which necessitate such an enormous amount of detail that prospective buyers refuse to take them. . . . These difficulties are so great that a number of important buyers of Treasury obligations are withdrawing altogether from the purchase of the bills, and I am convinced that unless the present law can be modified the Treasury may presently have difficulty in continuing this method of financing on a satisfactory basis.41

At the behest of Treasury officials, Congress acted to cure the design defect by exempting gains on sales of Treasury bills from taxation.42 The next succeeding auction, on July 10, 1930, attracted a record $329 million of bids for an offering of $50 million of bills. One dealer attributed the “great success” of the offering to the change in tax treatment.43 At the end of 1930 the Treasury further enhanced the attractiveness of Treasury bills when it authorized wire transfers of bills between Federal Reserve banks.44 40. Letter, April 17, 1930, from Ernest Wagner, president, Discount Corporation, to J. Herbert Case, chairman of the Board of Directors, Federal Reserve Bank of New York, Federal Reserve Bank of New York file 413.7. See also 1930 Treasury Annual Report, p. 23 (“the bookkeeping records required in order to calculate gains, as differentiated from exempt interest, were so complicated that a very real sales resistance resulted”) and Beckhart, Smith, and Brown (1932, p. 354) (noting “the large amount of burdensome bookkeeping necessary to comply with the terms of the Act of June, 1929”). 41. Letter, April 21, 1930, from J. Herbert Case, chairman of the Board of Directors, Federal Reserve Bank of New York, to Ogden Mills, Under Secretary of the Treasury, Federal Reserve Bank of New York file 413.7. 42. Act of June 17, 1930. See “Tax Bill to Senate,” Wall Street Journal, June 7, 1930, p. 12 (noting that the Treasury Department “has advocated modification of the law to prevent the difficulty in continuing [bill] financing on a satisfactory basis. In a communication to Congress, Treasury pointed out that on the last issue of Treasury bills there were no less than 17 different rates of discount, which represented the different competitive bids that were accepted.”). See also Treasury Circular no. 418, as amended, June 25, 1930, reprinted in 1930 Treasury Annual Report, p. 309, and Treasury Decision no. 4292, June 25, 1930, reprinted in 1930 Treasury Annual Report, p. 313. 43. Letter, July 15, 1930, from Ernest Wagner, president, Discount Corporation, to J. Herbert Case, chairman of the Board of Directors, Federal Reserve Bank of New York, Federal Reserve Bank of New York file 413.7. 44. Smith (1956, p. 88).

216

Chapter 14

Summary The early 1920s witnessed the evolution of the primary market for Treasury securities from a wartime market, where the Treasury sold unlimited quantities of securities in fixed-price offerings, to a peacetime market, where it sold limited quantities of securities on a regular quarterly basis and scaled allotments in favor of small investors and investors rolling over maturing investments. Three defects in the system came to light: (1) securities were chronically underpriced, (2) the Treasury sometimes had to borrow well in advance of its projected needs, and (3) allowing banks to pay for new issues by crediting War Loan Deposit Accounts allowed banks to substitute War Loan deposits for other, more expensive, sources of funds. The Treasury initiated auction offerings of Treasury bills in late 1929 to correct those defects.

15

Coda on Treasury Debt Management during the 1920s

The 1920s did not present a funding emergency like that faced by Treasury Secretary McAdoo during World War I and Secretary Mellon did not have the opportunities to demonstrate his ingenuity that McAdoo had. Mellon nevertheless made good use of his tenure by paying attention to detail and fostering the evolution of what had once been a “project finance” market into the sort of liquid national securities market needed to sustain a more or less permanent national debt. At the end of June 1919, interest-bearing Treasury debt amounted to $25.2 billion, including $3.5 billion in certificates of indebtedness, $4.5 billion in notes, and $17.2 billion in bonds.1 By mid-1930, the debt had been cut to $15.9 billion as a result of strong tax receipts (even in the face of repeated tax cuts) and tight expenditure controls. Mellon and his Under Secretaries, S. Parker Gilbert, Garrard Winston, and Ogden Mills, engineered three important developments in Treasury finance: 1. They refinanced (with intermediate-term notes) $2.5 billion of short-term debt that threatened to clog the Treasury market in early 1921. 2. They developed machinery appropriate for applying budget surpluses to debt reduction in an orderly fashion, redeeming three Liberty Loans between 1922 and 1928 with a thoughtfully constructed mix of early retirements, exchange offerings, cash redemptions at call or maturity, and regular quarterly refinancings. 3. They introduced Treasury bills in 1929 and, in the process, began to revive auction offerings of Treasury securities. The 1920s also witnessed the evolution of a bank-financed market for shortterm Treasury debt, rigidly constrained to trade at par, into a liquid and unconstrained secondary market. 1. 1919 Treasury Annual Report, pp. 186, 221, and 551.

IV

THE GREAT DEPRESSION

16

Treasury Finance during the Great Depression

The Great Depression was the central economic event of the 1930s. The National Bureau of Economic Research places a peak in business activity in August 1929 and the following trough more than three and a half years later, in March 1933. Between 1929 and 1933, gross national product fell 44 percent, from $204 billion to $142 billion (in constant, 1958, dollars), employment fell 18 percent, from 46 million to 38 million, and the unemployment rate rose from 3.2 percent to 25 percent. Over the next four years, production and employment recovered to 1929 levels—GNP (in 1958 dollars) was $203 billion and total employment was 46 million by 1937—but labor force growth left the unemployment rate mired at 14 percent. A second contraction between mid-1937 and mid-1938 reduced aggregate production to $193 billion and pushed the unemployment rate back up to 19 percent. The Great Depression fostered an unprecedented expansion in peacetime federal spending. Coupled with low tax receipts, the increased spending produced chronic budget deficits that changed the focus of Treasury finance from debt redemption to deficit finance. This chapter examines the consequences of the Great Depression for Treasury tax receipts, federal expenditure policies, and Treasury indebtedness. Treasury Tax Receipts Figure 16.1 shows that Treasury receipts followed the course of economic activity in the 1930s: declining sharply as a result of the Great Contraction of 1929 to 1933, rising during the subsequent recovery, and then falling modestly in the wake of the second, 1937 to 1938, contraction. Income tax receipts followed the course of the economy, albeit with a lag attributable to the one-year delay in collection. (The tax on income earned in a given calendar year was due in four quarterly installments during the following year.) Business activity did not turn down dramatically until late in 1929, so income tax receipts remained strong through the end of 1930. However, sharply lower incomes in 1930 led to a comparably sharp decline in 1931 receipts. The result was a significant (but not catastrophic) fall in fiscal year 1931 receipts (with strong receipts in the second half of calendar 1930 offset by weaker receipts in the first half of calendar 1931), followed by a further decline in fiscal 1932 receipts. Britain’s Departure from the Gold Standard and the Revenue Act of 1932 Treasury receipts rose marginally in fiscal 1933, and more sharply the following year, as a result of tax increases mandated by the Revenue Act of June 6,

Chapter 16

7 6 5

Billions of dollars

4 3 2 1

1939

1938

1937

1936

1935

1934

1933

1932

1931

1930

1929

1928

0 1927

222

Fiscal year Social Security wage taxes

Other receipts

Income taxes

Figure 16.1 Treasury receipts

1932. That act was a direct consequence of the September 1931 decision by Great Britain to go off the gold standard.1 Britain’s decision to abandon the gold standard provoked an internal currency drain from U.S. banks and precipitated a nationwide wave of bank failures as worried depositors sought to convert their bank deposits to cash.2 Foreign holders of dollar-denominated assets, fearing that America might follow Britain, began to convert their dollar assets to gold and to repatriate the gold,3 exposing the U.S. banking system to an external gold drain that compounded the internal currency drain. 1. Ahamed (2009, pp. 404–30) provides a compelling account of the sequence of events that began with the failure of Credit Anstalt, the oldest, largest, and most prestigious bank in Austria, in May 1931, continued with the collapse of the German banking system, and ended with Britain going off the gold standard. 2. There had been two earlier waves of bank failures, one between November 1930 and January 1931 triggered by the failure of a Tennessee investment bank (Wicker 1996, ch. 2) and another between April and August 1931 that affected banks in Chicago and Toledo, Ohio (Wicker 1996, ch. 3). Wicker (1996) states (p. 77) that “a full-fledged nationwide banking crisis erupted when Britain announced its departure from gold convertibility. . . . Prior to that date bank suspensions . . . had been region specific,” and (p. 62) that “Bank failures . . . and increased hoarding immediately accelerated after the British announcement.” 3. Wicker (1996, p. 86) (“The reduction in the U.S. monetary gold stock commenced immediately after Britain announced its departure from the gold standard . . ..”). See also Chandler (1971, ch. 11).

223

Treasury Finance during the Great Depression

In a classic response to the outflow of gold, the Federal Reserve Banks raised their discount rates.4 The New York Fed raised its discount rate from 1½ percent to 2½ percent on October 9, and then to 3½ percent on October 16. Open market rates rose in parallel. Yields on 3- to 6-month Treasury certificates of indebtedness rose from 0.42 percent in August 1931 to 2.48 percent in January 1932. Treasury bond yields rose from 3.18 percent to 4.26 percent over the same interval. President Herbert Hoover, anxious to avoid a reduction in credit available to business borrowers, proposed a tax increase (in his December 1931 budget message) in an attempt to arrest the growing federal budget deficit and restore confidence in the dollar.5 The resulting legislation, the Revenue Act of 1932, effected “one of the largest increases in taxes ever imposed by the Federal Government in peace times.”6 As shown in table 16.1 and figure 16.2, the act restored the normal individual income tax rates of the Revenue Act of 1918 and very nearly restored surtax rates on net incomes less than $100,000, thereby undoing most of what Andrew Mellon had accomplished in the 1920s. The act also raised the corporate income tax rate to 13¾ percent. The new taxes went into effect at the beginning of calendar year 1933, leading (as shown in figure 16.1) to a modest increase in “other” tax receipts in fiscal 1933 (due to a variety of excise tax increases in the 1932 act) and a larger increase the following year. Income tax receipts rose marginally in fiscal 1934 (due to the greater collections in the first half of calendar 1934 that resulted from increased assessments on income earned in calendar 1933) and continued to strengthen thereafter as the economy gradually recovered. Revenue Acts during the New Deal The most important federal tax legislation during the New Deal was the introduction of a wage tax to support the old-age insurance program authorized by the Social Security Act of 1935. Figure 16.1 shows the contribution of Social Security wage taxes in the second half of the 1930s. (Chapter 17 describes the 4. Meltzer (2003, p. 345). Meltzer (2003, p. 273) characterizes the discount rate increases as “required” by the “gold standard rules.” A 1935 Federal Reserve staff memo (cited in Meltzer 2003, p. 493n. 161) concluded that “The rate increase probably served more to add to the deflationary movement of succeeding months than to check the gold outflow,” but Meltzer (2003, p. 356) points out that the System’s actions “came when the gold outflow was large, and no one could predict how long the outflow would last or how large it would be.” 5. “Text of the President’s Budget Message with Recommendations for New Revenues,” New York Times, December 10, 1931, p. 16. See also Stein (1969, p. 34) (“Continued government borrowing to finance deficits, or even the prospect of it, held several dangers. It would depress security prices and raise interest rates further, increasing the difficulty of private businesses in borrowing, reducing the value of bank assets, and further straining confidence in the banks.”). 6. 1932 Treasury Annual Report, p. 21. See also Brown (1956, p. 868) (“The scope of the [Revenue Act of 1932] was clearly the equivalent of major wartime enactments.”) and Blakey and Blakey (1932).

224

Chapter 16

Table 16.1 Individual income tax assessments Revenue Act of 1918a

Revenue Act of 1926

Revenue Act of 1932

Normal tax

4% of first $4,000 of net income in excess of credits and exemptions, 8% of balance

4% of first $4,000 of net income in excess of credits and exemptions, 8% of balance

Personal exemptions

$1,000 individual, $2,000 married couple, $200 per dependent

1½% of first $4,000 of net income in excess of credits and exemptions, 3% of next $4,000, 5% of balance $1,500 individual, $3,500 married couple, $400 per dependent

$5,000–6,000 1%

$10,000–14,000 1%

$6,000–10,000 1%

52% 56 60 63 63 64 64 65

20% 20 20 20 20 20 20 20

48% 49 50 51 52 53 54 55

Surtaxes Lowest bracket, net income Rate in lowest bracket Upper bracket rates: $100,000–150,000 150,000–200,000 200,000–300,000 300,000–400,000 400,000–500,000 500,000–750,000 750,000–1,000,000 Over $1,000,000

$1,000 individual, $2,500 married couple, $400 per dependent

Source: Revenue Act of 1918, Revenue Act of 1926, and Revenue Act of 1932. a. Effective for 1919 tax year.

80

Percent

60

40

20

0 1

10

100

1,000

Net income, thousands of dollars Revenue Act of 1918

Figure 16.2 Marginal surtax rates

Revenue Act of 1926

Revenue Act of 1932

225

Treasury Finance during the Great Depression

Social Security wage tax in the course of a broader discussion of trust funds and special non-marketable Treasury securities.) Outside of the Social Security wage tax, Treasury tax receipts during the New Deal were driven primarily by fluctuations in aggregate economic activity; tax legislation had no more than a minor effect on aggregate revenues.7 Two revenue acts were nevertheless memorable, albeit more for political than economic reasons. The Revenue Act of 1935 The Revenue Act of August 30, 1935, was probably the most controversial tax measure signed into law during the New Deal—even though it has been characterized as “financially insignificant”8—because it substantially raised individual income surtax rates on net incomes in excess of $100,000. The higher upper-bracket surtax rates earned Roosevelt the lasting enmity of wealthy Americans. Historian William Leuchtenburg observed that the Revenue Act of 1935 was “the first which reached directly into the pockets of the wealthy [and] raised an outcry from business and the press such as had greeted none of the President’s previous recommendations.”9 Harold Ickes, Roosevelt’s Secretary of the Interior, concluded that the tax increase made “a bitter enemy out of practically everyone” among the “very rich.”10 The Revenue Act of 1936 The Revenue Act of June 22, 1936, introduced a surtax on undistributed corporate income that was wildly unpopular with business leaders because it represented direct federal intrusion into managerial decision-making. Historian Mark Leff concluded that: The 1936 Revenue Act challenged business’s most sacred prerogative, the employment of its own profits, the perceived engine of economic growth. It constricted the authority of the politically potent corporate managerial class. Each decision to obtain new equipment, redeem a debt, or restore backup funds as insurance against future slumps imposed a tax penalty. The insinuation of a government partner into such decisions seemed more onerous and disruptive than the previous simple assurance that higher profits would bring standard higher levies.11

7. In addition to the Revenue Acts of 1935 and 1936, discussed in the text below, there were revenue acts in 1934, 1937, 1938, and 1939. Leff (1984, pp. 65, 194, 263, and 275) concludes that the 1934 act “collected . . . little additional revenue,” that the 1937 act, aimed at closing loopholes in the tax law, “was decidedly minor legislation,” that the macroeconomic effects of the 1938 act were “marginal,” and that “in the big picture, the tax revisions made [by the 1939 act] were insignificant.” 8. Leff (1984, p. 93). 9. Leuchtenburg (1963, p. 152). 10. Quoted in Brownlee (2004, p. 96). 11. Leff (1984, pp. 208–209).

226

Chapter 16

Table 16.2 Corporate income tax rates prescribed by the Revenue Act of 1936 Tax on net income Net income $0–2,000 2,000–15,000 15,000–40,000 40,000+ Surtax on undistributed net income Undistributed net income 0–10% of net income 10–20 20–40 40–60 60+

Tax rate 8% of net income 11 13 15 Tax rate 7% of undistributed net income 12 17 22 27

Source: Revenue Act of 1936.

The surtax, shown in table 16.2, was intended to give firms an economic incentive to distribute profits to shareholders. Leff observes that New Dealer Rexford Tugwell favored the tax as a device “designed to ‘liberate’ purchasing power by rechanneling sterile corporate surpluses to stockholders.”12 Wealthy stockholders, would, of course, be subject to the high marginal surtax rates prescribed by the Revenue Act of 1935 on their dividend income. The undistributed profits tax failed to produce revenues anything like what Treasury officials had projected.13 Congress reduced the tax rate to 2½ percent of undistributed net income in the Revenue Act of May 28, 1938, and eliminated the tax altogether in the Revenue Act of June 29, 1939.14 Treasury Expenditures during the Great Contraction During the first two years of the Great Contraction, President Hoover attempted to deal with the decline in economic activity by encouraging greater state and local government spending on construction, with only a de minimis federal contribution. He was an active proponent of using public works to reduce unemployment, but he also believed that federal spending was too small and could not be expanded fast enough to have much impact. Federal expenditures consequently increased in fiscal 1930 and 1931 no faster than they had increased during the late 1920s. (Figure 16.3 shows Treasury expenditures other than for debt retirement from 1927 to 1939.) 12. Leff (1984, p. 56). 13. Blakey and Blakey (1937, p. 698). 14. Blakey and Blakey (1938, 1939).

227

Treasury Finance during the Great Depression

10

Billions of dollars

8

6

4

1939

1938

1937

1936

1935

1934

1933

1932

1931

1930

1929

1928

0

1927

2

Fiscal year RFC and Land Bank equity subscriptions (1932) and veterans' bonuses (1936 and 1937) Unemployment relief Other expenditures (other than for debt retirement)

Figure 16.3 Treasury expenditures (other than for debt retirement)

Birth of the Reconstruction Finance Corporation State and local construction expenditures increased in 1931, but the increase was more than offset by a concurrent decline in private investment. Hoover attributed the decline to the growing number of bank failures that cut off credit to business borrowers and demoralized credit officers at banks that remained open.15 In early October 1931, shortly after the beginning of the wave of bank failures that followed Britain’s decision to leave the gold standard, Hoover invited a group of prominent bankers to organize a private credit pool to lend to distressed and illiquid, but still solvent, banks, arguing that public confidence would be enhanced if the country’s banks were seen to be acting cooperatively.16 The bankers believed that the deepening depression was beyond their control and that federal action was needed,17 but nevertheless acceded to 15. Olson (1977, pp. 12, and 24–25). 16. Olson (1977, pp. 25–26) and Meltzer (2003, pp. 345–46). 17. Nash (1959, p. 463) and Olson (1977, p. 26).

228

Chapter 16

the President’s request and, on October 13, 1931, announced the formation of the National Credit Corporation. Any bank subscribing 2 percent of its net deposits could join the new credit agency, which was authorized to borrow up to $1 billion in the private credit markets, and apply for financial assistance.18 Hoover’s hope that the private effort would stimulate business activity foundered on a continuing wave of bank failures, and he was forced to accept the bankers’ demand for a federal credit agency.19 The Reconstruction Finance Corporation (RFC) was authorized, by the Reconstruction Finance Corporation Act of January 22, 1932, to lend to banks and other financial intermediaries, as well as to railroads.20 The new agency was capitalized with $500 million of equity subscribed by the Treasury and was authorized to borrow an additional $1.5 billion from the Treasury.21 Figure 16.3 shows the $500 million 18. Olson (1977, p. 27). Olson (1977, pp. 29 and 31–32) observes that officials of the National Credit Corporation were reluctant to extend credit on other than ample collateral and delayed making loans in the hope that the banking crisis would abate on its own. See also Nash (1959, p. 465) (“From the very start the operation of the National Credit Corporation was slow and cumbersome.”). 19. Nash (1959, pp. 465–66) and Olson (1977, p. 29). Hoover was hardly enthusiastic in his support of the new agency. In his December 1931 State of the Union message he stated that “It may not be necessary to use such an instrumentality very extensively. The very existence of such a bulwark will strengthen confidence,” and suggested that “it should be placed in liquidation at the end of two years.” “Text of President Hoover’s Annual Message as Read to Congress Yesterday,” New York Times, December 9, 1931, p. 21. 20. Mason (2001) gives a good overview of RFC programs in the 1930s. See also Ebersole (1933). Lending to railroads was included in order to prop up deficit-plagued rail lines and the prices of their bonds. See “Reconstruction Corporation: Its Aim and Plan of Work,” New York Times, January 17, 1932, p. XX4 (“One of the paramount features of the [RFC] concerns loans to the railroads, the only industry which would receive benefits independently of financial institutions. The railroads are made special exceptions because of the widespread ownership of railroad bonds by insurance companies, savings banks, national banks and trust companies, and individuals.”). Railroad bonds had been a popular investment in the 1920s and their falling market values jeopardized the balance sheets of numerous financial institutions. Friedman and Schwartz (1963, pp. 355–56), Chandler (1971, pp. 179–80), Olson (1977, pp. 10–11, 30, and 34), and Mason (2001, p. 178). Meltzer (2003) observes (p. 332) that “Many banks had been forced to close because the decline in the market value of their bond portfolios made them insolvent.” and (p. 346n. 80) that “Railroad bonds posed the main problem. During the 1920s, small banks with insufficient local loan demand bought railroad bonds to increase earnings. Also many banks invested savings deposits in bonds. . . . Interest payments became uncertain as railroad earnings declined, so bond prices fell.” Nash (1959) traces the origins of the RFC to the War Finance Corporation (WFC), created by the War Finance Corporation Act of April 5, 1918, to “provide credits for industries and enterprises . . . necessary or contributory to the prosecution of [World War I].” The wartime activities of the WFC were quite limited. By the end of the war the WFC had made only $68 million in loans and had only $38 million of loans outstanding ($30 million having been repaid); 1918 Treasury Annual Report, p. 58. The WFC was more important in the postwar period, particularly in purchasing Liberty Loans and certificates of indebtedness to support the market for Treasury securities. See 1919 Treasury Annual Report, pp. 82 and 106. See generally, 1918 Treasury Annual Report, p. 56, 1919 Treasury Annual Report, p. 103, and Olson (1977, pp. 12, 24–26, 41–43, and 45). 21. An early version of the Reconstruction Finance Corporation Act contemplated that RFC loans would be funded from public sales of RFC notes and bonds, but the final version of the act authorized the RFC to borrow directly from the Treasury. Olson (1977, pp. 36–37). See also “Glass

229

Treasury Finance during the Great Depression

Treasury subscription in fiscal 1932,22 as well as a concurrent $125 million equity subscription to the Federal Land Banks intended to support an expansion in Land Bank lending to farmers.23 By the end of July 1932, the RFC had lent more than $500 million to banks, more than $270 million to railroads, and about $150 million to insurance companies and other financial intermediaries.24 Bank lending remained weak in the late spring of 1932 and Hoover concluded that more direct federal action was needed.25 In mid-May he recommended expanding the RFC’s mandate to allow the agency to lend to states for unemployment relief and to finance construction of income-producing, “self-liquidating” public works such as toll bridges and tunnels.26 (Lending to state governments to finance unemployment relief was a break from Hoover’s previous position that the responsibility for relief rested solely with state and local governments. Financing “self-liquidating” public works was seen as a way for the federal government to more directly stimulate employment without appearing to sponsor uneconomical “make-work” projects.) Congress incorporated both proposals into the Emergency Relief and Construction Act of July 21, 1932. Primarily as a result of continuing growth in the original RFC loan programs, and secondarily as a result of the Emergency Relief and Construction Sees Death of Discount Clause,” New York Times, January 19, 1932, p. 4, and “$2,000,000,000 Bill for Reconstruction Ready for Passage,” New York Times, January 21, 1932, p. 1. 22. The Treasury also purchased between $200 million and $300 million of RFC debt in fiscal 1932, but—in a bit of budgetary legerdemain—this expenditure was not recorded as a budget expenditure. See 1932 Treasury Annual Report, p. 70 (“. . . payments against credits established through the purchase of the corporation’s notes by the Treasury were treated as public debt transactions, not chargeable against ordinary receipts”) and 1940 Treasury Annual Report, pp. 39–40 (“Pursuant to section 9 of the [Reconstruction Finance Corporation Act], expenditures made by the Corporation from funds derived from the sale of its obligations to the Secretary of the Treasury were treated as public debt transactions and, therefore, were not included as expenditures in the Budget. This provision was based upon the assumption that the money loaned by the Reconstruction Finance Corporation from the proceeds of such obligations, when repaid, would be available for the retirement of the Corporation’s obligations held by the Treasury and used by the Treasury to retire an equivalent amount of public debt obligations.”). The Treasury initially reported $268 million of RFC debt purchases in fiscal 1932, but later revised the figure down to $206 million. Compare 1932 Treasury Annual Report, p. 70 (reporting $268 million in note purchases) with 1934 Treasury Annual Report, p. 6 (reporting $206 million in purchases of RFC securities other than capital stock). The RFC quarterly report for the quarter ending June 30, 1932, records $350 million of notes payable to the U.S. Treasury and $51 million of cash balances held by the U.S. Treasury, for a net liability to the Treasury of $299 million. 23. The $125 million equity subscription to the Federal Land Banks was authorized by the Act of January 23, 1932. 24. 1933 Federal Reserve Board Annual Report, pp. 6–7. See also Upham and Lamke (1934, ch. 11). 25. Olson (1977, pp. 62 and 66). 26. “Hoover Urges 3-Point Relief Plan of $1,500,000,000 to Use as Loans,” New York Times, May 13, 1932, p. 1.

230

Chapter 16

Act, total RFC lending expanded significantly in fiscal 1933, with the Treasury supplying an additional $1.3 billion to the new agency.27 The Banking Crisis of 1933 and the Role of the RFC in Reopening the Banks The U.S. banking system had already collapsed by the time Franklin Roosevelt was sworn in as President on Saturday, March 4, 1933. What Allan Meltzer describes as “the final bank runs”28 started a month earlier, on February 14, when the collapse of negotiations over a loan from the RFC to a group of Michigan banks led the governor of Michigan to declare a statewide “bank holiday.”29 The unusual, but not unprecedented,30 action immobilized the payments system in Michigan31 and sparked a vicious cycle as depositors in other states hastened to withdraw deposits and hoard currency in anticipation of the imposition of similar restrictions on their own banks and as state banking authorities reacted by declaring the very holidays that depositors were trying to avoid.32 At the same time, the banking system faced a renewed external gold drain as foreigners accelerated their conversion of dollar-denominated assets 27. The Treasury initially classified the $1.3 billion as explicitly financing RFC operations (1933 Treasury Annual Report, p. 7) but later distributed the expenditures to functional categories (1938 Treasury Annual Report, p. 418), reflecting more accurately the role of the agency as an instrumentality of the federal government. (Further information on RFC accounting during the New Deal appears in the 1940 Treasury Annual Report, p. 40.) See also “The ‘R.F.C.’ Now Functioning as a Four-Billion-Dollar Bank,” New York Times, July 24, 1932, p. XX4. 28. Meltzer (2003, p. 377). 29. The events leading up to the Michigan bank holiday are described in Sullivan (1936), Ballantine (1948), Awalt (1969), Kennedy (1973), and Wigmore (1985). See also Wicker (1996, pp. 116–20). The Michigan bank holiday was originally scheduled to last for eight days but continued until the beginning of the national bank holiday on March 6. 30. Andrew (1908b, p. 498) and Sprague (1910, p. 286) report that governors of several states declared bank holidays during the panic of 1907. There was also a localized bank holiday in Nevada in the fall of 1932 and a brief, one-day, holiday in Louisiana on February 4, 1933. Upham and Lamke (1934, pp. 12 and 14) and Mason (2003, pp. 107–108). 31. Alter (2006, p. 155) (“the fallout [from the Michigan bank holiday] was instantaneous. Detroit and other Michigan cities defaulted on bonds. Food grew scarce, as shelves were cleaned out in anticipation.”). 32. See Wicker (1996, pp. 125–28). Wicker (1996, p. 126) points out the rapid spread of state banking holidays: “Legislation was drafted and passed at breakneck speed to meet the threat of the deteriorating banking situation. Governors and state banking officials were immediately empowered to close all banks within a state, limit deposit withdrawals, or allow directors of individual banks at their discretion to declare temporary moratoria and limit withdrawals. . . . Each state reacted to protect its own interest without regard to its effects on others . . ..” and characterizes (p. 108) the importance of the state holidays for the diffusion of the banking panic: “The bank holiday was the mechanism for transmitting banking unrest from state to state. The declaration of a banking holiday in one state motivated depositors to withdraw deposits from out-of-state banks to meet their immediate transactions needs thereby transmitting withdrawal pressures to contiguous states and to the New York and Chicago money markets. Moreover, depositors in surrounding states became alarmed that similar deposit restrictions would be imposed in their states and would therefore rush to withdraw deposits in anticipation of a bank moratorium.”

231

Treasury Finance during the Great Depression

into gold and their repatriation of the gold,33 and it even began to experience an internal gold drain as Americans lost confidence in their currency and began to convert paper dollars to gold.34 By sunrise on Inauguration Day, banks in all forty-eight states were either operating under state-imposed restrictions on deposit withdrawals or had been closed pursuant to state banking holidays.35 The Program to Reopen the Banks Roosevelt recognized that while uncoordinated state initiatives had closed the banks, reopening the banks would require a comprehensive national plan. The first step was closing, by presidential proclamation, the few banks that remained open.36 This unprecedented action was based on section 5(b) of the Trading with the Enemy Act of October 6, 1917, which provided, in relevant part, “that the President may investigate, regulate, or prohibit, under such rules and regulations as he may prescribe, by means of licenses or otherwise, any transactions in foreign exchange, export or earmarkings of gold or silver coin or bullion or currency . . ..”37 In addition to reciting this statutory authority,38 the proclamation asserted the factual basis for Roosevelt’s action: that “there have been heavy and unwarranted withdrawals of gold and currency from our banking institutions for the purpose of hoarding,” as well as “continuous and increasingly extensive speculative activity abroad in foreign exchange [that] has resulted in severe drains on the nation’s stocks of gold.”39 The program to reopen the banks was cobbled together in a matter of just a few days, primarily by holdovers from the Hoover administration who stayed 33. Wicker (1996, p. 129–30), Wigmore (1987), and Meltzer (2003, p. 381). 34. 1933 Federal Reserve Board Annual Report, p. 8. See also Friedman and Schwartz (1963, p. 326) (“For the first time also, the internal drain partly took the form of a specific demand for gold coin and gold certificates in place of Federal Reserve notes or other currency.”). 35. Wicker (1996, p. 128). Meltzer (2003, p. 382) reports that by March 4, thirty-five states had declared a bank holiday that closed all banks in the state. Silber (2009, p. 22) discusses the intricacies of identifying states that declared bank holidays. See also “Status of Banking Restrictions by States,” New York Times, March 5, 1933, p. F24. 36. “The President’s Bank Proclamation,” New York Times, March 6, 1933, p. 1. The Federal Reserve Board’s 1933 Annual Report noted (p. 10) that, “An important purpose of [the national bank holiday] was to attack the problem of bank failures comprehensively by reviewing at one time the condition of all banks . . ..” Roosevelt’s original proclamation provided that the holiday would end after Thursday, March 9, but the holiday was subsequently extended indefinitely. See “The President’s Proclamation,” New York Times, March 10, 1933, p. 1. 37. 1933 Treasury Annual Report, pp. 194–95. 38. Interestingly the proclamation misquoted the Trading with the Enemy Act, adding some significant additional language in saying that the act provided “that the President may investigate, regulate, or prohibit, under such rules and regulations as he may prescribe, by means of licenses or otherwise, any transactions in foreign exchange, and the export, hoarding, melting, or earmarkings of gold or silver coin or bullion or currency . . ..” 1933 Treasury Annual Report, p. 194, additional words in italics. 39. Awalt (1969, pp. 364–65n. 4) discusses why the Trading with the Enemy Act happened to include language that supported the declaration of a bank holiday.

232

Chapter 16

on to help resolve the crisis.40 Their efforts received added urgency from the need to refinance $700 million of Treasury debt maturing on Wednesday, March 15.41 The holdover Under Secretary of the Treasury, Arthur Ballantine, feared that missing the payment would “ruin the government.”42 The reopening program had two key elements: 1. it reopened solvent and well-capitalized banks promptly, and in the process controlled the currency and gold drains that had exacerbated the crisis,43 and 2. it added new capital to, and ultimately reopened, banks with impaired capital that were not insolvent. Hopelessly insolvent banks would either be liquidated or reorganized.44 In an Executive Order dated Friday, March 10, Roosevelt authorized Secretary of the Treasury William Woodin to license the reopening of national banks and state banks that were members of the Federal Reserve System and authorized “the appropriate authority having immediate supervision of banking institutions in each state” to license the reopening of state nonmember banks.45 The following day the President announced that Treasury officials would begin to license banks for reopening on Monday, March 13. The first banks to be reopened would be those located in the twelve Federal Reserve Bank cities. Licensing would begin the following day for banks in the 250 cities with 40. Alter (2006, p. 228) characterizes the effort as “a strange and unprecedented cooperation between administrations.” The holdovers included Ogden Mills (former Secretary of the Treasury), Arthur Ballantine (continuing as Under Secretary of the Treasury until May 15, 1933), and Francis Awalt (acting Comptroller of the Currency). George Harrison (governor of the Federal Reserve Bank of New York), Walter Wyatt (general counsel of the Federal Reserve Board), and George Davison (president of the Central Hanover Bank and Trust Company of New York and chairman of the New York Clearing House Association) also made important contributions to the reopening program. Ballantine (1948, p. 138), Olson (1988, p. 40), and Awalt (1969). 41. The Treasury refinanced the maturing securities with $470 million of 5-month certificates of indebtedness and $470 million of 9-month certificates that were issued on March 15 but that were not announced until Saturday, March 11. See “Treasury Financing After Usual Plan,” Wall Street Journal, March 14, 1933, p. 7 (noting a “delay in issuance of the call for bids”). $100 million of Treasury bills had matured and had been redeemed on Wednesday, March 1, before the major New York and Chicago banks closed, and there were no additional bills coming due before another $100 million matured on March 29. Settlement of a Friday, March 3, auction of $75 million of bills maturing June 7 was originally scheduled for Monday, March 6, but was postponed until the banks reopened. “Reserve Bank Loan for Treasury Seen,” New York Times, March 11, 1933, p. 23 (“$75,000,000 of ninety-three day Treasury bills . . . should have been paid for on March 6, but remained unpaid because of the banking holiday.”). 42. Quoted in Alter (2006, p. 228). 43. See Silber (2009). 44. See Alter (2006, pp. 229–30) (“The idea was to distinguish three classes of banks: A, B, and C. The “A” banks were basically healthy and could be reopened quickly, the “B” banks were shaky but could open in the weeks ahead if they were bolstered, and the “C” banks were insolvent and would stay closed until they were reorganized.”) 45. “Text of Regulations for Reopening Nation’s Banks,” New York Times, March 11, 1933, p. 4.

233

Treasury Finance during the Great Depression

recognized clearing house associations; licensing for all other banks would begin on Wednesday, March 15. The Treasury licensing program provided a system for assessing whether a bank was solvent and well capitalized and could be reopened promptly, or whether it needed additional capital before reopening, or whether it was hopelessly insolvent. However, the licensing program did nothing to solve the problem of the gold and currency drains, and it did not provide any additional capital for marginally solvent banks. Plugging the Gold Drain Roosevelt began to address the problem of the gold drain on Friday, March 10, when he restricted the export of gold coin, gold bullion, and gold certificates from the United States and restricted U.S. banks from paying out gold or gold certificates.46 Less than a month later, on April 5, the President went further and ordered all persons (including partnerships and corporations as well as individuals) within the United States to surrender any gold or gold certificates in excess of $100 to a Federal Reserve Bank against payment in “any other form of coin or currency coined or issued under the laws of the United States.”47 Roosevelt’s actions in plugging the gold drain took the United States off the gold standard that it had been on since before the Gold Standard Act of 1900: that the United States Treasury would coin gold bullion into coins with a gold content of 0.053750 ounces of gold (of 0.9 fineness) per dollar, and that it would deliver gold against payment in gold certificates, United States notes, or Federal Reserve notes at a price of $18.60 per ounce of gold of 0.9 fineness. (Early in 1934 the United States returned to an international gold bullion standard, whereby the Treasury undertook to buy gold bullion at a price of $35.00 per ounce (of pure gold) and to sell gold bullion to foreign central banks (but not to others) at the same price. The appendix to this chapter describes the events leading to the establishment of the international gold bullion standard.) Accommodating the Currency Drain and Preparing to Recapitalize the Banks The Emergency Banking Act of March 9, 1933—passed by Congress in seven hours as the first action of a hastily called special session—addressed the 46. Executive Order, March 10, 1933, reprinted in 1933 Treasury Annual Report, p. 196. The Secretary of the Treasury could authorize exceptions to the restrictions. On April 20 Roosevelt prohibited the export of gold and gold certificates other than as permitted by the Secretary of the Treasury to foreign official institutions. Executive Order, April 20, 1933, reprinted in 1933 Treasury Annual Report, p. 198. 47. Executive Order, April 5, 1933, reprinted in 1933 Treasury Annual Report, p. 197. The $100 exemption was revoked by order of the Secretary of the Treasury on December 28, 1933. 1933 Federal Reserve Board Annual Report, p. 27, and 1934 Treasury Annual Report, p. 194.

234

Chapter 16

problems of the currency drain and the need to provide capital for marginally solvent banks.48 Title IV of the Emergency Banking Act authorized the Federal Reserve Banks to issue Federal Reserve Bank notes against a pledge of Treasury securities or commercial paper and without any gold backing. This represented a material expansion of the existing power of the Reserve Banks to issue Federal Reserve notes against a pledge of not less than 40 percent gold and the balance in gold, eligible commercial paper, or Treasury securities. (Pledge of Treasury securities against issues of Federal Reserve notes had been only recently authorized, by the Glass–Steagall Act of February 27, 1932. Prior to passage of that act, Federal Reserve notes had to be backed by at least 40 percent gold and the balance in either gold or eligible commercial paper.49) The ability of a Reserve Bank to issue Federal Reserve Bank notes without any gold backing allowed the Banks to issue currency in virtually unlimited amounts and thus to meet whatever demands for currency might develop as commercial banks reopened for business. Title III of the Emergency Banking Act authorized national banks to issue preferred stock—something they were not previously allowed to do—in amounts approved by the Comptroller of the Currency,50 and it authorized the Secretary of the Treasury to request that the RFC purchase preferred stock from any national bank or state bank considered to be in need of additional capital.51 Title III further authorized the RFC to issue securities as needed to finance purchases of bank preferred stock.52 Reopening the Banks The initial phase of the program to reopen the banks moved ahead with remarkable speed. By the close of business on Wednesday, March 15, the Treasury had reopened 5,077 banks that accounted for $26 billion of deposits. 48. In addition to the provisions discussed below, the Emergency Banking Act confirmed the action of the President in declaring the bank holiday and extended the emergency powers of the President. This ex post congressional ratification was important because there was some question about whether the original statutory basis for Roosevelt’s proclamation of a national bank holiday, the Trading with the Enemy Act of 1917, was adequate. See Ballantine (1948, pp. 137, 138, and 139) and Awalt (1969, pp. 358). 49. See section 7 of the Act of June 21, 1917, amending section 16 of the Federal Reserve Act of 1913. 50. The preferred stock had to pay a cumulative dividend of 6 percent (of par value) per annum and had to confer unconditional voting rights on preferred shareholders. The Act of June 15, 1933, subsequently allowed national banks to issue multiple classes of preferred stock. 51. Olson (1988, pp. 37–39) and Awalt (1969) discuss the history of Title III. 52. The Emergency Banking Act also authorized the Comptroller of the Currency to appoint conservators to assist with the recapitalization and reopening of national banks with impaired capital structures and facilitated the reorganization of marginal banks by allowing a plan of reorganization to go into effect without the unanimous consent of all of a bank’s creditors. See section 207 of the act and Awalt (1969, p. 364)

235

Treasury Finance during the Great Depression

By April 12 all well-capitalized banks had been reopened, and reopenings of member banks, including both national banks and state member banks, had increased to 5,425 banks. By the end of 1933, reopenings of member banks had grown to almost 6,000 banks with deposits of $27 billion. By mid-1934 the numbers reached 6,379 banks and $29 billion of deposits.53 Congress had authorized two programs to facilitate bank reopenings: expanded issuance of Federal Reserve Bank notes and RFC purchase of bank preferred stock. The Federal Reserve Banks ended up issuing only about $200 million of Federal Reserve Bank notes.54 In his first fireside chat on Sunday, March 12, President Roosevelt bluntly stated that “Your government does not intend that the history of the past few years shall be repeated. We do not want and will not have another epidemic of bank failures.”55 Depositors quickly came to appreciate that the new administration meant to keep open the banks that it had allowed to reopen and that there was little need to hold currency beyond what was needed for hand-to-hand transactions.56 In contrast, RFC purchases of bank capital obligations (including preferred stock, capital notes, and notes secured by preferred stock) proved crucial to the reopening of marginally solvent banks. Cumulative RFC purchases of bank capital obligations rose from $43 million in mid-1933 to $815 million in mid1934.57 One study concluded that “In cases of reorganization the RFC has usually furnished half the new capital by purchasing preferred stock or similar obligations.”58 Treasury Expenditures during the New Deal Once the immediate problem of reopening the banks had been addressed, Roosevelt could turn to the more fundamental problem of alleviating economic distress. Although personally skeptical of the idea that the federal government could undertake economically justifiable public works projects significantly 53. Upham and Lamke (1934, pp. 255–58). See also 1933 Federal Reserve Board Annual Report, p. 22. 54. 1936 Board of Governors of the Federal Reserve System Annual Report, p. 108, and Friedman and Schwartz (1963, p. 421). 55. “The President’s Speech,” New York Times, March 13, 1933, p. 1. 56. See Silber (2009) on the reasons why depositors could be confident that reopened banks were unlikely to fail. In fact, only 9 reopened national banks and 6 reopened state member banks with aggregate deposits of $20 million had failed as of the end of 1933. Upham and Lamke (1934, p. 48) and 1933 Federal Reserve Board Annual Report, p. 26. 57. Upham and Lamke (1934, p. 195). See also 1933 Federal Reserve Board Annual Report, pp. 25–26. Mason (2001, p. 186) and Olson (1988, ch. 4) point out the importance of the RFC’s program for recapitalizing banks in preparation for deposit insurance coverage by the Federal Deposit Insurance Corporation. 58. Upham and Lamke (1934, p. 193).

236

Chapter 16

beyond the scale initiated during the last year of the Hoover administration, Roosevelt appreciated the importance of relieving the desperate plight of millions of Americans. In May 1933 he proposed a $3.3 billion public works program that became the basis for Title II of the National Industrial Recovery Act of June 16, 1933. Approvals for projects to be financed with NIRA funds proceeded slowly at first, largely because the economy seemed to be recovering.59 However, unemployment remained high and, with winter approaching, Roosevelt approved an emergency $800 million program of “light work” that would employ four million people.60 The work relief program introduced in the winter of 1933 to 1934 established the basic pattern of federal government spending for the balance of the 1930s: the government would provide relief to the needy but it would not seek to spend its way out of the depression.61 Figure 16.3 shows that expenditures— other than for relief and for an early distribution of bonuses promised to veterans of World War I (bonus payments are discussed in chapter 17)—did not rise much faster in the mid- and late-1930s than they had in the late 1920s and early 1930s, and that expenditures for unemployment relief did not subsequently advance much beyond the level reached in fiscal 1934. The only material variation came when spending for unemployment relief fell in fiscal 1938, as relief programs shrunk in response to the strengthening economy, and then increased again in fiscal 1939 in the wake of the 1937 to 1938 contraction.62 Deficits and Indebtedness Figure 16.4 summarizes Treasury receipts and expenditures from 1927 to 1939. The deficits between 1931 and 1933 were attributable primarily to depressed tax receipts, exacerbated in 1932 by equity subscriptions to the Reconstruction Finance Corporation and the Federal Land Banks. Tax receipts recovered between 1934 and 1937, but significantly higher relief expenditures, and the early distribution of veterans’ bonuses in 1936 and 1937, pushed the 59. See Stein (1969, pp. 55–56) (stating that, according to Harold Ickes, Roosevelt’s Secretary of the Interior, Roosevelt “did not want to be stuck with huge commitments for [ultimately unnecessary] public works spending stretched out into the future”). See also Alter (2006, p. 303) (stating that Harold Ickes slowed down spending by the National Recovery Administration because he was “terrified of corruption”). 60. Stein (1969, p. 56) (stating that Roosevelt recognized that the program “was a relatively costly way to provide assistance”). 61. Stein (1969, pp. 60–62). 62. On April 14, 1938, Roosevelt proposed a significant increase in expenditures to combat the 1937–38 recession. “Roosevelt Asks Expenditure of $5,000,000,000, War on Recession” New York Times, April 15, 1938, p. 1, and Stein (1969, pp. 109–12). See also Chandler (1971, pp. 323–25 and 333).

237

Treasury Finance during the Great Depression

10

Billions of dollars

8

6

4

2

1939

1938

1937

1936

1935

1934

1933

1932

1931

1930

1929

1928

1927

0

Fiscal year Receipts

Expenditures (other than for debt retirement)

Figure 16.4 Treasury receipts and expenditures (other than for debt retirement)

deficit higher. The deficit contracted in fiscal 1938 as a result of rising receipts and shrinking expenditures associated with the continuing recovery, but then expanded again the following year when the recession of 1937 to 1938 led to reduced income tax receipts and increased expenditures. Figure 16.5 shows that the persistent deficits of the 1930s reversed the steady decline in indebtedness that had characterized the 1920s. From mid1930 to mid-1939, aggregate indebtedness more than doubled, from $16 billion to $35 billion, and easily surpassed the previous high watermark of $25 billion reached in mid-1919. Appendix: Getting to a Gold Bullion Standard of $35 per Fine Ounce After taking the United States off the gold standard in the spring of 1933, Franklin Roosevelt had no clear idea what to do next—whether to try to restore the gold standard in its original form, reduce the gold content of the dollar to something less than the amount (0.053750 ounces of gold of 0.9 fineness) confirmed by Congress in 1900, leave the country with a nonconvertible dollar, or try something else entirely. Congress gave him room to exercise his best judgment—Title III of the Agricultural Adjustment Act of May 12, 1933, commonly known as the “Thomas amendment,” authorized (but did not require)

Chapter 16

50

40

Billions of dollars

30

20

10

1939

1938

1937

1936

1935

1934

1933

1932

1931

1930

1929

1928

0 1927

238

Figure 16.5 Treasury debt, mid-1927 to mid-1939

the President to reduce the gold content of the dollar by up to 50 percent—and he knew he could ask for additional legislation. Roosevelt’s gold policy evolved gradually during the summer and fall of 1933. The first step was his decision to put domestic recovery ahead of negotiating a return to a multinational gold standard. “The sound internal economic system of a nation,” Roosevelt told the World Monetary and Economic Conference in London in early July, “is a greater factor in its well-being than the price of its currency . . ..”63 Reflating domestic commodity prices to the level prevailing in the mid1920s soon became a leading goal of domestic recovery.64 The idea was to lighten the burden that deflation had imposed on debtors. (The consumer price index fell 27 percent between 1926 and 1933.) In his fourth “fireside chat” on 63. “Text of President’s Statement,” New York Times, July 4, 1933, p. 1. See also “Roosevelt Rebuke Stuns Gold Bloc, but Conference is Likely to Go On; President Turns to Domestic Drive,” New York Times, July 4, 1933, p. 1. The London conference is described in Chandler (1971, 277–82) and Meltzer (2003, 443–50). 64. Meltzer (2003, p. 449) (“Reflation of the domestic commodity price level became a key element in a policy of domestic recovery.”). Restoration of the price level to about the level prevailing in 1926 was frequently cited as an administration objective. See, for example, “To Set up a Gold Market,” New York Times, October 23, 1933, p. 1 (referring to restoration of “the 1926 price level” as a benchmark for assessing a proposed policy initiative), “Bankers View Plan with Deep Concern,” New York Times, October 24, 1933, p. 3 (reporting that “most monetary experts said yesterday that it was [Roosevelt’s] apparent intention to attempt to raise prices to about the 1926 level”), and “Bill Is Sent to Congress,” New York Times, January 16, 1934, p. 1 (stating that “the object of today’s message is the same that has animated the President since he took office—to restore the price scale of about 1926 . . ..”).

Treasury Finance during the Great Depression

36

35

Dollars per fine ounce

239

34

33

32

31

30 Oct 22, 1933

Nov 5, 1933

Nov 19, 1933 RFC price

Dec 3, 1933

Dec 17, 1933

Dec 31, 1933

Jan 14, 1934

London market price

Figure 16.6 Gold purchase prices quoted by the Reconstruction Finance Corporation and prices in the London gold market

Sunday, October 22, Roosevelt emphasized that “since last March the definite policy of the government has been to restore commodity price levels.”65 Following restoration, he planned to “establish and maintain a dollar which will not change its purchasing and debt-paying power during the succeeding generation.” As a first step, Roosevelt announced at the end of the October 22 fireside chat that the Reconstruction Finance Corporation would shortly begin buying gold “newly mined in the United States.” The RFC began posting a buying price three days later.66 At the end of the month Roosevelt announced that gold purchases would be extended to offshore markets as well.67 Figure 16.6 shows the buying price of gold posted by the RFC between October 25 and mid-January 1934 (when RFC gold purchases ended), as well as the U.S. dollar price of gold in the London gold market.68 (In the figure the RFC and London market prices are in dollars per fine ounce of gold, i.e., per ounce of gold free of impurities. The Gold Standard Act of 1900 specified the 65. “The President’s Speech,” New York Times, October 23, 1933, p. 1. 66. “Dollar’s Gold Value 66¢,” New York Times, October 26, 1933, p. 1. 67. “Experts at White House,” New York Times, October 30, 1933, p. 1. 68. Gold was quoted in the London market in terms of pounds sterling. The dollar price was derived by multiplying the quoted sterling price by the price of sterling in terms of dollars.

240

Chapter 16

price in terms of gold of 0.9 fineness.) The buying price posted by the RFC was frequently above the London market price because the RFC did not buy all the gold that it could have at its posted prices.69 During the life of the gold purchase program the RFC bought 695 thousand ounces of gold in the United States at a total cost of $23 million and 3.3 million ounces of foreign gold at a cost of $108 million.70 Roosevelt’s gold purchase program created immense uncertainty in a variety of markets because it was so loosely linked to the goal of reflation. Reflation required expansion of the money supply but the RFC (not being a central bank) had to finance its gold purchases by issuing short-term debt.71 The RFC gold purchases were, therefore, equivalent to sterilized, or credit-financed, gold purchases; they were better-calculated to raise the dollar price of gold than to raise the general price level.72 The admittedly experimental nature of the program exacerbated market uncertainty. The New York Times reported that “It has been suggested in wellinformed quarters that the President himself does not have a definite objective in mind, but is awaiting the result of the gold-purchase experiment with a view to using accumulated progressive experience as a guide from day to day.”73 In addition to the loose connection between program goals and program design, the gold purchase program was ineptly executed.74 Until midNovember, Roosevelt raised the RFC buying price every business day—a 69. See, for example, “First Gold Bought for RFC in France; Dollar Weakens,” New York Times, November 3, 1933, p. 1 (reporting that the administration intended to “carry on only small scale operations. Whether such relatively small purchases will accomplish the President’s purpose of keeping the gold value of the dollar at a low level, and bring about a gradual rising of commodity prices appeared uncertain to some officials.”) and “Roosevelt Holds 3-Hour Conference on Gold Purchases,” New York Times, November 13, 1933, p. 1 (reporting that “purchases of gold in foreign markets will be continued, it is understood, but with only small amounts bought from time to time.”). 70. “House Chiefs Back Gold Debate Curb; To Vote Tomorrow,” New York Times, January 19, 1934, p. 1. 71. “Dollar’s Gold Value 66¢,” New York Times, October 26, 1933, p. 1, and “Gold Price Raised 48¢ above London; Markets Decline,” New York Times, October 27, 1933, p. 1. 72. See, for example, “Experts Here Confused,” New York Times, October 23, 1933, p. 1 (reporting that “there was extreme skepticism among bankers as to the practicality of controlling the internal price level through changes in the gold value of the currency”). The idea for the gold purchase program is commonly attributed to Professor George Warren of Cornell University. See Pearson, Myers, and Gans (1957), “Warren’s Standing as Chief Treasury Tactician Clarified as Sprague, No Bench-Warmer, Quits,” Wall Street Journal, November 23, 1933, p. 8, “Laboratory of the Commodity Dollar,” New York Times, December 10, 1933, p. SM3, and “Dr. Warren Explains His Money Theory,” New York Times, January 21, 1934, p. SM3. See also Warren and Pearson (1935). 73. “Dollar’s Gold Value 66¢,” New York Times, October 26, 1933, p. 1. 74. The inept execution might have been exacerbated by the opposition of Secretary of the Treasury William Woodin and Under Secretary of the Treasury Dean Acheson to the gold purchase program. When Woodin took a leave of absence for health reasons on November 15, Roosevelt accepted Acheson’s resignation as well and appointed Henry Morgenthau, governor of the Farm

241

Treasury Finance during the Great Depression

strategy better designed to encourage forbearance on the part of sellers than actual sales.75 At first, the RFC price was above the London price, but the gap narrowed as the dollar began to depreciate against the pound. Around the middle of November market participants became concerned that Roosevelt’s gold purchase program would lead to chronic inflation in the United States and they began to flee the dollar. The result was a violent collapse of the dollar—sterling exchange hit $5.41 per pound on November 15; it had been $4.62 per pound in late October.76 The dollar price of gold in the London market spiked to $35 per fine ounce. The flight from the dollar quickly spilled over into the market for U.S. Treasury bonds. The New York Times noted a “sharp decline” in the price of government bonds and reported “open concern among Treasury officials who direct the raising of funds to carry on the recovery program.”77 Following a round of “heavy selling” of government bonds, market participants began to express anxiety about whether the Treasury could continue to finance the federal deficit.78 In an unprecedented move, the Federal Reserve Bank of New York initiated an inquiry into what The New York Times described as “unwarranted liquidation and short selling of government bonds.”79 Faced with unsettled conditions in both the foreign exchange and domestic bond markets, Roosevelt throttled back on the steady escalation of RFC Credit Administration and a backer of the gold purchase program, as Under Secretary and Acting Secretary of the Treasury. Morgenthau had no prior experience in banking or finance. “Treasury Headed by Morgenthau Jr.; Woodin on Leave,” New York Times, November 16, 1933, p. 1, “Transfer of Morgenthau to Treasury Logical Step,” New York Times, November 16, 1933, p. 22, and “Roosevelt Firm on Gold Policy,” Wall Street Journal, November 16, 1933, p. 1 75. Meltzer (2003, p. 453n. 74) states that “Each morning Morgenthau, Warren, and Jesse Jones, head of the RFC, met in Roosevelt’s bedroom. Morgenthau reported the previous day’s prices of gold and commodities. Roosevelt chose a new gold price for the day. The aim was to keep the gold price rising.” 76. “Gold Policy Drops Dollar to 66.74%,” New York Times, October 25, 1933, p. 2, and “Britain is Alarmed as Dollar Slumps,” New York Times, November 16, 1933, p. 3. 77. “Roosevelt Holds 3-Hour Conference on Gold Purchases,” New York Times, November 13, 1933, p. 1. 78. “Government List Leads Bonds Down,” New York Times, November 23, 1933, p. 34, and “Warning on Uncertainty,” New York Times, November 23, 1933, p. 1 (citing a resolution of the Federal Advisory Council of the Federal Reserve Board that “unless there is a monetary stabilization, it will become increasingly difficult for the government to finance its large commitments for reconstruction purposes and to refinance its maturing obligations.”) 79. “Short Bond Sales Traced by Reserve,” New York Times, November 23, 1933, p. 1, and “Federal Bonds Up on News of Inquiry,” New York Times, November 24, 1933, p. 31. The inquiry came on the heels of a complaint by the Committee for the Nation, a group of private citizens that supported Roosevelt’s reflation program, that some New York banks were lending Treasury bonds to short sellers to depress government bond prices. “Seeks Short Sale Probe on U.S. Bonds,” Wall Street Journal, November 20, 1933, p. 4. Subsequently the New York Fed was reported to have concluded that although banks did lend bonds for the purpose of settling sales, those loans did not contribute to the weakness in bond prices. “N.Y. Banks Report on Bond Position,” Wall Street Journal, November 24, 1933, p. 1.

242

Chapter 16

gold-buying prices. As shown in Figure 16.6, he kept the price unchanged at $33.56 per fine ounce from November 14 to November 18 and raised it to only $34.06 per fine ounce by the end of the year. The Wall Street Journal described the de-escalation of the rate of price increase as a matter of “testing the brakes.”80 In late 1933 Roosevelt began to explore a different route to reflation: reducing the gold content of the dollar in a single stroke and putting the United States on a gold bullion standard, whereby it would stand ready to buy and sell gold in bulk form—such as bars—at the new, higher, dollar price. The gold bullion standard would differ from a conventional gold standard because the Treasury would neither mint gold coins nor print gold certificates, and because the Treasury would not stand ready to convert its currency into gold coin. The dollar would be convertible into gold, but not of a form suitable for hoarding by any but the very wealthy. In preparation for devaluing the dollar and moving to a gold bullion standard, on January 15, 1934, Roosevelt asked Congress to limit his right (under the Thomas amendment) to reduce the gold content of the dollar to between 50 and 60 percent of the content specified in the Gold Standard Act of 1900, and to authorize him to subsequently vary the gold content of the dollar within those limits from time to time as he might deem necessary.81 The limitation was intended to give market participants some degree of certainty about the fate of the dollar, without removing all of the discretion previously granted to the President to respond to evolving conditions. Roosevelt also asked Congress for statutory authority to transfer title to the gold held by the Federal Reserve Banks to the U.S. Treasury (so the Treasury could receive the full benefit of any revaluation) and to give the Banks new “gold certificates” that would replace gold in backing Federal Reserve note and deposit liabilities.82 The bond market responded promptly, and favorably, to the President’s message. The New York Times reported that the message “lifted from the minds of traders and investors any uncertainty over the ultimate fate of the currency which may have lately operated to repress confidence . . .. It uncorked buying that had been bottled up by some fear and much bewilderment over the monetary policy of the country.”83 80. “Brakes Tested on Gold Policy,” Wall Street Journal, November 21, 1933, p. 1. 81. “Text of Roosevelt Monetary Message,” New York Times, January 16, 1934, p. 4. 82. The gold certificates would not be redeemable in gold except “at such times and in such amounts as, in the judgment of the Secretary of the Treasury, are necessary to maintain the equal purchasing power of every kind of currency of the United States.” See “Reserve Waives Profit on Gold,” Wall Street Journal, January 17, 1934, p. 6, and section 6 of the Gold Reserve Act of 1934. 83. “All Markets Join in Rise,” New York Times, January 16, 1934, p. 1.

243

Treasury Finance during the Great Depression

Box 16.1 Effect of the January 31, 1934, Revaluation on the Value of Treasury’s Gold Coin and Bullion At the end of 1933 the Treasury and the Federal Reserve Banks owned a total of $4 billion of metallic gold: 194 million fine ounces, valued at the statutory price of $20.67 per fine ounce.a Their holdings included $2.6 billion of Federal Reserve Bank gold (including gold coin and bullion held in Reserve Bank vaults and gold in several Federal Reserve gold funds held by the Treasury),b $1.2 billion of gold held by the Treasury against outstanding gold certificates,c $156 million of gold held by the Treasury against outstanding United States notes,d about $30 million of gold held by the Treasury against outstanding national bank notes,e and about $34 million of “free” gold owned by the Treasury.f When President Roosevelt revalued gold to $35 per fine ounce on January 31, 1934, the Treasury (having received title to the gold previously owned by the Federal Reserve Banks) marked up the value of the 194 million fine ounces of gold by $2.8 billion.g It allocated $2 billion to an exchange stabilization fund authorized by Congress in the Gold Reserve Act of 1934 and left the balance in its General Fund.h a. $20.67 = 1/(0.053750 ounces of gold of 0.9 fineness per dollar, times 0.9 fineness). b. Federal Reserve Bulletin, February 1934, p. 94. The entry for “Gold” includes gold coin and bullion held in Federal Reserve vaults, in the Federal Reserve agents’ fund, and in the gold settlement fund. c. 1934 Treasury Annual Report, p. 376. See also Federal Reserve Bulletin, February 1934, p. 94 ($946 million of gold certificates on hand at Federal Reserve Banks and due from Treasury), and “Black Tells of Reserve Banks’ Stand on Gold,” New York Times, January 19, 1934, p. 15 ($219 million of gold certificates in circulation outside of Federal Reserve Banks). d. 1934 Treasury Annual Report, p. 376 (labeled “Gold reserve”). e. “Black Tells of Reserve Banks’ Stand on Gold,” New York Times, January 19, 1934, p. 15. f. “Black Tells of Reserve Banks’ Stand on Gold,” New York Times, January 19, 1934, p. 15. g. $2.8 billion = 194 million fine ounces of gold, times price change of $14.33 per fine ounce. The markup is noted as a receipt in the 1934 Treasury Annual Report, p. 297. h. The transfer of $2 billion to the exchange stabilization fund is noted as an expenditure in the 1934 Treasury Annual Report, p. 297. See also Schwartz (1997). The 1934 Treasury Annual Report shows, on p. 12, an $811 million balance attributable to the revaluation in the Treasury’s General Fund at the end of fiscal year 1934.

Congress gave the President everything he asked for in Gold Reserve Act of January 30, 1934. The very next day, Roosevelt announced that he was reducing the gold content of the dollar to 59.06 percent of the content specified in the Gold Standard Act, to 0.031746 ounces of gold (of 0.9 fineness).84 The new standard amounted to revaluing gold to a price of $35 per fine ounce.85 (Box 16.1 describes how the revaluation resulted in a $2.8 billion markup in the value of the Treasury’s gold holdings.) The Treasury simultaneously announced that it would begin immediately to buy gold in unlimited quantities at a price of $35 per fine ounce (less one-quarter of one percent for handling) and that it would sell gold at $35 per fine ounce (plus one-quarter of one 84. Presidential Proclamation, January 31, 1934, reprinted in Federal Reserve Bulletin, February 1934, pp. 68–69. 85. $35 per fine ounce, times 0.031746 ounces of gold (of 0.9 fineness), times 0.9 fineness = $1.

244

Chapter 16

percent) to foreign central banks “whenever our exchange rates with gold standard countries reach gold export point.”86 Roosevelt’s announcement triggered a flood of gold into the United States as foreigners began to sell gold to the Treasury and to buy relatively underpriced American goods and services and to invest in American enterprises.87 More gold, more than $200 million worth, arrived in New York between February 15 and February 20 than the RFC had purchased between October 25 and January 15; the Treasury assay office in New York was “buried under the flood.”88 The January 31, 1934, devaluation of the dollar had two crucial advantages over the earlier gold purchase program. First, it fixed a definite band for the dollar price for gold—and thereby eliminated much of the uncertainty engendered by the earlier program. Second, payment for gold with devalued dollars led to a dramatic expansion in central bank liabilities. Table 16.3 shows that the total assets of the twelve Federal Reserve Banks increased from $7.0 Table 16.3 Assets and liabilities of Federal Reserve Banks, December 31, 1933, to December 31, 1936 (in $billions)

Assets Gold and gold certificates Treasury securities Other assets Total assets Liabilities Federal Reserve notes Member bank reserve balances Other liabilities and capital Total liabilities and capital

1933

1934

1935

1936

3.524 2.437 1.080 7.041

5.124 2.430 0.888 8.442

7.553 2.431 1.042 11.026

8.852 2.430 1.243 12.525

3.080 2.729 1.232 7.041

3.221 4.096 1.125 8.442

3.709 5.587 1.730 11.026

4.284 6.606 1.635 12.525

Source: Board of Governors of the Federal Reserve System (1943, p. 332).

86. “Statement to the Press by the President on January 31,” reprinted in Federal Reserve Bulletin, February 1934, pp. 67, “Statement to the Press by the Secretary of the Treasury on January 31” and “Statement to the Press by the Secretary of the Treasury on February 1,” reprinted in Federal Reserve Bulletin, February 1934, p. 69, and “Provisional Regulations Issued under the Gold Reserve Act of 1934,” January 30 and January 31, 1934, reprinted in Federal Reserve Bulletin, February 1934, pp. 82–88. The 1940 Treasury Annual Report noted (p. 123) that the willingness of the Treasury to sell gold at $35 per fine ounce to foreign central banks when the exchange rate reached the gold export point allowed gold “to perform its essential function, in the settlement of international balances,” but conveniently omitted mention of the historic role of the gold standard in limiting monetary expansion. 87. See, for example, the shipments noted in “$25,000,000 Gold Arrives,” New York Times, February 10, 1934, p. 27, and “Ships Quit France with More Gold,” New York Times, February 10, 1934, p. 21. 88. “Incoming Gold Tide Swamps Assayers,” New York Times, February 20, 1934, p. 5.

245

Treasury Finance during the Great Depression

billion at the end of 1933 to $12.5 billion at the end of 1936, entirely as a result of an expansion in Reserve Bank holdings of gold certificates. The Banks’ currency and deposit liabilities increased in parallel. The expansion in central bank liabilities occurred because, after January 1934, gold imports were financed with central bank liabilities rather than with interest-bearing debt. When the Treasury bought gold at $35 per fine ounce, it paid with a check on its account at the Federal Reserve Bank of New York.89 When the seller of the gold deposited the check, the Fed debited the Treasury’s account and credited the reserve account of the seller’s bank—thereby increasing reserves available to the banking system. (Whether the increased reserves appeared as Federal Reserve currency or Federal Reserve deposit liabilities depended on the preferences of banks and other private sector actors—the Fed stood ready to exchange one for the other on a dollar-for-dollar basis.) When the Treasury’s account at the New York Fed got too low, the Treasury would issue certificates on the additional gold in its vaults to the Federal Reserve (at the new statutory rate of $35 per fine ounce) against a credit to its Reserve Bank account. At the end of the day, Federal Reserve holdings of gold certificates were higher, reserves available to the banking system were higher, and Treasury balances at the Federal Reserve were unchanged.90 The expansion in central bank liabilities that followed devaluation contributed to the economic recovery that occurred between 1934 and 1936.91 89. The Treasury used the Federal Reserve Bank of New York as its fiscal agent in buying and selling gold. “Statement to the Press by the Secretary of the Treasury on January 31, 1934,” reprinted in Federal Reserve Bulletin, February 1934, p. 69. 90. See “Treasury Adopts Plan to ‘Sterilize’ Gold Acquisitions,” New York Times, December 22, 1936, p. 1, and “New Deal Eyes Gold Desterilizing to Halt Deflation,” Wall Street Journal, April 14, 1938, p. 1. 91. Meltzer (2003, p. 463) (stating that the gold purchases that followed devaluation “achieved the desired end of higher commodity prices and economic expansion that the administration sought”).

17

Nonmarketable Treasury Debt

Prior to 1936, federal deficits were financed almost exclusively with marketable Treasury securities, but figure 17.1 shows that deficits after fiscal 1935 were financed with increasing amounts of nonmarketable debt. The nonmarketable debt included savings bonds sold to individual investors and special issues sold to two trust funds, the Unemployment Trust Fund and the Old-Age Reserve Account, established by the Social Security Act of 1935. This chapter examines the issuance of nonmarketable Treasury debt in the second half of the 1930s. The chapter begins by describing an early example of a government trust fund and the first instance of a trust fund investing in special issue debt. The latter is important because it set an important precedent for later, much larger, trust funds and because it played a crucial role in the complicated history of bonus payments (during the 1930s) to veterans of World War I. The mechanics of the two trust funds established by the Social Security Act are described next. The chapter concludes with a discussion of the very different savings bond program. Government Trust Funds A trust fund is a pool of assets held for a specific purpose—for example, to meet the operating expenses of a college—and managed by a person or persons acting in a fiduciary capacity. The U.S. Treasury managed several modest trust funds in the mid-1920s, including the United States Government Life Insurance (USGLI) Fund that provided death and disability benefits for active duty and discharged servicemen and women.1 The USGLI fund operated much as it would have had it been managed by private sector bankers. Participants remitted monthly insurance premiums and Treasury officials used the premiums to acquire government bonds in conventional secondary market transactions.2 Payments to disabled participants and to the beneficiaries of deceased participants were funded with interest payments received from the fund’s investments. In May 1924 Congress authorized a new trust fund, the Adjusted Service Certificate Fund, to finance bonus payments to men and women who had served in the armed forces during World War I. The investments acquired by the Adjusted Service Certificate Fund broke new ground in federal finance and laid a foundation for the much larger trust funds established by the Social Security Act. 1. McGill (1949). 2. In mid-1925 the USGLI fund held $148.4 million of marketable bonds, including $6.6 million of First Liberty 4¼ percent bonds, $18.1 million of Second Liberty 4¼ percent bonds, $42.7 million of Fourth Liberty bonds, $49.2 million of other Treasury bonds, and $31.9 million of bonds issued by the Federal Land Banks. 1925 Treasury Annual Report, p. 122. See also McGill (1949, pp. 152–53).

248

Chapter 17

50

Billions of dollars

40

30

20

10

Marketable debt

1939

1938

1937

1936

1935

1934

1933

1932

1931

1930

1929

1928

1927

0

Nonmarketable debt

Figure 17.1 Treasury debt, mid-1927 to mid-1939

The Adjusted Service Certificate Fund After the end of World War I, discharged veterans argued that they deserved to be paid a bonus for their war service and to ease their transition back to civilian life.3 Democratic Congressmen generally supported their efforts, but many Republicans were reluctant to do anything that would jeopardize the tax cuts that were at the center of their party’s political agenda.4 After years of wrangling, Congress passed (over the veto of President Calvin Coolidge5) the World War Adjusted Compensation Act of May 19, 1924. The 3. Siegel and Taylor (1948, p. 529). See also Kettleborough (1924) and Dickson and Allen (2004). 4. Dickson and Allen (2004, p. 26) (stating that President Harding opposed a bonus because “it would increase the national debt and thwart his plan to cut taxes”), “Low Tax Crusade Favored by Borah,” New York Times, January 6, 1924, p. E1 (quoting Senator William Borah, Republican of Idaho, as saying “There is no avoiding or sidestepping this issue: one is either for tax reduction or he is against it. One has a right to be for the bonus or he has the right to be for tax reduction. But you can not be for both.” and reporting that Borah described the bonus as “the greatest obstacle to the plan of tax reduction”) and “$2,119,000,000 Bonus Will Be Reported in the House Today,” New York Times, March 17, 1924, p. 1 (reporting that Republicans on the House Ways and Means Committee opposed the bonus bill because it would limit tax reduction). 5. Dickson and Allen (2004, pp. 28–29) state that Congress was motivated to override the presidential veto by Coolidge’s “intemperate remarks” in his veto message: “Patriotism which is bought and paid for is not patriotism . . .. Service to our country in time of war means sacrifice. It is for that reason alone that we honor and revere it. To attempt to make a money payment out of the earnings of the people to those who are physically well and financially able is to abandon one of our most cherished American ideals.”

249

Nonmarketable Treasury Debt

act granted every veteran of the war a nontransferable “adjusted service certificate” that would pay its face amount in twenty years or upon the earlier death of the veteran. The face amount of a certificate was computed as the face amount of a 20-year endowment life insurance policy that could be purchased (assuming a 4 percent rate of interest and standard mortality statistics) with 125 percent of a veteran’s “adjusted service credit.” Adjusted service credits were computed as the sum of $1 per day of home service and $1.25 per day of overseas service recorded between April 5, 1917 and July 1, 1919. A 30-year-old veteran with an adjusted service credit of $625 would be entitled to an adjusted service certificate with a face value of $1,577.50.6 It was estimated that more than three million men and women would receive adjusted service certificates and that the bonus program would cost more than $2 billion.7 Ignoring, for simplicity, the relatively inexpensive provision for early payment in the event of death,8 the World War Adjusted Compensation Act authorized the distribution of 20-year zero-coupon Treasury bonds worth about $2 billion. Congress could have recorded the distribution as a $2 billion expenditure financed with $2 billion of new indebtedness, that is, financed with the adjusted service certificates themselves. However, that would have eliminated the budget surplus projected for fiscal 1925 and upset the plans of the Coolidge administration to proceed with the tax cuts that were ultimately authorized in the Revenue Act of 1924. Congress instead created the Adjusted Service Certificate Fund and authorized annual appropriations to the fund sufficient to finance redemption of the adjusted service certificates. The appropriations were to be determined “in accordance with accepted actuarial principles and . . . interest at 4 per centum per annum . . .”9 and were expected to run at a rate somewhat in excess of $100 million per year.10 The Secretary of the Treasury was authorized to invest the annual appropriations in Treasury securities. 6. This example appears in “Passage of Bonus and Threat of Veto Force Tax Changes,” New York Times, May 21, 1924, p. 1. 7. “Bonus Bill Passed by House, 355 to 54, in Forty Minutes,” New York Times, March 19, 1924, p. 1, “Bonus Bill Becomes Law,” New York Times, May 20, 1924, p.1, “Insurance Policies Will Go to 3,038,283 Veterans and Cash Payments to 389,583 Under Bonus Law,” New York Times, May 20, 1924, p. 1, and “Passage of Bonus and Threat of Veto Force Tax Changes,” New York Times, May 21, 1924, p. 1. 8. A 20-year zero-coupon Treasury bond with a face value of $1,577.50 and yielding 4 percent per annum would have “cost” $575.92 in adjusted service credits ($575.92 × 1.25 = $1,577.50 × 1.04−20), only $50 less than an adjusted service certificate with the same face value issued to a 30-year-old veteran. 9. Section 505 of the World War Adjusted Compensation Act of 1924. 10. “Bonus Bill Ready to Go to House,” New York Times, March 13, 1924, p. 4, “House Committee Reports Bonus Bill,” New York Times, March 14, 1924, p. 19, “$2,119,000,000 Bonus Will Be Reported in the House Today,” New York Times, March 17, 1924, p. 1, and “Bonus Bill Passed by House, 355 to 54, in Forty Minutes,” New York Times, March 19, 1924, p. 1.

250

Chapter 17

Treasury Secretary Mellon could have managed the Adjusted Service Certificate Fund like the earlier-established USGLI fund: purchasing outstanding marketable Treasury bonds in conventional secondary market transactions. He chose instead the novel course of issuing “special” nonmarketable securities directly to the fund. This meant that the Treasury would issue less marketable debt to the public (because it would not have to raise the money needed to make secondary market purchases of outstanding Treasury securities for the fund), but it had no effect on aggregate indebtedness (because the special nonmarketable debt would offset the reduced issuance of marketable debt on a dollar-for-dollar basis). Treasury officials pointed out that special issue debt would be more convenient than conventional debt because it could be issued with a 4 percent annual coupon that matched the interest rate assumption specified by Congress and because it could be redeemed as needed to satisfy claims.11 Table 17.1 shows the growth of the Adjusted Service Certificate Fund through the end of fiscal year 1930. Congress appropriated somewhat more than $100 million annually and most of the appropriations were allocated (in book-entry transactions that had no cash flow consequences) to special nonmarketable Treasury issues. A small working balance was maintained in a cash account. The only time the Treasury made a cash disbursement was when the holder of an adjusted service certificate passed away. (When Treasury officials needed to replenish the cash account they redeemed some of the special issues in the Adjusted Service Certificate Fund, funding the redemptions by selling conventional marketable debt to the general public.) Given that Congress chose to obfuscate (at least for purposes of public accounting) the full value of what it had granted to veterans, Mellon’s decision Table 17.1 Appropriations, interest receipts, payments, and assets of the Adjusted Service Certificate Fund, fiscal year 1925 to fiscal year 1930 ($millions) 1925 Assets, beginning of fiscal year 0.0 Appropriations 100.0 Interest 0.0 Payments −4.6 Assets, end of fiscal year 95.4 Composition of assets, end of fiscal year Special nonmarketable issues 95.4 Other assets 0.0

1926

1927

1928

1929

1930

95.4 120.0 3.9 −15.0 204.3

204.3 116.0 7.7 −13.9 314.1

314.1 112.0 12.3 −35.2 403.2

403.2 112.0 16.1 −16.7 514.6

514.6 112.0 20.5 −16.6 630.5

203.9 0.4

312.9 1.2

401.8 1.4

513.0 1.6

629.3 1.3

Source: Treasury Annual Reports.

11. 1925 Treasury Annual Report, pp. 118–20, and Peach (1951, p. 47).

251

Nonmarketable Treasury Debt

to invest the assets of the Adjusted Service Certificate Fund in special issues was as sensible as it was innovative. The common understanding of congressmen and Treasury officials was that the Adjusted Service Certificate Fund would continue to grow (as a result of continuing annual appropriations) until the mid-1940s, when the adjusted service certificates matured. At that time the Treasury would redeem the special issues to pay off the certificates, funding the redemptions with public sales of marketable debt.12 The Emergency Adjusted Compensation Act of 1931 As the country slid off into depression in 1930, veterans began to agitate for early payment of their adjusted service certificates. President Hoover opposed early payment but Congress, in a compromise measure (the Emergency Adjusted Compensation Act of February 27, 1931) passed over the President’s veto, decided to allow veterans to borrow up to 50 percent of the face amount of their certificates from the Adjusted Service Certificate Fund.13 The data in table 17.2 suggests that veterans borrowed more than $700 million in fiscal year 1931 and almost $200 million in fiscal 1932.14 The loans were funded Table 17.2 Appropriations, interest receipts, payments, and assets of the Adjusted Service Certificate Fund, fiscal year 1931 to fiscal year 1935 ($millions)

Assets, beginning of fiscal year Appropriations Interest Payments and loans Assets, end of fiscal year Composition of assets, end of fiscal Special nonmarketable issues Other assets

1931

1932

1933

1934

1935

630.5 224.0 34.3 −765.6 123.2 year 121.8 1.4

123.2 200.0 3.0 −214.4 111.8

111.8 100.0 5.6 −122.3 95.1

95.1 50.0 4.6 −30.2 119.5

119.5 50.0 5.7 −18.7 156.5

105.0 6.8

92.0 3.1

117.8 1.7

155.5 1.0

Source: Treasury Annual Reports.

12. 1925 Treasury Annual Report, pp. 118–20. 13. Section 502 of the original World War Adjusted Compensation Act of 1924 provided that holders of adjusted service certificates could pledge their certificates as collateral for commercial bank loans, but limited both the amount of and the interest rate on a loan. Subsequently, in the Act of March 3, 1927, Congress provided for government loans and authorized the USGLI Fund to finance the loans. By the end of fiscal 1930, the USGLI had lent $280 million to veterans, either in the form of policyholder loans or loans secured by adjusted service certificates. 1930 Treasury Annual Report, p. 120. The Emergency Adjusted Compensation Act of 1931 provided a third way for veterans to borrow on their adjusted service certificates—and on much more lenient terms. 14. Treasury annual reports between 1931 and 1935 do not distinguish between payments to beneficiaries following the death of a veteran and loans to veterans. Payments to beneficiaries in the late 1920s amounted to about $20 million per annum (see table 17.1), so it is not unreasonable to assume that the 1931 total of $766 million of loans and payments included at least $700 of loans. Brown (1956, p. 863) emphasizes the substantial fiscal stimulus supplied by this loan program. See also Stein (1969, pp. 24–26).

252

Chapter 17

with stepped-up appropriations (Congress appropriated $224 million in fiscal 1931 and $200 million in fiscal 1932) and by the redemption of special issue debt held in the Adjusted Service Certificate Fund. (The redemption of special issue debt was, in turn, funded by sales of conventional marketable debt to the general public.) Fund holdings of special issue debt declined from $629 million in mid-1930 to $105 million in mid-1932.15 The Adjusted Compensation Payment Act of 1936 As the Great Contraction came to an end in the last year of the Hoover administration and the first year of the Roosevelt administration, veterans renewed their demands for early payment of their adjusted service certificates. Hoover, and later Roosevelt, resisted but, in 1936, Congress passed—and sustained over yet another presidential veto—the Adjusted Compensation Payment Act of January 27, 1936.16 The act provided for immediate payment of the certificates in nonmarketable but immediately redeemable 19-year, 3 percent adjusted service bonds.17 As shown in table 17.3, the Adjusted Service Certificate Fund Table 17.3 Appropriations, interest receipts, payments, and assets of the Adjusted Service Certificate Fund, fiscal year 1936 to fiscal year 1939 ($millions) 1936 Assets, beginning of fiscal year 156.5 Appropriations 1,830.0 Interest 10.4 Payments for adjusted service bonds −1,745.2 Payment to USGLI Funda −0.0 Other payments −59.2 Assets, end of fiscal year 192.6 Composition of assets, end of fiscal year Special nonmarketable issues 126.8 Other assets 65.8

1937

1938

1939

192.6 500.2 4.1 −146.6 −500.2 −10.8 39.4

39.4 0.0 1.4 −13.3 −0.0 −0.7 26.8

26.8 0.0 1.1 −6.8 −0.0 −0.7 20.4

37.6 1.8

25.8 1.0

19.5 0.8

Source: Treasury Annual Reports. a. In fiscal 1937 the Adjusted Service Certificate Fund paid $500 million of a special nonmarketable issue of 4½ percent adjusted service bonds to the USGLI Fund in exchange for loans to veterans made by the latter Fund and secured with adjusted service certificates. McGill (1949, pp. 159-61 and 256).

15. The appropriations to the Adjusted Service Certificate Fund appear as expenditure items in the fiscal 1931 and 1932 budgets (1931 Treasury Annual Report, p. 443, and 1932 Treasury Annual Report, p. 357), but the loans funded by the redemption of nonmarketable debt and sale of marketable debt do not. For accounting purposes, the latter was a substitution of one form of Treasury debt (marketable debt held by public investors) for another form of Treasury debt (nonmarketable debt held in the Fund). The budget expenditures associated with the issuance of nonmarketable Treasury debt to the Fund had been recorded in earlier years—see table 17.1. 16. Dickson and Allen (2004) recount the long road to the Adjusted Compensation Payment Act. 17. 1936 Treasury Annual Report, pp. 21–23 and 253–61.

253

Nonmarketable Treasury Debt

redeemed $1.75 billion of the adjusted service bonds in fiscal year 1936 and $147 million the following year.18 The 1936 redemptions were funded with a $1.8 billion budget appropriation that, as shown in figure 16.3, materially increased federal expenditures in fiscal 1936. The 1937 redemptions were funded with the redemption of special issue debt held in the Fund.19 By mid1939 the assets in the Fund had shrunk to $20 million. Comment The operation of the Adjusted Service Certificate Fund illustrates two important advantages of special issues for government trust funds: they could be made redeemable on demand without jeopardizing the stability of the national debt and they could be issued at par with interest rates that matched congressional mandates. Both of these features reappeared when Congress authorized the Treasury to purchase special issues for the trust funds created by the Social Security Act. The Social Security Trust Funds The Social Security Act of August 14, 1935, established two new federal trust funds: the Unemployment Trust Fund and the Old Age Reserve Account (subsequently renamed the Federal Old-Age and Survivors Insurance Trust Fund).20 The two funds had distinct purposes but shared a common difference from the Adjusted Service Certificate Fund: they took in cash (from state unemployment funds and from private-sector employers, respectively) and turned the cash over to the Treasury in exchange for special issues of nonmarketable debt. (In contrast, the Treasury issued nonmarketable debt to the Adjusted Service Certificate Fund in satisfaction of book-entry appropriations to the fund.) Thus the Social Security trust funds provided net new cash for the Treasury and reduced the need to sell conventional, marketable Treasury debt to the general public.21 18. Brown (1956, p. 863) emphasizes the substantial fiscal stimulus associated with early payment of the adjusted service certificates. See also Stein (1969, pp. 58–59). 19. $500 million of the fiscal 1937 expenditures for veterans’ bonuses shown in figure 16.3 and table 17.3 resulted from the issuance of a special issue of 4½ percent adjusted service bonds to the USGLI Fund in exchange for loans made by the Fund to veterans (see note 13 above) and secured with the veterans’ adjusted service certificates. See McGill (1949, pp. 159–61 and 256). 20. See Altmeyer (1966), Witte (1962), Schlabach (1969), and Derthick (1979) for details on the origin and early development of the programs authorized by the Social Security Act of 1935. 21. Leff (1983, p. 377) notes that Secretary Morgenthau was “very much aware” of this feature.

254

Chapter 17

The Unemployment Trust Fund The Unemployment Trust Fund was set up as an interest-earning cash account for state unemployment insurance programs.22 Each state had an account in the fund. The Secretary of the Treasury was authorized to receive money from a state unemployment agency for the state’s account, invest the money in interest-bearing Treasury securities, and pay out account balances on demand to the state. Table 17.4 shows the growth in the Unemployment Trust Fund in the second half of the 1930s. In view of the “payable on demand” feature, the Unemployment Trust Fund either had to invest in short-term Treasury securities—and accept the very low interest rates prevailing on those securities—or invest in special Treasury issues that could be redeemed at any time. The Social Security Act provided for special issues, and further provided that they should “bear interest at a rate equal to the average rate of interest . . . borne by all interest-bearing obligations of the United States then forming part of the public debt . . ..”23 Thus state unemployment insurance funds got the benefit of prompt liquidity while earning interest at longer term rates. The Old-Age Reserve Account In its most famous sections the Social Security Act provided for a national program of old-age benefits and a system of wage taxes to support those benefits.24 The taxes were levied on employers and employees equally, at a Table 17.4 Deposits, interest receipts, withdrawals, and assets of the Unemployment Trust Fund, fiscal year 1936 to fiscal year 1939 ($millions) 1936

1937

1938

1939

Assets, beginning of fiscal year 0.0 Deposits by state agencies 18.9 Interest 0.1 Withdrawals by state agencies 0.0 Assets, end of fiscal year 19.0 Composition of assets, end of fiscal year Special nonmarketable issues 19.0 Cash balance 0.0

19.0 291.6 2.7 −1.0 312.3

312.3 747.7 15.2 −191.0 884.2

884.2 811.3 26.8 −441.8 1,280.5

311.0 1.3

872.0 12.2

1,267.0 13.5

Source: Treasury Annual Reports.

22. Only one state had an unemployment insurance program prior to 1935. Title IX of the Social Security Act gave other states a strong economic incentive to adopt similar programs. By mid-1937 every state had a program. Baicker, Goldin, and Katz (1998). 23. Section 904 of the Social Security Act of 1935. 24. The benefits were prescribed in Title II (“Federal Old-Age Benefits”) of the act and the taxes were prescribed in Title VIII (“Taxes with Respect to Employment”). See, generally, Miron and

255

Nonmarketable Treasury Debt

rate of 1 percent of a worker’s wages (up to a maximum annual wage of $3,000) in calendar years 1937, 1938, and 1939, and were to be levied at higher rates in subsequent years, reaching a maximum of 3 percent in 1949.25 Benefits were to be paid to individuals aged 65 and over beginning in 1942, computed on a sliding scale as a function of cumulative wages earned since the end of 1936. An individual would receive, monthly, one-half percent of his or her cumulative wages not in excess of $3,000, plus one-twelfth of 1 percent of the amount of cumulative wages in excess of $3,000 but not in excess of $45,000, plus one-twenty-fourth of 1 percent of the amount of cumulative wages in excess of $45,000, subject to a maximum monthly benefit of $85. There was no provision for payments to spouses or widows.26 The Social Security Act established an Old-Age Reserve Account to provide for future benefit payments, but it did not require that the new wage taxes be deposited to that account.27 It authorized instead an annual budget appropriation to the Reserve Account “sufficient . . . to provide for the [statutory benefits], such amount to be determined . . . in accordance with accepted actuarial principles, and based upon such tables of mortality as the Secretary of the Treasury shall from time to time adopt, and upon an interest rate of 3 per centum per annum . . ..”28 The Act further provided that balances in the Old-Age Reserve Account were to be invested in Treasury issues, including special issues bearing interest at the rate of 3 percent per annum. Table 17.5 Weil (1998). The provisions of the 1935 act were substantially amended by the Social Security Act Amendments of August 10, 1939. See, 1939 Treasury Annual Report, pp. 46–48, 1940 Treasury Annual Report, pp. 137–39, Leff (1983), Leff (1984, pp. 275–86), and Altmeyer (1966, ch. 4). 25. Brown (1956, p. 869) states that the wage taxes “began in [calendar year] 1937 to exert a pronounced effect” on aggregate demand. See also Stein (1969, pp. 99–100 and 114–15). The Social Security Act Amendments of 1939 canceled the tax increase scheduled to take effect in 1940 but left the balance of the schedule in place. 26. The Social Security Act Amendments of 1939 changed the structure of benefits in three ways: (1) they changed the formula to (1 + 0.01 × Y) × {0.3 × min[AMW, 50] + 0.1 × min[AMW, 200]}, where AMW is average monthly wages and Y is years of contribution, (2) they extended benefits to spouses and widows (at 50 percent and 75 percent, respectively, of the benefit to the primary worker), and (3) they advanced the starting year for benefits to 1940. Miron and Weil (1998, pp. 301–302) state that “the new formula awarded much higher benefits to those retiring in the early years of the system relative to those retiring when the system was mature than did the 1935 act.” 27. The act separated old-age benefits (in Title II) from the supporting taxes (in Title VIII) out of fear of a legal challenge to the new program. Miron and Weil (1998, pp. 298–99). (In 1937 the Supreme Court rejected a constitutional challenge to the program. See Helvering v. Davis, 301 U.S. 619 (1937).) Nevertheless, President Roosevelt and members of Congress understood that the benefits would be financed with the taxes. Roosevelt famously commented that, “We put those payroll contributions there so as to give the contributors a legal, moral, and political right to collect their pensions. . . . With those taxes in there, no damn politician can ever scrap my social security program.” Quoted in Schlesinger (1958, pp. 308–309) and Leff (1983, p. 374). The Social Security Act Amendments of 1939 explicitly appropriated 100 percent of the wage tax receipts to the renamed Federal Old-Age and Survivors Insurance Trust Fund. 28. Section 201 of the Social Security Act of 1935.

256

Chapter 17

Table 17.5 Wage tax receipts and appropriations, interest receipts, payments, and assets of the Old-Age Reserve Account, fiscal year 1937 to fiscal year 1939 ($millions)

Wage tax receipts Assets, beginning of fiscal year Appropriations Interest Payments Assets, end of fiscal year Composition of assets, end of fiscal year Special nonmarketable issues Cash balance

1937

1938

1939

256.2 0.0 265.0 2.3 0.0 267.3

510.0 267.3 500.0 15.4 −5.4 777.3

529.6 777.3 390.0 27.0 −13.9 1,180.4

267.2 0.1

662.4 114.9

1,177.3 3.1

Source: Treasury Annual Reports.

shows wage tax receipts and annual appropriations to the Old-Age Reserve Account beginning in 1937. Comment By mid-1939, $3 billion of Treasury debt was in the form of special nonmarketable debt held in trust funds administered by the Department of the Treasury (table 17.6). The nonmarketable debt accounted for 7.5 percent of all interestbearing Treasury debt and substantially reduced the amount of debt that Treasury officials had to sell in public offerings. Savings Bonds Treasury officials introduced savings bonds in 1935 to diversify their funding sources and broaden the distribution of Treasury debt by giving individual investors of modest means an opportunity to save without fear of loss. Secretary Morgenthau described the bonds as adding “another string to our bow” and The Wall Street Journal observed that the Treasury “is evidently continuing its efforts to gain broader distribution for its mounting total of securities outstanding.”29 The New York Times noted that “The motive behind the plan is plain enough. A very large part of the public debt—possibly as much a half—is represented by securities already in the hands of the banking system. In this instance the Treasury is proposing to go beyond the banks to seek a wider and fresher market.”30 In testimony before the Senate Finance Committee, Mor29. “Would Allow Issue of 45 Billion Bonds,” New York Times, January 22, 1935, p. 1, and “The Baby Bond Bill,” Wall Street Journal, January 24, 1935, p. 3. 30. “Baby Bonds,” New York Times, February 4, 1935, p. 14.

257

Nonmarketable Treasury Debt

Table 17.6 Special nonmarketable Treasury securities held in selected Federal trust funds, fiscal year 1931 to fiscal year 1939 ($millions)

Adjusted Service Certificate Fund United States Government Life Insurance Funda Unemployment Insurance Trust Fund Old-Age Reserve Account Total

1931

1932

1933

1934

1935

1936

121.8

105.0

92.0

117.8

155.5

126.8

19.0

121.8

105.0

92.0

117.8

155.5

145.8

1937

1938

1939

37.6

25.8

19.5

500.2

500.2

500.2

311.0

872.0

1,267.0

267.2

662.4

1,177.3

1,116.0

2,060.4

2,964.0

Source: Treasury Annual Reports. a. In fiscal 1937 the Adjusted Service Certificate Fund paid $500 million of a special nonmarketable issue of 4½ percent adjusted service bonds to the USGLI Fund in exchange for loans to veterans made by the latter Fund and secured with adjusted service certificates. McGill (1949, pp. 159-61 and 256).

genthau stated that the new securities were intended to “reach a small investors’ class who want to invest their money,” and that the securities gave the government “an additional place to go to borrow money.”31 Unlike conventional bonds, savings bonds did not pay a periodic coupon— they were discount securities, like Treasury bills—and they could be redeemed on demand.32 The first feature limited the cost of administering the savings bond program; the second feature eliminated the risk of fluctuations in value and, since the securities were not transferable, provided a source of liquidity. The first series of savings bonds, announced on February 25, 1935, matured ten years from the date of purchase and was offered at a price of 75 percent of face amount.33 If held to maturity, the bonds would yield 2.90 percent per 31. Committee on Finance (January 29, 1935, p. 18). In 1938 Morgenthau reiterated that savings bonds “were designed to furnish a type of government security which would be attractive to any citizen desiring to put aside funds for future use. They were particularly intended as a repository for savings in relatively small amounts.” “Treasury to Go On with Baby Bonds,” New York Times, March 21, 1938, p. 21. 32. Savings bonds bore a close similarity to the war savings stamps and certificates of World War I. See chapter 4. 33. Treasury Circular no. 529, February 25, 1935 (offering series A savings bonds), reprinted in 1935 Treasury Annual Report, p. 197. The Treasury introduced subsequent series of savings bonds after 1935—see Treasury Circular no. 554, December 16, 1935 (offering series B savings bonds), reprinted in 1936 Treasury Annual Report, p. 242; Treasury Circular no. 571, December 16, 1936 (offering series C savings bonds), reprinted in 1937 Treasury Annual Report, p. 238; and Treasury Circular no. 596, December 15, 1938 (offering series D savings bonds), reprinted in 1939 Treasury Annual Report, p. 248—but the structure of redemption prices remained unchanged until April 1941, when the Treasury introduced series E bonds—see Treasury Circular no. 653, April 15, 1941, reprinted in 1941 Treasury Annual Report, p. 304.

258

Chapter 17

Table 17.7 Redemption prices and yields to redemption for series A savings bonds Years held

Redemption pricea

Yield to redemptionb

1 1½ 2 2½ 3 3½ 4 4½ 5 5½ 6 6½ 7 7½ 8 8½ 9 9½ 10

76.00 77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00 85.00 86.00 87.00 88.00 90.00 92.00 94.00 96.00 98.00 100.00

1.33 1.76 1.97 2.09 2.16 2.21 2.24 2.27 2.28 2.29 2.29 2.30 2.30 2.45 2.57 2.67 2.76 2.84 2.90

Source: Treasury Circular no. 529, February 25, 1935 (offering series A savings bonds), reprinted in the 1935 Treasury Annual Report, p. 197. Note: The bonds were sold at a price of 75 percent of face amount. a. Percent of face amount. b. Percent per annum, compounded semiannually.

annum, compounded semiannually. However, if redeemed prior to maturity (on the schedule of redemption prices shown in table 17.7) they would earn a lower yield; redemption after one year would result in a yield of only 1.33 percent. The structure of redemption prices resulted in higher yields on bonds held for longer intervals and gave investors an incentive to hold, rather than redeem, their bonds. Savings bonds were targeted to investors of modest means. They were available in face amounts as small as $25 and they were offered continuously at fixed prices at post offices throughout the country. The Treasury did not offer savings bonds to institutional investors and individual investors could not purchase more than $10,000 face amount of savings bonds in a single calendar year. By mid-1939 there was a total of $1.8 billion of savings bonds outstanding.

259

Nonmarketable Treasury Debt

Concluding Remarks Nonmarketable issues of Treasury debt, to individual investors in the form of savings bonds and to government trust funds in the form of special issues, became an important source of funds for the Treasury in the second half of the 1930s. The issues materially reduced the amount of conventional, marketable debt that Treasury officials had to sell to finance chronic budget deficits. Savings bonds were straightforward: they were not substantively different from conventional bonds, except that they were nontransferable, putable, discount instruments. Special issues for the Unemployment Trust Fund were hardly more complicated. Although they were assets of the trust fund (rather than assets of the state unemployment funds), the states had well-identified claims on the trust fund that, in aggregate, equaled the value of the trust fund assets. The situation would not have been any more transparent if the states had held conventional, marketable Treasury debt in their individual accounts. The Old-Age Reserve Account was a different, more opaque, matter. Individual beneficiaries did not have legally enforceable claims on the Old-Age Reserve Account and their statutory entitlements were not more than loosely linked to the value of the nonmarketable securities held in the account. Only time would tell whether the government had made adequate provision for the benefits that it promised to pay and whether it would be able to redeem the nonmarketable securities in the Old-Age Reserve Account, either out of budget surpluses or by issuing conventional debt.

18

Treasury Debt Management during the Great Contraction

The success of a Treasury debt manager depends on three things: goals that are compatible with the economic environment, a rationally constructed program for achieving those goals, and the mental agility to respond to unanticipated developments. Andrew Mellon and his Under Secretaries were successful debt managers in the 1920s because they had a clear vision of what they wanted to accomplish: reduction of the war debt, as well as a congenial fiscal environment and a program of early retirements, cash redemptions at maturity, and refinancings calibrated to achieve their objective. In addition they demonstrated that they could take advantage of unanticipated developments, such as when they called the Second Liberty Loan for early redemption after tax revenues began to come in higher than expected in March 1927.1 After 1930, economic circumstances precluded further debt reduction, a fact acknowledged—somewhat obliquely—in the 1930 Treasury Annual Report: The policy consistently pursued by the Treasury Department has been to reduce our war debt as rapidly as possible in days of plenty. This, of course, implied as a corollary that in periods of depression, when the Government revenue is restricted, the rate of debt reduction should be slowed up . . ..2

Circumstances instead forced Treasury officials to reshape their policies toward a new goal of refinancing maturing debt at longer maturities and lower interest rates. This chapter traces the evolution of Treasury debt management actions between 1930 and early 1933. The first section examines the outlook for debt management in the second half of 1930. The following sections describe how unforeseen events first complicated, and then derailed, the refinancing program that Secretary Mellon and Under Secretary of the Treasury Ogden Mills had mapped out, and how those officials fashioned innovative responses. Most prominent, when faced with unexpected financing requirements in the first half of 1931 they introduced a program of regularly refinancing maturing Treasury bills with new bills, thereby turning bills from their originally intended use as an instrument of cash management into an instrument of debt management. The last section describes how Mills, who became Secretary of the Treasury in February 1932 after Mellon was confirmed as Ambassador to Great Britain, funded the activities of the Reconstruction Finance Corporation in 1932 and early 1933. 1. See chapter 12. 2. 1930 Treasury Annual Report, p. 38. See also the similar testimony of Under Secretary of the Treasury Ogden Mills before the House Appropriations Committee in late 1930, reported in “Discusses Public Debt Reduction,” Wall Street Journal, December 6, 1930, p. 9.

262

Chapter 18

Table 18.1 Refinancings and paydowns on tax dates in 1930 ($millions) Date

Maturing

Mar 15, 1930

100

Refinanced

Paid down Bill issued Dec 17, 1929, and maturing Mar 17, 1930 Certificate issued Jun 15, 1929 Certificate to mature Dec 15, 1930

404 483 21 Jun 16, 1930

550

Sep 15, 1930

51

Certificate issued Sep 16, 1929 Certificate to mature Jun 15, 1931

429 121

Bill issued Jul 14, 1930, and maturing Sep 15, 1930 Certificate issued Dec 16, 1929 Certificate to mature Sep 15, 1931

352 334 69 Dec 15, 1930

102

Bills issued Oct 15 and 16, 1930, and maturing Dec 16 and 17, 1930 Certificate issued Mar 15, 1930 Certificate to mature Jun 15, 1931 Certificate to mature Dec 15, 1931

483 160 268 157 Source: Treasury Annual Reports.

The Outlook for Treasury Debt Management in 1930 Treasury financing operations in 1930 were not materially different from operations in the second half of the 1920s. As shown in table 18.1, the Treasury sold a total of five certificates of indebtedness in four tax date financings. Tax receipts were large enough to allow some paydown of maturing indebtedness in each of the four quarters. However, officials were well aware that tax revenues would fall in 1931 and that the budget surplus would likely disappear.3 They began to plan for deficits and debt refinancing rather than surpluses and debt reduction. The existing structure of Treasury indebtedness gave little cause for concern. As of mid-1930, debt maturing within a year, including $156 million of Treasury bills and $1.3 billion of certificates of indebtedness, amounted to less than 10 percent of the government’s total interest-bearing debt. Only about 20 percent of the debt, including three series of notes in addition to the bills and 3. “$159,000,000 Drop in Revenue in Year,” New York Times, September 18, 1930, p. 3 (reporting that “the real worry on the part of experts now is that income tax collections will be greatly reduced in [1931] . . . for income taxes paid in that period will be assessed against corporate and individual incomes received in the calendar year 1930, during which many corporations have reduced or omitted dividends and suffered sharp curtailments of earnings, and individual incomes, because of unemployment and curtailed dividend payments, probably have shrunk sharply”).

263

Treasury Debt Management during the Great Contraction

Table 18.2 Marketable Treasury notes and war and postwar bonds outstanding on June 30, 1930 Coupon rate Notes A-1930-32 3½% B-1930-32 3½ C-1930-32 3½ Liberty bonds First Liberty Loan 3½, 4, 4¼ Fourth Liberty Loan 4¼ Postwar bonds (the Mellons) 33⁄8 33⁄8s of 1943 33⁄8s of 1947 33⁄8 4¼s of 1952 4¼ 4s of 1954 4 3¾s of 1956 3¾

Amount outstanding ($millions)

First call date

Maturity date

Mar 15, 1930 Sep 15, 1930 Dec 15, 1930

Mar 15, 1932 Sep 15, 1932 Dec 15, 1932

674 500 452

Jun 15, 1932 Oct 15, 1933

Jun 15, 1947 Oct 15, 1938

1,934 6,268

Jun 15, 1940 Jun 15, 1943 Oct 15, 1947 Dec 15, 1944 Mar 15, 1946

Jun 15, 1943 Jun 15, 1947 Oct 15, 1952 Dec 15, 1954 Mar 15, 1956

359 493 759 1,037 489

Source: 1930 Treasury Annual Report, pp. 551–52. Note: Excludes $725 million of pre-war bonds.

certificates, matured before the end of 1932. The $6.3 billion Fourth Liberty Loan, due to mature in 1938, was unquestionably the biggest concern, but it was relatively far out on the horizon and Treasury officials had ample experience in whittling down war loans before they reached final maturity. Beyond that, there was the $2 billion First Liberty Loan (maturing in 1947) and the five Mellons (maturing between 1943 and 1956).4 Table 18.2 shows that the three notes and all the war and postwar bonds were redeemable at the option of the Treasury before final maturity. Officials had made active use of the redemption options on the Second Liberty Loan and the Victory Liberty Loan during the 1920s, in part to redeem a portion of the securities as funds became available and in part to refinance a portion with smaller issues to later dates when additional funds were expected to become available. There was little prospect, in mid-1930, of redeeming, any time in the foreseeable future, any of the then-outstanding notes and bonds out of budget surpluses. The redemption options on the three notes and on the First and Fourth Liberty bonds were nevertheless valuable because they allowed the Treasury to refinance the securities before their final maturities at lower interest rates and longer terms. Figure 18.1 shows that short-term Treasury rates had fallen from a peak of just over 5 percent in May 1929 to less than 2 percent in June 1930. Bond yields had also declined, from 3.65 percent in the middle of 1929 to 3.25 percent in mid-1930, and were back at the lows reached in the spring of 1928. 4. The Mellons are discussed in the appendix to chapter 12.

Chapter 18

6.00

5.00

4.00

Percent

264

3.00

2.00

1.00

0 1923

1924

1925

1926

1927

1928

Certificates of Indebtedness

1929

1930

1931

Bonds

Figure 18.1 Yields on Treasury securities

This gave the Treasury an opportunity to refinance advantageously the three notes (which bore coupon rates of 3½ percent), the Fourth Liberty bond (which bore a coupon rate of 4¼ percent), and the three conversion bonds derived from the First Liberty bond (the First Liberty 4s, the First Liberty 4¼s, and the First Liberty second 4¼s).5 On September 10, 1930, Secretary Mellon announced that the Treasury would redeem on March 15, 1931, the $1.1 billion of notes maturing in March and September 1932.6 The announcement was interpreted as reflecting the belief of Treasury officials that interest rates were unlikely to rise during the next six months and that the notes could be refinanced at advantageous rates.7 5. The appendix to chapter 5 describes the conversion option in the First Liberty bond and explains how investors exercised that option. 6. The call notice appears in Treasury Circular no. 428, September 10, 1930, reprinted in 1930 Treasury Annual Report, p. 315. See also “Treasury Calls in $1,149,000,000 Notes,” New York Times, September 10, 1930, p. 24. 7. “Mellon’s Plan Index of Higher Bond Prices,” New York Times, September 11, 1930, p. 38 (“As the Treasury expects to take care of the notes with an issue bearing a lower rate of interest, the belief was reflected that money rates would continue easy or even rule easier. The announcement caused a sharp increase in the demand for United States Government obligations.”) and “Treasury Note Refunding Seen,” Wall Street Journal, September 16, 1930, p. 8 (“Calling of the [notes] at this time shows that the Treasury means to take full advantage of the prevailing ease in money rates and indicates the likelihood that easy money—although perhaps not as easy as at present— will continue over the next few months.”).

265

Treasury Debt Management during the Great Contraction

Three months later Mellon revealed more of the Treasury’s emerging debt management strategy when he requested, in his 1930 annual report to Congress, authority to issue up to $8 billion of new bonds.8 Bond issuance during the 1920s had relied on a provision in the Second Liberty Bond Act that (as amended by the Fourth Liberty Bond Act) authorized the Secretary of the Treasury to issue up to $20 billion of bonds. The fifth Mellon, the 31⁄8s of 1943 issued in the summer of 1928, left a balance of only $1.9 billion of unused authority9 and Mellon said he wanted greater latitude in future debt management decisions: While it is impossible to forecast at this time what form future refunding operations will take, it is obvious that the orderly and economical management of the public debt requires that the Treasury Department should have complete freedom in determining the character of securities to be issued . . ..10

Congress acceded to Mellon’s request in early March 1931.11 At the beginning of 1931 Treasury officials were poised to begin a calculated program of refinancing the nation’s debt at longer maturities and lower interest rates. However, that nascent program was first complicated by the veterans’ loan program authorized by the Emergency Adjusted Compensation Act of 1931 and then derailed by the September 1931 decision of Great Britain to abandon the gold standard. The Emergency Adjusted Compensation Act of 1931 Following passage of the Adjusted Service Compensation Act in 1924,12 the Veterans’ Bureau (the forerunner of today’s Veterans Administration) issued zero-coupon adjusted service certificates with an aggregate face value of about $3.4 billion to veterans of World War I. The certificates were scheduled to be paid in the mid-1940s, but when the Depression deepened in late 1930, some congressmen began to contemplate a faster payout. In November 1930 Senator Arthur Vandenberg (Republican of Michigan) wrote to President Hoover requesting that the administration study the 8. 1930 Treasury Annual Report, p. 39, and “Mellon Asks $8,000,000,000 Issue,” Wall Street Journal, December 5, 1930, p. 15. 9. The $18.1 billion of previously issued bonds included $3.8 billion of the Second Liberty Loan, $4.2 billion of the Third Liberty Loan, $7.0 billion of the Fourth Liberty Loan, $0.8 billion of the 4¼s of 1952 (issued in October 1922), $1.0 billion of the 4s of 1954 (issued in December 1924 and March 1925), $0.5 billion of the 3¾s of 1956 (issued in March 1926), $0.5 billion of the 33⁄8s of 1947 (issued in June 1927), and $0.4 billion of the 33⁄8s of 1943 (issued in the summer of 1928). The First Liberty Loan was issued pursuant to the provisions of the First Liberty Bond Act and did not count against the $20 billion limit. 10. 1930 Treasury Annual Report, p. 39. 11. See Act of March 3, 1931. 12. See chapter 17.

266

Chapter 18

possibility of paying off the certificates at full face value immediately. Hoover referred the request to Mellon, who replied that the Treasury could not “successfully sell three and a half billions of bonds for this purpose at this time.”13 The question of early payment of the adjusted service certificates attracted national attention in the course of hearings before the Senate Finance Committee and the House Ways and Means Committee in late January 1931.14 The possibility of a large bond issue to finance an early payment program was, according to The New York Times, “responsible for . . . widespread selling of Liberty bonds and long-term treasury issues.”15 Between Saturday, January 24, and Saturday, January 31, yields on outstanding Treasury bonds rose 20 basis points. The Times characterized the decline in bond prices as “the worst break in nearly ten years.”16 Concerned that investor uncertainty might affect the terms on which the $1.1 billion of notes called for redemption in March could be refinanced,17 Congress seized on a suggestion from Owen Young, chairman of the General Electric Company and prospective Democratic presidential candidate, that the Treasury lend certificate holders up to 50 percent of the face amount of their certificates.18 Loans were viewed as likely to be less expensive than outright payments because it was believed unlikely that every veteran would borrow 13. “Bonus Cashing Plan Opposed by Mellon,” New York Times, December 7, 1930, p. 1. Mellon subsequently opposed several other proposals for early payment. “Robinson Asks Cooperation on Relief, But He Condemns President’s Criticism,” New York Times, December 11, 1930, p. 1, “Mellon Opposes Fish Bonus Plan,” New York Times, December 16, 1930, p. 2, and “Mellon Condemns Veterans’ Cash Bill,” New York Times, January 22, 1931, p. 12. 14. “Veterans Cash Plan Totals 3½ Billions,” New York Times, January 27, 1931, p. 16, “Mellon Attacks Proposal,” Wall Street Journal, January 29, 1931, p. 15, and “Mellon Takes War on Bonus to House,” New York Times, January 30, 1931, p. 1. 15. “Peril to Financing by Treasury Feared,” New York Times, February 1, 1931, p. 41. See also, “Market Comment,” Wall Street Journal, January 29, 1931, p. 10 (reporting that “recent liquidation in [Treasury] bonds is attributed to selling by bankers who are impressed by progress in Washington of the movement to pass legislation to authorize payment of soldier bonus certificates. If the government was forced to pay the full value of these certificates it would cost between three and four billion dollars and undoubtedly make a large bond issue necessary.”), and “Bonus Bond Issue Would Close Banks, Mitchell Asserts,” New York Times, February 3, 1931, p. 1. 16. “Peril to Financing by Treasury Feared,” New York Times, February 1, 1931, p. 41. 17. “Peril to Financing by Treasury Feared,” New York Times, February 1, 1931, p. 41 (reporting “grave apprehension that a continued decline in [Treasury] bond prices [attributed to investor concern over early payment of the certificates] will jeopardize the treasury’s plan for refunding the two issues of 3½ per cent treasury notes . . ..”) and “Doubt Long Funding of Treasury Notes,” New York Times, February 3, 1931, p. 42. 18. “Young Favors Cash Bonus for Needy Veterans Only; Urges Federal Bank Curb,” New York Times, February 5, 1931, p. 1, “Bonus Agreement Plans $500,000,000,” New York Times, February 7, 1931, p. 1, and “Republicans Speed Bonus Plan Action,” New York Times, February 10, 1931, p. 12. Senator Vandenberg had offered a similar plan earlier (“Veterans Cash Plan Totals 3½ Billions, New York Times, January 27, 1931, p. 16, and “Bonus Bond Issue Would Close Banks, Mitchell Asserts,” New York Times, February 3, 1931, p. 1) but political support coalesced after Young backed the proposal.

267

Treasury Debt Management during the Great Contraction

the full amount allowed.19 Young’s proposal became the basis for the Emergency Adjusted Compensation Act of February 27, 1931, passed over a presidential veto in late February. Treasury officials estimated that the loan provisions of the Emergency Adjusted Compensation Act would require $1 billion in additional funding.20 New funding was required for two reasons. First, the act authorized the Administrator of Veterans’ Affairs to finance loans to veterans by redeeming the special nonmarketable Treasury securities issued during the 1920s to the Adjusted Service Certificate Fund. On the day the act became law, the Fund held $755 million of nonmarketable securities.21 To finance the redemptions, the Treasury would have to sell conventional debt to public investors.22 Second, the Treasury would have to finance whatever additional appropriations Congress provided to fund veterans’ loans. Since 1928, Congress had appropriated $112 million annually to the Fund, on the first day of January each year. In 1931, following the usual $112 million January appropriation, Congress appropriated an extra $112 million in March and another $200 million in December.23 The prospect of $1 billion in new financing quickly dampened the interest of investors in long-term Treasury debt. Treasury officials had planned to refinance most of the $1.1 billion of notes called for redemption in March 1931 with an offering of $1 billion of 12-year, 3¼ percent bonds.24 However, 19. “Loans on Bonus Up to 50% Voted by House Committee; Cost Is Put at $700,000,000,” New York Times, February 13, 1931, p. 1 (reporting an estimate that 60 percent of veterans would apply for loans). 20. “Mellon Warns Congress Bonus Means $1,000,000,000 in Government Financing,” New York Times, February 14, 1931, p. 1, and “House Overrides Hoover Bonus Veto; Vote is 328 to 79,” New York Times, February 27, 1931, p. 1 (reporting a comment from President Hoover that the act would require “approximately $1,000,000,000”). 21. 1931 Treasury Annual Report, p. 118. 22. 1931 Treasury Annual Report, p. 54 (“In order to provide funds for the loans to veterans . . . it was necessary for the Treasury to convert into cash securities held in the adjusted service certificate fund. The making of loans from this fund involves essentially the substitution of the veterans’ notes for United States obligations held in the fund as investment, and the sale of the latter in the open market; actually the special United States obligations held in the fund are not sold but are redeemed by the Treasury as loans are made to veterans, and other United States securities are sold in the market according to the Treasury’s cash requirements.”). 23. 1931 Treasury Annual Report, pp. 118 and 120, and 1932 Treasury Annual Report, pp. 107 and 108. 24. “Congress Concurs on Bond Refunding,” New York Times, March 3, 1931, p. 4, and “Bonus Deferred U.S. Refunding,” Wall Street Journal, March 4, 1931, p. 1. See also, “Treasury Issue Well Received,” Wall Street Journal, March 3, 1931, p. 20 (reporting that “had there been no disturbance in the market as the result of bonus agitation, the Treasury would have been able to place a much larger proportion of the present offering in the form of a bond issue, and at possibly lower interest”). And see further, “Government Loan to be Near Record,” New York Times, March 1, 1931, p. N9 (reporting the statement of Under Secretary Mills that the $1.1 billion of notes to be redeemed on March 15 had been called before the question of early payment of the certificates had come up, and remarking that many observers doubted that the Treasury would have called the notes had the additional financing requirements been envisioned in September of 1930).

268

Chapter 18

in the wake of the congressional override of the President’s veto of the Emergency Adjusted Compensation Act, Treasury officials announced that they would offer only $500 million of 12-year bonds, and at the higher rate of 33⁄8 percent. The balance of the notes would be refinanced with short-term certificates of indebtedness.25 Market conditions similarly led the Treasury to finance the veterans’ loan program with short-term debt. In addition to the 12-year bonds, the Treasury sold $300 million of 6-month certificates of indebtedness and $624 million of 1-year certificates in the March financing. The excess over the $1.1 billion needed to retire the notes went to fund loans to veterans.26 And within a month the Treasury was back in the market, raising more money for the loan program with an offering of $275 million of 8-month certificates.27 That offering was the first sale of certificates for settlement on other than a quarterly tax date in two and a half years and the first such sale other than in connection with the refinancing of a Liberty Loan in almost nine years.28 The Introduction of “Regular and Predictable” Treasury Bill Offerings The Treasury also raised substantial amounts of new money in 1931 with Treasury bills. The 1931 bill financings had a distinctly different tone than earlier bill sales and led to the first instance of “regular and predictable” sales of Treasury debt securities—a debt management policy that was extended to notes and bonds in the second half of the 1970s and that has since become one of the pillars of the Treasury market. Table 18.1 shows that the Treasury paid off four Treasury bills shortly after three of the four tax dates in 1930. The paydowns were consistent with the idea put forth by Under Secretary Mills in 1929 that bills would be used for cash management, to be sold when money was needed and timed to mature shortly after quarterly tax dates.29 At the beginning of 1931 there was only a single, $127 million Treasury bill outstanding.30 By the end of June, however, there were six bills outstanding 25. “Treasury Offers $1,400,000,000 Issue,” New York Times, March 2, 1931, p. 1. 26. “Treasury Offers $1,400,000,000 Issue,” New York Times, March 2, 1931, p. 1. 27. “Treasury to Obtain Large Loan at Once to Meet Bonus Drain,” New York Times, April 2, 1931, p. 1, and “Treasury Offers $275,000,000 Issue,” New York Times, April 8, 1934, p. 25. 28. “Treasury Certificate Issue,” New York Times, April 2, 1931, p. 41 (commenting that “it is extremely unusual for the treasury to issue certificates of indebtedness on other than the quarterly tax dates. Such a course has not been followed within recent years.”) and “Treasury to Obtain Large Loan at Once to Meet Bonus Drain,” New York Times, April 2, 1931, p. 1 (commenting that “rarely since the war has there been a certificate issue at any time except at quarterly tax payment dates”). 29. This idea was not scrupulously observed even in 1930. A $50 million bill matured on July 14, 1930, and was refinanced at maturity with another bill. However, the latter bill matured on a tax date (September 15, 1930) and was not itself refinanced with a third bill. 30. The bill was issued on November 17, 1930, and matured on February 16, 1931.

269

Treasury Debt Management during the Great Contraction

with an aggregate face amount of $445 million; by the end of the year there were eight bills with an aggregate face amount of $575 million. Treasury officials explained their enthusiasm for bill financing by noting that the government saved money borrowing in small amounts as the need arose, rather than relying exclusively on large quarterly financings. When the Treasury announced on May 4 an offering of $50 million of bills, The Wall Street Journal remarked that “the Treasury has been saving interest by borrowing in small lots, rather than borrowing full amount needed on quarterly financing dates and having it stay idle until needed.”31 The New York Times reported similarly that “the Treasury does not desire to keep any more cash than is absolutely necessary and therefore makes comparatively small borrowings frequently as the demands for cash are foreseen.”32 The idea of borrowing when funds were needed was a reasonable explanation for the Treasury’s use of bills in financing loans to veterans, and it was certainly consistent with the idea of bills as a cash management device. However, it does not explain three other features of bill issuance in 1931: none of the bills issued in 1931 were issued with a maturity date on or shortly after a tax payment date, all but one of the bills matured 13 weeks after issue, and all but one of the bills was refinanced with a new 13-week bill.33 These three features suggest that the Treasury began to use bills for debt management in 1931. In May 1931 The New York Times observed that “Temporary borrowings [in the bill market] have now become a matter of routine in the affairs of the treasury, and the old policy of restricting them closely to issues what would fall due on quarterly tax-payment dates has . . . been abandoned . . .. ”34 Figure 18.2 shows the structure of 13-week bill issuance in 1931 and 1932. Weeks when 13-week bills were issued are shown with a gray box. Weeks are classified into thirteen “cycles.” Cycle index 1 denotes the week beginning Monday, January 5, 1931, and every thirteenth week thereafter. Cycle index 2 denotes the week beginning January 12, 1931, and every thirteenth week thereafter. Subsequent cycles are defined analogously. If there was no previously issued 13-week bill maturing in the same week, the box is marked with a star (*). (There are eight such weeks.) If there was a 13-week bill maturing in the same week, the box is marked with a plus (+) to denote an increase in issuance (eight cases), a cross (×) to denote a decrease (one case), or left empty 31. “Treasury Offers $50,000,000 Bills,” Wall Street Journal, May 5, 1931, p. 13. 32. “Treasury Indicates Big Bond Issue Soon,” New York Times, May 5, 1931, p. 5. 33. The exceptions arose in the context of refinancing $154 million of bills maturing May 18, 1931. On May 18 the Treasury paid down (rather than refinanced) $54 million of the maturing bills and refinanced the balance with $50 million of 60-day bills maturing July 17, 1931, and $50 million of 91-day bills maturing August 18, 1931. 34. “Treasury Indicates Big Bond Issue Soon,” New York Times, May 5, 1931, p. 5. Emphasis added.

Chapter 18

13 12 11 10

Cycle index

270

9 8 7 6 5 4 3 2 1 0 Jan 1931

Jul 1931 None

New

Jan 1932 Increase

Jul 1932 No change

Jan 1933

Decrease

Figure 18.2 Weekly sales of 13-week Treasury bills. Cycle index 1 denotes the week beginning Monday, January 5, 1931, and every thirteenth week thereafter. Cycle index 2 denotes the week beginning Monday, January 12, 1931, and every thirteenth week thereafter. Subsequent cycles are defined analogously.

to denote a simple refinancing of the maturing bill (thirty-nine cases). Weeks when no 13-week bills were issued are shown with a circle. Figure 18.2 shows that Treasury officials introduced a new cycle of 13-week bills seven times in 1931.35 Contemporaneous accounts specifically noted that the proceeds of the offerings would go to finance loans to veterans.36 Officials did not subsequently terminate any of the cycles (by paying down a maturing bill), but regularly refinanced each bill with another bill. By the fourth quarter of 1931 the Treasury was regularly selling bills in eight out of thirteen weeks each quarter. This pattern continued in 1932, when the Treasury added one cycle and terminated a different cycle.37 35. The new cycles included cycle 5 in the week beginning Monday, February 2, 1931, cycle 13 in the week beginning March 30, cycle 4 in the week beginning April 27, cycle 6 in the week beginning May 11, cycle 9 in the week beginning June 1, cycle 2 in the week beginning July 13, and cycle 8 in the week beginning August 24. 36. See, for example, “Treasury Prepares to Meet Loan Needs,” Wall Street Journal, March 26, 1931, p. 1 (“issue is entirely for veterans’ loan purposes as Treasury has no other immediate need for funds, as next maturities come in May”) and “Treasury Sells $50,000,000 Bills,” Wall Street Journal, April 21, 1931, p. 14 (“There are no maturities to be met at this time; the Treasury is . . . borrowing to meet expenses and expenditures necessitated by veterans’ loans . . ..”). 37. The Treasury introduced cycle 3 bills in the week beginning April 18, 1932. It also terminated one bill cycle in 1932: the cycle 5 bill issued in the week beginning November 2, 1931, was paid down when it matured in the week beginning February 1, 1932.

271

Treasury Debt Management during the Great Contraction

Figure 18.2 also shows that the Treasury did not issue bills in weeks beginning at about the 15th of the third month of a quarter—cycle 11 denotes the week beginning Monday, March 16, 1931, and every thirteenth week thereafter—or in either of the two surrounding weeks (weeks with cycle index 10 or 12). Funds could be raised more conveniently at those times in the regular quarterly tax date financings. Debt Extension in 1931 The Emergency Adjusted Compensation Act complicated the efforts of Treasury officials to extend the maturity of Treasury debt because it led them to finance loans to veterans with short-term debt, including Treasury bills and certificates of indebtedness. It did not, however, derail their efforts to extend debt maturities. Even before the Emergency Adjusted Compensation Act was passed, bankers were quoted as saying that the federal government needed to reduce its short-term debt in preparation for refunding the Fourth Liberty Loan, which would become redeemable at the option of the Treasury in October 1933.38 Once the short-term character of the financing operations for the veterans’ loan program became clear, there was more frequent commentary on the increasingly “unwieldy” character of the short-term debt.39 Shortly after announcement of the March 1931 tax date financing, Congress acceded to Mellon’s request for authority to issue an additional $8 billion of Treasury bonds.40 The enlarged authority enabled Mellon and Mills to continue with their program of extending the maturity of the nation’s debt. As shown in table 18.3, the Treasury sold $821 million of 18-year, 31⁄8 percent bonds in the June 1931 financing. The bonds were so well-received— the Treasury received subscriptions for $6.3 billion of the bonds—that Mellon 38. “Bonus Financing Held Inescapable,” New York Times, February 17, 1931, p. 2 (reporting that refunding operations for the Fourth Liberty Loan “could not be undertaken, in the opinion of bankers, unless provision were first made for reducing the outstanding volume of short-term obligations . . ..”). See also “Government Loan to be Near Record,” New York Times, March 1, 1931, p. N9 (noting “the opinion of many bankers [that] the outstanding short-dated securities should be materially reduced. This is considered particularly desirable in view of the extensive refunding operations with which the Treasury will be faced in 1933.”). 39. “Treasury Offers $275,000,000 Issue,” New York Times, April 8, 1931, p. 25 (“Secretary Mellon and Under-Secretary Ogden L. Mills have repeatedly stated that [the short-term debt] is too large to be properly handled, and much heavier than should exist in view of the possible refunding of Liberty bond issues . . ..”), “Short Term Debt Excess Feared,” Wall Street Journal, April 10, 1931, p. 13 (“As the short term debt grows, it becomes unwieldy.”), and “Treasury Offers $50,000,000 Issue,” New York Times, April 21, 1931, p. 46 (“The short term debt has become unwieldy and, as soon as a favorable opportunity is presented, it will be greatly reduced with large refunding operations involving a bond issue.”). 40. Act of March 3, 1931, and “Congress Concurs on Bond Refunding,” New York Times, March 3, 1931, p. 4.

272

Chapter 18

Table 18.3 Refinancings and new financings on the first three tax dates in 1931 ($millions) Date

Maturing

Mar 16, 1931

626 483

Issued

New money Note issued Mar 15, 1927 Note issued Sep 15, 1927 Certificate to mature Sep 15, 1931 Certificate to mature Mar 15, 1932 33⁄8% Bond to mature Mar 15, 1943

300 624 594 409 Jun 15, 1931

429 160

Certificate issued Jun 16, 1930 Certificate issued Dec 15, 1930 31⁄8% Bond to mature Jun 15, 1949

821 232 Sep 15, 1931

334 300

Certificate issued Sep 15, 1930 Certificate issued Mar 16, 1931 Certificate to mature Sep 15, 1932 3% Bond to mature Sep 15, 1955

314 800 480

Source: Treasury Annual Reports and “Treasury Offers $1,400,000,000 Issue,” New York Times, March 2, 1931, p. 1.

announced on June 7 that the Treasury would redeem the last outstanding Treasury note, the 3½s of December 15, 1932, on December 15, 1931, expecting to refinance the note with a long-term bond at an interest rate of 3 or 31⁄8 percent.41 The Mellon/Mills program of debt extension seemed to be firmly in place when the Treasury offered a third tranche of bonds—$800 million of 24-year, 3 percent bonds—in the September tax date financing. However, the September bonds proved to be a tough sell. The Treasury had to keep the offering open longer than usual—for five days, whereas the March bond offering was closed the day after it was announced and the June bond offering was open for only two days—and ultimately had to apply what The New York Times later described as “official pressure” to some of the larger banks to avoid a failed offering.42 The poorly received offering left the bond market unprepared for what came next. 41. Treasury Circular no. 439, June 8, 1931, reprinted in 1931 Treasury Annual Report, p. 349, and “Mellon to Redeem Note Issue Dec. 15,” New York Times, June 8, 1931, p. 18. 42. 1931 Treasury Annual Report, pp. 312, 325, and 338, “Treasury Now Offering $300,000,000 in 3% Bonds, $500,000,000 in 21⁄8% Notes,” New York Times, June 4, 1934, p. 1 (remarking that “The September, 1931, 3 per cent issue … was floated after the books on subscriptions had been kept open for some time longer than had been customary.”), and “The Under-Subscribed Loan,” New York Times, September 1, 1935, p. E8 (noting that “voluntary subscriptions did not cover the full amount [of the September 1931 bond offering], and official pressure had to be applied to the larger banks to make up the deficiency”).

273

Treasury Debt Management during the Great Contraction

The Gold Crisis Derails the Debt Extension Program The public’s interest in buying Treasury bonds evaporated in the wake of the announcement by Great Britain on Sunday, September 20, 1931, that it was abandoning the gold standard. Concerned about the potential consequences of the ensuing demand for American gold by foreign investors, the Federal Reserve Banks began raising their discount rates. The Federal Reserve Bank of New York raised its discount rate from 1½ percent in mid-September to 3½ percent in mid-October. Yields on Treasury bonds rose to 3.93 percent in December (figure 18.3). Losses on previously issued bonds—the price of the 3 percent, 24-year bond issued in September dropped to the low 80s—soured investors on further purchases and officials were forced to limit their offerings in the December 1931 tax date financing to three short-term securities: $300 million of 6-month certificates, $400 million of 9-month certificates, and $600 million of 1-year notes.43 In addition to being unable to sell long-term debt, the Treasury had to offer substantially higher rates on its short-term debt. The 1-year notes carried a 3¼ percent coupon; three months earlier the Treasury had sold 1-year certificates with a 11⁄8 percent coupon. (The notes were also fully tax-exempt, the first fully tax-exempt note issue since the Victory Liberty Loan in 1919.) Bond yields rose still higher in January 1932, to 4.25 percent, and the Treasury continued to limit its financings to the front end of the yield curve. In late January officials offered a total of $350 million in 31⁄8 percent 6-month certificates and 3¾ percent 1-year certificates. They were so unsure of the demand for Treasury debt at the time of the offering that they specified that the two certificates would be issued in proportion to their subscriptions.44 A provision in the Second Liberty Bond Act (as amended by the Third Liberty Bond Act) restricted the interest rate on new bonds to not more than 43. “Treasury Offers $1,300,000,000 Issue, Largest Since War,” New York Times, December 7, 1931, p. 1. See also “Easy Treasury Sales Mislead,” Wall Street Journal, December 29, 1932, p. 1 (noting that in December 1931 “the Treasury found itself unable to float any new securities with more than a year’s maturity . . ..”) and “New Federal Loan Sells at a Premium,” New York Times, August 1, 1933, p. 23 (commenting that a mid-1933 bond issue was “the first that the Treasury has made since Sept. 15, 1931, when former Secretary of the Treasury Andrew Mellon put out a floatation of 3 per cent bonds. That ill-fated issue, launched just before the suspension of the gold standard by England, has sold at a discount ever since it was floated. The experience of the banks with the Mellon 3s has made them wary and has prevented renewed bond financing by the Treasury for two years.”). 44. Treasury Circular no. 454, January 25, 1932, reprinted in 1932 Treasury Annual Report, p. 226 (“The amount of each series [of certificates] will be in the proportion that the total subscriptions for that series bears to the total subscriptions received for both series.”). 61.3 percent of the total subscriptions were for the 6-month certificates, so 61.3 percent of the certificates were issued as 6-month certificates.

Chapter 18

5.00

4.00

Percent

274

3.00

2.00

1.00

0 Jan 1930

Jan 1931

Jan 1932

Certificates of Indebtedness

Jan 1933 Bonds

Figure 18.3 Yields on Treasury securities

4¼ percent and kept the Treasury from offering bonds in the March 1932 tax date financing. The New York Times reported that “Treasury officials would welcome the opportunity for an issue of bonds because of the unwieldy volume of the short-term debt. However, unless there is a radical change in the money market, this is considered impossible . . .. With the last issue of one-year certificates [i.e., the 1-year certificates sold in late January] commanding 3¾ per cent, bonds could hardly be floated within the [4¼ percent] interest limit.”45 Short-term interest rates began to fall sharply after the Federal Reserve initiated, during the spring of 1932, a program of aggressive open market purchases of Treasury securities.46 Bond yields receded to about 3.75 percent. The Treasury could have sold bonds with interest rates below the 4¼ percent ceiling that spring, but the wide spread between short- and long-term yields kept the Treasury issuing shorter term debt for the balance of the year. The Wall Street Journal reported in August that “so long as cheap [short-term] funds remain available, officials will be inclined to take advantage of them to save the Treasury as much in interest costs as possible.”47 45. “Short-Term Issue Announced by Mills,” New York Times, February 25, 1932, p. 31. 46. Chandler (1971, ch. 13) and Meltzer (2003, pp. 357–63) describe the Fed’s open market purchases in the spring of 1932. 47. “Low Rates Hint U.S. Conversion,” Wall Street Journal, August 25, 1932, p. 1.

275

Treasury Debt Management during the Great Contraction

Table 18.4 Selected balance sheet entries of the Reconstruction Finance Corporation ($millions)

Cash Loans Distributions for relief Capital (issued to Treasury) Notes (issued to Treasury)

Mar 31, 1932

Jun 30, 1932

Sep 30, 1932

Dec 31, 1932

Mar 31, 1933

116 185 — 350 —

51 729 — 500 350

32 995 14 500 600

8 1,145 80 500 810

4 1,389 201 500 1,220

Source: Quarterly reports of the Reconstruction Finance Corporation.

Debt Management in 1932: Financing the Reconstruction Finance Corporation Congress passed the Reconstruction Finance Corporation Act on January 22, 1932, in the wake of the failure of the voluntary National Credit Corporation to alleviate the growing economic distress.48 Congress authorized the RFC to borrow from the Treasury and to lend to banks, other financial institutions, and railroads. The Emergency Relief and Construction Act of July 21, 1932, extended the RFC’s mandate to include lending to states for unemployment relief and to finance self-liquidating projects. The RFC loan programs substantially enlarged Treasury financing requirements. By the end of 1932 the RFC had sold $500 million of capital stock and $800 million of notes to the Treasury and had dispersed substantially all of the proceeds (table 18.4). The need to finance the RFC led Treasury officials to make two changes in their debt management practices: they began selling coupon-bearing debt on a regular basis at times other than tax payment dates and, after the Federal Reserve accelerated its program of monetary expansion in April 1932,49 they began selling intermediate-term notes as well as certificates of indebtedness.50 Figure 18.4 shows the offerings of coupon-bearing Treasury debt between January 1932 and March 1933, inclusive. The frequency of financings was 48. See chapter 16. 49. Chandler (1971, p. 197) and Meltzer (2003, pp. 361–63) discuss the acceleration of the Federal Reserve program of monetary expansion. 50. 1932 Treasury Annual Report, p. 67 (“Beginning in May, 1932, Treasury notes with maturities from two to five years . . . were sold in large volume.”). The Treasury did not increase significantly the outstanding quantity of Treasury bills in 1932 because it concluded that, “In view of the already large amount of early maturities outstanding . . . the issuance of longer maturities as conditions warranted, even at somewhat higher rates, was in the interest of sound and orderly management of the public debt.” 1932 Treasury Annual Report, p. 67. At the end of 1931 there was $575 million of Treasury bills outstanding; a year later there was $641 million of bills outstanding.

Chapter 18

1,500

1,250

Millions of dollars

276

1,000

750

500

250

0 Jan 1932

Apr 1932

Tax date certificates

Jul 1932 Tax date notes

Oct 1932

Jan 1933

Other certificates

Other notes

Figure 18.4 Offerings of coupon-bearing Treasury debt, on tax payment dates and other dates

double that of previous years: twice a quarter instead of once a quarter, and the increased use of notes after the first quarter of 1932 is plainly evident. Notes accounted for almost 80 percent of the coupon-bearing debt sold in the second half of 1932 and commentators remarked on the strong investor demand for the securities.51 The Treasury did not omit notes from an offering until the March 1933 tax date financing, when the collapse of the banking system precluded anything other than short-term debt. (The March 1933 financing offered a total of $800 million of 5-month and 9-month certificates of indebtedness. The Treasury had not offered certificates with a maturity shorter than six months in more than a decade. Officials were again so unsure of the demand for the certificates that they specified that the certificates would be issued in proportion to their subscriptions.52) Figure 18.5 shows the term to maturity, and figure 18.6 shows the issue size, of the certificates and notes issued between January 1932 and March 1933. Except for the tendency to regularly offer 1-year certificates, there was little consistency in either term or size from offering to offering. There was, however, some consistency in the selection of offering dates. Offerings other than tax 51. See, for example, “Treasury to Offer $450,000,000 Notes,” New York Times, October 6, 1932, p. 31 (reporting “an excellent market for notes”). 52. Treasury Circular no. 431, March 13, 1933, reprinted in 1933 Treasury Annual report, p. 162.

Treasury Debt Management during the Great Contraction

5

Years

4

3

2

1

0 Jan 1932

Apr 1932

Tax date certificates

Jul 1932

Oct 1932

Tax date notes

Jan 1933

Other certificates

Apr 1933

Other notes

Figure 18.5 Term to maturity of new offerings of coupon-bearing Treasury debt, on tax payment dates and other dates

1,000

750 Millions of dollars

277

500

250

0 Jan 1932

Apr 1932

Tax date certificates

Jul 1932 Tax date notes

Oct 1932

Jan 1933

Other certificates

Apr 1933

Other notes

Figure 18.6 Issue size of new offerings of coupon-bearing Treasury debt, on tax payment dates and other dates

278

Chapter 18

date offerings were generally settled on the first business day of the second month of a quarter, about halfway between the preceding and following tax dates. (The only exception is the sale of notes for settlement on October 15, 1932.) The lack of consistency in the tenor of note offerings is somewhat surprising in view of the Treasury’s recent regularization of 13-week bill offerings, but it may be that officials were simply working up to regularly offering notes with a maturity of five years or thereabouts. The lack of consistency in issue size is more understandable, in view of the variability of RFC disbursements. Concluding Remarks For reasons beyond the control of Treasury officials, Treasury debt management during the Great Contraction was less proactive, and more reactive, than had been the case in the 1920s. In late 1930 Mellon and Mills mapped out a program of refinancing the nation’s debt at longer maturities and lower rates, but their efforts were thwarted, first by passage of the Emergency Adjusted Compensation Act, then by the financial turmoil that followed Britain’s decision to abandon the gold standard, and finally by the need to finance the Reconstruction Finance Corporation. Nevertheless, Mellon and Mills came away with some notable accomplishments, financing veterans’ loans with a novel program of regular and predictable bill issuance and financing the RFC with notes when they couldn’t sell bonds on sufficiently attractive terms. Perhaps the only criticism that can be lodged against their policies is that they failed to see that the severity of the Great Contraction made the national debt more or less permanent and that a permanent debt might be financed more efficiently with a program that offered certificates of indebtedness and notes as well as bills on a regular and predictable basis.

19

Treasury Debt Management during the New Deal

The strategic objectives of Treasury debt managers during the New Deal, from the spring of 1933 to mid-1939, hardly differed from those of Andrew Mellon and Ogden Mills during the Great Contraction. The need to finance chronic deficits, and the need to refinance increasing amounts of maturing debt, prompted continuing interest in extending maturities to avoid a pileup of shortterm obligations that might set the stage for a funding crisis. Officials were also mindful that the $6.3 billion Fourth Liberty Loan would mature in 1938. The loan was too large to pay off in a single operation and would have to be refinanced into more manageable issues prior to maturity, as had been the case with the Second and Third Liberty Loans and the Victory Liberty Loan a decade earlier. (The First Liberty Loan was less of a problem. It was relatively small, only $2 billion, and did not mature until 1947, although the low level of interest rates that prevailed after 1934—see figure 19.1—made refinancing the First Liberties fiscally attractive.) William Woodin, Roosevelt’s first Secretary of the Treasury, and Henry Morgenthau, who followed Woodin when Woodin resigned for health reasons in January 1934, did not break new ground in Treasury debt management, but they did restart the Mellon/Mills program of extending the maturity of Treasury debt and they successfully refinanced the Liberty Loans. Woodin and Morgenthau also continued the regular weekly sales of Treasury bills pioneered by Mellon and Mills. Total bill indebtedness rose to almost $2 billion at the beginning of 1935. (It was $640 million when Roosevelt took office.) In the course of expanding the bill program, Morgenthau experimented with 26- and 39-week bills in lieu of 13-week bills but returned to the shorter maturity in late 1937. Morgenthau additionally experimented with Treasury cash management policies. Most important, in 1935 he began auctioning long-term bonds in relatively small, $100 million lots between quarterly tax dates to raise money on an “as needed” basis. Unfortunately, the effort was marked by incompatible objectives—selling long-term bonds with an auction process that (as discussed below) depended on banks and dealers underwriting the Treasury’s offerings was inherently incompatible with raising money “as needed”—and quickly failed. In late 1935 Morgenthau reinstated, with much greater success, the original Mills program of raising money between tax dates by selling Treasury bills that matured on or shortly after future tax dates. The success of the tax date bill program demonstrated the utility of Mills’s conception, but the program became superfluous and was terminated in 1938 when other relatively steady sources of funds, including social security wage taxes and proceeds from sales of savings bonds, became available. By mid-1939 Treasury debt was at an all-time high of almost $40 billion, financed (as shown in figure 19.2) primarily with bonds, but also with bills, notes and nonmarketable securities. Bills were sold weekly, notes and bonds

Chapter 19

5.00

Percent

4.00

3.00

2.00

1.00

0 1933

1934

1935

1936

1937

Certificates of Indebtedness

1938 Bills

1939

1940

Bonds

Figure 19.1 Yields on Treasury securities

40

Billions of dollars

280

30

20

10

0

1933

1934

1935

Pre-war and war bonds Certificates

1936

1937

Postwar bonds Bills

Figure 19.2 Composition of Treasury debt, mid-1933 to mid-1939

1938 Notes Other debt

1939

281

Treasury Debt Management during the New Deal

were sold at quarterly tax dates, and nonmarketable securities, including savings bonds and special issues for federal trust funds, were issued as needed. Maturity Extension, 1933 and 1934 Debt extension was inconceivable when Franklin Roosevelt took office. Simply keeping the federal government adequately funded after the national bank holiday was quite enough of a miracle and it is hardly surprising that Treasury officials relied primarily on short-term debt in the spring of 1933. In the March tax date financing, the Treasury sold $469 million of unusually short 5-month certificates of indebtedness and $473 million of 9-month certificates. Between March 4 and May 6, Treasury bills outstanding expanded by more than $300 million, from $640 million to $980 million. Nevertheless, it wasn’t long before commentators began to revisit the topic of debt extension. As early as April 2, The New York Times reported “a welldefined opinion that the time is approaching when long-term financing—put aside since September 15, 1931, because of the unsettled conditions of the money market—will be resumed for the purpose of reducing the short-dated debt to more manageable proportions . . ..”1 In July, The Wall Street Journal reported that “Public debt is considered by Treasury officials as off-balance as between long and short term obligations . . ..”2 Between the spring and fall of 1933, the Treasury cautiously tested the market for longer-term securities. As shown in figure 19.3, the Treasury sold 3-year notes in May, 5-year notes in June, and 8-year bonds in August. The August offering was the first bond sale in almost two years. Secretary Woodin described the offering as “important” and marking “a further step in placing government finance on a broader and more stable base.”3 The Wall Street Journal characterized it as a “feeler”: “If the public response . . . is considerable, another, and larger bond issue will likely follow . . ..”4 The Treasury received almost $3 billion of subscriptions for the $500 million of bonds that it offered (The New York Times reported that Woodin “took no chances with his first bond offering” and that the 3¼ percent coupon rate was “highly pleasing to the market”5) and Woodin followed up in October with a 12-year bond. 1. “Recasting the National Debt: A Knotty Problem in Finance,” New York Times, April 2, 1933, p. XX5. 2. “Professors Join Treasury Forces,” Wall Street Journal, July 11, 1933, p. 1. See also, “Roosevelt Studies Long-Term Issue,” New York Times, July 27, 1933, p. 25 (reporting that “For some time the Treasury has been anxious to consolidate a portion of its short-term certificates into bonds . . ..”). 3. “Treasury Makes Popular Offer,” Wall Street Journal, July 31, 1933, p. 1. 4. “Treasury Makes Popular Offer,” Wall Street Journal, July 31, 1933, p. 1. 5. “New Federal Loan Sells at a Premium,” New York Times, August 1, 1933, p. 23.

Chapter 19

20

15

Years

282

10

5

0 Jan 1933

Jul 1933

Jan 1934

Jul 1934

Jan 1935

Figure 19.3 Term to maturity of new offerings of coupon-bearing Treasury debt

Treasury’s maturity extension program stalled after the October offering, when the ill-conceived RFC gold purchase program disrupted the market and led to a fall in bond prices.6 (As shown in figure 19.1, Treasury bond yields rose from 3.20 percent in October to 3.50 percent in January 1934.) Officials offered only a 1-year certificate of indebtedness in the December tax date financing and followed with a 7½-month certificate and a 13½-month note in a late January 1934 financing. Administration officials were reported to believe that the Treasury could not issue long-term bonds except at interest rates that would cause “a downward movement in [outstanding Treasury securities] and would have a very considerable adverse effect on the banks which own a large amount [of the securities].”7 As in the spring of 1933 the Treasury increased its reliance on bills at the same time that it was selling shorter term coupon-bearing debt: the volume of bills outstanding rose from $950 million in early December 1933 to $1.4 billion in late February 1934. Conditions in the bond market began to improve after the January 31, 1934, revaluation of gold to $35 per fine ounce, and Treasury officials gradually revived the maturity extension program, issuing 3-year notes in February, 4-year notes in March, 12-year bonds in April, 14-year bonds in June, and 18-year bonds in December. 6. The gold purchase program is discussed in the appendix to chapter 16. 7. “Treasury Charts Pulse of Market,” New York Times, January 26, 1934, p. 2.

283

Treasury Debt Management during the New Deal

Refinancing the Liberty Loans Treasury officials refinanced the First and Fourth Liberty Loans more or less as they had refinanced the Second and Victory Liberty Loans: by announcing a call for early redemption of some or all of the bonds and then offering to exchange new securities for the called bonds. Boxes 19.1 and 19.2 summarize the calls and exchange offerings.8 Treasury officials added two new features to the Liberty Loan refinancings. Simultaneously with the first call for early redemption of the Fourth Liberty Loan, the Treasury offered to exchange a 12-year bond for the called and uncalled Fourth Liberties.9 The new bond paid 4¼ percent interest the first year and 3¼ percent per annum thereafter. The extra one hundred basis points of interest in the first year was a “sweetener,” intended to mollify bondholders who might otherwise be reluctant to exchange Fourth Liberties, which paid a 4¼ percent coupon, for a new bond paying only 3¼ percent. The Treasury never again issued a split-coupon bond. The second innovation came in the course of setting the terms of the September 1934 exchange offer for the second tranche of Fourth Liberties called for early redemption (commonly referred to as the “second called Fourth Liberties”). Officials wanted to offer another intermediate-term bond but a weak market left then uncertain about the likely success of a bond offering.10 To limit the likelihood of a poorly received offer (and a large demand for cash redemption of the second called Fourth Liberties), officials offered to exchange either 4-year notes or 12-year bonds for the Fourth Liberties.11 This was the first dual-option exchange offer. The Treasury repeated the technique in a June 1935 exchange offer for First Liberty bonds, in a September 1935 8. In calling some, but not all, of the Fourth Liberties, the Treasury had to identify which bonds were being called. This was done by choosing at random the last digit of the serial number of the called bonds. See “Move Surprises Capital,” New York Times, October 12, 1933, p. 1 (“The formal call of the Fourth Liberty bonds took place with an elaborate ceremony, conducted by [Under Secretary of the Treasury Dean Acheson] in the outer offices of Secretary Woodin. High Treasury and Federal Reserve Board officials looked on as Mr. Acheson drew from a glass jar an envelope containing the numbers of the three series to be called. These were the series whose serial numbers of ten digits end on 9, 0 and 1 . . ..”) and “$1,870,000,000 More in Liberties Called,” New York Times, October 13, 1934, p. 23 (“The bonds included in today’s call were those bearing the serial numbers ending in the digits 5, 6, or 7. Secretary Morgenthau, in the presence of a few of his associates, drew from a glass the slips containing the numbers.”). 9. Treasury Circular no. 502, October 12, 1933, reprinted in 1933 Treasury Annual Report, p. 180. 10. “Choice of Bonds Offered by U.S.,” Wall Street Journal, September 10, 1934, p. 1 (“The Treasury attitude apparently is one of extreme skepticism concerning the issuance of long-term bonds at this time.”). See also “Treasury Offers to Refund Issues of $1,774,748,500,” New York Times, September 10, 1934, p. 1 (characterizing the bond market as “soft”). 11. Treasury Circular no. 523, September 10, 1934, reprinted in 1935 Treasury Annual Report, p. 182, and Treasury Circular no. 524, September 10, 1934, reprinted in 1935 Treasury Annual Report, p. 184.

284

Chapter 19

Box 19.1 Refinancing the Fourth Liberty Loan First Call • Announced October 12, 1933: $1.880 billion Fourth Liberty bonds called for early redemption on April 15, 1934.a • Offer announced October 12, 1933, to exchange 12-year bonds paying a 4¼ percent coupon the first year and 3¼ percent thereafter, for called and uncalled Fourth Liberty bonds.b $875 million called Fourth Liberties and $26 million uncalled Fourth Liberties were tendered in exchange.c • Offer announced April 4, 1934, to exchange 3¼ percent 12-year bonds for called Fourth Liberty bonds.d $827 million called Fourth Liberties were tendered in exchange.e • The balance of $178 million called Fourth Liberties were redeemed with cash on and after April 15, 1934.f Second Call • Announced April 13, 1934: $1.246 billion Fourth Liberty bonds called for early redemption on October 15, 1934.g • Offer announced September 10, 1934, to exchange either 2½ percent 4-year notes or 3¼ percent 12-year bonds for second called Fourth Liberty bonds.h $596 million second called Fourth Liberties were tendered in exchange for the notes, $457 million were tendered in exchange for the bonds.i • The balance of $193 million second called Fourth Liberties were redeemed with cash on and after October 15, 1934.j a. Treasury Circular no. 501, October 12, 1933, reprinted in the 1933 Treasury Annual Report, p. 176, and 1934 Treasury Annual Report, p. 9. b. Treasury Circular no. 502, October 12, 1933, reprinted in 1933 Treasury Annual Report, p. 180. c. 1934 Treasury Annual Report, p. 9, and 1940 Treasury Annual Report, p. 56. d. Treasury Circular no. 508, April 4, 1934, reprinted in 1934 Treasury Annual Report, p. 168. e. 1934 Treasury Annual Report, p. 170, and 1940 Treasury Annual Report, p. 56. f. 1934 Treasury Annual Report, pp. 9–10. g. Treasury Circular no. 509, April 13, 1934, reprinted in 1934 Treasury Annual Report, p. 170, and 1934 Treasury Annual Report, p. 10. h. Treasury Circular no. 523, September 10, 1934, reprinted in 1935 Treasury Annual Report, p. 182, and Treasury Circular no. 524, September 10, 1934, reprinted in 1935 Treasury Annual Report, p. 184. i. 1935 Treasury Annual Report, p. 187, and 1940 Treasury Annual Report, p. 56. j. 1935 Treasury Annual Report, p. 19.

exchange offer for the last tranche of Fourth Liberty bonds, and in eight subsequent exchange offers between December 1935 and September 1938. Treasury officials extended the idea further in December 1938, when they made a triple-option offer, offering to exchange a 5-year note, a 9-year bond, or a 27-year bond for a maturing note.12 The Wall Street Journal called the 12. Treasury Circulars no. 598, 599, and 600, December 5, 1938, reprinted in 1939 Treasury Annual Report, pp. 233, 235, and 236, respectively.

285

Treasury Debt Management during the New Deal

Box 19.1 Continued Third Call • Announced October 12, 1934: $1.869 billion Fourth Liberty bonds called for early redemption on April 15, 1935.k • Offer announced March 4, 1935, to exchange 27⁄8 percent 25-year bonds for third called Fourth Liberty bonds.l $1.558 billion third called Fourth Liberties were tendered in exchange.m • The balance of $311 million third called Fourth Liberties were redeemed with cash on and after April 15, 1935.n Fourth (and Final) Call • Announced April 14, 1934: $1.246 billion Fourth Liberty bonds called for early redemption on October 15, 1935.o • Offer announced September 3, 1935, to exchange either 1½ percent 3½-year notes or 2¾ percent 12-year bonds for fourth called Fourth Liberty bonds.p $429 million fourth called Fourth Liberties were tendered in exchange for the notes, $569 million were tendered in exchange for the bonds.q • The balance of $248 million fourth called Fourth Liberties were redeemed with cash on and after October 15, 1935.r k. Treasury Circular no. 525, October 12, 1934, reprinted in 1935 Treasury Annual Report, p. 188, and 1935 Treasury Annual Report, p. 19. l. Treasury Circular no. 531, March 4, 1935, reprinted in 1935 Treasury Annual Report, p. 201. m. 1935 Treasury Annual Report, p. 204, and 1940 Treasury Annual Report, p. 56. n. 1935 Treasury Annual Report, p. 20. o. Treasury Circular no. 539, May 13, 1935, reprinted in 1935 Treasury Annual Report, p. 215, and 1935 Treasury Annual Report, p. 20. p. Treasury Circular no. 550, September 3, 1935, reprinted in 1936 Treasury Annual Report, p. 222, and Treasury Circular no. 551, September 3, 1935, reprinted in 1936 Treasury Annual Report, p. 224. q. 1936 Treasury Annual Report, p. 226, and 1940 Treasury Annual Report, p. 56. r. 1936 Treasury Annual Report, p. 18, and 1940 Treasury Annual Report, p. 56.

triple-option offer an “innovation” that was “regarded in Government bond circles as a recognition by the Treasury of the varied investment tastes” of the owners of Treasury securities.13 (Contemporaneous commentators did not, however, take note of the fact that multiple-option exchange offers weakened the ability of Treasury debt managers to manage the maturity structure of the national debt.14) Multiple-option exchange offers remained a feature of Treasury exchange offers through the 1960s.15 13. “Treasury’s December Financing Plans Bring Rise in Bonds,” Wall Street Journal, December 3, 1938, p. 1. 14. Tilford Gaines (1962, p. 79) later pointed out that multiple-option exchange offers left “the maturity distribution of the debt . . . in the hands of . . . investors.” 15. Garbade (2004a, 2007).

286

Chapter 19

Box 19.2 Refinancing the First Liberty Loan • On March 15, 1935, the Treasury called all $1.933 billion of the First Liberty bonds for early redemption on June 15, 1935.a • On April 22, 1935, the Treasury offered to exchange either 15⁄8 percent 4¾-year notes or 27⁄8 percent 24¾-year bonds for First Liberty bonds.b $864 million First Liberties were tendered in exchange for the notes, $746 million were tendered in exchange for the bonds.c • The balance of $323 million First Liberties were redeemed with cash on and after June 15, 1935.d a. Treasury Circular no. 535, April 22, 1935, reprinted in the 1935 Treasury Annual Report, p. 211, and 1935 Treasury Annual Report, p. 21. b. Treasury Circular no. 536, April 22, 1935, reprinted in 1935 Treasury Annual Report, p. 218, and Treasury Circular no. 537, April 22, 1935, reprinted in 1935 Treasury Annual Report, p. 220. c. 1935 Treasury Annual Report, p. 222, and 1940 Treasury Annual Report, p. 57. d. 1935 Treasury Annual Report, pp. 22 and 223, and 1940 Treasury Annual Report, p. 57.

The 1935 Bond Auctions In late May 1935 the Treasury auctioned $100 million of 13-year bonds,16 the first auction offering of coupon-bearing securities since the 50-year Panama Canal bonds in 1911. Investors submitted tenders for $270 million of the new bonds and the Treasury accepted $98.8 million of the tenders, at prices ranging from 10326 (103 and 26–32nds percent of par value) down to 1031.17 The auction met—but did not materially exceed—the expectations of Treasury officials.18 The sale reopened an $825 million issue that was first sold in a joint cash and exchange offering in June 1934. The outstanding bonds traded in the secondary market at 10326 the day before the announcement of the auction and at 10316 the day before the auction. The average accepted auction price was 1034. The New York Times reported that “officials took into consideration the fact that it was a new venture in which the investors were not skilled and felt that under the circumstances the sale of about $100,000,000 of the bonds at about the market price was a favorable accomplishment. If any 16. Treasury Circular no. 541, May 27, 1935, reprinted in 1935 Treasury Annual Report, p. 223. 17. Tenders were received at Federal Reserve Banks and branches (but not at the Treasury Department in Washington, D.C.) up to 3:00 p.m., Eastern Standard time, on Wednesday, May 29. Auction results were announced on Thursday, May 30. The Treasury rejected bids for $22 million of bonds at a price of 103 rather than ration those bidding at that price to only 5.5 percent of what they bid for ($1.2 million = 5.5 percent of 22 million). “New Bond Bids Treble Offering; Treasury Accepts $98,779,000,” New York Times, May 31, 1935, p. 25. 18. “Treasury Offering of New 3% Bonds Oversubscribed,” Wall Street Journal, May 31, 1935, p. 1.

287

Treasury Debt Management during the New Deal

Table 19.1 Treasury bond auctions in 1935

Bond Coupon (%) Maturity Amount ($million) Announcement Auction Settlement Amount bid ($million) Amount issued ($million) Auction bid prices High Average Stop Secondary market price on day prior to auction

First auction

Second auction

Third auction

Fourth auction

Fifth auction

3 Jun 15, 1948 100 May 27 May 29 Jun 3 270 99

3 Jun 15, 1948 100 Jun 24 Jun 26 Jul 1 461 113

27⁄8 Mar 15, 1960 100 Jul 15 Jul 17 Jul 22 510 102

27⁄8 Mar 15, 1960 100 Jul 29 Jul 31 Aug 5 321 106

27⁄8 Mar 15, 1960 100 Aug 12 Aug 14 Aug 19 147 98

10326 1034 1031 10316

10324 10318 10317 10320

10127 10119 10119 10121

10124 10118 10117 10119

1018 10025 10021 101

Sources: Treasury Circular no. 541, May 27, 1935, reprinted in 1935 Treasury Annual Report, p. 223; 1935 Treasury Annual Report, p. 225; Treasury Circular no. 544, June 24, 1935, reprinted in 1935 Treasury Annual Report, p. 227; 1935 Treasury Annual Report, p. 228; Treasury Circular no. 546, July 15, 1935, reprinted in 1936 Treasury Annual Report, p. 217; 1936 Treasury Annual Report, p. 219; Treasury Circular no. 547, July 29, 1935, reprinted in 1936 Treasury Annual Report, p. 219; 1936 Treasury Annual Report, p. 220; Treasury Circular no. 548, August 12, 1935, reprinted in 1936 Treasury Annual Report, p. 220; 1936 Treasury Annual Report, p. 221, and New York Times, various dates.

disappointment was felt, it was because a large oversubscription had not been received at somewhat over 103.”19 The reintroduction of bond auctions was a major initiative in Treasury debt management. The Wall Street Journal enthusiastically described the new procedure as an “innovation,” but Secretary Morgenthau more guardedly characterized it as an “experiment.”20 Treasury officials were sufficiently encouraged by the results to bring two additional auction offerings in June and July 1935. As summarized in table 19.1, the second offering was for another $100 million of the same 13-year bond and attracted bids for $461 million. The third offering was for $100 million of a 25-year bond and attracted $510 million in 19. “New Bond Bids Treble Offering; Treasury Accepts $98,779,000,” New York Times, May 31, 1935, p. 25. 20. “Treasury to Sell $100,000,000 Bonds to Highest Bidders,” Wall Street Journal, May 27, 1935, p.1, and “Treasury Proposes $770,000,000 Refunding,” Wall Street Journal, May 28, 1935, p. 3. See also “A Treasury Experiment,” New York Times, May 29, 1935, p. 20 (characterizing the first bond auction as an “experiment, intended to test the popularity of such an issue . . ..”).

288

Chapter 19

bids.21 In both auctions the average sale price was only two 32nds of a percent of principal value below the secondary market price on the day prior to the auction. Secretary Morgenthau characterized the results of the third auction as “very satisfactory” and The New York Times reported that “there was every indication that the competitive bidding policy . . . would be continued.”22 Why Did the Treasury Reintroduce Auction Sales of Bonds? Although the Treasury auctioned bonds in seven out of ten offerings between 1894 and 1911,23 it switched to fixed-price offerings during World War I to effect a wider distribution of the war debt and it continued to rely on subscription sales of coupon-bearing securities thereafter. Thus the question naturally arises: Why did the Treasury reintroduce bond auctions in 1935? Secretary Morgenthau decided to auction bonds for two reasons: he wanted to issue notes and bonds on a tactical, discretionary basis, and he wanted a more efficient primary market pricing mechanism. It was clear by 1935 that selling Treasury notes and bonds in fixed-price offerings was costly to taxpayers. Treasury officials sold new issues at par but specified coupon rates that exceeded secondary market yields to limit the risk of a failed offering.24 Figure 19.4 is the proof of the pudding: investors commonly subscribed for vastly more securities than the Treasury offered. The Wall Street Journal said of the new auction process: “by letting purchasers set the prices . . ., it will be possible to come closer to the market than under the usual method of offering bonds at par.”25 Morgenthau stated similarly that “through the sale of bonds on a competitive basis a better rate could be obtained for the government . . ..”26 Morgenthau was also interested in moving away from the practice, introduced by Secretary Mellon in the 1920s, of selling coupon-bearing securities on a regular schedule. In late May 1935, The New York Times reported Morgenthau as saying that the program of issuing securities at only the 21. The third auction reopened a bond that was first sold in an exchange offering for the third called Fourth Liberties in March 1935 and that was reopened in May 1935 in an exchange offering for the First Liberties. At the time of the third auction, that bond was the largest ($2.3 billion) and (other than the Panama Canal bond maturing in 1961) longest Treasury bond in the market. 22. “New Federal Issue Subscribed 5 Times,” New York Times, July 19, 1935, p. 25. 23. See chapter 3. 24. See, for example, “Treasury Offers $100,000,000 Issue in Financing Test,” New York Times, May 27, 1935, p. 1 (reporting that, under the fixed-price subscription method, “it has been necessary for the Treasury so to gauge the market’s appetite as to assure the success of an offering, with the result that the interest rate usually has been slightly above the market.”). 25. “Treasury Proposes $770,000,000 Refunding,” Wall Street Journal, May 28, 1935, p. 3. 26. “Treasury to Sell $100,000,000 Issue,” New York Times, June 24, 1935, p. 25. See also, Joint Economic Committee (July 1959, p. 1157) (Treasury staff report stating that, in 1935, proponents of the auction process believed that auctions would allow the Treasury to obtain funds “at a minimum interest cost”).

289

Treasury Debt Management during the New Deal

10

Billions of dollars

8

6

4

2

0 Jan 1933

Jan 1934 Unfilled subscriptions

Additional issued

Jan 1935 Offered

Figure 19.4 Amount offered, issued, and subscribed in fixed-price Treasury cash subscription offerings

quarterly tax dates had been abandoned and that securities “would be ordered when it appeared that the Treasury needed the money to finance governmental activities . . ..”27 The idea of borrowing money “as needed” was cited repeatedly during the early summer of 1935. In mid-July The New York Times took note of the new Treasury policy of avoiding “heavy lump borrowings” and The Wall Street Journal stated that “The Treasury will conduct its cash borrowing operations in coming months as far as possible through small and frequent issues rather than through large sales of securities on quarterly financing dates . . .. This more flexible method will have the advantage over quarterly sales of cumbersome size in that it will enable the Treasury to keep its borrowings strictly in line with spending needs.”28 Auction Weakness The second Treasury bond auction followed the first by four weeks, the third followed the second by three weeks, and the gap narrowed to two weeks when 27. “Treasury Plans Large Refinancing,” New York Times, May 28, 1935, p. 39. 28. “Steady Financing is Treasury’s Aim,” New York Times, July 12, 1935, p. 27, and “Treasury to Depend More on Auctions for Bond Financing,” Wall Street Journal, July 12, 1935, p. 1. See also, “Treasury Offers $100,000,000 Issue,” New York Times, July 15, 1935, p. 25 (“It is the understanding that in the future the Treasury may at such intervals as additions to the cash balance by borrowings are deemed necessary, vary its programs by relatively small offerings of bonds or shorter term issues as the market conditions dictate.”).

290

Chapter 19

Morgenthau announced the fourth auction: a second tranche of $100 million of 25-year bonds. The results of the fourth auction were about the same as the third in terms of price, but bidding was more restrained. The Treasury received tenders for only $321 million of bonds in the fourth auction. The fifth auction—for $100 million of a third tranche of 25-year bonds— followed the fourth auction by two weeks and produced results that were distinctly less favorable. The announcement that the Treasury would offer more 25-year bonds caught market participants by surprise—they had been expecting a shorter maturity—and led to sharply lower prices in secondary market trading. Outstanding 25-year bonds traded down 11–32nds, to 1014, following early news of the forthcoming auction.29 The accelerating pace of new bond issues dampened investor interest—the bond market was described as “tired”—and dealers, still holding significant positions from the fourth auction, were reluctant to bid for more of the same.30 The Treasury received tenders for only $147 million of bonds and ended up selling $98 million at an average price of 10025, almost a quarter of a point below where outstanding bonds closed in secondary market trading on the day before the auction. Morgenthau claimed to be “satisfied” with the results of the fifth auction but The Wall Street Journal suggested the sale “was not a brilliant success.”31 A 1940 Treasury staff report stated that the auction method “became increasingly unpopular [in August 1935], as indicated by the criticism which developed in the market and also by the fact that both the total tenders and the number of tenders received for the last two offerings were sharply lower than for the two immediately preceding.”32 A Failed Auction At the beginning of August the new Treasury auction process appeared to be a success. It banished the chronic over-subscriptions that plagued fixed-price offerings and it produced primary market prices comparable to secondary market prices. By the middle of the month, however, market participants were beginning to complain about the unpredictability of the offerings. Then, in late August, the Treasury pushed the auction experiment a little too hard and produced a failed auction that ended the experiment. 29. “Federal Bonds Dip in Active Treading,” New York Times, August 9, 1935, p. 26, and “Bond Sales on the New York Stock Exchange,” New York Times, August 9, 1935, p. 26. 30. “Federal Bonds Dip in Active Trading,” New York Times, August 9, 1935, p. 26, and “Virtually All Bonds Decline: Governments Off on Larger Volume,” Wall Street Journal, August 9, 1935, p. 6. 31. “Bids Show Decline on Federal Bonds,” New York Times, August 16, 1935, p. 23, and “Treasury Takes $98,465,000 Bid of $100,000,000 Bond Offering,” Wall Street Journal, August 16, 1935, p. 2. 32. Joint Economic Committee (July 1959, p. 1158).

291

Treasury Debt Management during the New Deal

On Monday, August 26, the Treasury announced an auction offering of $100 million of 4-year Federal Farm Mortgage Corporation (FFMC) bonds that were fully and unconditionally guaranteed as to both principal and interest by the United States.33 Deciding where to bid the new offering was complicated by the absence of secondary market trading in any comparable FFMC bonds.34 To make matters worse, the auction closed on Wednesday, August 28, in the middle of the last week before the Labor Day weekend, at a time when many market participants were on vacation.35 The Treasury received tenders for only $85.6 million of the bonds. Officials tried to put a good face on the outcome, but Morgenthau admitted that “it wasn’t so good.”36 The failure of the FFMC auction triggered a cascade of criticism of the new auction process. The New York Times reported that “government bond dealers have frowned for some time upon the Treasury Department’s policy of distributing securities by auction sales” and stated that “the result of the sale definitely reflected the disapproval by the banking world to this method of offering securities.”37 The Wall Street Journal was explicit about one source of dealer dissatisfaction: Bond dealers who ordinarily subscribe heavily to issues priced at par [in subscription offerings] and who obtain a considerable profit through quick turnovers did not like the auction system. Compelled to bid at the market, they were unable to get the usual profit provided by the Treasury in issuing securities at par with coupon rates slightly above market yields.38

The mid-August Treasury bond auction was the last auction offering of coupon-bearing Treasury securities for more than a quarter of a century. Morgenthau was initially reluctant to abandon the auction system for bonds,39 but by late October The New York Times was reporting that he had decided to rely solely on fixed-price bond offerings in the future.40 33. Treasury Circular no. 549, August 26, 1935, reprinted in Federal Reserve Bank of New York Circular no. 1579, August 26, 1935. 34. “Treasury Offers $100,000,000 Bonds,” New York Times, August 26, 1935, p. 23. 35. “The Under-Subscribed Loan,” New York Times, September 1, 1935, p. E8. 36. “Federal Bond Sale Fell Short of Goal,” New York Times, August 30, 1935, p. 1. 37. “Federal Bond Sale Fell Short of Goal,” New York Times, August 30, 1935, p. 1. 38. “Treasury Offers $1,750,000,000 Bonds and Notes,” Wall Street Journal, September 3, 1935, p. 1. 39. See, “Dip in U.S. Bonds Raises Question of Money Trend,” New York Times, August 30, 1935, p. 1 (reporting that Morgenthau felt the auction method “had worked very well so far as the Treasury was concerned, and it would not be abandoned”), and “Treasury Closes Books on Sale of Notes for Cash,” Wall Street Journal, September 4, 1935, p. 1 (stating that “the Treasury made it clear that it had no intention of giving up the auction method”). 40. “Treasury Announces $50,000,000 Bill Issue,” New York Times, October 25, 1935, p. 31 (reporting that “the Treasury intends to drop, for the time being at least, the auction system of selling bonds”).

292

Chapter 19

Why Did the Auction Process Fail for Bonds? Morgenthau’s attempt to reintroduce bond auctions to the Treasury market ended in failure because it tried to fuse two inherently incompatible mechanisms: auctioning long-term securities and raising funds “as needed.” Treasury bond auctions are strikingly different from fixed-price offerings of bonds with yields purposely set above market-clearing levels. Auction market participants have to be far more knowledgeable about the state of investor demand if they are to place bids that are rich enough to be accepted, but not so rich that they end up buying bonds at prices in excess of where the bonds will trade in subsequent secondary market transactions. As a practical matter, the Treasury had to rely on government bond dealers and large banks—the only market participants with the requisite knowledge—to underwrite its auction offerings. A 1940 Treasury staff report acknowledged that auction participation was “sharply restricted” because “many banks and investors outside of the largest centers felt that they were not in a position to gage the market with any degree of accuracy . . ..”41 The report further observed that market participants outside of the money centers who did submit bids “generally paid the highest prices. The largest portion of the new issues awarded above the average price . . . went to bidders outside of New York City, while most of the amounts awarded at or below the average went to banks, brokers, and dealers in New York.” The need to rely on dealers and large banks was not, in itself, fatal to the auction process; the Treasury had been auctioning bills successfully since 1929, primarily through dealers and large banks.42 But bill auctions had an important characteristic not shared by the bond auctions: they came on a regular (weekly) basis and bill dealers could manage their inventories to avoid being caught with large positions when the Treasury came to market (as happened in the fifth bond auction). In addition bills were inherently less risky than bonds (because they had shorter maturities), so dealers could be more tolerant of unexpected Treasury decisions with respect to bill sales. Despite the sensitivity of auction market participants to surprises concerning bond 41. Joint Economic Committee (July 1959, p. 1157). Even before the first auction, the Wall Street Journal reported that “In asking the market directly to set the issue price on the securities, [Treasury] officials expect banks to be the largest buyers of the bonds.” “Treasury to Sell $100,000,000 Bonds to Highest Bidders,” Wall Street Journal, May 27, 1935, p. 1. This was a clear indication that officials recognized that auction participation was going to be far narrower than participation in subscription offerings. 42. W. Randolph Burgess, “Notes on the Mechanism of the Market for Treasury Bills,” typescript dated July 16, 1937, Federal Reserve Bank of New York, Federal Reserve Bank of New York file 413.7 (“The principal initial market for Treasury bills is found in the large New York City banks and the government security dealers. Banks in other principal centers have been bidders from time to time, as have also business corporations and foreign banks of issue, but over a period more than 85 per cent of the bills have been originally issued in New York, and of this amount a large proportion has been tendered for by the principal banks and dealers.”).

293

Treasury Debt Management during the New Deal

offerings, Treasury officials exercised substantial discretion in deciding when to sell bonds. The 1940 Treasury staff report observed that, “After the first issues, the market became somewhat nervous over the extent to which the tender method was to be employed. Due to uncertainty as to the time, size, and frequency of such offerings, they had the same effect on the market as if a known seller was waiting to dispose of a very substantial block of bonds at any time.”43 Morgenthau erred when he tried to combine (1) discretionary offerings to raise money “as needed” with (2) auction sales of bonds. He could have raised money on a discretionary basis with bill auctions (and, as noted below, did so soon after he abandoned bond auctions), and he could have auctioned bonds on a regular and predictable basis (as was ultimately done by Paul Volcker in the first half of the 1970s44), but he couldn’t raise money on a discretionary basis with bond auctions. A Return to Tax Date Offerings of Notes and Bonds In April 1936 Secretary Morgenthau announced that the Treasury would no longer issue notes and bonds to raise funds “as needed,” but rather would confine sales of coupon-bearing securities to the quarterly tax dates. “That makes the situation more orderly,” he said. “It gives investors a certainty as to when issues will be made.”45 Figure 19.5 shows the maturities of note and bond offerings from 1935 to mid-1939. Except for a pause following a rise in bond yields in mid-1937 that resulted from an increase in Federal Reserve reserve requirements,46 the Treasury persisted in its efforts to lengthen the maturity of the debt by selling intermediate- and long-term bonds.47 Tax Date Bills Morgenthau’s decision to abandon bond auctions as a source of “as needed” financing reopened the question of how to raise money between quarterly tax 43. Joint Economic Committee (July 1959, p. 1157). 44. Garbade (May 2004, 2007). 45. “Treasury Expands Bonus Borrowing,” New York Times, April 28, 1936, p. 1. 46. Meltzer (2003, pp. 490–534) discusses the increase in reserve requirements. 47. See, for example, “U.S. Debt Spread Over Longer Term,” New York Times, September 21, 1936, p. 31 (Morgenthau announcement that the average maturity of Treasury debt had increased from seven years and four months in March 1933 to nine years and eight months in August 1936), and “Treasury Refundings of $14,000,000,000 Due Next Five Years,” Wall Street Journal, September 21, 1936, p. 1.

Chapter 19

30

25

20

Years

294

15

10

5

0 1935

1936

1937

1938

1939

1940

Figure 19.5 Term to maturity of new offerings of coupon-bearing Treasury debt (through mid-1939). Auction offerings shown with a cross (×), fixed-price cash offering for settlement on other than a tax date shown with a plus (+). All other offerings are fixed-price tax date offerings, either for cash or in exchange for outstanding securities.

dates. He resolved the problem even before it was clear that he had abandoned bond auctions. On Thursday, September 26, 1935, the Treasury announced that it would auction $50 million of a 166-day Treasury bill maturing on Monday, March 16, 1936. Auction tenders were due at or before 2 p.m. on Monday, September 30, 1935, and the bills would be issued two days later, on October 2. The Treasury had, of course, been auctioning bills since 1929. The distinctive feature of the new bill was its maturity date: the issue date of the (as yet unannounced) March 1936 tax date offering. Over the next eight weeks the Treasury offered an additional $400 million of bills, $50 million per week, all maturing on March 16, 1936. The bills was refinanced in the March tax date offering when—in the largest cash sale of securities since the Liberty Loan campaigns—the Treasury sold $629 million of 5-year notes and $727 million of 15-year bonds.48 48. In the March 1936 tax date financing, Treasury officials also made an exchange offering of the 5-year note and the 15-year bond to holders of $559 million of a maturing note. Note holders took $48 million of the 5-year note, $496 million of the 15-year bond, and redeemed the remaining $15 million of the maturing note. Officials contemplated giving holders of the maturing bills a similar option to exchange their bills, but decided in the end to do a cash redemption and a cash refinancing of the bills. “Treasury States New Issues’ Terms,” New York Times, March 2, 1936, p. 1 (“The large amount of cash sought was due only to the fact that the [Treasury] department decided against a policy of permitting holders of the maturing block of Treasury bills to exchange them . . ..”).

Treasury Debt Management during the New Deal

273

182

Days

295

91

0 Jul 1935

Jan 1936

Jul 1936

Jan 1937

Jul 1937

Jan 1938

Jul 1938

Figure 19.6 Term to maturity of new offerings of tax date Treasury bills

The October and November 1935 program to raise money by issuing bills maturing in mid-March 1936 was a direct application of Ogden Mills’s original, 1929, vision of how bills could be used to enhance the efficiency of Treasury cash management. In April 1936 Morgenthau announced a second series of tax date bills: $50 million per week, to mature on December 15, saying that “this method of financing gives us the elasticity we need” in funding expenditures between the tax dates.49 Between the fall of 1935 and the spring of 1938, the Treasury issued eight series of tax date bills. As shown in figure 19.6 and table 19.2, bill maturities varied from 71 days to 223 days and the aggregate amount of a series varied from $250 million to $450 million. After the second series the Treasury staggered the maturity dates of the bills to days after taxes were due. For example, $100 million of the March 1937 tax date bills matured on Tuesday, March 16, $100 million matured on March 17, and $100 million matured on March 18. This facilitated bill redemptions out of tax receipts at times when the Treasury chose not to offer any notes or bonds for cash.50 49. “Treasury Expands Bonus Borrowing,” New York Times, April 28, 1936, p. 1. 50. In the March 1937 tax date financing the Treasury offered only to exchange a 16¾-year bond for a maturing note and did not offer any securities for cash. Spreading out the bill maturities over days when tax payments were actually received eliminated the need for the Treasury to borrow from the Federal Reserve on a day-to-day basis to fund the redemption of maturing debt and facilitated compliance with Section 206 of the Banking Act of August 23, 1935, which provided, for the first time, that the Federal Reserve could not buy Treasury securities directly from the Treasury.

296

Chapter 19

Table 19.2 Tax date Treasury bills Series maturing March 1936 • Issued between October 2 and November 27, 1935, maturing March 16, 1936 • Nine issues, $50 million each, total of $450 million Series maturing December 1936 • Issued between May 6 and June 24, 1936, maturing December 15, 1936 • Eight issues, $50 million each, total of $400 million Series maturing March 1937 • Issued between December 2, 1936, and January 6, 1937, maturing March 16, 17, and 18, 1937 • Six issues, $50 million each, total of $300 million Series maturing June 1937 • Issued between March 3 and April 7, 1937, maturing June 16, 17, and 18, 1937 • Six issues, $50 million each, total of $300 million Series maturing September 1937 • Issued between April 21 and June 2, 1937, maturing September 16, 17, and 18, 1937 • Seven issues, $50 million each, total of $350 million Series maturing December 1937 • Issued between July 14 and September 8, 1937, maturing December 16, 17, 18, 20, and 21, 1937 • Nine issues, $50 million each, total of $450 million Series maturing March 1938 • Issued between October 27 and December 15, 1937, maturing March 16, 17, 18, and 19, 1938 • Eight issues, $50 million each, total of $400 million Series maturing June 1938 • Issued between March 2 and March 30, 1937, maturing June 16, 17, and 18, 1936 • Five issues, $50 million each, total of $250 million Source: Treasury Annual Reports.

By the end of 1937 the Treasury was receiving a significant flow of funds at times other than quarterly tax dates from social security wage taxes and proceeds from sales of savings bonds.51 The value of raising money more or less continuously in relatively small amounts in the bill market declined and the Treasury did not issue tax date bills after the last series matured in June 1938. Regular Bills Even while it was experimenting with bond auctions and tax date bills as sources of “as needed” financing, the Treasury continued to sell bills on a regular weekly basis as part of its debt management operations. 51. See, for example, “Wall Street Studies Treasury Moves,” New York Times, September 12, 1937, p. F1 (commenting on “the steady inflow of funds from the Social Security taxes”), and “Treasury to Sell Additional Bills,” New York Times, February 4, 1938, p. 27 (noting that “The amount of the national debt in the hands of the public is expected to decline gradually . . ., principally because of the investment of social security tax money in government obligations . . ..”). See also, “U.S. Not to Borrow More ‘New Money,’” New York Times, March 1, 1938, p. 29, and “Wall Street Ponders March Financing,” New York Times, February 19, 1939, p. 55 (commenting on the contribution of savings bonds to keeping down sales of marketable securities).

Treasury Debt Management during the New Deal

2.5

2.0

Billions of dollars

297

1.5

1.0

0.5

0 1933

1934

1935

1936

1937

1938

1939

1940

Figure 19.7 Treasury bills outstanding (exclusive of tax date bills)

Figure 19.7 shows the variation in the quantity of Treasury bills outstanding during the New Deal, exclusive of tax date bills. As was noted earlier in this chapter, the Treasury expanded its bill financings in the spring of 1933 (during the initial recovery from the bank holiday), in the winter of 1933–34 (when the RFC gold purchase program disrupted the bond market), and again in the fall of 1934 (during a third episode of rising bond yields). However, it engineered the expansions in three different ways. The spring 1933 expansion was accomplished by increasing the number of “cycle weeks” in which the Treasury issued 13-week bills. By the last quarter of 1932 the department was issuing bills in eight out of every thirteen cycle weeks (see figure 18.2). As shown in figure 19.8, the Treasury added four more cycle weeks in the spring of 1933. The additional cycle weeks allowed the Treasury to raise $335 million in new money.52 By mid-1933 the only week the Treasury wasn’t selling 13-week bills was the 11th cycle week, which fell in the middle of the third month of a quarter, when quarterly tax payments were due and the Treasury regularly issued notes and bonds. The winter 1933–34 expansion, when bills outstanding rose from less than $1 billion in early December to $1.4 billion in late February, was accomplished by increasing the quantity of bills issued each week. In some weeks the Treasury simply issued more 13-week bills than it redeemed, but it other weeks the Treasury issued both 13- and 26-week bills. 52. $75 million of new money was raised with bills issued on March 6 (cycle week 10), $100 million with bills issued March 22 (cycle week 12), $100 million with bills issued April 5 (cycle week 1), and $60 million with bills issued May 3 (cycle week 5).

298

Chapter 19

13 12 11 10

Cycle index

9 8 7 6 5 4 3 2 1 0 Jan 1933

Apr 1933 None

New

Jul 1933 Increase

Oct 1933 No change

Jan 1934

Decrease

Figure 19.8 Weekly sales of 13-week Treasury bills. Cycle index 1 denotes the week beginning Monday, January 2, 1933, and every thirteenth week thereafter. Cycle index 2 denotes the week beginning Monday, January 9, 1933, and every thirteenth week thereafter. Subsequent cycles are defined analogously.

In June 1934, the Treasury stopped issuing 13-week bills and issued only 26-week bills (see figure 19.9). From mid-1934 to early 1935, the Treasury regularly sold $75 million of 26-week bills every week and total bills outstanding grew to about $2 billion ($75 million per week for 26 weeks = $1.95 billion). In early 1935, the Treasury decided to extend bill maturities further, to 39-weeks, while cutting weekly issue sizes to $50 million and maintaining total outstandings at the $2 billion level.53 (The Treasury also issued eight 19-week bills between May 22 and July 10, 1935, inclusive, to fill in maturities missed during the transition to 39-week bills.) The program of funding $2 billion of Treasury debt with weekly issues of $50 million of 39-week bills continued until the fall of 1937. Reinstating 13-Week Bills In the summer and fall of 1937 Treasury and Federal Reserve officials undertook a fundamental reexamination of the Treasury’s short-term financing program. 53. Simmons (1947, p. 335) remarks that “The episode of long-dated Treasury bills has never been adequately explained in official statements. It may represent nothing more than experimentation on the part of the Treasury.”

Treasury Debt Management during the New Deal

52

39

Weeks

299

26

13

0 1934

1935

1936

1937

1938

1939

Figure 19.9 Term to maturity of new offerings of regular Treasury bills

As shown in figure 19.10, average auction discount rates moved up sharply early in 1937, primarily as a result of increases in Federal Reserve reserve requirements. Federal Reserve officials became increasingly concerned with a decline in the liquidity of the bill market and conjectured that 39-week bills might be “too long” to be attractive to many market participants seeking a liquid investment.54 W. Randolph Burgess, manager of the Federal Reserve’s System Open Market Account (and a vice president of the Federal Reserve Bank of New York), observed that bill dealers like Discount Corporation and Salomon Brothers were prepared to “bid freely for substantial amounts of [3- or 6-month bills], but they would deal cautiously for limited amounts of nine months bills. They said that at all times the demand for the nine months bills is spotty.”55 54. Memo dated June 14, 1937, from W. Randolph Burgess, vice president, Federal Reserve Bank of New York, to George Harrison, president, Federal Reserve Bank of New York, Federal Reserve Bank of New York file 413.7 (stating that “a nine months’ bill is too long”), and memo dated June 18, 1937, from Burgess to Harrison, Allan Sproul, first vice president, Federal Reserve Bank of New York, and Walter Matteson, assistant vice president, Federal Reserve Bank of New York, Federal Reserve Bank of New York file 413.7 (stating that “the nine months bill is not a proper market instrument but both ninety day bills and six months bills conform reasonably to market usage and requirements”). See also, Miller (1938, p. 91) and Larkin (1951, p. 3, fn. 2). 55. Memo dated June 18, 1937, from W. Randolph Burgess, vice president, Federal Reserve Bank of New York, to George Harrison, president, Federal Reserve Bank of New York, Allan Sproul, first vice president, Federal Reserve Bank of New York, and Edward Matteson, assistant vice president, Federal Reserve Bank of New York, Federal Reserve Bank of New York file 413.7. See also, W. Randolph Burgess, “Notes on the Mechanism of the Market for Treasury Bills,” typescript dated July 16, 1937, Federal Reserve Bank of New York, Federal Reserve Bank of New York file 413.7 (noting that “the nine months bills particularly have not found as ready a market after their issuance, as have the shorter bills.”).

Chapter 19

1.00

0.75

Percent

300

0.50

0.25

0 1933

1934

1935 13-week bills

1936

1937

26-week bills

1938

1939

1940

39-week bills

Figure 19.10 Average Treasury bill auction discount rates, 1933 to mid-1939. Auctions on March 3, 1933, at 4.26 percent, March 20, 1933, at 1.83 percent, March 27, 1933, at 1.72 percent, and April 3, 1933, at 1.35 percent, not shown.

Concurrently Treasury officials expressed concern that participation in bill auctions was limited to the large New York banks and government securities dealers and that small banks had “steadily refused to bid on Treasury bills.”56 Burgess suggested that there were two reasons for the concentration.57 First, bill yields were so low that only money center banks (who used bills to adjust their reserve positions from day to day and who consequently placed a high value on liquidity) found them attractive.58 Second, . . . the mechanism of bidding is difficult for those outside of the money market. The rates at which bills are sold change from time to time with considerable rapidity, and only those who are

56. “Treasury Studies Means of Tapping Idle Bank Funds,” Wall Street Journal, September 14, 1937, p. 1. 57. W. Randolph Burgess, “Notes on the Mechanism of the Market for Treasury Bills,” typescript dated July 16, 1937, Federal Reserve Bank of New York, Federal Reserve Bank of New York file 413.7. 58. E.A. Goldenweiser, Director of the Division of Research and Statistics at the Board of Governors of the Federal Reserve System, observed that “country banks generally adjust their reserve positions by drawing on their balances with city correspondents, and the ultimate adjustment has to be made by city banks. It is for this reason that city banks hold relatively large volumes of short-term liquid assets, such as bankers’ acceptances, call loans, and Treasury bills.” E. A. Goldenweiser, “Buying Rates for Treasury Bills,” mimeograph dated July 17, 1937, Federal Reserve Bank of New York file 413.7A.

301

Treasury Debt Management during the New Deal

in close touch with changes in money conditions feel able to estimate these trends. Out-of-town banks which have tried bidding for bills report that their bids have been unsuccessful or that they have bid too high and have paid more for the bills than more well-informed bidders.59

In mid-September 1937 Secretary Morgenthau announced a program aimed at identifying whether “the kind of bills we are selling are the best for the country.”60 The Wall Street Journal reported that the Treasury was considering offering “weekly certificate issues carrying fixed interest rates rather than weekly bill offerings” in an attempt to attract greater participation by small banks in the primary market for short-term Treasury securities.61 The debate within the Treasury and the Federal Reserve continued through November, with a growing number of officials concluding that, on balance, bills provided an inexpensive source of funds and were useful to market participants.62 The question was finally settled in early December when Morgenthau announced that the department would begin issuing 13-week bills on a regular weekly basis at the end of the month. Bills, he said, “are a good kind (of securities) to use. There is a great shortage of that type of paper on the market.”63 By mid-1938 the Treasury was regularly issuing $100 million of 59. The similarity with the concentration of bidders in bond auctions in 1935 at the large banks and dealer firms—see Joint Economic Committee (July 1959, p. 1157)—is striking. 60. “Treasury Studies Means of Tapping Idle Bank Funds,” Wall Street Journal, September 14, 1937, p. 1. 61. “Treasury Studies Means of Tapping Idle Bank Funds,” Wall Street Journal, September 14, 1937, p. 1. The Treasury had stopped issuing certificates of indebtedness in early 1934. The Treasury later cited four reasons for its decision to abandon certificates and to rely strictly on bills, including (1) the lower cost of issuing bills through competitive bidding, (2) the convenience of not having to set fractional interest rates in a market where yields were on the order of basis points rather than percentage points, (3) the liquidity of the bill market in major financial centers that allowed the Treasury to bring new issues when cash was needed, and (4) the convenience of setting maturities on tax bills to dates when tax payments would be received. 1940 Treasury Annual Report, pp. 58–59. 62. “Treasury Sentiment Crystallizes Against Shift in Financing Methods,” Wall Street Journal, October 6, 1937, p. 1, and “Morgenthau May Explain Policy Today; He Talks to Reserve’s Open Market Body,” New York Times, October 7, 1937, p. 41 (“Recently Mr. Morgenthau indicated that one of the questions to be discussed with the [Federal Reserve’s] Open Market Committee was the advisability of replacing the present 273-day discount bill issues by another security with a fixed interest rate and a longer maturity. At that time the rate at which the Treasury was marketing the discount bills had stiffened and there were predictions in some financial circles that it might rise as high as 1 per cent. Since then, however, . . . the Treasury has sold [them] at less than one-half of 1 per cent. The belief here, therefore, is fairly general that the suggestion of substituting another type of security for the bills will be dropped . . ..”). In early November the Investment Bankers Association, a trade association of broker-dealer firms, argued that bills had become increasingly important to market participants during 1937 and urged that bill issuance be continued. “Treasuries Firm as Market Waits Quarter Financing,” Wall Street Journal, November 15, 1937, p. 6. See also, chapter 9 (“Commercial Banks and Treasury Bills”) in Miller (1938). 63. “Treasury Loans to be Refunding,” New York Times, December 3, 1937, p. 35, and “Morgenthau Announcement,” Wall Street Journal, December 3, 1937, p. 8.

302

Chapter 19

13-week bills every week (figure 19.9) and the quantity of outstanding bills stabilized at $1.3 billion in early September (figure 19.7).64 Concluding Remarks Perhaps the most surprising aspect of Treasury debt management at the end of the 1930s is the absence of any significant evolution after early 1933. One might have reasonably expected some substantial innovation over the course of a six-year period during which, for the first time, the Federal government ran chronic deficits in the absence of war. Nevertheless, by mid-1939 the Treasury was issuing notes and bonds on quarterly tax dates (a Mellon innovation of the 1920s) and 13-week bills weekly (the structure Mellon and Mills were moving toward in 1931 and 1932). However, Morgenthau did accomplish the two debt management objectives that were most clearly evident at the start of the New Deal: lengthening the maturity of the debt and refinancing the last two Liberty Loans. Additionally he recognized the importance of introducing a more flexible policy at the interface between Treasury cash management and Treasury debt management, and he developed a bond auction initiative to address that need. When the initiative failed (on account of the attempt to fuse two incompatible objectives), he fell back on Mills’s original program of issuing tax date bills. Tax date bills might have been a lasting legacy of New Deal cash management, but were ultimately rendered superfluous by the introduction of other sources of relatively continuous funding. In mid-1939 the two most significant shortcomings of Treasury debt management were (1) the continued reliance on fixed-price sales of notes and bonds and (2) the failure to offer bonds with predictable maturities. With respect to the latter shortcoming, the Treasury probably regressed during the 1930s. Although officials offered bonds on a regular schedule, they aggressively varied the maturities of their offerings in light of market conditions, shortening up when bond yields rose in the winter of 1933–34 and in mid-1937 and lengthening when yields declined. 64. The reduced issuance size was consistent with a recommendation of W. Randolph Burgess that the Treasury bill market should be “kept small enough so the Treasury is not too dependent on this narrow market.” Memo dated June 5, 1937, “Ways and Means of Facilitating a Better Distribution of Treasury Bills throughout the Country,” from W. Randolph Burgess, vice president, Federal Reserve Bank of New York, to George Harrison, president, Federal Reserve Bank of New York, Federal Reserve Bank of New York file 413.7.

20

The Primary Market during the Great Depression

Large, sometimes even massive, oversubscriptions for fixed-price offerings of coupon-bearing Treasury securities (certificates, notes, and bonds) were commonplace in the 1930s. Only five out of sixty-four cash offerings between 1930 and 1938 generated subscriptions smaller than twice what was offered, only fifteen generated subscriptions smaller than three times the amount offered, and—as shown in figure 20.1—on thirteen occasions subscriptions were an order of magnitude larger than what was offered. The average “cover ratio” of debt subscribed to debt offered was 6.9. Oversubscribed offerings were more than mere curiosities. They were indicative of an inefficient primary market (because they were a direct consequence of underpricing by Treasury officials) and they were risky: an unexpected decline in secondary market prices during an offering could leave a new issue largely in the hands of investors who had expected to receive much smaller allotments and who might be forced to sell at distress prices. This chapter describes how Treasury officials tried to contain oversubscriptions. Oversubscriptions during the Great Contraction Oversubscriptions for fixed-price offerings of Treasury securities first captured public attention in the March 1931 tax date financing, when the Treasury offered $300 million of 6-month certificates of indebtedness, $600 million of 1-year certificates, and $500 million of 12-year, 33⁄8 percent bonds. Officials had planned to offer $1 billion of 12-year bonds (with a coupon rate of 3¼ percent) but halved the offer (and bumped the coupon rate) as a result of investor disquiet with the recently passed veterans’ bonus legislation.1 The Treasury received subscriptions for a modest $400 million of the shorter certificate, a more significant $1.2 billion of the longer certificate, and—contrary to expectations—a remarkable $2.1 billion of the bond. The New York Times reported that “the big oversubscription for the . . . bonds indicated to some experts that . . . an interest rate ¼ per cent less might have been justified.”2 The bonds rose a point and a half above the original offering price when they began trading on Monday, March 16. Three months later, in the June tax date financing, the Treasury offered $800 million of an 18-year, 31⁄8 percent bond and received tenders for $6.3 billion 1. “Congress Concurs on Bond Refunding,” New York Times, March 3, 1931, p. 4, and “Bonus Deferred U.S. Refunding,” Wall Street Journal, March 4, 1931, p. 1. See also “Treasury Issue Well Received,” Wall Street Journal, March 3, 1931, p. 20 (reporting that “had there been no disturbance in the market as the result of bonus agitation, the Treasury would have been able to place a much larger proportion of the present offering in the form of a bond issue, and at possibly lower interest”). 2. “Mellon Gets Bids of $3,734,000,000,” New York Times, March 6, 1931, p. 2.

Chapter 20

35 30 25

Cover ratio

304

20 15 10 5 0 1930

1931

1932

1933

1934

1935

1936

1937

1938

1939

Figure 20.1 Cover ratios (ratio of amount subscribed to amount offered) for fixed-price cash subscription offerings of Treasury securities

of the bonds; the largest oversubscription up to that time. The New York Times reported that the subscriptions were heavily padded: “banks, in order to meet their own investment requirements and those of customers, subscribed to considerably larger amounts of bonds than they actually desired, as it is the custom of the treasury to allot only part of the larger subscriptions.”3 Padding was risky for both individual investors and the market as a whole because there was no upper limit on the subscription that an investor might submit. An investor who wanted to buy $1 million of a new note and who anticipated a 20 percent allocation might subscribe for $5 million of the securities. However, if the investor believed that other investors were also enthusiastic about the offering and thought the Treasury might only allot 10 percent of a tender, she might subscribe for ten times what was wanted. This created the risk that some subscribers might be allotted many more bonds than they were prepared to hold if demand proved unexpectedly weak.4 3. “Bond Subscriptions Total $6,000,000,000,” New York Times, June 5, 1931, p. 1. 4. “Free riding” was a related practice of subscribing for a new issue with the intention of selling promptly whatever was awarded. Since the Treasury offered new issues with above-market coupon rates at par, free riding was usually quite profitable. See “Treasury to Depend More on Auctions for Bond Financing,” Wall Street Journal, July 12, 1935, p. 1 (“In the past, the Treasury has sold bonds at par and has offered interest coupons sufficiently above market yields to attract buyers. Purchasers of government bonds under this system obtain an immediate profit since new issues sold at par almost invariably go to premiums in the market.”), and see also Porter (1938).

305

The Primary Market during the Great Depression

Criticism of the apparently generous terms of the June 1931 financing may have made Treasury officials atypically aggressive in pricing an offering of $800 million of 24-year bonds three months later.5 They put a 3 percent coupon on the issue, even though many market participants thought a 31⁄8 percent coupon would have been more in line with market conditions, and barely avoided a failed offering. Even after keeping the subscription books open longer than usual and applying “official pressure” to large banks, the Treasury received tenders for only $940 million of the bonds.6 Investor interest in long-term Treasury offerings evaporated in the wake of Britain’s departure from the gold standard in late September 1931. Responding to the shift in demand, the Treasury offered only short-term debt in its December tax date financing: $300 million of a 6-month certificate, $400 million of a 9-month certificate, and $600 million of a 1-year note. It received subscriptions for a total of $1.8 billion of the securities, making for an unusually low cover ratio of about 1.4. A January 1932 offering of $350 million of 6-month and 1-year certificates elicited subscriptions for only $646 million of the securities. Demand for Treasury securities began to improve in the spring of 1932, when the Federal Reserve initiated a more expansionary monetary policy. Treasury officials responded by offering intermediate-term notes as well as certificates, and demand for the notes exploded. An April offering of $225 million of 2-year notes generated subscriptions for $2.5 billion of the securities (a cover ratio of 11.1), a July offering of $325 million of 4-year notes led to subscriptions for $3.8 billion (a cover ratio of 11.7), and offerings of 4½-year notes in October and 4-year notes in December produced cover 5. “Eager Market Seen for Treasury Issue,” New York Times, September 1, 1931, p. 7, and “National Debt Burden Shifted to the Future,” New York Times, September 6, 1931, p. 101 (“There was some criticism because of [the] tremendous oversubscriptions [for the 18-year bonds offered in June] and this may have influenced the treasury experts in quoting . . . 3 per cent on the $800,000,000 issue offered this month.”). 6. “Treasury Now Offering $300,000,000 in 3% Bonds, $500,000,000 in 21⁄8% Notes,” New York Times, June 4, 1934, p. 1 (remarking that “The September, 1931, 3 per cent issue . . . was floated after the books on subscriptions had been kept open for some time longer than had been customary”), and “The Under-Subscribed Loan,” New York Times, September 1, 1935, p. E8 (noting that “voluntary subscriptions did not cover the full amount [of the September 1931 bond offering], and official pressure had to be applied to the larger banks to make up the deficiency”). See also “Eager Market Seen for Treasury Issue,” New York Times, September 1, 1931, p. 7 (“Perhaps the most significant point in the reception of the new financing announcement was that no market was established for the bonds on a ‘when issued’ basis. Last June, when the $800,000,000 31⁄8s were offered, a market was quickly established, and on the very day of the announcement the issue was quoted at a substantial premium, and as a result the issue was oversubscribed nearly eight times.”).

306

Chapter 20

ratios in excess of 18. Padding was repeatedly identified as the source of the oversubscriptions.7 The oversubscription phenomenon peaked in January 1933 when the Treasury offered $250 million of 5-year notes and received subscriptions for $7.8 billion of the securities—a cover ratio of 31.2. Treasury Secretary Mills explained that “This great volume of subscriptions is due in large measure to the fact that many subscribers are deliberately applying for amounts far in excess of their requirements, anticipating that under the Treasury’s percentage allotment they will receive a reduced amount approximating their actual needs.”8 Treasury officials appreciated the gravity of the problem.9 The New York Times identified three separate sources of official concern: (1) oversubscriptions gave “a completely erroneous impression of the borrowing capacity of the government,” (2) oversubscriptions made it “difficult for the Treasury to gauge the ability of the market to take up future issues,” and (3) oversubscriptions opened up the possibility of “serious financial difficulties for the subscribers themselves.”10 The Times provided an illustration of the third concern: Early in November [1932] the British Treasury offered . . . £300,000,000 of 3 per cent bonds. Previous issues . . . had been greatly oversubscribed and allotments had been drastically scaled down. In the case of the 3 per cent loan, however, the low coupon, coupled with the large amount offered, failed to bring as great an oversubscription as had earlier offerings. Much to their dismay, some subscribers received allotments of 70 per cent on their subscriptions . . .. Unable to take up so large an amount, they were forced hurriedly to sell the excess, with the result that the market for the new issues went at once to a discount of 2½ per cent, large losses resulted, and the gilt-edge bond market suffered a case of “indigestion” from which it did not recover for a full month.11

7. “New Treasury Issue ‘Oversold’ 10 Times,” New York Times, April 28, 1932, p. 32 (“many financial interests seeking to obtain a part of the unexpectedly small total of the two issues had padded their subscriptions several times to get the amount they wished”), “$450,000,000 Offering Oversubscribed in Day; Treasury to Announce Allocations Tuesday,” New York Times, October 7, 1932, p. 29 (“banks always place subscriptions of several times their requirements”), “Treasury Issue 18 Times Oversubscribed; $8,368,000,000 Is Offered for New Notes,” New York Times, October 11, 1932, p. 31 (“banks pad their subscriptions in order to be assured of obtaining their requirements”), and “Minimize Demand for Treasury Paper,” New York Times, December 14, 1932, p. 35 (“banks ‘padded’ their subscriptions, bidding for many more of the notes and certificates than they desired or expected to be awarded”). 8. “$7,550,000,000 Oversubscription on Notes Brings Treasury Warning on Padding Bids,” New York Times, January 27, 1933, p. 23. 9. “$7,550,000,000 Oversubscription on Notes Brings Treasury Warning on Padding Bids,” New York Times, January 27, 1933, p. 23 (“padding has steadily increased until it has now reached such proportions that the department must consider measures to deal with it in the interests of other subscribers and the Treasury”). See also, “Hit Padded Offers for Federal Loans,” New York Times, January 29, 1933, p. N7 (“The Treasury has been aware of [the] drawbacks to the padding of subscriptions . . . but it has not yet found a way to prevent the practice.”). 10. “Hit Padded Offers for Federal Loans,” New York Times, January 29, 1933, p. N7. 11. “Hit Padded Offers for Federal Loans,” New York Times, January 29, 1933, p. N7.

307

The Primary Market during the Great Depression

Preferential Awards to Subscribers Tendering Maturing Securities The dangers of padding were exacerbated by the policy introduced by the Treasury in December 1922 of usually (but not always) giving preferential allotments to subscribers who indicated that they would pay with maturing securities instead of with cash. The Treasury had initiated the policy to reduce turnover in War Loan Deposit Account balances when it was simultaneously paying off maturing securities and issuing new securities. Prior to the onset of the Depression, preferential allotments had little impact on primary market activity. However, as oversubscriptions began to expand, and as Treasury allotments fell, investors realized that tendering maturing securities in lieu of cash would eliminate uncertainty about what they might receive. In twenty-five cash offerings between December 1929 and February 1933 in which the Treasury gave preferred allotments to subscribers tendering maturing securities, subscribers tendering maturing securities accounted for an average of 44 percent of the amount issued.12 In eight of the twenty-five offerings they accounted for more than half of the amount issued. And in three cases—the bond offering in March 1931, an offering of 1-year certificates in September 1931, and an offering of 6-month certificates in December 1931— they accounted for the entire issue; subscribers proposing to pay cash were completely shut out. The Treasury policy of giving preferential allotments to subscribers tendering maturing securities increased the difficulty of subscribing with the intent of paying cash. An investor not only had to judge the intensity of interest from other investors who would pay cash, he or she also had to judge the intensity of interest from investors who could move to the front of the line by tendering maturing securities. Oversubscriptions during the New Deal Treasury officials in the incoming Roosevelt administration were acutely aware of the problems with fixed-price subscription offerings. After struggling through a March 1933 tax date financing that was announced only four days before it settled and that settled only two days after the largest American banks reopened, Secretary Woodin announced a late-April offering of $500 million of 3-year notes. Subscribers (other than banks and trust companies and recognized dealers in Treasury securities) were, for the first time, required to make a down payment of 10 percent of the amount subscribed.13 The cover ratio on 12. In contrast, subscribers tendering maturing securities in seven cash offerings between March 1928 and September 1929 accounted for an average of only 20 percent of the amount issued. 13. Treasury Circular no. 482, April 24, 1933, reprinted in 1933 Treasury Annual Report, p. 165.

308

Chapter 20

the notes was a modest 2.4. Treasury officials thereafter continued to require down payments, sometimes reducing the amount required to 5 percent. Cover ratios ranged between 2.7 and 6.9, and averaged 4.5, over the next twelve months. In January 1934, Treasury officials ended the practice of giving preferential allotments to investors tendering maturing issues and shortly thereafter began to conduct joint but separate cash and exchange offerings. In June 1934, the Treasury offered to sell $300 million of 14-year bonds for cash and separately and simultaneously offered to exchange more of the same bonds for a maturing certificate of indebtedness and for a note due to mature in August. The Treasury issued $172 million of bonds in exchange for certificates ($3 million of the certificates were redeemed for cash) and $317 million in exchange for notes ($28 million of the notes were redeemed for cash).14 The new program of joint cash and exchange offerings allowed the Treasury to continue to benefit from rolling over outstanding debt without churning War Loan Deposit Account balances while simultaneously simplifying the process of raising new money (because a cash subscriber did not have to judge the intensity of demand for a new issue from investors who could move to the front of the line). However, neither the requirement of a downpayment nor the introduction of joint cash and exchange offerings addressed the fundamental problem: the Treasury was chronically underpricing its new issues. The June 1934 cash offering of $300 million of 14-year bonds produced $2.5 billion of subscriptions and a cover ratio of 8.4. The New York Times characterized the strong response as “surprising.”15 The next cash subscription offerings came in December 1934, when the Treasury offered $450 million of 1½-year notes and $450 million of 18-year bonds. Officials acted aggressively to limit oversubscriptions, asking money center banks to cut down their tenders so that a “fair” allotment could be achieved.16 The New York Times described officials as believing that “a wider distribution of the securities throughout the country, which would be obtained on a rationing basis if offerings from the large financial centres were reduced, was desirable.”17 In spite of their efforts, the Treasury received 14. 1934 Treasury Annual Report, pp. 177–78. 15. “Bids of 7 Billions on Treasury Issues,” New York Times, June 7, 1934, p. 1. 16. “$5,400,000,000 Bid in Treasury Offers,” New York Times, December 5, 1934, p. 35 (reporting that “banks have been asked to revise and cut down their bids, so that a fair allotment may be achieved”), and “Treasury Allots Bonds and Notes,” New York Times, December 13, 1934, p. 38 (reporting that “cash subscriptions, particularly from the New York district and other financial centres, would have run much higher but for the fact that the Treasury turned them back for revision when it became evident that many of the banks and other financial institutions were padding their bids heavily in the hope of obtaining a larger share of the total allotments”). 17. “Treasury Allots Bonds and Notes,” New York Times, December 13, 1934, p. 38.

309

The Primary Market during the Great Depression

subscriptions for $3.0 billion of the notes (cover ratio of 6.7) and $2.3 billion of the bonds (cover ratio of 5.2). The New York Times reported that officials were “amazed.”18 Treasury officials next tried to impose direct quantitative limits on bank subscriptions and substantially increased the down payment required on small subscriptions. In a July 1935 offering of $500 million of 4½-year notes, the Treasury provided that banks could not subscribe for notes in excess of onehalf of their capital and surplus.19 Subscribers (other than banks and trust companies) for amounts less than $5,000 had to make a down payment equal to the full amount of their subscriptions; a non–bank subscriber for a larger amount had to make a downpayment of 5 percent of the amount of its subscription, but not less than $5,000. Despite the new restrictions, the cover ratio on the July offering was 5.9. The oversubscription problem would have been eliminated by the bond auctions that Secretary Morgenthau introduced in June 1935, but officials abandoned the auction system after a weak auction of 25-year Treasury bonds in mid-August and a failed auction offering of Federal Farm Mortgage Corporation bonds at the end of August.20 After terminating the auction experiment, the Treasury returned to direct limits on bank subscriptions and down payments for others. In December 1935 the Treasury offered $450 million of 5-year notes and got subscriptions for $2.5 billion (cover ratio of 5.5), and it offered $450 million of 11¾-year bonds and got subscriptions for $2.0 billion (cover ratio of 4.5). Morgenthau claimed that oversubscriptions would have been greater in the absence of the Treasury’s restrictions: “If there hadn’t been that limitation, this thing would have gone I don’t know where. The sky would have been the limit on subscriptions.”21 Oversubscriptions began to edge up again in 1936. In the March 1936 tax date financing, the Treasury offered $600 million of 5-year notes and got subscriptions for $3.4 billion (cover ratio of 5.6), and it offered $650 million of 15-year bonds and got subscriptions for $5.1 billion (cover ratio of 7.9). In the June financing, Morgenthau deputized the Federal Reserve Banks, acting in their role as fiscal agents for the United States, to “examine applications for cash offerings of interest-bearing securities . . . and to report to the Secretary of the Treasury any which appear to be excessive, with recommendation as to the reduction or rejection of any such 18. 19. 20. 21.

“$5,400,000,000 Bid in Treasury Offers,” New York Times, December 5, 1934, p. 35. Treasury Circular no. 545, July 8, reprinted in 1936 Treasury Annual Report, p. 215. See chapter 19. “Treasury Loan Bids 5 Times $900,000,000,” New York Times, December 4, 1935, p. 8.

310

Chapter 20

applications . . ..”22 Morgenthau claimed that the purpose of the new screening process was to “provide for an equitable allotment and distribution of securities to all classes of subscribers . . ..”23 Nevertheless, oversubscriptions remained elevated. Officials offered $400 million of a 5-year note and got subscriptions for $2.8 billion of the notes (cover ratio of 6.9), and they offered $600 million of an 18-year bond and got subscriptions for $4.3 billion of the bonds (cover ratio of 7.1). The New York Times reported that “The large oversubscription . . . was regarded as . . . noteworthy in view of the fact that all precautions were taken to prevent the padding of subscriptions and to avoid the chances of speculation in government securities.”24 Padding and oversubscriptions continued unabated through the end of the 1930s. The cover ratio averaged 11.0 for ten cash offerings between September 1936 and December 1938.25 The failed attempt to introduce auction sales of notes and bonds in 1935 had left the Treasury with no effective means of mitigating the ineluctable consequences of their reluctance to price cash offerings closer to secondary market prices. Exchange Offerings The oversubscription phenomena were rooted in chronically underpriced cash offerings. The most efficient solution was to set offering prices closer to secondary market prices, where demand would more nearly equal supply. However, the best route to that end—auctions—had been tried and judged a failure. Within the narrower confines of fixed-price offerings, closer pricing ran a greater risk of a failed offering and Treasury officials had had enough close brushes with failed offerings to make them leery of that strategy.26 A less efficient solution was to limit demand for underpriced securities by some sort of non–price rationing. Morgenthau tried placing direct limits on 22. Letter from Henry Morgenthau, Secretary of the Treasury, to the presidents of the Federal Reserve Banks, May 27, 1936, reprinted in 1936 Treasury Annual Report, pp. 264–65. See also “Treasury Hopes to Halt Pyramiding On Its Offerings,” New York Times, May 29, 1936, p. 2 (reporting that five large Treasury dealers in New York had agreed with the Federal Reserve Bank of New York that they would not buy “allotment rights” to a cash subscription offering until the subscription books for the offerings had been closed, and that the Reserve Bank had been indicating “to [New York] Stock Exchange houses amounts for which they might ‘reasonably’ bid”). 23. Letter from Henry Morgenthau, Secretary of the Treasury, to the presidents of the Federal Reserve Banks, May 27, 1936, reprinted in 1936 Treasury Annual Report, pp. 264–65. 24. “Treasury Loan Oversold in Day,” New York Times, June 2, 1936, p. 1. 25. There were no cash subscription offerings in the first half of 1939. 26. Including, most recently, the bond offering in September 1931, but including also the offerings identified in box 14.1.

The Primary Market during the Great Depression

10

8

Billions of dollars

311

6

4

2

0 1934

1935

1936

1937

1938

1939

Fiscal year Exchange (for other securities)

Exchange (for Liberty loans)

Cash

Figure 20.2 Notes and bonds sold in fixed-price cash offerings and in exchange offerings

bank subscriptions and increasing (with down payments) the economic cost of non–bank subscriptions, but the attempt proved futile. A different approach to the same end was to limit subscriptions to investors who could pay in maturing securities. Figure 20.2 shows the volume of notes and bonds sold in fixed-price cash offerings and in exchange offerings for maturing securities between fiscal year 1934 and fiscal 1939. Exchange offerings are further separated into exchange offerings for Liberty bonds and exchange offerings for other securities. Secretary Mellon used exchange offers to refinance the Victory Liberty notes and the Second and Third Liberty Loans in the 1920s, but otherwise relied on cash offerings (albeit with preferential allotments to subscribers tendering maturing securities in exchange for the new debt). In contrast, Morgenthau made frequent use of exchange offers, not only in refinancing the First and Fourth Liberty Loans but also in refinancing other securities. (He made greater use of cash offerings than exchange offerings in fiscal years 1936 and 1937, when he had to raise money for the bonus payments authorized by the Adjusted Compensation Payment Act of January 27, 1936,27 and for a Treasury gold 27. See chapter 17.

312

Chapter 20

sterilization program.28 After fiscal year 1937, he increasingly relied on sales of savings bonds to individual investors and special issues to the Social Security trust funds to raise new money.29) Exchange offerings had an advantage over cash offerings in terms of limiting padding and oversubscriptions: as long as the Treasury needed to refinance all of a maturing issue, there was no need to allocate the supply of the new issue. Officials could fill all subscriptions, secure in the knowledge that they would be called upon to issue no more than what was maturing. In essence, exchange offerings became a device to sell securities in fixed-price offerings against payment in a “special currency”: maturing securities. Exchange offers obfuscated the consequences of, but did not eliminate, underpricing. The excess demand that would have been manifest in oversubscriptions had an offering been for cash was channeled into greater secondary market demand for maturing securities. It was common in the 1930s for maturing Treasury securities to trade at a price, inclusive of accrued interest, that exceeded the coupon and principal payment promised at maturity, i.e., to trade at a negative yield.30 The excess value represented the value of the right to exchange the maturing security for a new, underpriced, security. 28. The gold sterilization program was a program to finance Treasury purchases of gold bullion with the proceeds from sales of Treasury securities (instead of with central bank balances, as had been the case since early 1934—see the appendix to chapter 16). The sterilization program was intended to slow or stop the growth of reserves in the banking system. The program started in December 1936, continued aggressively through July 1937, and continued on a less active basis through March 1938. Morgenthau terminated the program in April 1938. See “Treasury Steps In to Stem Gold Flow into Credit Banks,” New York Times, December 22, 1936, p. 1, “Bankers Here Urged Step,” New York Times, December 22, 1936, p. 19, “Treasury Adopts Plan to ‘Sterilize; Gold Acquisitions,” Wall Street Journal, December 22, 1936, p. 1, “$1,392,065,461 Gold Freed for Spending,” New York Times, April 15, 1938, p. 1, “Treasury Gives Up Sterilizing Policy,” New York Times, April 19, 1938, p. 31, Chandler (1971, pp. 313–14 and 329–31), and Meltzer (2003, pp. 504–507 and 531–33). 29. See chapter 17. 30. Cecchetti (1988). See also, “New Federal Loan Sells at Premium,” New York Times, August 1, 1933, p. 23 (noting that a certificate of indebtedness exchangeable for new bonds was trading at a negative yield), “Sept. 15 Date to Present 4th 4¼s,” Wall Street Journal, April 16, 1934, p. 13 (noting that the second called Fourth Liberty bonds were trading at a negative yield to their October 15, 1934, redemption date, “presumably . . . on the basis of an exchange offer in due course which may be profitable”), and “March Financing of U.S. Treasury in Spotlight Now,” Wall Street Journal, February 23, 1937, p. 8 (pointing out that short-term notes which were “expected to get rights to subscribe to . . . new bonds” were trading at a negative yield).

21

Statutory Control of Treasury Indebtedness

Whether issuing securities to finance a current deficit or to refinance maturing debt, Treasury officials have to comply with whatever restrictions Congress chooses to impose. There were three statutory restrictions on Treasury debt management actions at the beginning of the Great Depression:1 1. a limit of $10 billion on outstanding bills and certificates of indebtedness,2 2. a limit of $7.5 billion on outstanding notes,3 and 3. authority to issue no more than $20 billion of Treasury bonds.4 In contrast, by mid-1939 there was only a single statutory limit of $45 billion on total outstanding indebtedness. This chapter examines why Congress gradually moved away from controlling individual categories of Treasury debt to controlling aggregate indebtedness.5 Statutory Controls before the Great Depression In the decades prior to the First World War, Congress treated long-term debt differently from short-term debt. The War Revenue Act of 1898, for example, authorized the Secretary of the Treasury to issue up to $400 million of bonds “to defray expenditures authorized on account of” the Spanish-American War and also authorized the Secretary to issue certificates of indebtedness, subject to a limit of $100 million on outstanding certificates.6 Thus, while Congress limited the amount of bonds that could be issued, it more loosely restricted the amount of short-term debt that could be outstanding. The difference arose because Congress thought of bonds as instruments for financing specific projects—waging a war or constructing a canal—and expected that the indebtedness represented by the bonds would be paid down, rather than refinanced, at maturity. In contrast, short-term debt greased the day-to-day operations of government and was likely to expand and contract as a function of the ebb and flow of receipts and expenditures. During World War I, Congress continued to distinguish between limiting the quantity of bonds that could be issued and limiting the amount of shortterm debt that could be outstanding. By the end of the war the Treasury was 1. Congress also stipulated (in the Third Liberty Bond Act) that the interest rate on a Treasury bond could not exceed 4¼ percent per annum. The limitation affected Treasury debt management only once during the Great Depression, in the March 1932 tax date financing. See chapter 18. 2. Act of June 17, 1929. 3. Revenue Act of 1921. 4. Fourth Liberty Bond Act. This limit applied to cumulative bond sales following the First Liberty Loan in 1917. The First Liberty Loan was authorized separately, by the First Liberty Bond Act. 5. See also Cooke and Katzen (1954). 6. War Revenue Act of 1898.

314

Chapter 21

authorized to issue up to $20 billion of bonds and to have up to $10 billion of certificates of indebtedness outstanding.7 Also during World War I, Congress authorized the Treasury to issue up to $7 billion of notes.8 By mid-1919, the Treasury had issued $14.9 billion of bonds out of the $20 billion authorized, of which $14.3 billion remained outstanding, and had $3.4 billion of marketable certificates and $3.5 billion of marketable notes outstanding.9 Statutory limits on Treasury indebtedness changed in two respects during the 1920s. First, to facilitate the refinancing of short-term debt into longer term securities, Congress changed (in 1921) the restriction on notes to a limit of $7.5 billion on outstanding notes.10 The change suggests that Congress came to view notes as an instrument of debt management rather than as an instrument for project finance. Second, in authorizing the Treasury to issue bills in 1929, Congress limited (to $10 billion) the total quantity of bills and certificates outstanding,11 thereby allowing officials to choose whether to sell bills or certificates in a given operation. The provision allowed the Treasury to exercise some limited (but nevertheless important) discretion in the type of instrument sold, but it did not expand the Treasury’s discretion over the maturity structure of the debt (because both bills and certificates were limited to a maturity not in excess of one year). By mid-1929, the Treasury had issued $18.1 billion of bonds out of the $20 billion authorized, of which $9.4 billion remained outstanding, and had $1.6 billion of marketable certificates and $2.3 billion of marketable notes outstanding.12 (The Treasury did not issue any bills until December 1929.) Broadening the Scope of Managerial Discretion Statutory limits on Treasury indebtedness evolved more rapidly during the Great Depression. The first post-1929 request for expanded authority came in the 1930 Treasury Annual Report, when Secretary Mellon requested authority to issue up to $28 billion of bonds, $8 billion more than the existing authority. Mellon was interested in preparing for the early refinancing of $1.6 billion of notes and $6.3 billion of Fourth Liberty bonds. Although his request was nar7. Fourth Liberty Bond Act and Victory Liberty Loan Act. 8. Victory Liberty Loan Act. 9. 1919 Treasury Annual Report, pp. 186, 221, and 260. There was also $0.9 billion of pre-war bonds, $2.0 billion of First Liberty bonds, and $1.1 billion of other debt (of which $0.95 billion was war savings stamps and certificates) outstanding. 10. Revenue Act of 1921. 11. Act of June 17, 1929. 12. 1929 Treasury Annual Report, p. 453. There was also $0.8 billion of pre-war bonds, $1.9 billion of First Liberty bonds, and $0.6 billion of other debt outstanding.

315

Statutory Control of Treasury Indebtedness

rowly phrased in terms of raising the ceiling on bond issuance, Mellon left no doubt that what he really wanted was the ability to choose what maturity debt to issue. He had ample room to issue bills, certificates of indebtedness, and notes, and wanted comparable room to issue bonds. “[I]t is obvious,” he wrote, “that the orderly and economical management of the public debt requires that the Treasury Department should have complete freedom in determining the character of securities to be issued and should not be confronted with any arbitrary limitation which was not intended to apply to these circumstances.”13 Congress acceded to Mellon’s request in March 193114 and the Treasury sold $2.2 billion of bonds in three offerings before Britain’s abandonment of the gold standard in September 1931 led to a hiatus in bond sales that lasted until mid-1933. Secretary Morgenthau’s first request for expanded authority to issue Treasury debt was similarly modest: an early 1934 request to raise the limit on outstanding notes from $7.5 billion to $10 billion. Congress acceded to his request,15 so that by early 1934 the limits on Treasury indebtedness had expanded to (1) a limit of $10 billion on total outstanding bills and certificates, (2) a limit of $10 billion on outstanding notes, and (3) authority to issue not more than $28 billion of bonds. In the middle of 1934 the Treasury had $2.9 billion of certificates and bills outstanding, $6.7 billion of notes, and had issued $24.4 billion of bonds out of the $28 billion authorized, of which $13.7 billion remained outstanding.16 Morgenthau’s second foray into relaxing statutory constraints on Treasury debt management was more aggressive. In early 1935 he proposed replacing the $28 billion limit on bond issuance with a $25 billion limit on outstanding bonds, saying that “This is the more satisfactory manner of fixing the maximum amount to be borrowed upon bonds and follows the laws authorizing the other classes of securities since 1917.”17 The proposed action would remove the last vestige of the pre-war notion that the Treasury sold bonds to finance specific projects. Bonds, like bills, certificates, and notes, would become a vehicle for financing deficits (regardless of how those deficits arose) and for refinancing maturing debt. At the time the Treasury was only $2.5 billion under the $28 billion issuance ceiling, but only $13.5 billion of the bonds remained outstanding, so the department would be able to issue an additional $11.5 billion of new bonds.18 13. 14. 15. 16. 17. 18.

1930 Treasury Annual Report, p. 39. Emphasis added. Act of March 3, 1931. Act of January 30, 1934. 1934 Treasury Annual Report, p. 337. Committee on Finance (January 29, 1935, p. 3). Committee on Finance (January 21, 1935, p. 2).

316

Chapter 21

Morgenthau also requested in early 1935 that the $10 billion limit on total outstanding bills and certificates be combined with the $10 billion limit on outstanding notes into a single $20 billion limit on total outstanding bills, certificates, and notes. This action would materially broaden the scope for discretion in Treasury debt management because it would allow the Treasury to refinance short-term bills with intermediate-term notes, and vice versa. Testifying before the Senate Finance Committee, Morgenthau expressed his belief that “the Government may be saved substantial amounts and may work toward lengthening the maturities of a larger portion of its public debt if more flexibility is granted in the issuance of the notes, certificates, and bills.”19 Both the House Ways and Means Committee and the Senate Finance Committee characterized Morgenthau’s requests as intended “to permit more flexible and economical Government financing.”20 Congress acceded to both requests,21 so that by the spring of 1935 there was a $20 billion limit on total outstanding bills, certificates, and notes, and a separate $25 billion limit on outstanding bonds. The last steps in the evolution of statutory control of Treasury indebtedness came in the late 1930s. In 1938, Congress raised the limit on outstanding bonds from $25 billion to $30 billion (at a time when there was $23.3 billion of bonds outstanding22) and also imposed a limit of $45 billion on total outstanding bills, certificates, notes, and bonds in lieu of the existing limit of $20 billion on total outstanding bills, certificates, and notes.23 The new limits did not allow more than the $45 billion of total indebtedness allowed previously, but did introduce some fungibility of bonds with shorter term debt. The only reason the four classes of securities were not fully fungible was the $30 billion subceiling on outstanding bonds. In requesting the new limits, Morgenthau expressed his view that restrictions on different categories of Treasury debt should be relaxed “to give the Treasury more latitude as to the kind of securities it can issue.”24 Finally, in mid-1939, Congress removed the $30 billion subceiling on outstanding bonds,25 leaving only the $45 billion limit on total outstanding debt. 19. Committee on Finance (January 29, 1935, p. 4). 20. Committee on Ways and Means (1935, p. 1) and Committee on Finance (January 21, 1935, p. 1). 21. Act of February 4, 1935. 22. “Would Lift Limits on U.S. Bond Total,” New York Times, May 7, 1938, p. 21. 23. Act of May 26, 1938. 24. “Treasury Plans Sale of Bonds Up to Billion,” Wall Street Journal, May 7, 1938, p. 1. See also, “Would Lift Limits on U.S. Bond Total,” New York Times, May 7, 1938, p. 21. 25. Act of July 20, 1939. See also “Flexibility Asked in National Debt,” New York Times, March 19, 1939, p. 68, and “President Urges Ending of Limit on Bonded Debt,” New York Times, March 21, 1939, p. 1.

317

Statutory Control of Treasury Indebtedness

The House Ways and Means Committee noted Morgenthau’s assertions that the “limitation of the face amount of bonds which may be outstanding at any one time may seriously interfere with the efficient and economical financing of Government requirements” and that removing the subceiling on outstanding bonds was “essential to enable the Treasury to continue its efforts to finance the needs of the Government in the most economical manner possible.”26 The committee concluded that removing the subceiling would “permit the Secretary of the Treasury to issue securities best suited at the time to meet the conditions of the market and the needs of the Government . . ..”27 The 1939 legislation left Treasury officials free to decide the maturities of new issues. In particular, they no longer had to return to Congress for expanded authority to issue particular types of debt—as Mellon had had to do in the winter of 1930–31 and as Morgenthau had had to do in early 1934, in early 1935, and again in the spring of 1938. However, the Treasury was not entirely free of congressional oversight, because Congress continued to limit total indebtedness. During the coming decades Treasury officials would return repeatedly to Congress seeking increases in the debt ceiling whenever that ceiling threatened to become a binding constraint.28 26. Committee on Ways and Means (1939, pp. 1 and 2). See similarly, Committee on Finance (1939). 27. Committee on Ways and Means (1939, p. 2). 28. Since the Treasury can not expend funds without congressional authorization, some commentators have questioned why Congress persists in limiting total indebtedness. Krishnakumar (2005, p. 137) suggests that the debt limit “encourages legislators to consider the interests of the general public and future generations, rather than those of special interest groups, and thus acts as an important check on party and interest group politics.” More particularly, it provides “an independent and focused opportunity for Congress to step back and consider the consequences of its deficit-spending decisions, to evaluate its fiscal policies, and even to implement fiscal reform if it decides that is has been borrowing too much too fast.” Krishnakumar concludes (p. 138) that debt limits have acted “as a catalyst for budget-reform and budget-balance measures aimed at reducing national borrowing.” See also Robinson (1959).

22

The Brief Revival and Subsequent Extinction of National Bank Notes

From the opening of the Federal Reserve in the fall of 1914 to the beginning of the New Deal in 1933, the United States used (simultaneously) five different types of currency: gold certificates, silver certificates, United States notes, Federal Reserve notes, and national bank notes. The first three currencies were direct emissions of the federal government. Federal Reserve notes were, for all practical purposes, direct emissions as well, but national bank notes were private liabilities. In adopting the Federal Reserve Act in 1913, Congress intended that Federal Reserve notes would, over time, replace national bank notes. Section 18 of the act provided a specific scheme for replacement. In 1916 and early 1917 the Federal Reserve and the Treasury, acting pursuant to that section, retired almost $56 million of bonds securing outstanding national bank notes. The redemption effort ended when the United States entered the First World War. This chapter traces the national bank note program in the 1920s and describes the dramatic, but brief, expansion of the program in 1932 and its final extinction in 1935. National Bank Notes between World War I and the Great Depression In mid-1919 there was $719 million of national bank notes in circulation. $683 million of the notes was secured by four issues of Treasury bonds. (The four bonds, shown in table 22.1, were the only bonds eligible to be used as collateral for national bank notes, meaning they were the only bonds that bore the “currency privilege.”) The remaining $36 million of national bank notes had been defeased and was secured with “lawful money”—Treasury currency or Federal Reserve notes.1 Figure 22.1 shows the subsequent variation in national bank notes. The falloff in notes secured by Treasury bonds between 1924 and 1925 followed the decision of Treasury officials to call the 4 percent bonds redeemable in 1925 for redemption on February 2, 1925.2 At the time of the call (on October 15, 1924), $119 million of the 4 percent bonds remained outstanding, of which $78 million was pledged against outstanding national bank notes. Prior to redemption, $16 million of the pledged bonds were replaced with 2 percent Consols and 2 percent Panama Canal bonds. When the 4 percent bonds were redeemed, the $62 million that remained with the Treasury was replaced with the redemption proceeds and the proceeds retained as collateral against the subsequent redemption of a matching quantity of national bank notes.3 1. 1922 Comptroller of the Currency Annual Report, p. 188. 2. The call notice appears in Treasury Circular no. 346, October 15, 1924, reprinted in 1924 Treasury Annual Report, p. 218. 3. 1925 Treasury Annual Report, p. 35. See also Treasury Circular no. 346, October 15, 1924, reprinted in 1924 Treasury Annual Report, p. 218.

Chapter 22

Table 22.1 Treasury bonds pledged against national bank notes, June 30, 1919 ($millions) Collateral

Amount pledged

4% Bonds redeemable in 1925 no maturity date redeemable on or after February 1, 1925 2% Consols redeemable in 1930 no maturity date redeemable on or after April 1, 1930 2% Panama Canal bonds of 1936 due August 1, 1936 redeemable on or after August 1, 1916 2% Panama Canal bonds of 1938 due November 1, 1938 redeemable on or after November 1, 1918 Total

Amount outstanding

55.7

118.5

563.9

599.7

47.6

49.0

25.1

25.9

692.3

793.1

Source: 1919 Treasury Annual Report, pp. 186 and 717.

1,000

Millions of dollars

750

500

250

Secured by lawful money

35 19

33 19

31 19

29 19

27 19

25 19

23 19

21 19

19

0 19

320

Secured by Treasury bonds

Figure 22.1 National bank notes outstanding, mid-1919 to mid-1936. Notes secured by lawful money are notes defeased by deposits of lawful money pursuant to section 4 of the Act of June 20, 1874.

321

Brief Revival and Subsequent Extinction of National Bank Notes

Table 22.2 Treasury bonds pledged against national bank notes, June 30, 1925 ($millions) Collateral

Amount pledged

Amount outstanding

2% Consols redeemable in 1930 no maturity date redeemable on or after April 1, 1930 2% Panama Canal bonds of 1936 due August 1, 1936 redeemable on or after August 1, 1916 2% Panama Canal bonds of 1938 due November 1, 1938 redeemable on or after November 1, 1918 Total

591.2

599.7

48.1

49.0

25.8

25.9

665.1

674.6

Source: 1925 Treasury Annual Report, pp. 35 and 528.

Following the redemption of the 4 percent bonds redeemable in 1925, a total of $675 million of Treasury bonds bore the circulation privilege. On June 30, 1925, $665 million (98.6 percent) of those bonds were pledged against outstanding national bank notes (table 22.2). The amount of pledged bonds hardly varied during the balance of the decade and the stock of bank notes secured by Treasury bonds remained virtually stationary at about $660 million. National Bank Notes during the Great Depression Figure 22.1 shows two significant changes in national bank notes after mid1932: bank notes secured by Treasury bonds increased sharply between mid1932 and mid-1933, but then declined to insignificance by mid-1936. The figure raises two questions: How could the quantity of bank notes secured by Treasury bonds increase (between mid-1932 and mid-1933) if virtually all of the bonds bearing the currency privilege were already pledged? And why did the stock of national bank notes secured by Treasury bonds decline so dramatically in the middle of the decade? New Life for National Bank Notes during the Great Contraction Acute and chronic deflation was a central feature of the Great Contraction. The consumer price index fell 24 percent between 1929 and 1933.4 Loans that called for repayment in nominal dollars became increasingly burdensome in 4. The consumer price index (based on a value of 100 in 1967) fell from 51.3 in 1929 to 38.8 in 1933 U.S. Department of Commerce (1975, p. 211).

322

Chapter 22

real terms and borrowers defaulted in increasing numbers as the contraction and deflation wore on. Monetary historians attribute much of the deflation to a failure of the Federal Reserve to prevent an unprecedented contraction in the money supply, from $26 billion in mid-1929 to $19 billion four years later.5 Milton Friedman and Anna Schwartz, for example, wrote that Prevention or moderation of the decline in the stock of money, let alone the substitution of monetary expansion, would have reduced the contraction’s severity and almost as certainly its duration. The contraction might still have been relatively severe. But it is hardly conceivable that money income could have declined by over one-half and prices by over one-third in the course of four years if there had been no decline in the stock of money.6

The contraction in the money supply was, in part, a result of a shift in the composition of the demand for money. In the aftermath of two waves of regional bank failures, one in the winter of 1930–31 triggered by the failure of a Tennessee investment bank and a second between April and August 1931 centered on Chicago and Toledo, and after the larger national tsunami of failures that followed the September 1931 decision by Great Britain to abandon the gold standard, Americans increasingly took to hoarding cash.7 Currency in circulation increased from $3.6 billion in mid-1929 (14 percent of the money supply) to $4.7 billion at the end of 1932 (23 percent of the money supply).8 The Fed’s mistake, in the face of the shift in the composition of the demand for money, was its failure to expand the monetary base—the sum of (1) Reserve Bank deposit liabilities to member banks and (2) currency outside the Treasury and Federal Reserve Banks—at a sufficiently rapid pace. Between June 1929 and December 1932, the Fed allowed the base to expand by only $1 billion, from $6.8 billion to $7.8 billion.9 Currency outside the Treasury and the Fed expanded by $930 million and excess reserves, the bankers’ equivalent of cash in a piggy bank,10 increased by $480 million over the same period,11 leading to a $360 million contraction in reserves supporting member 5. The money supply data are from Board of Governors of the Federal Reserve System (1943, p. 34), here measured as total demand deposits other than interbank and U.S. government deposits, less cash items in process of collection, plus currency outside banks. 6. Friedman and Schwartz (1963, p. 301). See also Meltzer (2003, p. 271) (“It is now generally accepted that the depth of the [Great Depression], its duration, and its spread through the world economy are mainly the result of monetary actions or inactions.”). 7. Friedman and Schwartz (1963) and Wicker (1966, 1996). 8. Board of Governors of the Federal Reserve System (1943, p. 34). 9. Total reserves: Board of Governors of the Federal Reserve System (1943, p. 396), total currency in circulation: Board of Governors of the Federal Reserve System (1943, pp. 411–12). 10. See, Meltzer (2003, p. 332) (“The rise in currency and in excess reserves are related. Both reflect the increased number of bank failures during the period.”). 11. Board of Governors of the Federal Reserve System (1943, p. 396).

323

Brief Revival and Subsequent Extinction of National Bank Notes

bank deposit liabilities. The attention of the monetary authorities was focused elsewhere, on nominal interest rates and member bank borrowings at the discount window.12 On the basis of those indicators, monetary policy was appropriately accommodative.13 Americans were well acquainted with the thesis that monetary expansion can cause inflation—the gold standard was widely accepted as a device to limit the prospect of inflation—and by 1932 some commentators had concluded that appropriately controlled increases in the money supply could reverse the deflation of the preceding three years. In the absence of leadership from the Federal Reserve, the argument for reflating the economy was left to politicians like Senator Arthur Capper (Republican of Kansas) who, in a radio address in January 1932, suggested that “a moderate inflation” was necessary.”14 Capper argued that “we need some more dollars in circulation. More dollars in circulation mean cheaper money; that means higher values for commodities, for labor, for evidence of wealth. More dollars in circulation will give more jobs, will enable debtors to pay interest on their debts, will enable taxpayers to pay their taxes without forfeiting their property. Congress’s next job is to insure an expansion that will be the equivalent of a moderate inflation of the currency.” In the same month, Senator David Walsh (Democrat of Massachusetts) suggested that the United States should issue $1 billion of new bonds bearing the currency privilege so that national banks could expand their circulation.15 The Hoover administration refused to expand the currency. Responding to Senator Walsh, Treasury Secretary Mellon said he could see no need for an increase in national bank notes: “If the country were confronted with a currency shortage, or if the established provision for currency supply were deemed inadequate, it might be urged with very good reason that, as an emergency measure, provision be made for increasing the national bank circulation. I do not find the conditions now existing would warrant such action.”16 (Administration officials were careful to describe the proposed Reconstruction Finance Corporation as a debt-financed vehicle aimed at 12. Meltzer (2003, p. 282) (“The main reason for the failure of monetary policy in the depression was the reliance on an inappropriate set of beliefs about speculative excesses and real bills. This set of beliefs, embodied in the Riefler-Burgess framework, directed attention to short-term market interest rates and member bank borrowing and encouraged their use as indicators of the magnitude and direction of monetary stimulus.”) 13. Meltzer (2003, p. 280) (“Once borrowing and short-term market rates had fallen below the range familiar to governors and commercial bankers, policy was ‘easy.’”). 14. “Hoover’s Program to Liquefy Credit; Not Cash Inflation,” New York Times, January 27, 1932, p. 1. 15. “Treasury Rejects Money Inflation,” New York Times, January 28, 1932, p. 1. 16. “Treasury Rejects Money Inflation,” New York Times, January 28, 1932, p. 1.

324

Chapter 22

liquifying private credit rather than a currency-financed program aimed at reflating the economy.17) The Anti-hoarding Campaign It was, nevertheless, hard to ignore the contractionary consequences of hoarding. On February 6, 1932, President Hoover announced the opening of an anti-hoarding campaign18 aimed at “educating the people to the necessity of sustaining the credit of their respective communities, and to convince them that banks are essential to their welfare.”19 However, the impracticality of an educational campaign became obvious as soon as the campaign director was asked whether he would “ask a man who has lost money in a bank failure to redeposit his money in some other bank.” The director replied that he was “not prepared at this time to answer that question.”20 Anti-hoarding campaign officials focused instead on a novel issue of 1-year, 2 percent Treasury certificates of indebtedness. To make the certificates attractive to people hoarding cash, the Treasury provided that they could be redeemed, at the option of a holder, on 60 days’ notice, and offered them in a minimum denomination of $50.21 Officials claimed that the certificates would satisfy a widespread demand for a “safe and secure substitute for currency now cached in the teapot or safety-deposit box.”22 To accomplish the intended purpose of releasing into circulation currency that had been hoarded, the Treasury provided that banks could pay for customer subscriptions to the new certificates with War Loan Deposit Account credits.23 Secretary of the Treasury Ogden Mills observed that if the certificates 17. “Hoover’s Program to Liquefy Credit; Not Cash Inflation,” New York Times, January 27, 1932, p. 1 (“In Europe inflation generally means the printing of additional currency to pay for government expenses. The United States, it was declared on behalf of the administration, contemplates no such inflation of the currency as part of its program to end the business depression. . . . Instead, the administration, it was explained, proposes to make up the government deficit of more than $200,000,000 for the current fiscal year by the issuance of bonds. . . . [T]he creation of the $2,000,000,000 Reconstruction Finance Corporation, soon to be organized, will be taken to liquidate credit in American financial institutions.”) 18. “Nation-Wide Pledge is Given to Hoover to Battle Hoarding,” New York Times, February 7, 1932, p. 1. See also “Hoover Asks Public to Cease Hoarding and Put Cash to Use,” New York Times, February 4, 1932, p. 1, and “Knox Opens Drive to Stop Hoarding,” New York Times, February 9, 1932, p. 13. 19. “Plan Debentures to Stop Hoarding,” New York Times, February 10, 1932, p. 2. 20. “Plan Debentures to Stop Hoarding,” New York Times, February 10, 1932, p. 2. 21. Treasury Circular no. 456, March 5, 1932, reprinted in 1932 Treasury Annual Report, p. 229. See also “‘Baby Bond’ Issue Agreed on to Halt Hoarding of Cash,” New York Times, February 19, 1932, p. 1, “Anti-Hoarding Bond a $50 Certificate,” New York Times, February 20, 1932, p. 2, “New Features Seen in ‘Baby’ Bond Issue,” New York Times, February 21, 1932, p. N5, and “Arranges for Sale of New Baby Bonds,” New York Times, March 5, 1932, p. 25. 22. “ ‘Baby Bond’ Issue Agreed on to Halt Hoarding of Cash,” New York Times, February 19, 1932, p. 1. 23. “Anti-Hoarding Bond a $50 Certificate,” New York Times, February 20, 1932, p. 2, and “Arranges for Sale of New Baby Bonds,” New York Times, March 5, 1932, p. 25.

325

Brief Revival and Subsequent Extinction of National Bank Notes

were purchased “with currency held outside of banks, the banks receiving the subscriptions will gain the cash deposited by the subscriber, while they may pay for the certificates delivered to the subscriber by means of a deposit credit for the account of the government.”24 Thus the certificates offered banks a way to borrow from the Treasury at the rate of interest on War Loan Deposit Accounts: ½ percent per annum,25 while investors got a credit risk-free certificate paying 2 percent per annum that was redeemable on 60 days’ notice. President Hoover launched the certificate offering with a nationwide radio address on March 7, 1932, in which he appealed for an end to hoarding: I believe that the individual American has not realized the harm he has done when he hoards even a single dollar away from circulation. He has not realized that his dollar compels the bank to withdraw many times that amount of credit from the use of borrowers. . . . Multiply this simple example by nearly a billion and a half of dollars of idle money now hidden in the country and you may get somewhere near a true picture of the enemy of our national security that we vaguely call “hoarding.” It strangles our daily life, increases unemployment and sorely afflicts our farmers.26

Secretary Mills, speaking after Hoover, similarly urged Americans to restore to active use “the dollars that have been withdrawn from circulation,” taking care to make clear that he was speaking of “money in safe-deposit boxes, or in socks, or under mattresses, or in a tin can,” and not funds on deposit in a bank.27 Whether because of the low, 2 percent, interest rate on the certificates, or because of the 60-day notice required for redemption, or because of a lack of a Treasury commitment to leave investor subscriptions in War Loan Deposit Accounts, the anti-hoarding certificates were a bust.28 The Treasury received subscriptions for only about $35 million of the securities.29 There was no follow-on offering and the anti-hoarding campaign disappeared from sight.30 The Federal Reserve Open Market Purchase Program in 1932 A simpler and more direct response to hoarding was for the Federal Reserve to accommodate the expanding demand for currency and excess reserves by 24. “Arranges for Sale of New Baby Bonds,” New York Times, March 5, 1932, p. 25. Mills also noted that “should funds for the purchase of certificates be withdrawn by depositors of the subscribing banks, they will automatically be replaced by a government deposit . . ..” 25. The interest rate on War Loan Deposit Account balances was reduced to ½ percent on June 1, 1931. 1931 Treasury Annual Report, p. 20. Interest on War Loan Deposit Account balances was eliminated on June 15, 1933. 1933 Treasury Annual Report, p. 267. 26. “Appeals by the President and Other National Leaders for an End of Hoarding,” New York Times, March 7, 1932, p. 4. 27. “Appeals by the President and Other National Leaders for an End of Hoarding,” New York Times, March 7, 1932, p. 4. 28. Nadler and Bogen (1933, p. 113) (“The Baby Bonds were another flop in the series of devices to rebuild public confidence by magic tricks.”). 29. 1932 Treasury Annual Report, p. 230. 30. “Sale of Baby Bonds to End Tomorrow,” New York Times, April 12, 1932, p. 31.

326

Chapter 22

purchasing Treasury securities in open market operations. Beginning in late February 1932 and continuing until the summer, that is exactly what the Fed did. From the end of February to the end of July, Reserve Bank holdings of Treasury securities expanded by $1.1 billion, from $740 million to $1.8 billion.31 Alan Meltzer, in his definitive history of the Federal Reserve, identifies three reasons for the Fed’s purchase program:32 1. Passage of the Glass–Steagall Act in February 1932 gave the Federal Reserve Board authority to allow the Reserve Banks to pledge Treasury securities as well as gold and commercial paper against issues of Federal Reserve notes. This relaxed the so-called free gold constraint on the Fed’s notes.33 2. Member bank borrowings from the Reserve Banks were high, over $800 million in January and February 1932,34 a signal of an undesirably tight monetary policy. 3. System officials were anxious to fend off congressional calls for more overtly inflationary policy initiatives.35 The Fed’s purchase program began to slow in June 1932 because some Reserve Bank governors believed the program was ineffective, because others became concerned about protecting their dwindling gold reserves, and because discount window borrowings had fallen back to a level consistent with an “easy” monetary policy.36 The ultimate impact of the program was blunted by several factors, including a $290 million runoff in member bank borrowings, a $65 million reduction in Reserve Bank investments in assets other than Treasury securities, and a $380 million gold outflow from the United States between the end of February and the end of July.37 The program accommodated a $200 million increase in Federal Reserve notes in circulation and a 31. Board of Governors of the Federal Reserve System (1943, p. 376). 32. Meltzer (2003, pp. 359–61). See also Friedman and Schwartz (1963, pp. 384–89). 33. Prior to passage of the Glass–Steagall Act, Federal Reserve notes could only be secured with commercial paper and gold. The dearth of commercial paper during the Great Contraction forced the Reserve Banks to pledge more gold than the minimum 40 percent required. Some officials were afraid that the Banks might not have enough ‘free,” or unpledged, gold to meet a sustained demand for gold from foreign investors, such as had occurred in October 1931 and January 1932 following the departure of Great Britain from the gold standard. Meltzer (2003, pp. 355–357), Friedman and Schwartz (1963, pp. 399–406), and Wicker (1966, pp. 165, 168, and 173). 34. Board of Governors of the Federal Reserve System (1943, p. 397). 35. Meltzer (2003, p. 360) reports that some Congressmen sought immediate payment of World War I bonuses with $2.4 billion of new Federal Reserve Bank notes, secured by Treasury bonds issued to the Reserve Banks. See also, Friedman and Schwartz (1963, p. 322) (“there was growing support [in Congress] for . . . monetary expansion, proposals widely castigated by the business and financial community as ‘greenbackism’ and ‘inflationary’ ”). 36. Meltzer (2003, pp. 363–71). 37. Board of Governors of the Federal Reserve System (1943, p. 376).

327

Brief Revival and Subsequent Extinction of National Bank Notes

$260 million increase in member bank excess reserves (between the end of February and the end of July), but it did not prevent a further decline, by $60 million, in required reserves.38 The Glass–Borah Amendment Following the failure of the anti-hoarding campaign and the waning of the Federal Reserve open market purchase program, Congress turned to more direct methods to reflate the economy. In July 1932 Senator William Borah (Republican of Idaho) proposed an amendment to the Federal Home Loan Bank Act then under consideration in the Senate that would add an entirely new section to the act, quite unrelated to the system of home loan banks that was at the heart of the legislation. Borah’s amendment, a variant of legislation introduced earlier by Senator Carter Glass (Democrat of Virginia), would allow national banks to issue, for three years, national bank notes against a pledge of Treasury bonds with coupon rates not in excess of 33⁄8 percent. The Borah amendment was a scaleddown version of Glass’s proposal, which would have extended the circulation privilege to all Treasury bonds for a period of five years. Borah’s version became known as the Glass–Borah amendment and was variously characterized as a “currency expansion plan” and, more frankly, as a “currency inflation scheme.”39 The amendment was approved by the Senate and remained (as section 29) in the final version of the Federal Home Bank Act of July 22, 1932.40 The Glass–Borah amendment extended the circulation privilege to $3 billion of Treasury bonds that did not previously bear that privilege (table 22.3). However, the volume of national bank notes could not expand to anything like that figure because an individual bank could not issue notes in excess of its paid-in capital.41 In mid-1932 national banks had an aggregate capitalization of $1.6 billion and outstanding notes secured by Treasury bonds of $670 million, so they could expand their note issues by no more than about $900 million.42 The actual expansion in national bank notes secured by Treasury 38. Board of Governors of the Federal Reserve System (1943, pp. 376 and 412). 39. “Borah Plan Draws Fire in the Senate,” New York Times, July 10, 1932, p. 2, and “Senate Approves Home Loan Bill,” New York Times, July 13, 1932, p. 2. Senator George Norris (Republican, Nebraska) argued that the measure would “not inflate to a dangerous degree.” “Home-Loan Measure In an All-Day Snarl,” New York Times, July 17, 1932, p. 1. 40. “Glass Inflation Bill Is Passed by Senate,” New York Times, July 12, 1932, p. 1, and “Dissension in Final Hours: House Recedes on Bank Bill, Accepting Glass Provision,” New York Times, July 17, 1932, p. 1. 41. “Recent Legislation on National Bank Note Circulation,” Federal Reserve Bulletin, August 1932, pp. 478–80. 42. The capitalization figure is from Federal Reserve Bulletin, August 1932, p. 475. The figure on national bank notes secured by Treasury bonds is from the 1932 Comptroller of the Currency Annual Report, p. 37.

328

Chapter 22

Table 22.3 Treasury bonds accorded the circulation privilege by the Glass–Borah Amendment (in $millions) Collateral

Amount outstanding, June 30, 1932

2½% postal savings bonds, various series 3% conversion bonds of 1946 and 1947a 3% Panama Canal bonds of 1961 (issued 1911) 33⁄8% bonds of March 15, 1943 (issued 1931) 33⁄8% bonds of June 15, 1943 (issued 1928) 33⁄8% bonds of June 15, 1947 (issued 1927) 31⁄8% bonds of June 15, 1949 (issued 1931) 3% bonds of September 15, 1955 (issued 1931) Total

36.2 28.9 49.8 544.9 353.0 454.1 821.4 800.4 3,088.8

Source: 1932 Treasury Annual Report, p. 72. a. Thirty-year bonds issued to Federal Reserve Banks in 1916 and 1917 in exchange for bonds acquired from national banks that had been pledged against outstanding national bank notes. See chapter 2.

bonds between mid-1932 and mid-1933 was a far more modest $184 million.43 Table 22.4 shows the composition of pledged collateral in mid-1932 and mid-1933. There is more than a little irony in the fact that almost twenty years after passing the Federal Reserve Act and thinking it had provided the country with a modern, elastic currency, Congress felt compelled to facilitate the expansion of an older currency that had been slated for extinction. The Extinction of National Bank Notes The Glass–Borah amendment extended the currency privilege to Treasury bonds with coupon rates not in excess of 33⁄8 percent for a period of three years that would end on July 22, 1935. By that date, all of the Glass–Borah bonds pledged against issues of national bank notes would have to be replaced with lawful money or with one of the three bonds that bore the currency privilege 43. National bank notes secured by Treasury bonds increased from $670 million in mid-1932 to $854 million in mid-1933. 1933 Comptroller of the Currency Annual Report, p. 46. National banks were less anxious to issue currency under the provisions of the Glass–Borah amendment after the Attorney General of the United States expressed his opinion that a bank would have to defease any notes issued under the amendment after three years. Some had thought that the three year limitation specified in the amendment referred to the deposit of new collateral rather than to the maintenance of outstanding notes. See Opinion of the Attorney General relating to the circulation privilege granted certain United States bonds under section 29 of the Federal home loan bank act of July 22, 1932, August 12, 1932, reprinted in 1932 Treasury Annual Report, p. 316, “Doubts Hold Up Borah Currency,” Wall Street Journal, August 10, 1932, p. 1 (“banks have hesitated to make extensive use of the privilege, pending official interpretations of the provisions”), and “Three-Year Limit for New Bank Notes,” New York Times, August 13, 1932, p. 19 (“Either there will be little use of [the currency privilege provided by the Glass–Borah amendment] or Congress must extend the privilege to a longer period.”).

329

Brief Revival and Subsequent Extinction of National Bank Notes

Table 22.4 Treasury bonds pledged against national bank notes, June 30, 1932, and June 30, 1933 (in $millions) Collateral

Amount pledged, June 30, 1932

Amount pledged, June 30, 1933

2% Consols redeemable in 1930 2% Panama Canal bonds of 1936 2% Panama Canal bonds of 1938 2½% postal savings bonds, various series 3% conversion bonds of 1946 and 1947 3% Panama Canal bonds of 1961 33⁄8% bonds of March 15, 1943 33⁄8% bonds of June 15, 1943 33⁄8% bonds of June 15, 1947 31⁄8% bonds of June 15, 1949 3% bonds of September 15, 1955 Total

596.0 48.8 25.7 — — — — — — — — 670.5

565.9 45.5 22.7 0.0 1.0 0.0 44.8 19.3 27.0 53.1 77.0 856.4

Source: 1932 Comptroller of the Currency Annual Report, p. 38, and 1933 Comptroller of the Currency Annual Report, p. 47.

as a contractual matter: the 2 percent Consols and the 2 percent Panama Canal bonds of 1936 and 1938. In early 1935 Treasury officials concluded that for three reasons, the time was ripe for completing the program of replacing national bank notes with Federal Reserve notes. First, in addition to the imminent expiration of the Glass–Borah amendment, the Consols and the Canal bonds could be redeemed at any time (because call protection on those bonds had lapsed44), so it was feasible to eliminate every bond that could be used to secure national bank notes. Second, the monetary base was expanding rapidly as a result of the gold flows set in motion by the January 1934 reduction of the gold content of the dollar,45 so there was no compelling need to maintain national bank notes as a distinct currency. Third, the Treasury had $642 million of unallocated profits from the 1934 revaluation of gold that it could use to finance the redemptions without contracting the monetary base.46 Mechanically the Treasury could issue gold certificates (on the unallocated profits) to the Federal Reserve in exchange for credits to its general account, and then use those credits to redeem the bonds. 44. The Panama Canal bonds of 1936 had been redeemable since August 1, 1916, the Panama Canal bonds of 1938 had been redeemable since November 1, 1918, and the Consols had been redeemable since April 1, 1930. 45. Discussed in the appendix to chapter 16. 46. When the dollar value of gold was reset from $20.67 per fine ounce to $35.00 per fine ounce in January 1934, the value of the Treasury’s inventory of gold rose by $2.812 billion. $642 million of that increase remained unappropriated in 1935 and available to finance the bond redemptions. “Last of the ‘Free Gold,’ $675,000,000 Consols and Panama Bonds to Be Redeemed,” New York Times, March 11, 1935, p. 1, and 1935 Treasury Annual Report, p. 205.

330

Chapter 22

On March 10, 1935, the Treasury announced that it was calling the Consols for redemption on July 1, 1935, and that it was calling the Panama Canal bonds of 1936 and 1938 for redemption on August 1, 1935.47 $490 million of Consols (out of the $600 million outstanding) and $42 million of Canal bonds (out of the $75 million outstanding) was pledged against national bank notes. An additional $145 million of other Treasury bonds was pledged against national bank notes pursuant to the Glass–Borah amendment.48 Thus, by the beginning of August, the Treasury would have to redeem $675 million of called bonds, and it would have to supervise the defeasance of $677 million of national bank notes secured with Treasury bonds. The fortuitous near-equality of the principal value of the bonds to be redeemed and the face amount of the bank notes to be defeased simplified the Treasury’s job. In broad terms, the Treasury could pay off the $675 million of called bonds with funds credited to its general account following the deposit of gold certificates with the Fed and simultaneously offset the contraction in the monetary base caused by the redemption of $532 million of pledged Consols and Canal bonds and the termination of the currency privilege on $145 million of pledged Glass–Borah bonds. The actual process was a little more complicated. During the spring of 1935 national banks began defeasing their bank notes with deposits of lawful money, withdrawing their previously pledged Glass–Borah bonds as well as, in some cases, Consols and Canal bonds. To offset what would have otherwise been a loss of reserves from the banking system (due to the banks’ deposits of lawful money with the Treasury), the Treasury issued gold certificates to the Federal Reserve and disbursed the balances that were credited to its general account. The Consols and Canal bonds that remained on deposit with the Treasury in July and August were replaced with redemption payments funded with additional Treasury issues of gold certificates. Consols and Canal bonds that were not on deposit with the Treasury and that were not securing any national bank notes were redeemed with Treasury funds called from War Loan Deposit Accounts.49 47. “Last of the ‘Free Gold,’ $675,000,000 Consols and Panama Bonds to Be Redeemed,” New York Times, March 11, 1935, p. 1. The call notices appear in Treasury Circulars no. 533 and 534, April 1, 1935, reprinted in 1935 Treasury Annual Report, pp. 206 and 208, respectively. 48. “Last of the ‘Free Gold,’ $675,000,000 Consols and Panama Bonds to Be Redeemed,” New York Times, March 11, 1935, p.1. 49. “Called Bank Bills Retiring Slowly,” New York Times, May 5, 1935, p. F1, “National Bank Notes Slowly Disappearing,” New York Times, May 22, 1935, p. 29, “Will Use Deposits to Retire Consols,” New York Times, June 25, 1935, p. 31, “Treasury Redemption July 1 Will Not Boost Banks’ Excess Reserves,” Wall Street Journal, June 26, 1935, p. 7, “Treasury Calls $352,869,100 Funds,” New York Times, June 28, 1935, p. 31, “Treasury Retiring $599,724,000 Bonds,” New York Times, July 2, 1935, p. 7, and “Gold Certificates at Record in Banks,” New York Times, July 6, 1935, p. 22.

331

Brief Revival and Subsequent Extinction of National Bank Notes

Following the expiration of the Glass–Borah amendment and the redemption of the Consols and Canal bonds, hundreds of millions of dollars of national bank notes remained in circulation, but all of those notes were secured with lawful money. National bank notes that came into the possession of Treasury officials in the normal course of business were replaced with the currencies that remained in use: Federal Reserve notes, United States notes, and silver certificates. Concluding Remarks The expiration of the Glass–Borah amendment and the redemption of the Consols and the Panama Canal bonds of 1936 and 1938 marked the end of a seventy-year national bank note program originally intended to enhance the demand for U.S. Treasury bonds. The end of the program had little effect on the market for Treasury bonds because that market had long since grown beyond the point where it could be supported by public demand for currency. Perhaps the most instructive aspect of the last years of national bank notes is that even as they were on the verge of passing into history, Congress reached out to them as an instrument of monetary policy. In an emergency the most unlikely tools may be put to use if they are readily at hand.

23

Coda on Treasury Debt Management during the Great Depression

As a long-run proposition, the most important advance in Treasury debt management during the Great Depression was the introduction of regular and predictable auctions of 13-week Treasury bills. Regular weekly auctions weren’t what Ogden Mills had in mind when he and J. Herbert Case worked together to introduce Treasury bills to the American money markets in 1929, and it certainly wasn’t a straight line from the sporadic bill auctions of 1930 to the regular weekly auctions of 13-week bills in 1938 and 1939, but the program nevertheless satisfied a need for a short-term, highly liquid security and contributed to low-cost Treasury financing. Of equal importance, the regular and predictable bill auctions that emerged during the Great Depression would serve as a model for the regular and predictable note and bond auctions that came to characterize Treasury debt management in the last quarter of the twentieth century. Treasury officials did not accomplish nearly as much with notes and bonds. Andrew Mellon and Ogden Mills had little opportunity to think about the merits of extending the bill model to notes and bonds during the turmoil of the Great Contraction, when they needed to fund loans to veterans in the first half of 1931, get the Treasury through the tsunami of bank failures that followed the abandonment of the gold standard by Great Britain in September 1931, and fund the newly authorized Reconstruction Finance Corporation in 1932. William Woodin and Henry Morgenthau hardly had it any easier in 1933 and early 1934, funding the Treasury in the aftermath of the bank holiday and then during the market volatility that preceded the devaluation of the dollar in January 1934. By 1935 Secretary Morgenthau was in a position to develop a Treasury finance program adapted to financing chronic budget deficits but missed an opportunity when he tried to raise funds “as needed” with bond auctions. He ended up abandoning bond auctions and returning to the Mellon program of offering notes and bonds at fixed prices on a regular quarterly schedule and to the Mills program of using bills—tax date bills, not 13-week bills—to raise funds “as needed.” Had he reversed the process and introduced regular auction sales of notes and bonds after developing tax date bills to satisfy his short-term funding requirements, the United States might have begun auctioning notes and bonds 35 years earlier than it did.

V

L’ENVOI

24

Treasury Debt Management since 1939

World War I marked the beginning of an important change in Treasury debt management. During the two decades prior to the war the Treasury sold securities on a sporadic basis pursuant to specific congressional authority to finance two narrowly defined projects: the Spanish–American War and the Panama Canal. In contrast, during the 1920s the Treasury sold securities on a regular basis to refinance maturing debt; during the 1930s it similarly sold securities to finance the budget deficits of the Great Depression as well as to refinance maturing debt. As a result of the change in the reason for—and frequency of—issuing securities, Congress gradually ceded control over both issuance and the composition of the debt. By the end of the 1930s Treasury officials were routinely financing budget deficits and refinancing maturing debt with weekly bill auctions and quarterly tax-date sales of 5-year notes and longer term bonds—issuance strategies designed and executed with little congressional participation. This concluding chapter examines several aspects of how Treasury officials exercised their authority with respect to debt management policy after 1939. The chapter begins with a brief survey of how the Treasury financed World War II (by reworking the Liberty Loan drives of World War I), explains how a wartime interest rate stabilization program solved a problem faced by Secretary McAdoo twenty-five years earlier, and describes the difficult process of terminating that program following the end of hostilities. The chapter concludes with a discussion of how the Treasury arrived at the strategy of regular and predictable auction offerings that has been in place since the mid-1970s,1 a strategy that grew out of the innovations of Secretary Mellon and Secretary Mills in the 1920s and early 1930s.2 Treasury Indebtedness after 1939 Figure 24.1 shows the growth in Treasury debt to the end of the twentieth century. The debt ballooned from $40 billion in mid-1939 to $268 billion in mid-1946 (as a result of the financing requirements of World War II) and then expanded more gradually to $313 billion in 1965. After 1965 indebtedness began to grow at a materially faster pace, to $370 billion in 1970, $876 billion in 1980, $3.1 trillion in 1990, and $5.7 trillion in 2000. 1. See Garbade (2004a, 2007). 2. The chapter is not, however, a complete history of the Treasury market since 1939. Many important debt management initiatives (e.g., advanced refundings and Operation Twist), securities (e.g., tax anticipation bills, cash management bills, TIPS (“Treasury Inflation Protected Securities”), and STRIPS (“Separate Trading of Registered Interest and Principal of Securities”)), and institutional arrangements (e.g., the Federal Reserve book-entry system and the Treasury Tax and Loan system) are not mentioned. Neither do we examine here the post–World War II growth of the dealer market for Treasury securities, dealer financing, the Fed’s primary dealer system, or the role of interdealer brokers.

338

Chapter 24

6

Trillions of dollars

5

4

3

2

1

0 1910

1920

1930

1940

1950

1960

1970

1980

1990

2000

Figure 24.1 Treasury debt, mid-1910 to mid-2000

Unlike the post–World War I experience, the United States never paid down any significant fraction of the debt incurred in the Second World War. Nevertheless, as shown in figure 24.2, the economic significance of the debt declined steadily from the late 1940s to 1980. Generally strong economic growth and persistent inflation—initially modest but increasingly virulent during the 1970s—reduced the ratio of Treasury indebtedness to aggregate national production from a high of 1.2 in 1946 to a postwar low of 0.3 in 1981, after which accelerating indebtedness began to push the ratio back up. Financing World War II World War II began in September 1939 when Germany invaded Poland. Few Americans were prepared, at the outset, to abandon the isolationist posture adopted during the 1930s.3 The conflict was something for Europeans to sort out among themselves. Isolationist sentiment began to soften in the spring of 1940, when the German army occupied Norway and Denmark, conquered France, and forced the evacuation of British forces from the beaches of Dunkirk. It softened further as the Battle of Britain raged during the summer of 1940. Americans 3. Popular demand for separation from European affairs led to the adoption of the Neutrality Acts of August 31, 1935, February 29, 1936, May 1, 1937, and November 4, 1939.

339

Treasury Debt Management since 1939

125

Percent

100

75

50

25

0 1910

1920

1930

1940

1950

1960

1970

1980

1990

2000

Figure 24.2 Treasury indebtedness as a percentage of gross national product

became concerned that if Britain surrendered and the British navy fell into German hands, Germany would be able to project naval power far into the North Atlantic. Acutely aware of the situation in Europe, President Roosevelt gradually stepped up the pace of defense preparedness. To pay the bills, Congress raised the debt limit—from $45 billion to $49 billion in June 1940 and then to $65 billion in February 1941. A month later Congress approved the Lend–Lease Act of March 11, 1941, authorizing the President to “sell, transfer . . ., exchange, lease, [or] lend” weapons and munitions to “any country whose defense the President deems vital to the defense of the United States.” This was widely understood to refer in particular to Great Britain. Two weeks after that, Congress appropriated $7 billion for the purchase and manufacture of weapons and munitions to be distributed pursuant to the terms of the LendLease Act. Defense preparedness and assistance to friendly countries turned to active engagement following the bombing of Pearl Harbor on December 7, the American declaration of war on Japan on December 8, and the German declaration of war on the United States on December 11, 1941. The Wartime Interest Rate Stabilization Program The key Treasury debt management action of World War II was the 1942 agreement between the Treasury and the Federal Reserve that interest rates on long-term Treasury debt would be capped at 2½ percent for the duration of

340

Chapter 24

the war.4 The cap was intended to solve a problem that Treasury Secretary McAdoo had faced during World War I: investors were reluctant to buy longterm fixed-rate bonds when the duration of the war was uncertain and there was a risk that an unexpectedly lengthy war would result in higher bond yields in the future. Conversely, as E. A. Goldenweiser, Director of Research at the Board of Governors, noted in a mid-1941 memo to the Federal Open Market Committee (FOMC), “When the public is assured that the rate will not rise, prospective investors will realize that there is nothing to be gained by waiting, and a flow [of funds] into Government securities . . . may be confidently expected.”5 A 2½ percent maximum was deemed appropriate because that was the rate at which the Treasury had been able to sell $2.7 billion of long-term bonds in October and December 19416 and because senior Federal Reserve officials believed that higher rates would not prevent inflation, would increase the cost of government borrowing, and would burden bondholders with capital losses.7 In addition to capping the rate on long-term Treasury bonds, the Fed agreed, in the spring of 1942, to cap the 3-month bill rate at 3⁄8 percent8—a rate marginally above the contemporaneous market rate on 3-month bills. When wartime expenditures subsequently began to stimulate rapid growth in economic activity, Federal Reserve officials tried to get Treasury officials to agree to an increase in the bill rate but the latter rejected their entreaties and the 3⁄8 4. See Murphy (1950), Walker (1954), and Wicker (1969). 5. Minutes of the Federal Open Market Committee, June 10, 1941, p. 27. See also, Thomas and Young (1947, p. 91) (policy of stable rates adopted to “encourage prompt buying of securities by investors, who might otherwise have awaited higher rates”) and Thomas (1951, p. 622) (policy of stable rates adopted out of “recognition that with almost unlimited borrowing always in prospect, interest rates would tend to rise and that such a prospect would tend to hold back purchases of securities”). 6. Murphy (1950, p. 93) 7. Minutes of the Federal Open Market Committee, September 27, 1941, p. 43 (summarizing the view of E. A. Goldenweiser that higher rates “would not be a feasible policy for the reason that it would increase the cost of Government borrowing without being effective in preventing price rises, and it could not be carried out with the approval of other agencies of Government. It would also raise serious problems about the decline in the capital value of outstanding securities.”). See also Thomas and Young (1947, p. 91) (maintaining interest rates at the level prevailing prior to U.S. entry into the war served to “keep down the interest cost on the Government’s war debt”) and Meltzer (2003, p. 580) (noting Secretary Morgenthau’s “passionate attachment to low interest rates”). 8. Federal Reserve Bank of New York Circular no. 2430, May 8, 1942 (“As an added means of assuring the liquidity of investments in Treasury bills, aside from the short maturity and ready marketability of the securities, the Board of Governors of the Federal Reserve System announced on April 30 that the Federal Open Market Committee had directed the Federal Reserve banks to purchase for the System Open Market Account all Treasury bills that may be offered to them, on a discount basis at the rate of 3⁄8 per cent per annum. This arrangement gives assurance to purchasers of Treasury bills that, in case they have a need for cash before the maturity of the bills, they can obtain it by selling Treasury bills to the Federal Reserve Bank, if necessary.”), “Treasury to Issue New Type of Bond,” New York Times, May 1, 1942, p. 29, and “Treasury Plans New ‘Tap’ Issue: An Innovation,” Wall Street Journal, May 1, 1942, p. 6. See also Murphy (1950, p. 98).

341

Treasury Debt Management since 1939

percent ceiling held for the duration of the war.9 Maximum yields at maturities between 3 months and 20 to 25 years were interpolated to give a smooth curve: yields on 1-year certificates were capped at 7⁄8 percent and yields on 7- to 9-year bonds were capped at 2 percent. There were two important consequences of the Fed’s commitment to limit interest rates to not more than pre-war levels and to stabilize the yield curve with a positive slope that reflected pre-war expectations of rising interest rates.10 First, the levels were too low for wartime and the Federal Reserve ended up buying, and thus monetizing, a significant amount of war debt, with adverse consequences for postwar inflation. Second, the shape of the curve was inconsistent with the decision to stabilize interest rates. As market participants gradually became aware of the inconsistency, they increasingly evidenced a preference to hold higher yielding—but hardly more risky—long-term debt in lieu of short-term debt.11 Financing the War Table 24.1 shows that 24 percent of the $228 billion increase in Treasury debt from mid-1939 to mid-1946 was financed with bills and certificates of indebtedness and that 62 percent was financed with marketable bonds and nonmarketable savings bonds. The remaining 14 percent was spread over marketable notes, tax and savings notes (similar to the certificates sold in anticipation of bond sales and tax collections during World War I), and special trust fund issues. Treasury Secretary Morgenthau initially hoped to finance the bulk of the pre-war defense program, and later the war itself, with savings bonds sold individually and through payroll savings plans to a wide spectrum of buyers. He wanted to attract some of the rising incomes of wage earners in order to contain inflationary pressures and he wanted to do so without the intense—and sometimes coercive—bond drives of World War I.12 However, it became clear by mid-1942 that savings bonds were not attracting nearly enough money and that the Treasury would have to reintroduce bond drives and sell marketable 9. Wicker (1969, p. 453) (stating that Federal Reserve officials did not interpret their support of a 3⁄8 percent bill rate in 1942 as an indefinite commitment: “The evidence on this point is overwhelming—the record of continuous requests made to the Treasury during the war period to approve an increase in the bill rate.”) and Wicker (1969, p. 457) (stating that Treasury officials “repeatedly rejected requests by the [Federal Open Market Committee] to increase bill rates.”). 10. See, for example, Lutz (1940, p. 59) (stating that “In recent years we have had a situation where future short rates . . . were expected to rise, which accounts for the fact that we have a series of yields which ascends with the length of the maturities. . . . Direct evidence [of such expectations] is to be found in the financial journals, which are full of warnings that present interest rates are unusually low.”) 11. Walker (1954, pp. 31–33), Wicker (1969, pp. 455–456), and Meltzer (2003, pp. 580 and 596). 12. Morse (1971) and Samuel (1997).

342

Chapter 24

Table 24.1 Treasury debt in mid-1939 and mid-1946 Mid-1939 ($billions) Marketable debt Treasury bills Certificates of indebtedness Notes Bonds Nonmarketable debt Savings bonds Tax and savings notes Special issues Other Total

Mid-1946 ($billions)

Increase ($billions)

Percent of total increase

1.3 0.0 7.2 25.4

17.0 39.7 13.4 119.5

15.7 39.7 6.1 94.1

6.9 17.4 2.7 41.2

1.9 0.0 3.8 0.3 39.9

49.0 6.7 22.3 0.4 268.1

47.2 6.7 18.5 0.1 228.2

20.7 2.9 8.1 0.0 100.0

Source: 1939 Treasury Annual Report, p. 7, and 1946 Treasury Annual Report, p. 61.

bonds as well as savings bonds.13 The Treasury ultimately launched seven War Loan drives during the course of the war and one postwar Victory Loan drive (table 24.2). Unlike the Liberty Loan drives of World War I, each of which offered only a single security, the loan drives of the Second World War offered a variety of securities, reflecting a marketing strategy of “something for everyone.”14 Monetizing the War Debt Virtually all of the $15.7 billion increase in Treasury bills between mid-1939 and mid-1946 ended up in the vaults of the twelve Federal Reserve Banks (table 24.3). Of the $39.7 billion increase in certificates of indebtedness, $6.8 billion ended up in the same place. In contrast, only a modest fraction of the wartime increase in notes ended up with the Fed, and the Fed actually reduced its holdings of Treasury bonds during the war. The disproportionate expansion of Federal Reserve holdings of short-term Treasury debt reflects (1) the decision of the Treasury to issue across the curve during the war and (2) the preference of public investors for intermediate-term notes and longer term bonds in lieu of short-term debt. In keeping to its 1942 commitment to stabilize interest rates, the Fed had to purchase all of the bills that the public did not want to hold at the target interest rate of 3⁄8 percent— which, as it turned out, was essentially all the bills that the Treasury issued. 13. Gaines (1962, p. 51). 14. See “Report to Congress by Secretary Morgenthau,” July 21, 1945, reprinted in 1945 Treasury Annual Report, p. 397, stating (p. 410) that “the second major objective of the Treasury in its war borrowing—second only to the objective of avoiding inflation—has been to adapt the securities which it has offered to the public to the requirements of the various classes of investors.”

343

Treasury Debt Management since 1939

Table 24.2 Marketable security sales in the World War II war loan drives Term and type

Coupon rate (%)

First war loan, November 30 to December 23, 1942 7⁄ 1-yr certificate 8 5½-yr bond 1¾ 26-yr bond 2½ Second war loan, April 12 to May 1, 1943 7⁄ 1-yr certificate 8 9½-yr bond 2 26¼-yr bond 2½ Third war loan, September 9 to October 2, 1943 7⁄8 1-yr certificate 10-yr bond 2 26¼-yr bond 2½ Fourth war loan, January 18 to February 15, 1944 7⁄ 1-yr certificate 8 15½-yr bond 2¼ 26-yr bond 2½ Fifth war loan, June 12 to July 8, 1944 7⁄8 1-yr certificate 2¾-yr note 1¼ 10-yr bond 2 25¾-yr bond 2½ Sixth war loan, November 20 to December 16, 1944 7⁄8 1-yr certificate 2¾-yr note 1¼ 10-yr bond 2 26¼-yr bond 2½ Seventh war loan, May 14 to June 30, 1945 7⁄8 1-yr certificate 5½-yr bond 1½ 17-yr bond 2¼ 27-yr bond 2½ Victory loan, October 29 to December 8, 1945 7⁄ 1-yr certificate 8 17-yr bond 2¼ 27-yr bond 2½

Sales ($billions)

3.80 3.06 2.83 5.25 4.94 3.76 4.12 5.26 3.78 5.04 3.73 2.21 4.77 1.95 5.83 2.91 4.41 1.55 7.92 3.45 4.80 2.64 5.24 7.97 3.77 3.47 11.69

344

Chapter 24

Table 24.3 Assets and liabilities of Federal Reserve banks, mid-1939 and mid-1946 ($billions)

Assets Gold and gold certificatesa Loans and advances Treasury securities: Bills Certificates Notes Bonds Other assets Total assets Liabilities Federal Reserve notes Member bank reserve balances Other liabilities and capital Total liabilities and capital

Mid-1939

Mid-1946

Increase

13.5 0.0

18.1 0.2

4.6 0.2

0.5 0.0 1.2 0.9 1.0 17.1

14.5 6.8 1.7 0.8 2.7 44.7

14.0 6.8 0.6 −0.2 1.6 27.6

4.5 10.0 2.6 17.1

24.2 16.1 4.4 44.7

19.7 6.1 1.8 27.6

Source: All components except “Other assets” and “Other liabilities and capital” from Federal Reserve Bulletin, August 1939, p. 649, and August 1946, p. 883. “Other assets” and “Other liabilities and capital” constructed from data for June 28, 1939, and June 26, 1946, from Federal Reserve Bulletin, August 1939, pp. 650–52, and August 1946, pp. 884–86. “Total assets” and “Total liabilities and capital” constructed from components by summation. a. Includes redemption fund for Federal Reserve notes.

The resulting monetization of a nontrivial fraction of war debt fueled a $25.8 billion expansion in high-powered money (the sum of Federal Reserve notes and member bank reserve balances in table 24.3) and laid the foundation for a sharp postwar spike in prices.15 (The consumer price index rose 34 percent between 1945 and 1948.16) From the End of the War to 1958 World War II ended with the surrender of Germany on May 7 and Japan on August 15, 1945. The key debt management issue in the immediate postwar period was how the Federal Reserve should exit from the wartime interest rate stabilization program.17 Fed officials felt a continuing responsibility for avoiding precipitous action that might destabilize the now-gigantic Treasury market 15. Meltzer (2003, p. 586). 16. U.S. Department of Commerce (1975, p. 210). 17. See, for example, Sproul (1964, p. 228) (“In the reconversion period, at the end of the war in 1945, the problem facing the Federal Reserve System was how to proceed, and at what speed, to recapture [the initiative on creation of reserves], and to restore the ability of the Federal Reserve Banks to place a price upon reserve credit and a check on its availability which could be varied to meet changes in economic circumstances.”).

345

Treasury Debt Management since 1939

and did not want to abandon the program abruptly,18 but they did want to restrain the expansionary consequences of growing bank reserves. Some— including, most prominently, Allan Sproul, president of the Federal Reserve Bank of New York from January 1941 to June 1956—were willing to accept higher interest rates in the process.19 The first substantive change in the interest rate stabilization program occurred in July 1947, when the Treasury agreed to abandon the posted rate for Treasury bills.20 Under the new policy the FOMC expected the bill rate to “find its level in the market in proper relation to the yields on certificates of indebtedness.”21 Bill rates rose to 0.75 percent in August and the increase forced a reconsideration of certificate yields.22 The ceiling on 1-year certificate rates was raised to 1 percent in August and to 11⁄8 percent in November.23 Higher yields on bills and certificates, as well as the prospect of increased ceiling rates for longer term securities, incentivized banks and others to move in on the yield curve, selling bonds and buying shorter term securities. The Fed accommodated the reallocation, reducing its bill portfolio to $8.6 billion and expanding its holdings of bonds to $6.2 billion by mid-1948. At the same time the Treasury reframed from issuing bonds and did all of its refinancing with bills and certificates of indebtedness. (As a result the average maturity of the debt began to recede from the high watermark of 10 years and 5 months reached in June 1947.) The ceiling on 1-year certificate rates was raised to 1¼ percent in October 1948,24 but that was as high as the ceiling got before the second half of 1950. Although the Federal Reserve continued to seek further relaxation of the interest rate stabilization program, the recession that began in November 1948 undercut the urgency of its requests: the Treasury sold 1-year certificates with a 11⁄8 percent coupon rate in September and December, 1949. Following the resumption of economic growth in early 1950, the rate on new 1-year certificates returned to 1¼ percent. In mid-June the FOMC voted 18. See, for example, Thomas and Young (1947) (noting on p. 100 the Fed’s “responsibilities for maintaining an orderly and stable market for Government debt” and on p. 101 the System’s “new responsibility, inherited from war finance, of maintaining a stable market for the public debt”). Gaines (1962, p. 63) states that “the principal officials of [the Fed and the Treasury] were in substantial agreement on basic polices through most of the first four postwar years, differing principally on the levels of interest rates that should be maintained not on the root question of whether any level of rates should be maintained.” [Emphasis in the original.] 19. Meltzer (2003, pp. 634–35 and 637). Meltzer observes (p. 637) that the Board of Governors “favored control of specific uses of credit instead of more general policies.” 20. Meltzer (2003, pp. 642–43). See also, Thomas (1951, p. 626), and Sproul (1964, p. 228). 21. Federal Reserve Bank of New York Circular no. 3230, July 3, 1947. 22. Meltzer (2003, p. 643). 23. Sproul (1964, pp. 228–29). 24. Sproul (1964, p. 229).

346

Chapter 24

to raise the ceiling rate to 13⁄8 percent, to become effective following the conclusion of a late June offering of 13-month Treasury notes.25 However, the increase was put on hold when, on June 25, 1950, North Korean soldiers crossed the 38th parallel into South Korea and the Korean War began.26 The Accord The concern of Treasury and Federal Reserve officials that the United States might be on the verge of another major war,27 and the desire of Fed officials to avoid repeating the policy mistakes of the preceding war28 precipitated the final dismantling of the interest rate stabilization program. In light of the rapidly expanding American combat operations in Korea and the prospect of substantial increases in defense expenditures, the FOMC voted on Friday, August 18, 1950, to reinstate its previous decision to raise the ceiling on 1-year certificate rates to 13⁄8 percent.29 The FOMC’s action was not publicly disclosed, but the concurrent publicly announced decision of the Board of Governors to approve a request from the Federal Reserve Bank of New York to raise its discount rate to 1¾ percent, effective Monday, August 21, signaled unambiguously that short-term rates were going up.30 Treasury Secretary John Snyder did not accept passively the Fed’s decision to raise the ceiling on short-term interest rates.31 Shortly after being informed of the decision, he announced that the Treasury would offer 1¼ percent 13-month notes at par in exchange for $13.6 billion of securities due to be redeemed on September 15 and October 1.32 Since investors were unlikely to 25. Minutes of the Federal Open Market Committee, June 13, 1950, p. 13. 26. Minutes of the Executive Committee of the Federal Open Market Committee, July 10, 1950, p. 9 (stating in a letter subsequently sent to the Secretary of The Treasury on July 13, 1950, that “in view of the Korean situation, we have been holding in abeyance our previous decision . . ..”). 27. 1951 Treasury Annual Report, p. 263 (statement of Secretary of the Treasury John Snyder that “when aggression broke out in Korea, the Treasury visualized the possibility of a third world war”). 28. Minutes of the Executive Committee of the Federal Open Market Committee, July 10, 1950, p. 6 (“Mr. Sproul expressed the opinion that in the event of major military commitments a more nearly horizontal structure of rates than existed during the last war was highly desirable.”). 29. Minutes of the Federal Open Market Committee, August 18, 1950, p. 23. 30. “Federal Reserve and U.S. Treasury Act Separately to Curb Inflation,” New York Times, August 19, 1950, p. 26, and “Reserve Board Hikes Discount Rate to Curb Credit; Treasury Still Backs Cheap Money on Refunding Bonds,” Wall Street Journal, August 19, 1950, p. 2. 31. See, Hetzel and Leach (2001, p. 38) (stating that “when told that the Fed planned to raise short-term interest rates, Secretary Snyder reacted angrily”) and Minutes of the Federal Open Market Committee, February 6–8, 1951, pp. 34 and 36 (summaries of statements of Board of Governors Chairman Thomas McCabe and Allan Sproul that Snyder interpreted the Fed’s decision to raise interest rates as “an ultimatum”). 32. “Federal Reserve and U.S. Treasury Act Separately to Curb Inflation,” New York Times, August 19, 1950, p. 26. Meltzer (2003, p. 692) discusses the chronology of the Fed’s decision to raise short-term interest rates and Snyder’s announcement of the terms of the September refunding operations.

347

Treasury Debt Management since 1939

pay par for a 1¼ percent 13-month note when they could buy outstanding 1-year securities at a 13⁄8 percent yield, his action risked a failed offering of Treasury securities in wartime. Market participants quickly realized that the Fed and the Treasury were in open conflict.33 The Fed met Snyder’s challenge by bidding aggressively for the maturing securities and by exchanging those securities for the new notes for its own account.34 The Fed was unable to buy all of the maturing securities, but largely as a result of its efforts, 81 percent of the securities due to be redeemed on September 15 and 85 percent of the securities due to be redeemed on October 1 were exchanged;35 the balance was redeemed for cash. (In the late 1940s, cash redemptions, or “attrition,” usually ran about 5 percent of the maturing debt in an exchange offering.36) In mid-October 1950 the FOMC raised the ceiling on 1-year interest rates to 1½ percent.37 This time Treasury officials accepted the increase and, on November 22, announced that they would refinance $8 billion of debt maturing in mid-December and early January with a 5-year note bearing a 1¾ percent rate of interest that was not out of line with market yields in the 5-year sector.38 However, this offering also required substantial Federal Reserve support after the Chinese army began to engage American forces in North Korea on November 25. The Fed bought and exchanged $2.7 billion of the maturing issues, but even with that assistance 14 percent of the maturing securities were redeemed for cash.39 33. “Reserve Board Hikes Discount Rate to Curb Credit; Treasury Still Backs Cheap Money on Refunding Bonds,” Wall Street Journal, August 19, 1950, p. 2 (remarking that the Fed and Treasury “are working at cross purposes,” and reporting the view of “a top government fiscal expert” that the announcements “very plainly signify a difference of opinion on what government money policies are needed now”), and “Firm Money Apparently Wins First Round in Clash of Fiscal Policies,” Wall Street Journal, August 28, 1950, p. 7 (noting “the break into the open of the feud between the Federal Reserve System and the Treasury on fiscal and monetary policy”). 34. “Board Doesn’t Like Terms of Treasury’s New 13-Month Notes, But It’s Still Expected to Be a Major Subscriber,” Wall Street Journal, August 23, 1950, p. 2, and “Chapter Is Added to Fiscal History,” New York Times, September 10, 1950, p. 125. 35. “Treasury Reports 81% Response to Offer to Exchange Certificates,” New York Times, September 15, 1950, p. 45, “New Notes Find Fewer Takers Than Any Issue in Recent Years: Reserve Buying Prevents a Serious Failure,” Wall Street Journal, September 15, 1950, p. 3, and “Treasury Notes Only 85% Taken,” New York Times, September 30, 1940, p. 27. 36. 1951 Treasury Annual Report, p. 269. 37. Minutes of the Federal Open Market Committee, October 11, 1950, p. 8, and Minutes of the Executive Committee of the Federal Open Market Committee, October 11, 1950, p. 3. 38. 1951 Treasury Annual Report, p. 269 (1¾ percent rate chosen “to price the new issue in line with the market”), “Rapprochement is Seen,” New York Times, November 23, 1950, p. 62 (“By choosing a 1¾ percent rate . . . the Treasury reversed its repeated publicly-avowed disagreement with the persistent moves of the central bank over the last year to raise short-term interest rates”; the 1¾ percent rate “was taken in financial circles to reflect a full meeting of minds for the first time in a year between the Treasury and the Federal Reserve System”), and “Treasury Gives in to Reserve Board Demands for Rise: to Offer 1¾%, Five-Year Notes for $8 Billion Issue,” Wall Street Journal, November 24, 1950, p. 2 (rate important “in that it recognizes a situation that already exists”). 39. “Treasury Offers Undersubscribed,” New York Times, December 15, 1950, p. 54.

348

Chapter 24

Faced with the prospect of a bigger and longer war, and concomitantly larger Treasury financings, the Fed began, for the first time, to seek to free itself from its commitment to keep long-term Treasury yields below 2½ percent.40 At the same time, and for the same reason, Secretary Snyder and President Harry Truman sought a reaffirmation of the Fed’s commitment to the 2½ percent ceiling, a reaffirmation that the Fed declined to provide.41 The impasse continued until mid-February 1951, when Secretary Snyder went into the hospital and left Assistant Secretary of the Treasury William McChesney Martin to negotiate what has since become known as “the Treasury–Federal Reserve Accord” or, more simply, as “the Accord.”42 Late on Saturday, March 3, 1951, the Treasury and Federal Reserve rather tersely announced that they had “reached full accord with respect to debt management and monetary policies to be pursued in furthering their common purpose to assure the successful financing of the Government’s requirements and, at the same time, to minimize monetization of the public debt.” Privately the Fed agreed that it would, 1. maintain orderly conditions in the Treasury market, albeit without reference to any maximum rate of interest (other than as noted in item 3 below); 2. consult with the Treasury if it wanted to raise the discount rate above 1¾ percent before the end of 1951; and 3. provide limited support for long-term Treasury bonds until the Treasury could complete an exchange offering of 2¾ percent nonmarketable long-term bonds for outstanding 2½ percent marketable bonds.43 Allan Meltzer concludes that the Accord “ended ten years of inflexible [interest] rates” and was “a major achievement for the country.”44 40. Minutes of the Federal Open Market Committee, November 27, 1950, p. 9 (remark of Allan Sproul that the Committee should “look toward unfreezing the long end of the rate pattern”) and 1951 Treasury Annual Report, pp. 269–70 (remark that “Early in 1951, . . . officials of the Federal Reserve System outlined to the Treasury a program which would involve a reorientation of debt management policy. The program included proposals for further increases in interest rates, including increases in the long-term area.”). 41. Hetzel and Leech (2001, p. 40) and Meltzer (2003, pp. 700 and 705). 42. The Accord and the events leading up to it are recounted in 1951 Treasury Annual Report, pp. 263–73, Sproul (1964), Hetzel and Leach (2001), and Meltzer (2003, pp. 681–712). The negotiation of the terms of the Accord is recorded in Minutes of the Federal Open Market Committee, March 1–2, 1951, and March 3, 1951. 43. The terms of the Accord are noted by Meltzer (2003, p. 711). See also 1951 Treasury Annual Report, p. 272. The Fed’s support for long-term bonds was limited to the offering period for the 2¾ percent nonmarketable bonds and to purchases of no more $200 million. Treasury officials repeatedly stressed the importance of keeping secret the limitations on Federal Reserve support for long-term Treasury bonds during the period of the offering. Minutes of the Federal Open Market Committee, March 1–2, 1951, pp. 10, 15, 18, 19, and 38, and March 3, 1951, p. 8. 44. Meltzer (2003, p. 712).

349

Treasury Debt Management since 1939

After the Accord Long-term Treasury yields rose through the 2½ percent level in early April 1951 but the Accord did not immediately free monetary policy from the constraints that had bound it since 1942, and it did not immediately free Treasury debt management policy. The Fed maintained the discount rate at 1¾ percent through the end of the year and continued to support Treasury financings through the end of the following year.45 For its part the Treasury continued to rely on offerings of short-term bills and certificates.46 As a result the average maturity of the debt continued to recede, to 6 years and 7 months in mid-1951 and to 5 years and 8 months in mid-1952. The relentless shortening of the maturity structure of the debt raised rollover financings to levels that began to be perceived as a problem,47 and in 1952 the Treasury returned to selling marketable bonds for the first time since the 1945 Victory Loan drive. It brought a total of four bond offerings in 1952 and the first half of 1953, during a period of steady economic expansion. (Following the 1948–49 recession, the economy expanded until July 1953.) In the process Treasury officials found that, to quote Tilford Gaines, “the peak of a boom, when capital markets already were seriously strained, was a most difficult time to sell longer term Treasury securities.”48 When the economy began to slide off into recession in the second half of 1953, the Fed adopted a more accommodative monetary policy and investors became more interested in long-term fixed-income securities, but a new consideration appeared. Again quoting Gaines: “To sell long-term bonds in a recession would compete with business demands for capital and . . . conflict with other Government and Federal Reserve policies aimed at encouraging recovery.”49 “For the first time,” Gaines observed, “it became apparent that ‘there is no good time to sell long-term Government bonds,’ a conundrum that was to continue to plague Treasury administration of the debt . . ..”50 45. Gaines (1962, p. 66). 46. Gaines (1962, pp. 66 and 69). 47. Gaines (1962, p. 69) (noting that “virtually all Treasury offerings through the end of 1952 were of shorter maturity securities, and the debt management and credit policy problems created by a debt pressing constantly toward the shorter end of the maturity spectrum became steadily more urgent”). 48. Gaines (1962, p. 70). Gaines was an officer in the Securties Department (since renamed the Markets Group) of the Federal Reserve Bank of New York at the time he wrote the first draft of his pathbreaking book. He and Paul Volcker are cited in Roosa (1956, p. 7n.) for their “valiant help” in bringing to fruition what a generation of fixed-income market participants came to know as “the red book.” 49. Gaines (1962, p. 70). 50. Gaines (1962, p. 70). A decade later, Under Secretary of the Treasury Paul Volcker (1972) would recall the same cliché: “No time seems to be a good time for offering long-term Treasury securities—either rates are too high or there is a desire to maximize the flow of funds to other borrowers.”

350

Chapter 24

In addition to the policy problem of when to sell long-term Treasury debt, there was the institutional problem of how to sell it. The terms of an offering, including price and coupon rate, were typically announced about a week before the subscription books for the offering closed. Shortly before the books closed, when most large subscriptions were tendered, interest rates could be higher or lower than they were when the terms were announced. The offering was consequently liable to appear either too expensive, thereby risking a failed offering (in the case of a cash offering) or unexpectedly high attrition and a need for emergency financing (in the case of an exchange offering), or too cheap, leading to either a vast oversubscription (in the case of a cash offering) or unexpectedly low attrition and excessive—and costly—Treasury cash balances (in the case of an exchange offering). Under Secretary of the Treasury Robert Roosa suggested, in December 1951, that the Treasury could “minimize the risks of miscalculating investor response” by reviving the practice, introduced by Secretary Morgenthau, of offering holders of maturing debt a choice of any of several issues, including short-term as well as intermediate- and long-term debt.51 The idea was that investors with longer term investment horizons could subscribe to the longer term issues; the balance of the offering would be placed as short-term debt with investors with short-term horizons.52 Multi-option exchange offerings were “markedly successful” during the second half of 1953 and 1954; more than half of all exchange subscriptions by private investors took the longer of the offered securities.53 As a result the average maturity of Treasury debt rose for the first time since the end of World War II, from 5 years and 4 months in mid-1953 to 5 years and 6 months in mid-1954. However, when the economy began to recover in the second half of 1954 and interest rates began to rise, Treasury officials proved reluctant to advance the offering rates on long-term bonds in parallel with open market rates, and average maturity again began to fall (to 4 years and 9 months in mid-1957).54 51. Roosa (1952, p. 234). The first postwar multi-option exchange offer was extended in the February 1953 refunding, when the Treasury offered to exchange either a 1-year certificate of indebtedness or a 5-year, 10-month bond, for $8.9 billion of maturing certificates. See “Treasury Taking First Step on Debt,” New York Times, January 28, 1953, p. 37, “Treasury Issues Refunding Terms,” New York Times, January 30, 1953, p. 29, and “New Refinancing Held Big Success,” New York Times, February 10, 1953, p. 37. 52. See, Gaines (1962, pp. 70–71) (noting that the program of “optional exchange offerings . . . would, presumably, place only as many [Treasury] securities outside the short-term market as investors demanded and would . . . avoid forcing [long-term Treasury] securities into the market”). 53. Gaines (1962, p. 71). See also 1954 Treasury Annual Report, pp. 22 and 26–28. 54. See Gaines (1962, p. 79) (stating that “the Treasury allowed the rate on interest on its longterm debt to get out of touch with rates of interest on competitive investments” and that “there was a persistent tendency for a larger and larger proportion of the debt to move into shorter maturities”).

351

Treasury Debt Management since 1939

The July 1958 Refunding The economy entered another recession in August 1957 and economists predicted a long and fairly severe contraction; a mid-November reduction in the discount rate from 3½ percent to 3 percent was widely interpreted as only the first of several easing moves by the Fed.55 From a high of 3.99 percent in October 1957, yields on intermediate-term (3- to 5-year) Treasury debt fell to 2.25 percent in May 1958. On May 29, 1958, Treasury officials announced a $1 billion cash offering of 27-year bonds and a multi-option exchange offering of 11-month certificates and 6½-year bonds to refinance $9.6 billion of maturing debt. Investors expected the recession to continue to deepen and interest rates to continue to fall56 and, anxious to participate in the exchange offering, bid up the price of the maturing debt to 1003⁄8 percent of par57—a reflection of the value of the embedded exchange rights. Following the close of the subscription books on June 6 the Treasury announced that investors had agreed to exchange almost 96.3 percent of the maturing debt for new securities, leaving attrition at an unusually low 3.7 percent. The 6½-year bond was particularly popular: investors took $7.4 billion of the bond, compared to only $1.8 billion of the certificate.58 Around the time of the settlement of the exchange offering, on Monday, June 16, information began to appear that, contrary to expectations, the economy was strengthening.59 By Friday, June 20, the price of the new 6½-year bond was down to 9922 (99 and 22/32nds) and significant numbers of leveraged market participants were forced to liquidate their positions.60 To cushion the deleveraging, the Treasury began to buy the bond for its trust funds and (even though it had just issued the bond) for outright retirement.61 The price of the bond declined more slowly to 9915 on July 3 and to 997 on July 11. To avoid further roiling a fragile market, the Treasury announced on July 17 that it would offer only a 1-year certificate in its July refunding of $16.3 billion of maturing debt.62 However, it soon became evident that even that 55. U.S. Treasury and Federal Reserve System (1960, pp. 5 and 7). The discount rate was ultimately reduced to a low of 1¾ percent in mid-April, 1958. 56. U.S. Treasury and Federal Reserve System (1960, pp. 22 and 25). 57. U.S. Treasury and Federal Reserve System (1960, p. 40). 58. “Attrition Down to 3.7% in Treasury Financing,” New York Times, June 14, 1958, p. 26. 59. U.S. Treasury and Federal Reserve System (1960, p. 25). 60. U.S. Treasury and Federal Reserve System (1960, p. 60 and ch. 7). 61. “Treasury Lessens Squeeze on Bonds,” New York Times, July 10, 1958, p. 37, and “Treasury Buys Back $589.5 Million of 25⁄8% Bond Issue Sold in June,” Wall Street Journal, July 10, 1958, p. 5. 62. “Treasury Offers a Big 1-Year Issue,” New York Times, July 18, 1958, p. 29, and “Treasury Offers 1-Year, 15⁄8% Issue to Refund $16.3 Billion Securities,” Wall Street Journal, July 18, 1958, p. 2.

352

Chapter 24

short-term security was not attracting much interest. To avoid massive cash redemptions, the Fed purchased (and exchanged) $110 million of the maturing securities, and it purchased $1.1 billion of the new 1-year certificates in the when-issued market.63 Despite that support, 29 percent of the privately held portion of the maturing debt was presented for redemption.64 The unusual market volatility experienced between mid-June and late July, 1958, and the need for large-scale support operations by the Federal Reserve, caught the attention of government officials, members of Congress, and others.65 The episode triggered a major study of the Treasury market66 and precipitated several critiques of Treasury debt management. Critiques of Treasury Debt Management The critiques of Treasury debt management that appeared in the wake of the July 1958 refunding made two important points: Treasury debt operations were unpredictable and a source of market instability, and fixed-price offerings were excessively costly. Milton Friedman’s Critique Professor Milton Friedman of the University of Chicago suggested that Treasury debt operations should be simplified and streamlined “in such a way as to keep [them] from themselves being a source of instability.”67 He argued that: In the attempt to keep down the interest cost, and to achieve such other objectives as a wide distribution of securities and lengthened maturities, the Treasury has sought to “tailor” securities to the supposed demands of special groups of potential purchasers, and to time the issue of securities to fit into slack periods in the money market. The result has been a bewildering variety of securities of different maturities and terms, and lumpiness and discontinuity in debt operations, with refunding of major magnitude occurring on a few dates in the year. Instead of proceeding at a regular pace and in a standard way to which the market could adjust, debt management operations have been jerky, full of expedients and surprises, and unpredictable in their impact and outcome.68

63. “Reserve to Prop U.S. Bond Prices,” New York Times, July 19, 1958, p. 19, “Federal Reserve System to Buy Longer-Term Issues in Move to Bolster Sagging Government Bond Market,” Wall Street Journal, July 21, 1958, p. 3, “Reserve Backing U.S. Financing,” New York Times, July 23, 1958, p. 37, and “Reserve Buying of U.S. Issue Big,” New York Times, July 25, 1958, p. 25. 64. “U.S. Offer Meets Heavy Rejection,” New York Times, July 26, 1958, p. 18. 65. U.S. Treasury and Federal Reserve System (1960, p. 1). 66. U.S. Treasury and Federal Reserve System (1959 and 1960). 67. Friedman (1960, p. 60). See also, Joint Economic Committee (October 1959, pp. 3023–26 and 3051) (testimony of Milton Friedman). 68. Friedman (1960, p. 60). Friedman also noted (p. 61) the argument that “the maturity of the debt should be lengthened in [an expansion] to reduce the liquidity of the economy, and reduced in recession to increase liquidity,” but pointed out that, in practice, the exact opposite had occurred.

353

Treasury Debt Management since 1939

Friedman characterized the mid-1958 refundings as an “extreme example” of such unpredictability. Friedman advanced several recommendations:69 • There should be a small number of marketable issues, possibly limited to 3-month bills and 10-year bonds. • Offerings should be conducted at regular intervals, such as weekly or monthly, and advertised well in advance. • Securities should be sold only by auction, allowing the market to determine prices and yields.70 Freidman concluded that if the Treasury adopted his recommendations, “Debt operations would be regular in timing, reasonably stable in amount, and predictable in form. . . . There would be no need for the succession of redemption crises that have become the pattern in recent years, or for the extensive consultation, crystal gazing, and plain guesswork that now go into setting coupon rates, maturities, and the like.”71 Treasury’s Expansion of Bill Auctions and Defense of Fixed-Price Offerings of Notes and Bonds Even before Friedman advanced his recommendations, the Treasury had expanded its auction offerings of Treasury bills, introducing weekly auctions of 26-week bills in December 1958 and quarterly auctions of 1-year bills in March 1959.72 It did so to place a larger part of the short-term debt on a “routine rollover basis,” replacing sporadic fixed-price offerings of certificates of indebtedness with regular auction offerings of longer-term bills.73 Testifying before the Joint Economic Committee in July 1959, Secretary of the Treasury Robert Anderson readily acknowledged that auctions were “an efficient mechanism” for pricing Treasury bills.74 He further acknowledged that auction sales of notes and bonds would “relieve [the Treasury] of a major responsibility in pricing and selling coupon issues.”75 But Anderson 69. Friedman (1960, pp. 63–64). 70. Carson (1959) similarly recommended that the Treasury should auction its offerings. 71. Friedman (1960, p. 65). 72. Federal Reserve Bank of New York Circular no. 4663, November 18, 1958 (announcing new program for regular weekly offerings of 26-week Treasury bills) and Federal Reserve Bank of New York Circular no. 4715, March 19, 1959 (announcing new program of regular quarterly offerings of 1-year Treasury bills). 73. Gaines (1962, p. 81). 74. Joint Economic Committee (July 1959, p. 1150). 75. Joint Economic Committee (July 1959, p. 1148).

354

Chapter 24

nevertheless claimed that fixed-priced offerings were preferable to auction offerings of notes and bonds.76 Anderson premised his claim on the observation that many of the small banks, corporations, and individuals who subscribed to fixed-price offerings of notes and bonds did not have the “professional capacity” to bid in auctions.77 Lacking the requisite expertise, they were liable to either bid too high and pay too much or bid too low and be shut out, and were therefore likely to avoid bidding altogether and to buy new securities in the post-auction secondary market. (This is exactly what happened when Morgenthau attempted to auction bonds in 1935.) Anderson suggested that the withdrawal of small institutional and individual investors from the primary market would have several adverse consequences, including reducing the Treasury’s ability to distribute its debt broadly (an objective dating back to before the Spanish– American War bond issue in 1898) and increasing the risk that, for lack of sufficient participants, auctions might not be competitive and might even fail to attract sufficient interest to cover the Treasury’s financing requirements. He concluded that “the present practice of offering Treasury certificates, notes, and bonds at prices and interest rates determined by the Treasury . . . result[s] in an effective distribution of new Treasury issues at minimum cost to the taxpayer.”78 Tilford Gaines’s Critique Tilford Gaines did not advocate a particular Treasury debt management strategy but rather criticized the absence of any strategy. Like Friedman, he concluded that Treasury debt operations were unpredictable and a source of market instability. Gaines believed that Treasury officials needed to constrain their exercise of discretion in individual cash and refunding operations by designing, communicating, and adhering to a broad strategic debt management policy. He emphasized the need for discipline in individual operations and consistency across operations: • The practice of considering a full range of maturity alternatives as each financing approached created a degree of uncertainty that sometimes seriously complicated the financing itself.79

76. Joint Economic Committee (July 1959, pp. 1148–53). 77. Joint Economic Committee (July 1959, p. 1149). 78. Joint Economic Committee (July 1959, p. 1153). Friedman responded to Anderson’s defense of fixed-price offerings by suggesting that the Treasury could adopt a single-price auction format, where all successful auction participants pay the lowest accepted price, instead of the multipleprice format used for bill auctions, where each accepted tender is invoiced at the price proposed in the tender. Friedman (1960, p. 65). The idea of single-price auctions has a long history. See, for example, Adams (1887, pp. 168–69). 79. Gaines (1962, p. 77).

355

Treasury Debt Management since 1939

• Each financing operation seemed to be an independent crisis, requiring new and largely unpredictable decisions as to terms and maturities, rather than one part of a carefully planned debtmanagement program. Since the Treasury found it necessary to come to market nearly every month, either for cash or to refund, the result was a nearly continuous state of crisis in the Government securities market.80 • Debt operations [between World War II and the late 1950s have] been characterized by drift and expediency, and individual debt operations have generally been adjusted to “market conditions” (which is another way of saying they were guided by expediency) rather than to long-range policy objectives.81

Gaines was also critical of the techniques of Treasury debt management: “The absence of a well-articulated and generally understood set of techniques, developed with a view to making the management of the debt as technically simple as possible, helped to create the impression that each of the many operations was a minor crisis.”82 Like Friedman, he recommended smaller, more regular operations, rather than “huge operations [that] sometimes created almost impossible pressures on the market.”83 Gaines was particularly critical of multi-option exchange offers in which holders of maturing securities “were given an exchange option among two or three issues of different maturities . . .. In a very real sense, the maturity distribution of the debt was left in the hands of investors.”84 And in a related vein, he noted that because the Treasury had allowed the rate of interest on its long-term debt offerings to get “out of touch with rates of interest on competitive investments [during periods of rising interest rates], there was a persistent tendency for a larger and larger proportion of the debt to move into shorter maturities.”85 Gaines concluded that the “critical lack [in the postwar period] has been the absence of an over-all plan, and the perfection of techniques to effectuate the plan, that would have retained control of the structure of the debt in the Treasury’s hands so that the maturity arrangement that emerged would have reflected conscious policy objectives.”86 80. Gaines (1962, p. 79). 81. Gaines (1962, p. 89). 82. Gaines (1962, p. 77). 83. Gaines (1962, p. 77). See also p. 276 (suggestion to “routinize financing in order to eliminate, so far as possible, all uncertainty as to the maturity of the securities the Treasury will offer in any given financing”) and p. 289 (recommendation for “Treasury to come to market at regular and predictable intervals with offerings of long-term bonds to replace those that are moving down in the maturity structure”). 84. Gaines (1962, p. 79). 85. Gaines (1962, p. 79). In view of his assessment of the ability of Treasury officials to price new issues competitively, it is not surprising that Gaines generally favored auction offerings over fixed-price offerings. See, Gaines (1962, pp. 285 and 292). 86. Gaines (1962, p. 89). Gaines (p. 1) was also critical of those who debated debt management policies without adequately considering the matter of implementation: Public debt management has not been a neglected area of study. A great deal of unsolicited advice has been available to the Treasury from the extensive literature on debt management

356

Chapter 24

The Advent of Regular and Predictable Auction Offerings of Notes and Bonds Advantageous institutional change requires a reasoned understanding of the type of change that can enhance efficiency, but sometimes also requires a precipitating event that provides the impetus necessary to overcome the inertia of past patterns of behavior and the active resistance of those who benefit from the status quo. Friedman and Gaines provided reasoned explanations for why the Treasury should expand its program of regular and predictable auction offerings from bills to notes and bonds. Change, however, had to await the occurrence of precipitating events. The advent of regular and predictable auction offerings of notes and bonds in the 1970s can be traced to rising interest rate volatility and growing budget deficits. The rising volatility of interest rates in the 1960s (table 24.4) made fixed-price offerings increasingly risky for the Treasury—by the time the subscription books closed on an offering, the offering was increasingly likely to be either materially cheap or materially expensive compared to secondary market prices—and ultimately forced Treasury officials to expand the auction Table 24.4 Volatility of Treasury yields Year

3-Year note

5-Year note

Standard deviation of yield change over one business day, basis points 1963 1.32 1.07 1964 1.43 1.07 1965 1.76 1.58 1966 4.59 3.59 1967 4.27 3.96 1968 4.66 4.16 1969 5.46 4.59 1970 6.44 5.89

10-Year bond

0.92 0.84 1.11 3.24 3.31 3.34 4.13 5.53

Source: Computed from data in Federal Reserve Statistical Release H.15.

policies that has blossomed in the postwar years. By and large, however, this literature has been directed toward the broad economic consequences of alternative debt policies or toward developing the hypothetical role of debt management as one of the instruments of public policy. It has generally not recognized that the development of debt management as a positive policy instrument or, for that matter, the mere control of the debt to prevent its being a disturbing influence in the economy, must rest first upon the development of technical debt management procedures that enable the Treasury to control the structure and composition of the debt. The failure to develop techniques that would bring the debt under administrative control has been the most important deficiency in postwar debt management. Until this elementary deficiency is corrected, discussion of the economic policy role of public debt management is of only academic interest.

357

Treasury Debt Management since 1939

process. Growing budget deficits initially forced Treasury officials to come to market with increasing frequency (and increasing unpredictability), but ultimately forced it to regularize its coupon offerings. The Advent of Auction Offerings of Notes and Bonds Paul Volcker was sworn in as Under Secretary of the Treasury for Monetary Affairs on January 27, 1969. Within just a few days, he experienced first-hand the problem with bringing fixed-price offerings in a volatile market. The February 1969 Refunding On Wednesday, January 29, 1969, Under Secretary Volcker announced the terms of the February refunding of $14.5 billion of maturing securities: a multi-option exchange offering of 15-month and 7-year notes.87 Subscription books were set to close on Wednesday, February 5. The new notes were priced none too generously compared to outstanding Treasury notes when they were announced,88 and the refunding looked even less attractive when secondary market yields increased by about 15 basis points before the close of the subscription books. As a result only 64 percent of the $5.4 billion of maturing securities held by the public were exchanged for new securities.89 In subsequent testimony before the Joint Economic Committee, Volcker noted that the offering had “failed to attract much enthusiasm among potential investors.” He suggested that it might be desirable to examine “financing techniques that avoid . . . exposure to market fluctuations.”90 The May 1970 Refunding Fifteen months later, on Wednesday, April 29, 1970, Volcker announced the terms of the May 1970 refunding: a fixed-price cash offering of $3.5 billion of 18-month notes and a multi-option exchange offering of 3-year and 6¾-year 87. Federal Reserve Bank of New York Circular no. 6284, January 29, 1969. Until 1967 the maximum maturity of a Treasury note was, as prescribed in the Victory Liberty Loan Act of 1919, five years. Congress extended the maximum maturity to seven years in the Debt Limit Act of June 30, 1967. It subsequently extended the maximum maturity to ten years (in the Debt Limit Act of March 15, 1976). 88. “Treasury $14.47 Billion Refinancing Has Highest Coupon Rate Since 1865,” Wall Street Journal, January 30, 1969, p. 3 (noting that the 7-year note “is priced fairly tight to older issues in the 7-year range” and that “there were instances in which the new [15-month notes] were offered unofficially . . . at a price . . . well below the [Treasury offering price]”). 89. Federal Reserve Bank of New York Circular no. 6294, February 19, 1969. 90. Statement of Under Secretary of the Treasury Paul Volcker before the Joint Economic Committee, February 19, 1969.

358

Chapter 24

notes for $16.6 billion of maturing securities, $4.9 billion of which was owned by the public.91 The subscription books for the cash offering would be open for one day only, on Tuesday, May 5; the books for the exchange offering were set to close on Wednesday, May 6. In a nationally televised speech on the evening of Thursday, April 30, President Richard Nixon announced that American armed forces had crossed over from South Viet Nam into Cambodia in a large-scale combat operation aimed at eliminating Communist sanctuaries in the latter country.92 By the following Monday, antiwar protests had erupted at dozens of American colleges, four students had been killed by National Guard troops at Kent State University in Ohio, and (in the words of The Wall Street Journal) “the bond markets were battered.”93 Treasury yields were up about 25 basis points since the announcement of the refunding and the refunding was in danger of failing. The Wall Street Journal reported that ”several Government securities specialists thought the Federal Reserve System would consider entering the open market [on Tuesday, May 5] in an attempt to raise prices and lower yields, thereby possibly shoring up the Treasury’s operation.”94 The Fed reacted as expected, buying what was described as “large quantities” of Treasury bills.95 Even so, the Treasury barely covered the cash offering, receiving subscriptions for only $3.6 billion of the 18-month note.96 Contrary to expectations, the Treasury filled all subscriptions in full and left investors with more notes than anticipated.97 To prevent unprecedented attrition on the 91. Federal Reserve Bank of New York Circular no. 6531, April 29, 1970. See also, “Treasury Offers 3 Issues in Refinancing Operation,” New York Times, April 30, 1970, p. 55. 92. “Nixon Sends Combat Forces to Cambodia to Drive Communists from Staging Zone,” New York Times, May 1, 1970, p. 1. 93. “Prices Take Battering from the Shockwaves of Cambodian Invasion,” Wall Street Journal, May 5, 1970, p. 31. 94. “Prices Take Battering from the Shockwaves of Cambodian Invasion,” Wall Street Journal, May 5, 1970, p. 31. 95. “Treasury and Federal Reserve Join to Place $3.5 Billion Securities,” Wall Street Journal, May 6, 1970, p. 31. See also “Reserve Open Market Purchases Came to Aid of Treasury Notes,” New York Times, May 8, 1970, p. 50 (“The Federal Reserve System was forced to make ‘massive’ purchases of securities in the open market to prevent the Treasury’s $3.5 billion sale of notes on Tuesday from failing . . ..”) and “Refunding Cash Runoff Seemingly High at 31.5% But Treasury Satisfied,” Wall Street Journal, May 11, 1970, p. 20 (reporting that “just prior to the Treasury’s offering of its 18-month notes, department officials were said to have used some hard-sell tactics in an effort to influence several major commercial banks to participate. ‘We felt compelled to support the Treasury,’ was the diplomatic explanation from the spokesman of a large New York City bank . . ..”). 96. “Bond Note Issue is Barely Sold,” New York Times, May 8, 1970, p. 50, and “Response to $3.6 Billion U.S. Refunding is Believed Unusually Poor by Dealers,” Wall Street Journal, May 8, 1970, p. 17. 97. “Bond Note Issue is Barely Sold,” New York Times, May 8, 1970, p. 50 (reporting that “Wall Street generally expected allotments [on the cash offering] to fall somewhat between 50 per cent and 70 per cent of subscriptions”), and “Response to $3.6 Billion U.S. Refunding is Believed

359

Treasury Debt Management since 1939

exchange offering (that was still in progress), the Federal Reserve undertook what was described as a “massive” support operation.98 Following the close of the subscription books, the Treasury announced that, in spite of that support, 31.5 percent of the $4.9 billion of publicly held maturing securities had been redeemed for cash.99 A month after the May refunding, The New York Times asked Milton Friedman whether Federal Reserve “rescue operations” for Treasury financings might be distorting monetary policy. The paper reported that Friedman “said the fault in these cases lay in the Treasury’s funding methods. The Treasury should change to issuing all its financing through open auction instead of attempting to pre-assess market rates.”100 The First Note Auction On Thursday, October 22, 1970, Under Secretary Volcker announced the terms of the November 1970 refunding: a multi-option exchange offering of 3½-year and 5¾-year notes for $7.7 billion of maturing securities, $6.0 billion of which was owned by the public.101 The New York Times reported that “many Wall Street bond dealers regarded [the terms of the offering] as not particularly attractive.” One dealer described the rate on the 5¾-year note as “chintzy.”102 Attrition on the maturing debt was expected to run about 15 to 20 percent. Volcker said he would announce a cash offering of short-term securities (to raise new cash and to cover attrition) after the exchange offering had been completed.103 Between the time Volcker announced the exchange offering and the close of the books (on Thursday, October 29), Treasury yields fell about 15 basis points and the offering ended up looking quite attractive. Attrition was limited to 8.5 percent of the publicly owned securities ($652 million).104 On Friday, October 30, Volcker announced the second leg of the refunding: a cash offering of $2 billion of 18-month notes. He further announced that, Unusually Poor by Dealers,” Wall Street Journal, May 8, 1970, p. 17 (reporting the comment of one New York banker that he guessed “orders [for the cash offering] above $200,000 might receive as much as 60% of their desired amount—but 100% never entered my mind”). 98. “Reserve Offers to Buy U.S. Notes,” New York Times, May 9, 1970, p. 35, 99. “Reserve Offers to Buy U.S. Notes,” New York Times, May 9, 1970, p. 35, and “Refunding Cash Runoff Seemingly High at 31.5% But Treasury Satisfied,” Wall Street Journal, May 11, 1970, p. 20. 100. “Friedman Expecting New Money Rein by the Reserve,” New York Times, June 11, 1970, p. 65. 101. Federal Reserve Bank of New York Circular no. 6623, October 22, 1970. 102. “Treasury Notes Register Upturn,” New York Times, October 30, 1970, p. 59. 103. “Treasury to Seek Fresh $4.5 Billion Before Year-End,” Wall Street Journal, October 23, 1970, p. 2. 104. Federal Reserve Bank of New York Circular no. 6633, November 6, 1970.

360

Chapter 24

breaking with tradition, the notes would be sold on an auction basis. Volcker said that auctioning the securities would give the Treasury “a little more flexibility in a variety of market circumstances,” pointing out that market conditions often change between the time an offering is announced and when the subscription books are closed.105 A Treasury press release noted that “this test of the auction technique as a way of selling Treasury notes is part of a continuing effort of the Treasury to develop more efficient debt management techniques.”106 Treasury officials were careful in structuring the first public auction offering of coupon-bearing securities in 35 years.107 Rather than auction long-term bonds from the get-go, they chose to begin by auctioning less-risky short-term notes and they repeatedly reminded market participants of the similarity between the forthcoming auction and the Department’s well-established bill auctions. In its October 30 auction announcement, the Treasury stated that “the use of the auction method of sale represents an adaption of the technique used successfully for many years in marketing Treasury bills.”108 Three days later the Treasury noted that “bidding and other procedures in the Treasury’s new $2 billion cash financing will very closely follow the standard procedure used in regular Treasury bill auctions.”109 Following the close of the auction on November 6 the Treasury announced that it had received tenders for $5.2 billion of notes—2.6 times the amount offered. It accepted bids ranging from 100.93 percent of principal (to yield 6.09 percent) down to a stop-out price of 100.69 (to yield 6.26 percent), where there was a 32 percent pro rata allocation. The average accepted price was 100.76 (to yield 6.21 percent).110 The Wall Street Journal characterized the auction as “highly successful.”111 Subsequent Note and Bond Auctions Treasury officials brought five auction offerings of notes in 1971, with maturities ranging from 15 months to 5 years and 2 months (table 24.5). In a speech 105. “Treasury to Sell in Competitive Auction $2 Billion of 6¾% Notes Due in 1½ Years,” Wall Street Journal, November 2, 1970, p. 3. 106. Federal Reserve Bank of New York Circular no. 6629, October 30, 1970. 107. See “U.S. Is Revamping Funding Method,” New York Times, October 31, 1970, p. 39, and “Treasury to Sell in Competitive Auction $2 Billion of 6¾% Notes Due in 1½ Years,” Wall Street Journal, November 2, 1970, p. 3. However, the Treasury had auctioned bonds to competing syndicates of securities dealers on two occasions in 1963. Garbade (2004a). Those auctions were on an all-or-none basis and were not open to the general public. The first offering went well but the second was less successful and the Treasury did not repeat the experiment a third time. 108. Federal Reserve Bank of New York Circular no. 6629, October 30, 1970. 109. Federal Reserve Bank of New York Circular no. 6631, November 2, 1970. 110. Federal Reserve Bank of New York Circular no. 6633, November 6, 1970. 111. “Treasury Note Sale Is Highly Successful,” Wall Street Journal, November 6, 1970, p. 20.

361

Treasury Debt Management since 1939

Table 24.5 Auction offerings of Treasury notes, 1971 Auction date Jun 22 Aug 5 Aug 31 Oct 15 Nov 9

Term

Quantity offered ($billions)

Quantity bid ($billions)

Range of accepted yields (%)

Average accepted yield (%)

16 mo 18 mo 5 yr, 2 mo 3 yr, 4 mo 15 mo

2.25 2.50 1.25 2.00 2.75

4.0 4.1 3.4 4.6 4.0

5.71–6.05 6.44–6.59 5.92–6.02 5.46–5.61 4.79–4.96

6.00 6.54 5.98 5.58 4.91

Source: Garbade (2004a, tab. 2)

Table 24.6 Auction offerings of Treasury notes and bonds, 1972 Auction date Mar 28 May 2 May 2 Oct 11 Nov 1 Dec 20

Term

Quantity offered ($billions)

Quantity bid ($billions)

Range of accepted yields (%)

Average accepted yield (%)

3 yr 1 yr 9 yr, 9 mo 2 yr 4 yr 2 yr

1.75 1.25 0.50 2.00 3.00 2.00

3.8 3.3 1.3 4.8 7.1 5.6

5.69–5.80 4.23–4.47 6.23–6.32 5.77–5.89 6.16–6.21 5.72–5.85

5.78 4.44 6.29 5.86 6.20 5.83

Source: Garbade (2004a, tab. 3)

to a group of bond market participants on March 7, 1972, Volcker characterized the offerings as a “striking innovation”: I cannot claim that approach has yet been fully tested in adversity. But I can say it has met or surpassed every expectation so far, to the advantage of the Treasury and the market. I am confident it will pass further testing with larger amounts and longer maturities.112

The Treasury brought six additional auction offerings of coupon-bearing securities in 1972 (table 24.6), including a 9¾-year bond in May. The Treasury began auctioning long-term bonds in January 1973. To meet the objection articulated in 1959 by Treasury Secretary Anderson—that many investors who subscribed to fixed-price Treasury offerings did not have the “professional capacity” to bid in auctions—the Treasury implemented a singleprice format in which all tenders bidding at or above the stop-out price were 112. Volcker (1972). Volcker’s speech was reported in “Proposals on Reform of Debt Management Offered by Volcker,” New York Times, March 8, 1972, p. 57, and “Treasury Seeking to Put More Borrowing on Regular Basis, as with Bill Auctions,” Wall Street Journal, March 8, 1972, p. 2.

362

Chapter 24

Table 24.7 Auction offerings of long-term Treasury bonds in a single-price format, 1973 and 1974 Auction date

Term

Quantity offered ($billions)

Quantity bid ($billions)

Yield (%)

Jan 4, 1973 May 2, 1973 Aug 1, 1973 Oct 1, 1973 Feb 7, 1974 May 8, 1974

20 25 20 19 19 25

0.63 0.65 0.50 0.30 0.30 0.30

1.7 1.2 0.3 1.3 1.1 0.9

6.79 7.11 8.00 7.35 7.46 8.23

yr yr yr yr, 9 mo yr, 6 mo yr

Source: Garbade (2004a, tab. 4)

awarded securities at the stop.113 The Treasury observed that “this procedure will provide an incentive to bid at prices sufficiently high to be sure of awards, while also assuring each bidder that, if he bids at a price within the range of accepted bids, he will be awarded bonds at the same price as every other bidder.”114 Over the next fifteen months the Treasury offered long-term bonds in single-price auctions five more times (table 24.7). (As described in the appendix to this chapter, the Treasury switched its long-term bond auctions to a multiple-price format in mid-1974 and subsequently moved all of its auction offerings to a single-price format in the 1990s.) The last fixed-price Treasury offerings came in 1976.115 Since then, all offerings of marketable Treasury debt have been sold on an auction basis. Why Did Note and Bond Auctions Succeed in the 1970s? The Treasury’s success in institutionalizing auction sales of notes and bonds in the early 1970s is remarkable because the initiative came at a time when interest rates had become more volatile and auction bidding concomitantly more risky. That the Treasury even made the attempt testifies to the enduring 113. Friedman (1960, p. 65) had recommended the single-price format as a way to avoid limiting the primary market to professional dealers. 114. Federal Reserve Bank of New York Circular no. 7071, December 27, 1972. In March 1972, Volcker had hinted at the possibility of auctioning long-term securities in a single-price format, saying that “we are prepared to explore further variants [of the auction process], including (as the maturity is extended) the possibility of awarding all bids at the stop-out price to encourage wider investor participation.” Volcker (1972). 115. The three fixed-price offerings in 1976, all of which were for cash, included 7-year notes in the February refunding ($3.5 billion offered, $29.2 billion subscribed, $6.0 billion sold), 10-year notes in the May refunding ($3.5 billion offered, $8.9 billion subscribed, $4.7 billion sold), and 10-year notes in the August refunding ($4.0 billion offered, $24.4 billion subscribed, $7.6 billion sold). These offerings were made on a fixed-price basis because Treasury Secretary William Simon believed that “given the absorptive capacity of the market, auctions of much more than $2.5 billion at one time result in disproportionately high interest costs.” Committee on Ways and Means (1976, p. 9). The latest preceding fixed-price cash offering was in the August 1970 refunding. The last fixed-price exchange offering was in the February 1973 refunding.

363

Treasury Debt Management since 1939

significance of Friedman’s criticism of fixed-price offerings. At the same time, however, the failed 1935 attempt to auction coupon-bearing Treasury securities stood as a reminder that the advantages of market-driven auctions did not guarantee success: the process of introducing auction offerings also had to be carefully managed. There were two important differences between the auctions of couponbearing securities in the early 1970s and the 1935 auctions. First, the auctions of the early 1970s started with short-term issues and were extended gradually to longer maturities. Unlike the case in 1935, the Treasury did not immediately auction long-term bonds. Second, the auctions of the early 1970s were closely patterned on the successful and familiar bill auctions and did not introduce novel issuance patterns (as in 1935). These two features gave dealers an opportunity to build up their sales and risk management programs in a familiar setting. The Advent of Regular Offerings of Notes and Bonds During the 1960s the Treasury regularly sold notes and bonds in the second month of each calendar quarter, for settlement on the fifteenth of the month, either for cash or in exchange for maturing debt.116 It also sold notes and bonds on an as-needed basis in stand-alone cash offerings to replenish its cash balances. As shown in figure 24.3, in the first half of the 1960s the maturities of the mid-quarter offerings were quite unpredictable, sometimes limited to a single short-term certificate or note (when Treasury officials wanted to maintain upward pressure on short-term interest rates for balance-of-payments reasons, as in February 1961, August and November 1963, and November 1964) and sometimes including bonds with as much as thirty years to maturity (when officials wanted to mitigate further contraction of the maturity structure of the debt, as in August 1962).117 Beginning in August 1965, maturities became more predictable, but only because secondary market bond yields exceeded 4.25 percent and the Treasury lacked statutory authority to issue bonds with a coupon rate in excess of 4¼ percent.118 This limited the Treasury to selling 116. Regular mid-quarter offerings were introduced in the mid-1950s as a way to reduce interference with the Federal Reserve’s open market operations and with other borrowers. Federal Reserve Bank of New York Circular no. 4663, November 18, 1958 (“For some time, the Treasury has been working toward scheduling its maturities on [the mid-quarter dates] to reduce the number of times each year its financing will interfere with other borrowers such as corporations, states, municipalities, etc.; to minimize the ‘churning’ in the money markets on the major quarterly corporate income tax dates; and to facilitate the effective execution by the Federal Reserve of its monetary policy.”). 117. Garbade (2007, tab. 1). 118. Until 1971, the interest rate on a Treasury bond was, as provided by the Third Liberty Bond Act of 1918, limited to 4¼ percent. In 1971, Congress provided (in the Debt Limit Act of March 17, 1971) that up to $10 billion of bonds could be issued at higher interest rates. The exemption was expanded several times and the 4¼ percent ceiling was abolished altogether by the Technical and Miscellaneous Revenue Act of November 10, 1988.

Chapter 24

30 25 20

Years

364

15 10 5 0 1960

1962

1964

Mid-quarter refunding

1966

1968

1970

Stand-alone cash offering

Figure 24.3 Term to maturity of new offerings of coupon-bearing Treasury debt

certificates of indebtedness (with a maximum maturity of one year) and notes (with a maximum maturity of five years). (In June 1967 Congress extended the maximum maturity of a note to seven years,119 and the Treasury began issuing at that maturity in November 1967.) Also as shown in figure 24.3, stand-alone offerings were relatively rare. The Treasury brought only eleven such offerings in the 1960s. The Introduction of Two-Year Cycle Notes Between 1970 and 1972, mid-quarter exchange offerings exhibited unusually high attrition. In contrast to the 10 percent attrition that was considered normal in the 1960s,120 the average rate of attrition in nine exchange offerings between February 1970 and February 1972 was 24.3 percent. The high attrition forced the Treasury to rebuild its cash balances by issuing additional securities, including a total of $7.25 billion of new notes in four stand-alone offerings between June 1971 and April 1972.121 119. Act of June 30, 1967. 120. The average rate of attrition for all exchange offerings in mid-quarter exchange offerings between February 1960 and November 1969 was 13.4 percent. The average falls to 9.8 percent if the refundings in the third quarter of 1964 and the first quarter of 1965 are excluded. Those two refundings were accelerated several weeks, to mid-July and mid-January, respectively, and were not representative of other refundings. They had attrition rates of 67.0 and 54.5 percent, respectively. 121. Garbade (2007).

365

Treasury Debt Management since 1939

The spate of stand-alone offerings led Treasury officials to begin to think about regularizing coupon offerings outside of the mid-quarter refundings. In March 1972, Under Secretary Volcker revealed that he was considering whether “to routinize or regularize the handling of more of our debt, as we have done for many years in the bill area.”122 In particular, he was considering whether, “in contrast to building up the present concentration of note and bond maturities at quarterly intervals [on the 15th of the second month of each quarter], to be [refinanced] flexibly at the Treasury’s discretion at maturity,” it might not be better to adopt a scheme of “more frequent but also more routine rolling over of relatively short-term notes.” Such a scheme might “reduce market uncertainties . . . caused by large intermittent financing operations.” Treasury officials took the first step toward putting short-term note offerings on a regular schedule when they announced in early October 1972 that they would soon begin to auction 2-year notes at regular quarterly intervals. The first tranche—$2 billion of notes maturing on September 30, 1974—was auctioned on October 11.123 The 2-year note program broke new ground in two ways: it was the first program of regular and predictable sales of coupon-bearing securities with a specified term to maturity and it broke the pattern of coupon-bearing securities maturing on the fifteenth of the second month of a quarter. Debt management officials put 2-year notes on a self-sustaining cycle separate and apart from the mid-quarter refundings that had dominated Treasury finance in the 1960s. The introduction of 2-year cycle notes put short-term note sales on a regular schedule but did not signal that longer term notes and bonds would soon be sold on a similarly regular and predictable basis. Volcker commented in March 1972 that “regularization and routinization are nice sounding words; straightjacket and rigidity are not. From where I sit, I cannot help but be conscious of the number of times in which particular market or economic objectives may influence the Treasury’s thinking as to the form of a particular financing.”124 Treasury officials were not yet ready to abandon tactical discretion. Embracing Regular and Predictable Issuance as a Debt Management Strategy The rate of growth of marketable Treasury debt increased dramatically in fiscal year 1975. Outstanding notes and bonds increased by $25 billion between mid-1974 and mid-1975, an increase substantially in excess of increases in prior fiscal years. The rapid expansion led Treasury officials to regularize note sales beyond the two-year sector. 122. Volcker (1972). 123. Federal Reserve Bank of New York Circular no. 7013, October 5, 1972. 124. Volcker (1972).

366

Chapter 24

Forecasts of the federal budget deficit deteriorated significantly during the winter of 1974–75. In November 1974, officials estimated that the deficit for the fiscal year ending June 30, 1975, would be about $9 billion125 and that the deficit for fiscal year 1976 would be $10 to $20 billion.126 By March 1975, the deficit projections had grown to $45 billion and $80 billion, respectively.127 The fivefold growth in the two-year deficit (from $25 billion to $125 billion) meant that the Treasury would have to raise an unprecedented (for a peacetime economy) amount of new money. As early as December 1974, economists were predicting that stand-alone offerings would be made in nearly every month of the coming year.128 The Treasury ultimately brought a total of nine such offerings in fiscal 1975,129 more than double the previous record of four stand-alone offerings in fiscal 1972. Treasury officials struggled to cope with the expanding financing requirements. In January 1975 they announced an offering of 2-year notes outside of the quarterly cycle established in 1972. Under Secretary of the Treasury Jack Bennett (who had replaced Volcker on July 9, 1974) stated that “in the coming months, we will be studying the possibility of establishing regular month-end, rather than quarter-end, two-year cycle notes.130 Officials confirmed the new monthly frequency in April.131 Despite their efforts, Treasury officials soon reached the limit of what could be accommodated within the existing debt management framework. On March 20, 1975, the Treasury auctioned $1.25 billion of 15-year bonds at the same time that an underwriting syndicate led by Morgan Stanley & Co. brought to market the largest industrial debt offering in history: $300 million of 10-year notes and $300 million of thirty-year debentures from AAA-rated General Motors Corporation. The simultaneous offerings left the bond market in “chaos.” One dealer described the market as “a disaster,” another said it was “a shambles,” and The New York Times later reported that the “head-on competition between the most credit-worthy borrowers from the public and private 125. “Estimate of Fiscal ’75 Deficit Raised by Ford Aides as Recession Cuts Revenues,” Wall Street Journal, November 21, 1974, p. 2. 126. “Fiscal ’76 Budget Deficit is Now Likely, in a Range of $10 Billion to $20 Billion,” New York Times, November 11, 1974, p. 3. 127. “$37-Billion Rise in Deficit Is Seen,” New York Times, March 18, 1975, p. 15, and Committee on the Budget (1975, pp. 996, 1030, and 1033) (testimony of Secretary of the Treasury William Simon). 128. “Treasury Plans Big Borrowings,” New York Times, December 30, 1974, p. 39. 129. Garbade (2007). 130. Committee on Ways and Means (1975, p. 16) (transcript of January 22, 1975, news conference on Treasury financing plans). 131. “Official of Treasury Discloses Need for $41-Billion,” New York Times, April 1, 1975, p. 62.

367

Treasury Debt Management since 1939

sectors left the bond market in disarray.”132 The chairman of the Joint Economic Committee, Senator Hubert Humphrey (Democrat, Minnesota), criticized Treasury debt management as “being conducted in an inexplicable and seemingly highly inappropriate fashion.”133 The deficit had to be financed, but Treasury officials appreciated that closely spaced stand-alone offerings and head-to-head competition with private-sector borrowers could be reduced by replacing stand-alone sales with regular and predictable offerings. In June 1975 they announced $1.75 billion of 4-year notes that “might be the first of a ‘cycle’ of 4-year notes maturing at the end of a quarter.”134 “Might” turned to “would” when they announced a second tranche of 4-year notes in August.135 Four months later Under Secretary of the Treasury Edwin Yeo (who had replaced Bennett on August 5, 1975) announced that officials were “seriously considering” adopting a new series of 5-year notes.136 By mid-1976 the Treasury was issuing 2-year notes monthly and 4-year and 5-year notes on a regular quarterly basis. By 1982 the Treasury had added a 7-year note series and a 20-year bond series and had standardized mid-quarter refundings with regular offerings of 3-year and 10-year notes and 30-year bonds. Since 1980 the Treasury has offered coupon-bearing securities outside of a regular and predictable framework on only two occasions, in October 2001 and October 2008, when settlement fails threatened to impair the liquidity of secondary market trading.137 The Enduring Value of Regular and Predictable Issuance During the 1970s Treasury officials changed the framework within which they made debt management decisions, moving from tactical issuance of notes and bonds to a regular and predictable schedule. The emergence of regular and predictable issuance reduced the element of surprise in Treasury debt operations and allowed investors to plan future commitments of funds with greater confidence. 132. “Treasury Bond Auction Creates Chaos; Supply of Money Shows a Record Rise,” New York Times, March 21, 1975, p. 53, and “Financier for the U.S. Debt,” New York Times, April 20, 1975, p. F7. 133. “Financier for the U.S. Debt,” New York Times, April 20, 1975, p. F7. 134. “2 New Notes, More Bills but No Long-Term Issue,” New York Times, June 19, 1975, p. 63. 135. “Treasury Boosts Earlier Estimate of Its Cash Needs,” Wall Street Journal, August 7, 1975, p. 3 (referring to the second tranche of 4-year notes as “the second four-year cycle note”). 136. “Treasury Plans Heavy Borrowing,” New York Times, January 28, 1976, p. 58, and “Treasury to Sell $13.8 Billion Bills, Notes and Bonds,” Wall Street Journal, January 28, 1976, p. 25. 137. Fleming and Garbade (2002) and Garbade et al. (2010).

368

Chapter 24

Treasury officials have repeatedly asserted the advantages of regular and predictable issuance. In 1982, Deputy Assistant Secretary Mark Stalnecker expressed the view that “regularity of debt management removes a major source of market uncertainty, and assures that Treasury debt can be sold at the lowest possible interest rate consistent with market conditions at the time of sale.”138 In 1998, Assistant Secretary Gary Gensler observed that “consistency and predictability in [the Treasury’s] financing program . . . reduces uncertainty in the market and helps minimize our overall cost of borrowing.”139 And in 2002, Under Secretary Peter Fisher stated that “the Treasury’s continuing commitment to a schedule of regular and predictable auction dates is a means, over time, to the end of lowest cost borrowing.”140 Regular and predictable issuance was not a novel concept in the 1970s: the Treasury had been issuing bills regularly and predictably for decades. Nevertheless, debt managers had kept note and bond offerings on a tactical basis—in part because financing at least cost was not the only objective of Treasury debt management. (Debt managers sometimes chose to issue short-term debt to maintain upward pressure on short-term interest rates and sometimes chose to issue longer term debt to limit contraction in the maturity structure of the debt.) Regular and predictable issuance of notes and bonds became more attractive after Treasury officials had to bring four stand-alone cash offerings in fiscal year 1972 as a result of unusually high attrition in mid-quarter refundings. They introduced 2-year cycle notes to put short-term note financings on a more routine basis. The much larger and more significant need to fund a rapid expansion of the deficit in 1975 led them to phase in additional cycle notes and, ultimately, to abandon tactical issuance altogether. Appendix: The Death and Rebirth of the Single-Price Auction Format The single-price auction format that the Treasury introduced in 1973 for offerings of long-term bonds was not popular with Treasury dealers. Henry Kaufman, a well-known economist at Salomon Brothers, bluntly stated that the format “provides no incentives to . . . dealers to help in the distribution process.”141 Treasury officials nevertheless persisted in using the single-price format through May 1974. In response to complaints that the format deprived dealers of an opportunity to buy bonds slightly cheaper than other auction 138. Committee on Banking, Finance, and Urban Affairs (1982, p. 5). 139. Testimony before the House Committee on Ways and Means, June 24, 1998. Gensler went on to note that “in keeping with this principle, Treasury does not seek to time markets; that is, we do not act opportunistically to issue debt when market conditions appear favorable.” 140. Remarks to the Futures Industry Association, Boca Raton, Florida, March 14, 2002. 141. Kaufman (1973, p. 170).

369

Treasury Debt Management since 1939

participants, one official pointed out that “the objective is to encourage widespread and confident bidding,” and a “broader distribution of our securities.”142 “We’re appealing to a type of investor,” the official said, “who will be able to bid what he thinks the bond is worth to him without worrying about whether somebody else may get it cheaper.” On May 8, 1974, Deputy Secretary of the Treasury William Simon, formerly a senior partner at Salomon Brothers in charge of the government and municipal bond departments and past president of the Association of Primary Dealers in U.S. Government Securities,143 was sworn in as Secretary of the Treasury.144 In the very next refunding, in August 1974, the Treasury switched to a multiple-price format in auctioning long-term bonds. The Treasury did not state publicly the reason for the change, although one money market newsletter reported at the time that “debt managers found no evidence that [the singleprice format] was attracting enough additional or different bidders for the bonds to make its use worthwhile.”145 The then-Under Secretary of the Treasury for Monetary Affairs, Jack Bennett, subsequently testified that Secretary Simon “made the decision to discontinue the [single-price format] as a result of his judgment, based on his extensive experience in the market for Treasury securities, that the [single-price format] would bring in fewer dollars to the Treasury.”146 142. “Price of Treasury Bonds Decline in Light Trading,” New York Times, December 29, 1972, p. 39. 143. The Association of Primary Dealers in U.S. Government Securities was organized in the spring of 1969 “to foster high standards of commercial honor and business conduct” in the Treasury securities market following several instances of “improper” activities. See “U.S. Bond Market is Troubled,” New York Times, February 16, 1969, p. F1 (reporting a letter from Alan Holmes, Senior Vice President of the Federal Reserve Bank of New York and Manager of the System Open Market Account, expressing concern about “the adequacy of procedures used by dealer firms to guard against the possibility of improper activity on the part of any employee”), “Dealers Weigh Forming Group,” New York Times, April 16, 1969, p. 55, and “Dealers in Securities of U.S. Government Form an Association,” Wall Street Journal, April 24, 1969, p. 25. 144. See “Dent Named Commerce Chief as Expected; Simon of Salomon Bros. No. 2 in Treasury,” Wall Street Journal, December 7, 1972, p. 3, and “Simon Nominated as Treasury Secretary; Nixon to Expand His Own Economic Role,” Wall Street Journal, April 18, 1974, p. 3. 145. The Goldsmith–Nagan Bond and Money Market Letter, August 3, 1974. 146. Committee on Banking, Housing, and Urban Affairs (1991, p. 409). Two papers, Tsao and Vignola (1977) and Simon (1994), examine whether the Treasury received more aggressive bids in the six single-price auctions or the ten multiple-price auctions of long-term bonds held between February 1973 and August 1976. Neither paper comments on why the Treasury abandoned the single-price format. In 1982 Deputy Assistant Secretary of the Treasury Mark Stalnecker testified that the Treasury “analyzed or . . . tried to do some studies on the six [single-price] auctions that we held back in the mid-seventies and the results were inconclusive. It did not appear that there were significant cost savings and frankly our view is that we receive enough bids under our current auction mechanism, and it is well received by both investors and market professionals, so that after selling six securities by the [single-price] auction mechanisms with mixed results we ended that experiment.” Committee on Banking, Finance and Urban Affairs (1982, p. 24). Chari and Weber (1992, p. 4) state that the Treasury “abandoned the experiment [with single-price auctions] as largely inconclusive,” but do not provide a source.

370

Chapter 24

Treasury auction processes came under intense scrutiny in 1991 following several violations of Treasury auction rules by employees of Salomon Brothers.147 In the course of the ensuing investigations and congressional hearings, public officials became interested in alternative auction formats that might appeal to more investors and that might lead to lower financing costs, including the single-price format. Whether the Treasury would be better off selling securities in a single-price format or a multiple-price format was a matter that could only be resolved by empirical analysis.148 In September 1992 the Treasury announced that, in an experiment, it would begin to auction 2-year and 5-year notes in a single-price format.149 It subsequently produced two empirical studies analyzing the results of the experiment.150 Although the results were not unambiguous, the Treasury decided in October 1998 that they justified extending the single-price format to all auction offerings.151 147. See Committee on Banking, Housing, and Urban Affairs (1991), Department of the Treasury, Securities and Exchange Commission, and Board of Governors of the Federal Reserve System (1992), Jegadeesh (1993), and Jordan and Jordan (1996). 148. Department of the Treasury, Securities and Exchange Commission, and Board of Governors of the Federal Reserve System (1992, p. 14). See, similarly, Adams (1887, p. 169) (“The only financial defense for advertising to equalize bids, lies in the suggestion that the amount which a government must pay for its capital will be less when treating its creditors in a body than when accepting or rejecting each offer by itself. . . . [T]here has been no attempt to place a loan under [the single-price auction format] which tests its merits in a satisfactory manner”). 149. “New Process for Auctions to be Tested,” New York Times, September 4, 1992, p. D1, and “Treasury to Try ‘Dutch’ System at Its Auctions,” Wall Street Journal, September 4, 1992, p. C1. 150. Malvey, Archibald, and Flynn (1995) and Malvey and Archibald (1998). 151. “Dutch-Auction Format to be Adopted by the Treasury in More Sales of Issues,” Wall Street Journal, October 27, 1998, p. A24.

References

Adams, Henry. 1887. Public Debts; An Essay in the Science of Finance. New York: Appleton. Adams, Thomas. 1917. Customary war finance. New Republic (April 7): 292–94. Adams, Thomas. 1921a. Should the excess profits tax be repealed. Quarterly Journal of Economics 35 (3): 363–93. Adams, Thomas. 1921b. Fundamental problems of federal income taxation. Quarterly Journal of Economics 35 (4): 527–56. Ahamed, Liaquat. 2009. Lords of Finance: The Bankers Who Broke the World. New York: Penguin Press. Alter, Jonathan. 2006. The Defining Moment: FDR’s Hundred Days and the Triumph of Hope. New York: Simon and Schuster. Altmeyer, Arthur. 1966. The Formative Years of Social Security. Madison: University of Wisconsin Press. Amar, Akhil. 2005. America’s Constitution: A Biography. New York: Random House. Anderson, Frank. 1917. Fundamental Factors in War Finance. Journal of Political Economy 25 (9): 857–87. Andrew, A. Piatt. 1906. The influence of the crops upon business in America. Quarterly Journal of Economics 20 (3): 323–52. Andrew, A. Piatt. 1907. The Treasury and the banks under Secretary Shaw. Quarterly Journal of Economics 21 (4): 519–68. Andrew, A. Piatt. 1908a. The United States Treasury and the money market. The partial responsibility of Secretaries Gage and Shaw for the crisis of 1907. American Economic Association Quarterly (3rd series) 9, no. 1, Papers and Discussions of the Twentieth Annual Meeting (April): 218–31. Andrew, A. Piatt. 1908b. Substitutes for cash in the panic of 1907. Quarterly Journal of Economics 22 (4): 497–516. Awalt, Francis. 1969. Recollections of the banking crisis in 1933. Business History Review 43 (3): 347–71. Baicker, Kathrine, Claudia Goldin, and Lawrence Katz. 1998. A distinctive system: Origins and impact of U.S. unemployment compensation. In Michael Bordo, Claudia Goldin, and Eugene White, eds., The Defining Moment: The Great Depression and the American Economy in the Twentieth Century. Chicago: University of Chicago Press. Ballantine, Arthur. 1948. When all the banks closed. Harvard Business Review 26 (2): 129–43. Bank, Steven, Kirk Stark, and Joseph Thorndike. 2008. War and Taxes. Washington, DC: Urban Institute Press. Barnes, James. 1931. John G. Carlisle: Financial Statesman. New York: Dodd, Mead. Beckhart, Benjamin. 1924. The Discount Policy of the Federal Reserve System. New York: Henry Holt. Beckhart, Benjamin, James Smith, and William Brown. 1932. The New York Money Market: External and Internal Relations, vol. 4. New York: Columbia University Press. Biais, Bruno, and Richard Green. 2005. The microstructure of the bond market in the 20th century. Unpublished paper. Toulouse University and Carnegie Mellon University. Blakey, Roy. 1914. The new income tax. American Economic Review 4 (1): 25–46. Blakey, Roy. 1916. The new Revenue Act. American Economic Review 6 (4): 837–50. Blakey, Roy. 1917. The War Revenue Act of 1917. American Economic Review 7 (4): 791–815. Blakey, Roy. 1918. Shifting the war burden upon the future. In E. M. Patterson, ed., Financing the War: Annals of the American Academy of Political and Social Science 75 (January): 90–104. Blakey, Roy. 1922. The Revenue Act of 1921. American Economic Review 12 (1): 75–108. Blakey, Roy. 1924. The Revenue Act of 1924. American Economic Review 14 (3): 475–504. Blakey, Roy. 1926. The Revenue Act of 1926. American Economic Review 16 (3): 401–25. Blakey, Roy. 1928. The Revenue Act of 1928. American Economic Review 18 (3): 428–48.

372

References

Blakey, Roy, and Gladys Blakey. 1932. The Revenue Act of 1932. American Economic Review 22 (4): 620–40. Blakey, Roy, and Gladys Blakey. 1937. The Revenue Act of 1937. American Economic Review 27 (4): 698–704. Blakey, Roy, and Gladys Blakey. 1938. The Revenue Act of 1938. American Economic Review 28 (3): 447–58. Blakey, Roy, and Gladys Blakey. 1939. Federal Tax Legislation, 1939. American Economic Review 29 (4): 695–707. Blakey, Roy, and Gladys Blakey. 1940. The Federal Income Tax. New York: Longmans, Green. Bloomfield, Arthur. 1959. Monetary Policy under the International Gold Standard: 1880–1914. New York: Federal Reserve Bank of New York. Board of Governors of the Federal Reserve System. 1943. Banking and Monetary Statistics. Washington, DC: Board of Governors of the Federal Reserve System. Board of Governors of the Federal Reserve System.1976. Banking and Monetary Statistics: 1941–1970. Washington, DC: Board of Governors of the Federal Reserve System. Bogart, Ernest, Charles Bullock, Fred Fairchild, Henry Gardner, Robert Haig, Jacob Hollander, Edwin Kemmerer, Alexander Noyes, Carl Plehn, William Scott, H. Parker Willis, and Edwin Seligman. 1919. Report of the Committee on War Finance. American Economic Review 9 (1, suppl 2): 1–128. Broesamle, John. 1973. William Gibbs McAdoo. Port Washington, NY: Kennikat Press. Brown, E. Cary. 1956. Fiscal policy in the ‘thirties’: A reappraisal. American Economic Review 46 (5): 857–79. Brownlee, W. Elliot. 2004. Federal Taxation in America: A Short History, 2nd ed. Cambridge: Cambridge University Press. Brunner, Robert, and Sean Carr. 2007. The Panic of 1907: Lesson’s Learned from the Market’s Perfect Storm. New York: Wiley. Bullock, Charles. 1917. Financing the war. Quarterly Journal of Economics 31 (3): 357–79. Cannadine, David. 2006. Mellon: An American Life. New York: Knopf. Cannon, James. 1910. Clearing Houses. National Monetary Commission. Senate Document 491. U.S. Senate, 61st Cong., 2nd Sess. Carosso, Vincent. 1987. The Morgans: Private International Bankers, 1854–1913. Cambridge: Harvard University Press. Carson, Deane. 1959. Treasury open market operations. Review of Economics and Statistics 41 (4): 438–42. Cecchetti, Stephen. 1988. The case of negative nominal interest rates: New estimates of the term structure of interest rates during the Great Depression. Journal of Political Economy 96 (6): 1111–41. Chandler, Lester. 1958. Benjamin Strong, Central Banker. Brookings Institution: Washington, DC. Chandler, Lester. 1971. American Monetary Policy, 1928–1941. New York: Harper and Row. Chapman, John. 1923. Fiscal Function of the Federal Reserve Banks. New York: Ronald Press. Chari, V., and Robert Weber. 1992. How the U.S. Treasury should auction its debt. Federal Reserve Bank of Minneapolis Quarterly Review 16 (4): 3–12. Chernow, Ron. 1990. The House of Morgan: An American Banking Dynasty and the Rise of Modern Finance. New York: Simon and Schuster. Childs, Charles. 1947. Concerning U.S. Government Securities. Chicago: C.F. Childs. Committee on Banking and Currency. 1913. Changes in the banking and currency system of the United States. Report 69, U.S. Senate, 63rd Cong., 1st Sess., September 9. Committee on Banking and Currency. 1926. Stabilization. Hearings before the Committee on Banking and Currency on H.R. 7895, U.S. House of Representatives, 69th Cong., 1st Sess., March 24, 25, 30, 31, April 1, 6, 8, 9, 12, 13, and 14.

373

References

Committee on Banking and Currency. 1926 and 1927. Stabilization. Hearings before the Committee on Banking and Currency on H.R. 7895, U.S. House of Representatives, 69th Cong., 1st Sess., April 20, 21, 22, 27, 30, May 3, 4, 5, 6, June 10, 1926, and February 4, 1927. Committee on Banking and Currency. 1931. Operation of the National and Federal Reserve Banking Systems. Hearings before a Subcommittee of the Committee on Banking and Currency Pursuant to S. Res. 71, Appendix, Part 6, Federal Reserve Questionnaires, U.S. Senate, 71st Cong., 3rd Sess. Committee on Banking and Currency. 1960. Federal Reserve direct purchases—Old Series Currency Adjustment Act. Hearing before the Committee on Banking and Currency on S. 3702 and S. 3714, U.S. House of Representatives, 86th Cong., 2nd Sess., June 24. Committee on Banking, Finance and Urban Affairs. 1982. Problems associated with Federal debt management. Hearings before the Subcommittee on Domestic Monetary Policy of the Committee on Banking, Finance and Urban Affairs. U.S. House of Representatives, 97th Cong., 2nd Sess., March 23 and 24. Committee on Banking, Housing, and Urban Affairs. 1991. The activities of Salomon Brothers, Inc. in Treasury auctions. Hearings before the Subcommittee on Securities of the Committee on Banking, Housing, and Urban Affairs. U.S. House of Representatives, 102nd Cong., 1st Sess., September 11 and 12. Committee on the Budget. 1975. The 1976 First concurrent resolution on the budget. Hearings before the Committee on the Budget, U.S. Senate, 94th Congress, 1st Sess., March 11–14, 17, and 21. Committee on Finance. 1935. Amend Second Liberty Bond Act, As Amended, Etc. Report 32, U.S. Senate, 74th Cong., 1st Sess., January 21. Committee on Finance. 1935. To amend the Second Liberty Bond Act. Hearing before the Committee on Finance on H.R. 4304, U.S. Senate, 74th Cong., 1st Sess., January 29. Committee on Finance. 1939. Amending Second Liberty Bond Act, As Amended, Report 501, U.S. Senate, 76th Cong., 1st Sess., May 26. Committee on Ways and Means. 1917. Second Emergency Bond Issue. Hearing before the Committee on Ways and Means on H.R. 5901, U.S. House of Representatives, 65th Cong., 1st Sess., August 28 and 29. Committee on Ways and Means. 1918. Third Liberty Bond Bill. Hearings before the Committee on Ways and Means, U.S. House of Representatives, 65th Cong., 2nd Sess., March 27. Committee on Ways and Means. 1918. Supplementary Bond Legislation. Hearing before the Committee on Ways and Means, U.S. House of Representatives, 65th Cong., 2nd Sess., September 12. Committee on Ways and Means. 1919. Fifth Liberty Bond Bill. Hearings before the Committee on Ways and Means, U.S. House of Representatives, 65th Cong., 3rd Sess., February 13 and 14. Committee on Way and Means. 1920. Letter from the Secretary of the Treasury. House Document 905, U.S. House of Representatives, 66th Cong., 3rd Sess. Committee on Ways and Means. 1929. Issuance of Treasury bills. Hearings before the Committee on Ways and Means, U.S. House of Representatives, 71st Cong., 1st Sess., May 29. Committee on Ways and Means. 1935. Amend Second Liberty Bond Act, As Amended, Etc. Report 22, U.S. House of Representatives, 74th Cong., 1st Sess., January 23. Committee on Ways and Means. 1939. Amending Second Liberty Bond Act, As Amended. Report 610, U.S. House of Representatives, 76th Cong., 1st Sess., May 10. Committee on Ways and Means. 1975. Public debt ceiling. Hearings before the Committee on Ways and Means. U.S. House of Representatives, 94th Cong., 1st Sess., June 2–3. Committee on Ways and Means. 1976. Administration request for increase in debt ceiling. Hearing Before the Committee on Ways and Means. U.S. House of Representatives, 94th Cong., 2nd Sess., February 17. Cooke, H., and M. Katzen. 1954. The public debt limit. Journal of Finance 9 (3): 298–303. Davis, Andrew. 1910. The origin of the National Banking System. National Monetary Commission. Senate Document 582. U.S. Senate, 61st Cong., 2nd Sess.

374

References

Department of the Treasury, Securities and Exchange Commission, and Board of Governors of the Federal Reserve System. 1992. Joint Report on the Government Securities Market. Derthick, Martha. 1979. Policymaking for Social Security. Washington, DC: Brookings Institution. Dewey, Davis. 1934. Financial History of the United States. 12th ed. New York: Longmans Green. Dickson, Paul, and Thomas Allen. 2004. The Bonus Army: An American Epic. New York: Walker. Durand, E. Dana. 1917. Taxation versus bond issues for financing the war. Journal of Political Economy 25 (9): 888–916. Ebersole, J. Franklin. 1933. One year of the Reconstruction Finance Corporation. Quarterly Journal of Economics 47 (3): 464–92. Federal Reserve Bank of New York. 1940. The place of the dealer in the government security market. Unpublished paper. Federal Reserve Bank of New York. Fleming, Michael, and Kenneth Garbade. 2002. When the back office moved to the front burner: Settlement fails in the Treasury market after 9/11. Federal Reserve Bank of New York. Economic Policy Review 8 (2): 35–57. Friedman, Milton. 1960. A Program for Monetary Stability. New York: Fordham University Press. Friedman, Milton, and Anna Schwartz. 1963. A Monetary History of the United States, 1867– 1960. Princeton: Princeton University Press. Gaines, Tilford. 1962. Techniques of Treasury Debt Management. New York: Graduate School of Business, Columbia University/Free Press of Glencoe. Garbade, Kenneth. 2004a. The institutionalization of Treasury note and bond auctions, 1970–75. Federal Reserve Bank of New York. Economic Policy Review 10 (1): 29–45. Garbade, Kenneth. 2004b. Origins of the Federal Reserve book-entry system. Federal Reserve Bank of New York. Economic Policy Review 10 (3): 33–50. Garbade, Kenneth. 2006. The evolution of repo contracting conventions in the 1980s. Federal Reserve Bank of New York. Economic Policy Review 12 (1): 27–42. Garbade, Kenneth. 2007. The emergence of “regular and predictable” as a Treasury debt management strategy. Federal Reserve Bank of New York. Economic Policy Review 13 (1): 53–71. Garbade, Kenneth. 2008. Why the U.S. Treasury began auctioning Treasury bills in 1929. Federal Reserve Bank of New York. Economic Policy Review 14 (1): 31–47. Garbade, Kenneth, and Jeff Ingber. 2005. The Treasury auction process: Objectives, structure, and recent adaptations. Federal Reserve Bank of New York. Current Issues in Economics and Finance 11 (2, February). Garbade, Kenneth, Frank Keane, Lorie Logan, Amanda Stokes, and Jennifer Wolgemuth. 2010. The introduction of the TMPG fails charge for U.S. Treasury securities. Federal Reserve Bank of New York. Economic Policy Review 16 (2): 45–71. Garbade, Kenneth, John Partlan, and Paul Santoro. 2004. Recent innovations in Treasury cash management. Federal Reserve Bank of New York. Current Issues in Economics and Finance 10 (11, November). Garbade, Kenneth, and William Silber. 1979. The payment system and domestic exchange rates: Technological versus institutional change. Journal of Monetary Economics 5 (1): 1–22. Glass, Carter. 1927. An Adventure in Constructive Finance. Garden City, NY: Doubleday, Page. Goodhart, C. A. E. 1969. The New York Money Market and the Finance of Trade, 1900–1913. Cambridge: Harvard University Press. Hackley, Howard. 1973. Lending Functions of the Federal Reserve Banks: A History. Washington, DC: Board of Governors of the Federal Reserve System. Haig, Robert. 1919. The Revenue Act of 1918. Political Science Quarterly 34 (3): 369–91. Hammond, Bray. 1957. Banks and Politics in America, from the Revolution to the Civil War. Princeton: Princeton University Press. Harding, W. P. G. 1925. The Formative Period of the Federal Reserve System. Boston: Houghton Mifflin.

375

References

Hardy, Charles. 1932. Credit Policies of the Federal Reserve System. Washington, DC: Brookings Institution. Harris, Marcus. 1948. The over-the-counter market for United States government securities. Unpublished paper. Federal Reserve Bank of New York. Harris, S. E. 1933. Twenty Years of Federal Reserve Policy. Cambridge: Harvard University Press. Hendricks, George. 1933. The Federal Debt: 1919–1930. Washington, DC: Mimeoform Press. Hetzel, Robert, and Ralph Leach. 2001. The Treasury-Fed accord: A new narrative account. Federal Reserve Bank of Richmond Economic Quarterly 87 (1): 33–55. Hollander, Jacob. 1918a. Holdings by the banks of Treasury certificates. Federal Reserve Bulletin 4 (9): 845–47. Hollander, Jacob. 1918b. Certificates of indebtedness in our war financing. Journal of Political Economy 26 (9): 901–908. Hollander, Jacob. 1919. War Borrowing: A Study of Treasury Certificates of Indebtedness of the United States. New York: Macmillan. Jegadeesh, Narasimhan. 1993. Treasury auction bids and the Salomon squeeze. Journal of Finance 48 (4): 1403–19. Joint Committee on the Economic Report. 1952. Monetary policy and the management of the public debt: Replies to questions. Senate Document 123, Part 1. U.S. Senate, 82nd Cong., 2nd Sess. Joint Economic Committee. 1959a. Employment, growth, and price levels. Hearings before the Joint Economic Pursuant to S. Con. Res. 13, Part 6A, The Government’s Management of its Monetary, Fiscal, and Debt Operations, Congress of the United States, 86th Cong., 1st Sess., July 24, 27, 28, 29, and 30. Joint Economic Committee. 1959b. Employment, growth, and price levels. Hearings before the Joint Economic Pursuant to S. Con. Res. 13, Part 9A, Constructive Suggestions for Reconciling and Simultaneously Obtaining the Three Objectives of Maximum Employment, An Adequate Rate of Growth, and Substantial Stability of the Price Level, Congress of the United States, 86th Cong., 1st Sess., October 26, 27, 28, 29, and 30. Jordan, Bradford, and Susan Jordan. 1996. Salomon Brothers and the May 1991 Treasury auction: Analysis of a market corner. Journal of Banking and Finance 20: 25–40. Kang, Sung, and Hugh Rockoff. 2006. Capitalizing patriotism: The Liberty Loans of World War I. Working paper 11919. National Bureau of Economic Research, Cambridge, MA. Kaufman, Henry. 1973. Federal debt management: An economist’s view from the marketplace. In Issues in Federal Debt Management. Boston, MA: Federal Reserve Bank of Boston. Keegan, John. 1999. The First World War. New York: Knopf. Kemmerer, Edwin. 1910. Seasonal variations in the relative demand for money and capital in the United States. National Monetary Commission. Senate Document 588. U.S. Senate, 61st Cong., 2nd Sess. Kennedy, David. 2004. Over Here: The First World War and American Society. Oxford: Oxford University Press. Kennedy, Susan. 1973. The Banking Crisis of 1933. Lexington, KY: University of Kentucky Press. Kettleborough, Charles. 1924. Soldiers’ bonus. American Political Science Review 18 (3): 559–65. Kinley, David. 1910. The independent Treasury of the United States. National Monetary Commission. Senate Document 587. U.S. Senate, 61st Cong., 2nd Sess. Krishnakumar, Anita. 2005. In defense of the debt limit statute. Harvard Journal on Legislation 42 (1): 135–85. Larkin, John. 1951. United States Treasury bills, 1938–1951. Unpublished paper. Federal Reserve Bank of New York. Leff, Mark. 1983. Taxing the “forgotten man”: The politics of social security in the New Deal. Journal of American History 70 (2): 359–81. Leff, Mark. 1984. The Limits of Symbolic Reform: The New Deal and Taxation, 1933–1939. Cambridge: Cambridge University Press.

376

References

Leuchtenburg, William. 1963. Franklin D. Roosevelt and the New Deal, 1932–1940. New York: Harper and Row. Lockhart, Oliver. 1921. The development of interbank borrowing in the national system, 18691914. Journal of Political Economy 29 (2): 138–60. Love, Robert. 1925. The use of Liberty bonds in payment of estate taxes. American Economic Review 15 (2): 275–83. Lutz, F. A. 1940. The structure of interest rates. Quarterly Journal of Economics 55 (1): 36–63. Malvey, Paul, and Christine Archibald. 1998. Uniform Price Auctions: Update of the Treasury Experience. Washington, DC: U.S. Department of the Treasury, Office of Market Finance. Malvey, Paul, Christine Archibald, and Sean Flynn. 1995. Uniform-Price Auctions: Evaluation of the Treasury Experience. Washington, DC: U.S. Department of the Treasury, Office of Market Finance. Mason, Joseph. 2001. Reconstruction Finance Corporation assistance to financial intermediaries and commercial and industrial enterprises in the United States, 1932–37. In Stijn Claessens, Simeon Djankov, and Ashoka Mody, eds., Resolution of Financial Distress. Washington, DC: World Bank. Mason, Joseph. 2003. The political economy of Reconstruction Finance Corporation assistance during the Great Depression. Explorations in Economic History 40 (2): 101–21. McAdoo, William. 1931. Crowded Years: The Reminiscences of William G. McAdoo. Boston: Houghton Mifflin. McCullough, David. 1977. The Path between the Seas: The Creation of the Panama Canal, 1870–1914. New York: Simon and Schuster. McGill, Dan. 1949. An Analysis of Government Life Insurance. Philadelphia: University of Pennsylvania Press. Meeker, J. Edward. 1930. The Work of the Stock Exchange, rev. ed. New York: Ronald Press. Mellon, Andrew. 1924. Taxation: The People’s Business. New York: Macmillan. Meltzer, Allan. 2003. A History of the Federal Reserve: 1913–1951, vol. 1., Chicago: University of Chicago Press. Miller, Adolph. 1921. Federal Reserve policy. American Economic Review 11 (2): 177–206. Miller, Silas. 1938. The origin, procedure, development and economic value of United States Treasury bills. Unpublished paper. Federal Reserve Bank of New York. Miron, Jeffrey, and David Weil. 1998. The genesis and evolution of social security. In Michael Bordo, Claudia Goldin, and Eugene White, eds., The Defining Moment: The Great Depression and the American Economy in the Twentieth Century. Chicago: University of Chicago Press. Morse, Jarvis. 1971. Paying for a world war. Unpublished paper. U.S. Department of the Treasury. Murnane, M. Susan. 2004. Selling scientific taxation: The Treasury Department’s campaign for tax reform in the 1920’s. Law and Social Inquiry 29: 819–56. Murphy, Henry. 1950. The National Debt in War and Transition. New York: McGraw-Hill. Nadler, Marcus, and Jules Bogen. 1933. The Banking Crisis: The End of an Epoch. New York: Dodd, Mead. Nash, Gerald. 1959. Herbert Hoover and the origins of the Reconstruction Finance Corporation. Mississippi Valley Historical Review 46 (3): 455–68. Noyes, Alexander. 1895. The Treasury Reserve and the Bond Syndicate. Political Science Quarterly 10 (4): 573–602. Noyes, Alexander. 1909. Forty Years of American Finance. New York: Putnam’s. Noyes, Alexander. 1910. History of the national-bank currency. National Monetary Commission. Senate Document 572. U.S. Senate, 61st Cong., 2nd Sess. Olson, James. 1977. Herbert Hoover and the Reconstruction Finance Corporation, 1931–1933. Ames: Iowa State University Press. Olson, James. 1988. Saving Capitalism: The Reconstruction Finance Corporation and the New Deal, 1933–1940. Princeton: Princeton University Press.

377

References

Patterson, E. M., Arthur Young, Edwin Seligman, Henry Bond, Carl Barth, Roy Blakey, N.I. Stone, and R.R. Bowker. 1918. Federal taxes upon income and excess profits—Discussion. American Economic Review 8 (1, suppl.): 36–54. Patterson, Robert. 1954. Federal Debt-Management Policies, 1865–1879. Durham, NC: Duke University Press. Patton, Eugene. 1907. Secretary Shaw and precedents as to Treasury control over the money market. Journal of Political Economy 15 (2): 65–87. Peach, W. Nelson. 1941. The Security Affiliates of National Banks. Baltimore: Johns Hopkins University Press. Peach, W. Nelson. 1951. Treasury investment funds and open-market operations. Journal of Finance 6 (1): 46–53. Pearson, F., W. Myers, and A. Gans. 1957. Warren as presidential adviser. Farm Economics (December): 5598–5676. Plehn, Carl. 1918. Substance and shadow in war finance. American Economic Review 8 (3): 564–78. Plehn, Carl. 1920. War profits and excess profits taxes. American Economic Review 10 (2): 283–98. Porter, Sylvia. 1938. Gambling in governments. Scribner’s Magazine (December): 43–46. Reed, Harold. 1922. The Development of Federal Reserve Policy. Boston: Houghton Mifflin. Robinson, Marshall. 1955. Federal debt management: Civil War, World War I, and World War II. American Economic Review 45 (2): 388–401. Robinson, Marshall. 1959. The National Debt Ceiling: An Experiment in Fiscal Policy. Washington, DC: Brookings Institution. Rockoff, Hugh. 2004. Until it’s over, over there: The U.S. economy in World War I. Working Paper 10580. National Bureau of Economic Research, Cambridge, MA. Roosa, Robert. 1952. Integrating debt management and open market operations. American Economic Review 42 (2): 214–35. Roosa, Robert. 1956. Federal Reserve Operations in the Money and Government Securities Markets. New York: Federal Reserve Bank of New York. Samuel, Lawrence. 1997. Pledging Allegiance: American Identity and the Bond Drive of World War II. Washington, DC: Smithsonian Institution Press. Schiff, Mortimer. 1918. War time borrowing by the government. In E. M. Patterson, ed., Financing the War, Annals of the American Academy of Political and Social Science 75 (January): 38–51. Schlabach, Theron. 1969. Edwin E. Witte: Cautious Reformer. Madison: State Historical Society of Wisconsin. Schlesinger, Arthur. 1958. The Age of Roosevelt: The Coming of the New Deal. Boston: Houghton Mifflin. Schwartz, Anna. 1997. From obscurity to notoriety: A biography of the Exchange Stabilization Fund. Journal of Money, Credit and Banking 29 (2): 135–53. Seligman, Edwin. 1914. The federal income tax. Political Science Quarterly 29 (1): 1–27. Seligman, Edwin. 1918a. Loans versus taxes in war finance. In E. M. Patterson, ed., Financing the War, Annals of the American Academy of Political and Social Science 75 (January): 52–82. Seligman, Edwin. 1918b. The War Revenue Act. Political Science Quarterly 33 (1): 1–37. Siegel, Irving, and Maurice Taylor. 1948. Public expenditures for veterans’ assistance. Journal of Political Economy 56 (6): 527–32. Silber, William. 2007. When Washington Shut down Wall Street: The Great Financial Crisis of 1914 and the Origins of America’s Monetary Supremacy. Princeton: Princeton University Press. Silber, William. 2009. Why Did FDR’s bank holiday succeed? Federal Reserve Bank of New York. Economic Policy Review 15 (1): 19–30. Simmons, Edward. 1947. The position of the Treasury bill in the public debt. Journal of Political Economy 55 (4): 333–45.

378

References

Simmons, Edward. 1954. Sales of government securities to Federal Reserve Banks under repurchase agreements. Journal of Finance 9 (1): 23–40. Simon, David. 1994. The Treasury’s experiment with single-price auctions in the mid-1970s: Winner’s or taxpayer’s curse? Review of Economics and Statistics 76 (4): 754–60. Simon, Matthew. 1960. The hot money movement and the private exchange pool proposal of 1896. Journal of Economic History 20 (1): 31–50. Simon, Matthew. 1968. The Morgan–Belmont Syndicate of 1895 and intervention in the foreignexchange market. Business History Review 42 (4): 385–417. Smiley, Gene, and Richard Keehn. 1995. Federal personal income tax policy in the 1920s. Journal of Economic History 55 (2): 285–303. Smith, George. 1956. The Federal Reserve leased wire system: Its origin, purposes, and functions. Unpublished paper. Federal Reserve Bank of New York. Spahr, Walter. 1926. The Clearing and Collection of Checks. New York: Bankers Publishing. Sprague, O.M.W. 1910. History of crises under the National Banking System. National Monetary Commission. Senate Document 538. U.S. Senate, 61st Cong., 2nd Sess. Sprague, O. M. W. 1914. The war and the financial situation in the United States. Quarterly Journal of Economics 29 (1): 181–86. Sprague, O. M. W. 1915. The crisis of 1914 in the United States. American Economic Review 5 (3): 499–533. Sprague, O. M. W. 1917. The conscription of income a sound basis for war finance. Economic Journal 27 (105): 1–15. Sproul, Allan. 1964. The “Accord”—A landmark in the first fifty years of the Federal Reserve System. Federal Reserve Bank of New York. Monthly Review (New York, NY) 46 (11): 227–36. St. Clair, Labert. 1919. The Story of the Liberty Loans. Washington, DC: James William Bryan Press. Stein, Herbert. 1969. The Fiscal Revolution in American. Chicago: University of Chicago Press. Stevenson, David. 2004. Cataclysm: The First World War as Political Tragedy. New York: Basic Books. Sullivan, Lawrence. 1936. Prelude to Panic. Washington, DC: Statesman Press. Synon, Mary. 1924. McAdoo. Indianapolis: Bobbs-Merrill Company. Taus, Esther. 1943. Central Banking Functions of the United States Treasury, 1789–1941. New York: Columbia University Press. Taussig, F. W. 1918. The financial situation: A general survey. In E. M. Patterson, ed., Financing the War, Annals of the American Academy of Political and Social Science 75 (January): 1–11. Thomas, Woodlief. 1951. Lessons of war finance. American Economic Review 41 (4): 618–31. Thomas, Woodlief, and Ralph Young. 1947. Problems of Post-War Monetary Policy. Postwar Economic Studies 8: Federal Reserve Policy. Washington, DC: Board of Governors of the Federal Reserve System. Timberlake, Richard. 1978. The Origins of Central Banking in the United States. Cambridge: Harvard University Press. Tsao, Che, and Anthony Vignola. 1977. Price discrimination and the demand for Treasury’s longterm securities. Unpublished paper. U.S. Department of the Treasury. Tuchman, Barbara. 1958. The Zimmerman Telegram. New York: Macmillan. Upham, Cyril, and Ewin Lamke. 1934. Closed and Distressed Banks: A Study in Public Administration. Washington, DC: Brookings Institution. U.S. Department of Commerce. 1975. Historical Statistics of the United States, Colonial Times to 1970, Part I. Washington, DC: Government Printing Office. U.S. Treasury and Federal Reserve System. 1959. Treasury-Federal Reserve Study of the Government Securities Market, Part I. Washington, DC: Government Printing Office. U.S. Treasury and Federal Reserve System. 1960. Treasury-Federal Reserve Study of the Government Securities Market, Part II. Washington, DC: Government Printing Office.

379

References

Vanderlip, Frank. 1898. Lessons of our war loan. Forum (September): 27–36. Volcker, Paul. 1972. A new look at debt management. Speech to the Money Marketeers at the Bankers Club, New York, March 7. U.S. Department of the Treasury Library. Walker, Charles. 1934. The working of the pre-war gold standard. Review of Economic Studies 1 (3): 196–209. Walker, Charls. 1954. Federal Reserve policy and the structure of interest rates on government securities. Quarterly Journal of Economics 68 (1): 19–42. Wallace, Neil, and Tao Zhu. 2007. Float on a note. Journal of Monetary Economics 54 (2): 229–46. Warburg, Paul. 1910. The discount system in Europe. Reprinted in Paul Warburg. [New York: Macmillan.] Federal Reserve System II: 1930. Warburg, Paul. 1930. The Federal Reserve System, vol. 1. New York: Macmillan. Warren, George, and Frank Pearson. 1935. Gold and Prices. New York: Wiley. Watson, Jesse. 1926. The Bureau of the Mint: Its History, Activities and Organization. Baltimore: John Hopkins University Press. West, Robert. 1977. Banking Reform and the Federal Reserve: 1863–1923. Ithaca: Cornell University Press. Wicker, Elmus. 1966. Federal Reserve Monetary Policy, 1917–1933. New York: Random House. Wicker, Elmus. 1969. The World War II policy of fixing a pattern of interest rates. Journal of Finance 24 (3): 447–58. Wicker, Elmus. 1996. The Banking Panics of the Great Depression. Cambridge: Cambridge University Press. Wigmore, Barrie. 1985. The Crash and Its Aftermath: A History of Securities Markets in the United States, 1929–1933. Westport, CT: Greenwood Press. Wigmore, Barrie. 1987. Was the bank holiday of 1933 caused by a run on the dollar? Journal of Economic History 47 (2): 739–55. Willis, H. Parker. 1915. The Federal Reserve. Garden City, NY: Doubleday, Page. Willis, H. Parker. 1917. What the Federal Reserve system has done. American Economic Review 7 (2): 269–88. Witte, Edwin. 1962. The Development of the Social Security Act. Madison: University of Wisconsin Press.

Index

Accord, the, 348 Acheson, Dean, 240n Adams, Thomas, 150 Adjusted Compensation Payment Act (1936), 252–53, 311 Adjusted Service Certificate Fund (1924), 247, 248–51, 252–53, 267 Adjusted Service Compensation Act (1924), 265 Agricultural Adjustment Act (1933), Title III of (“Thomas amendment”), 237, 242 Aldrich–Vreeland Act (1908), 22, 23–24 All-or-none bidding, 36, 37 Allocation problem, for underpriced securities, 200–201 initial target-plus-oversubscription scheme, 201–202 and pricing of new issues, 205–207 and scaled allotments, 202–205 Anderson, Frank, 100 Anderson, Robert, 353–54, 361 Anticipatory financing, through certificates of indebtedness, 110–11, 112–13, 114–17, 121 Association of Primary Dealers in U.S. Government Securities, 369, 369n Auction sales (offerings), ix, 4–5, 209–10 advantage of, 209 Friedman’s recommendation of, 353 and Mellon’s program, 217 multiple-price format for, 362, 369, 370 need for sophistication restricts participation in, 292 of 1935 bonds, 286–88, 333 and reason for reintroduction of auction sales, 288–89 weakness and failure of, 289–93 of Panama Canal bonds, 45–46 rejection of for Liberty bonds, 91 reverse, 210, 211 single-price format for, 361–62, 368–70 of Treasury bills (1929), 8, 212–13, 216 of Treasury bills (New Deal years), 292, 294, 300 of Treasury bills (1959) expanded, 353 of Treasury bonds abandoning of (1935), 5, 309 1894 sales, 34–36 1896 sales, 39–41 in 1970s (regular and predictable), 356–68 of Treasury notes (1970s), 5, 356–57, 359, 365 advent of regular offerings, 363–65 and February 1969 refunding, 357 and first note auctions, 359–60 and May 1970 refunding, 357–59 question of reason for success of, 362–63 subsequent auctions, 360–62 and underpricing problem, 310 WWI switch from, 7, 143 Awalt, Francis, 232n

Ballantine, Arthur, 232, 232n Bank demand deposits, 17–18 Bankers Trust Company, 192 Bank failures (Great Depression), 222, 222n, 227, 228 hoarding in wake of, 322–23 Roosevelt’s promise to avoid repeat of, 235 Banking crisis (1933), 230–31 and plugging of gold drain, 233 and recapitalization of banks, 233–34 and reopening of banks, 231–33, 234–35 and short-term debt, 276 Banking holidays, 230, 230n Roosevelt’s proclamation of, 231 Banking panics, 22, 230n Banking system, collapse of, 9, 230, 276 Bank money, 16–21 Banks of discount, Federal Reserve Banks as, 26–27 Bearer bonds, 29–31 Beckhart, Benjamin, 142 Belmont, August, 37 Bennett, Jack, 366, 369 Bills, Treasury. See Treasury bills Bi-metallic standard, 13, 13n Blakey, Roy, 55 Bland-Allison Act (1878), 13n, 15–16, 16nn Bond purchase fund, 78–79, 164 and Liberty bonds, 197–98 and Victory notes, 163, 164 Bonds. See also Treasury bonds bearer, 29–31 registered, 29–31 Book-entry system, ix “Borrow and buy” arrangements, 136–37 for Liberty bonds, 98, 136 British government gold standard abandoned by, 9, 222–23, 265, 273, 278, 305, 315, 333 and hoarding of cash in U.S., 322 Second Liberty Bonds acquired by, 197 Bryan, William Jennings, 39 Budget deficits during Great Depression, 221, 223, 237, 241, 247, 259, 262, 279, 302 in 1970s, 356, 366–67, 368 Burgess, W. Randolph, 299 Call option, for certificates of indebtedness, 114 Capper, Arthur, 323 Carlisle, John, 35, 36, 37, 38, 40 Case, J. Herbert, 333 Cash balances, in Treasury, 32 Cash management borrowing small amounts as need arises, 269 Morgenthau’s “as needed” approach, 279, 288–89, 293, 333

382

Index

Cash management (cont.) through certificates of indebtedness (WWI), 109–10, 121 in anticipation of bond sales, 110–11, 112–13, 121 in anticipation of tax receipts, 115–17, 121 contractual innovations facilitating use of, 111, 114 interest rates on, 117–20 question of coercion in campaign for, 120 credit from Federal Reserve Banks for, 133 debt management integrated with, 4, 5, 302 Treasury bills as tool for, 8, 295 WWI expansion of, 144 Central banking and Federal Reserve, 24–25, 26 return to (1914), 26n Central reserve cities. pre-WWI, 20–21 Certificates of indebtedness, 2, 7, 109–10, 121, 143, 144, 154, 163, 168–69, 185–86 in anticipation of bond sales, 110–11, 112–13 in anticipation of tax receipts (WWI), 115–17, 121 in anti-hoarding campaign, 324–25 and “borrow and buy,” 136 contractual innovations facilitating use of, 111, 114 for debt management (WWI), 114–15 during Great Contraction and gold crisis uncertainty, 273, 274 1930s issuance of, 262 for RFC financing, 275, 276, 277 Great Depression (1931–1932) rise in yields of, 223 interest rates on, 117–20, 170 for New Deal financing, 282 in 1920s and allocation problem, 200–202 as amount of debt, 217 issue size of new offerings of, 171 longer term sales of (1921–1924), 199 1920 sales of, 190, 199 open market for, 186–96, 198 sales volumes of, 200 scaled allotments for, 203, 204 statutory limitations on, 167 term to maturity of, 171 vs. Treasury bills (Mills), 213 notes issued in same way as, 167 notes secured by (1919–1921 discount rates), 190 post-WWII, 345–46 question of coerciveness in campaign for, 120 and retirement of Third Liberty bonds, 179 sold to Federal Reserve banks, 131–33 and veterans’ certificates financing, 268 and War Revenue Act (1898), 313

in World War II, 341, 342 yields on, 264 Chandler, Lester, 95 Check collection, 18–19 Childs, C.F., 192, 197 Circulation accounts, 33, 33n Civil War and auction offerings, 34 financing for, 69, 70 and National Banking system, 16n New York Stock Exchange as bond market during, 31 U.S. notes issued during, 15 Cleveland, Grover, 37, 40, 41 Coercion and certificates of indebtedness, 120 and Liberty Loan campaigns, 100–102, 108 Commercial paper, 24 dearth of during Great Contraction, 326n discount rates on, 190 Commercial paper rates (1917–1919), 139 Committee for the Nation, 241n Commodity prices, reflating of (1933), 238–39. See also Reflation Congress characteristics of bonds specified by (1898), 29 and statutory control on Treasury, 313 (see also Statutory control of Treasury debt management) Constitution, U.S. Article I, Section 8 of, 13 Sixteenth Amendment of, 57 Conversion option, for Liberty Loans, 72-73, 74–75, 77–78, 83–84, 85, 85–89, 144 Coolidge, Calvin, 2, 152, 248n Coolidge administration, and tax cuts, 249 Corporate income taxes rate on raised (1926), 153 rate on raised (1932), 223 and Revenue Act (1936), 225–26 for WWI financing, 61 Cortelyou, George, 33 “Country banks,” pre-WWI , 20 “Cover ratio,” 205, 206 in 1930s, 303 Credit policy, Federal Reserve, and Treasury market, 140. See also Treasury market Credit stringencies, seasonal, 23 Currency, inelastic, 21–23 Currency privilege, 319, 323, 328 Curtis, William, 36 “Cycle” securities, 8, 269–71, 297 4-year notes, 367 2-year notes, 365, 366 Davison, George, 232n Debt, national. See National debt Debt, Treasury. See Treasury indebtedness

383

Index

Debt ceiling, 317, 317n broadening of management authority over, 315–17 before World War II, 339 Debt management. See also Primary market; Treasury indebtedness; Treasury financing and Adjusted Service Certificate Fund, 250 cash management integrated with, 4, 5, 302 critiques of, 352 by Milton Friedman, 352–53, 359 by Tilford Gaines, 354–56 factors in success of, 261 during Great Contraction, 261, 278, 279 and Emergency Adjusted Compensation Act, 265–68 and gold crisis effects, 273–74 1930 outlook for, 262–65 and 1931 debt extension, 271–72 and Reconstruction Finance Corporation, 275–78 refinancing as chief concern, 261, 262 “regular and predictable” Treasury bill offerings, 268–71 during Great Depression, 8–9, 333, 337 during New Deal, 279–81, 302 on “as needed” basis, 279, 281, 289, 292, 293, 296 maturity extension (1933–1934), 9, 281–82 maturity extension (1935 and following), 293 and 1935 bond auctions, 286–93 refinancing of Liberty Loans, 283–85 regular weekly bills, 296–302 reintroduction of bond auctions (1935), 287 and tax date bills, 279, 293–96, 302, 307 in 1920s, 154–55, 179–80, 217, 337 and “Mellons,” 180–83, 263, 265 paying off Treasury notes and reducing short-term debt, 168–72 payments from foreign governments, 154, 157–58 refinancing over new project financing, 5 refinancing short-term debt (1921–1923), 161–63, 168 retiring Second and Third Liberty bonds, 172–79 retiring Victory notes, 163–66 and scaled allotments, 205 sinking fund, 154, 156–57 surplus receipts, 158-159 and “regularization” of operations on quarterly tax dates, 168, 168n, 169–70, 180, 210 Volcker on, 365 regular and predictable issuance in, ix, 4, 6, 278, 333 (see also Regular and predictable issuance) statutory control of, 313 before Great Depression, 5, 166–67, 313–14

managerial authority expanded (Depression years), 5–6, 314–17, 337 technical procedures lacking for, 356n during World War I, 7, 143–44, 337, 337n certificates of indebtedness for, 114–15 in WWI aftermath, 7–8 in World War II, 341–42, 343 and monetizing debt, 342, 344 in WWII aftermath (to 1958), 344–50 and the “Accord,” 348 July 1958 refunding, 351–52, 353 Debt refinance. See Refinancing of debt Deficit financing, during Great Depression, 8–9, 221. See also Budget deficits and permanent debt, 278 Deflation during Great Contraction, 321–22 and money supply, 323 Demand deposits, 17–18 Demand for money, shift in composition of (during Great Contraction), 322–23 Devaluation of dollar (1934), 242, 243–44, 245, 245n, 333 Discount(s), by Reserve banks, 131 Discount Corporation, 197, 299 Discount privilege, from Federal Reserve banks, 26–27 Discount rates, 26n by Federal Reserve Bank of New York, 138, 139, 186–89 rise in (1919–1920), 191 rise in (1928), 207 rise in (1931), 223, 273 in Treasury–Federal Reserve Accord, 348 wartime discount rate policy, 137–40 Distributive justice and Revenue Act (1935), 225 and WWI financing, 55, 55n Dollar devaluation of (1934), 242, 243–44, 245, 245n, 333 flight from, 241 gold content of reduced, 237–38, 242–43, 329 Drexel & Co., 41 Durand, Roy, 55 Emergency Adjusted Compensation Act (1931), 251–52, 265–68, 278 Emergency Banking Act (1933), 233–34, 234n Emergency Relief and Construction Act (1932), 229–30, 275 Entitlement felt by investors, McAdoo on, 76n Estate tax privilege, for Third Liberty bonds, 78 Excess profit taxes, for WWI financing, 61–64, 147 elimination of, 148, 150, 151

384

Index

Exchange offerings during Great Depression, 310–12 preferential allotments for (during Great Contraction), 307, 308, 311 high attrition of (1970–1972), 364, 364n and Liberty Loan campaigns, 163 multi-option, 350 Gaines’s criticism of, 355 in refundings, 357, 358, 359 of 1900 bonds, 44 in 1920s, 7, 180 Exchange rates, and international movements of capital (pre-WWI), 23n Expenditures. See Treasury expenditures Federal Farm Mortgage Corporation (FFMC), 291, 309 Federal Home Bank Act (1932), 327 Federal Home Loan Bank Act (original title), 327 Federal Land Banks, 229, 236 Federal Old-age and Survivors Insurance Trust Fund, 253, 255n Federal Reserve Act (1913), 23, 24–27, 319 amendment of (1917), 25n section 13 of, 131 section 14 of, 194 section 15 of, 67–68 Federal Reserve Bank Credit, 185 Federal Reserve bank notes, 28n, 235 Federal Reserve Bank of New York, 8 and certificates of indebtedness, 119 discount rate raised by (1931), 273 loan and discount rates raised by, 186–89 Treasury gold purchases through, 245 Federal Reserve Banks certificates of indebtedness sold to, 131–33 as fiscal agents of U.S., 67–68, 110, 132, 136, 195 examination of securities applications called for, 309–10 New York bank as agent in buying and selling gold, 245, 245n and Treasury primacy during WWI, 141n gold certificates held by (1933–1936), 244–45 gold owned by (1933), 243 and interest rates (post-WWI), 186 lending to, 131 Liberty Loans campaign through, 95 and periodic loans to Treasury, 172 reopening of (1933), 232 repurchase agreements used by (WWI), 193 WWI fluctuations in Treasury accounts at, 109, 121 and WWII debt, 342, 344 Federal Reserve notes (money), 25, 27, 28n, 233, 234, 244, 319, 331 free gold constraint on, 326

Federal Reserve System, 24–28 assets of (1917–1919), 136–37 creation of, 13, 24–27 and discount rate, 26n (see also Discount rates) increase in reserve requirements for, 293, 299 in interest rate stabilization program (WWII), 339–41 in July 1958 refunding, 352 and May 1970 refunding, 358–59 monetary expansion program of, 275 open market purchase program of (1932), 325–27 Treasury market supported by (WWI), 131, 140–42 and “borrow and buy” program, 136–37 first financing, 131–32 and First Liberty Loan, 133–36 and wartime discount rate policy, 137–40 Treasury securities purchased by (1932), 274 Fifth war loan (WWII), 343 First Liberty Bond Act (1917), 69–72, 96, 109, 110 and certificates of indebtedness, 119 First Liberty Loan. See under Liberty Loans First war loan (WWII), 343 First World War. See World War I Fisher, Peter, 368 Fixed-price offerings, 4, 6, 8, 199 allocation problem for (underpricing), 200–201 initial target-plus-oversubscription scheme, 201–202 and scaled allotments, 202–205 Anderson’s defense of (1959), 354 failed offerings, 8, 201, 310 limiting risk of, 288 last instance of (1970s), 362 in 1920s, 180 oversubscription of, 303 (see also Oversubscriptions) and pricing of new issues, 205–207 reliance on as shortcoming of debt management, 302 return to (1935), 291, 333 tax date (term to maturity), 294 during World War I, 143, 288 for Liberty bonds, 91–92 Foreign governments debt payments, 154, 157–58, 161 Fourth Liberty Bond Act (1918), 79–80, 167, 265 Forth Liberty Loan. See under Liberty Loans Fourth war loan (WWII), 343 Friedman, Milton, 24, 322 Treasury debt management critiqued by, 352–53, 356, 359

385

Index

Gage, Lyman, 43 Gaines, Tilford, 179, 349 Treasury debt management critiqued by, 354–56, 356 General Motors Corporation, 366 Gensler, Gary, 368 Gilbert, S. Parker 217 Glass, Carter, 2, 82–84, 88n, 99, 140, 147, 327 Glass-Borah amendment, 327–28, 329, 330 expiration of, 331 Glass-Steagall Act (1932), 326 Gold revaluation of (1934), 282 RFC’s purchase of, 239–42, 282 Gold bullion standard, 233, 237–45 Gold certificates, 14–15, 319 Gold coins, 14 Gold content of dollar, reduction of, 237–38, 242–43, 329 Gold drain (1933), 230–31, 231 plugging of, 233 Goldenweiser, E. A., 340 Gold reserve bond auctioning to replenish, 34–35, 39–40 private sale of bonds to replenish, 37 Gold Reserve Act (1934), 243 Gold standard, 13–14, 27 Britain’s departure from, 9, 221–23, 265, 273, 278, 305, 315, 333 and hoarding of cash in U.S., 322 European countries on, 23 as inflation protection, 323 and Treasury money, 14 U.S. departure from, 233 Gold Standard Act (1900), 13–14, 233, 239–40, 242, 243 bonds authorized by, 30 and gold content of coins, 14 Section 2 of, 15 Section 11 of, 44 and silver dollars or certificates, 16 Gold sterilization program, 311–12, 312n Gore, Thomas, 101 Government spending. See Treasury expenditures Government trust funds, 247, 253. See also Social Security Trust Funds Adjusted Service Certificate Fund, 247, 248–51, 252–53, 267 United States Government Life Insurance (USGLI) Fund, 247, 251n Great Britain. See British government Great Contraction, 221, 226, 252, 333 dearth of commercial paper during, 326n debt management during, 261, 278, 279 and Emergency Adjusted Compensation Act, 265–68 and gold crisis effects, 273–74

1930 outlook for, 262–65 and 1931 debt extension, 271–72 and Reconstruction Finance Corporation, 275–78 refinancing as chief concern, 261, 262 “regular and predictable” Treasury bill offerings, 268–71 deflation in, 321–22 national bank notes during, 321–24 and anti-hoarding campaign, 324–25 and Federal Reserve open market purchase program (1932), 325–27 and Glass–Borah amendment, 327–28, 329, 330 oversubscriptions during, 303–306 (see also Oversubscriptions) Great Depression, 5–6, 8–9, 221 banking crisis (1931), 230–31 and plugging of gold drain, 233 and recapitalization of banks, 233–34 and reopening of banks, 231–33, 234–35 debt management during, 8–9, 333, 337 deficits and indebtedness during, 236–37, 238 and gold bullion standard, 237–45 national bank notes during, 321–24 New Deal, 235–36 primary market during, 303 and exchange offerings for maturing securities, 307, 310–12 and oversubscription during Great Contraction, 303–306 and oversubscription during New Deal, 307–10 tax receipts during, 221 and Britain’s departure from gold standard, 222–23 and New Deal revenue acts, 223, 225–26 and 1932 Revenue Act, 221–22, 223 Treasury expenditures during, 226–27 and Reconstruction Finance Corporation, 228–30 Gross national product. U.S., 1, 3 during Great Depression, 221 Harding, Warren, 2, 150, 153, 248n Harding administration, and refinancing of short-term debt, 162 Harrison, George, 232n Hawley, Willis, 212 Hoarding of cash, during Great Contraction, 322 campaign against, 324–25 Hoover, Herbert, 2 and anti-hoarding campaign, 323, 325 and early payment of veterans’ bonuses, 251, 252, 265 and Great Depression, 226, 227

386

Index

Hoover, Herbert (cont.) and Reconstruction Finance Corporation, 228n, 229 tax increase proposed by, 223 and zero coupon bills, 212 Hoover administration, currency expansion refused by, 323 Houston, David, 2, 147, 161–62 Hull, Cordell, 74 Humphrey, Hubert, 367 Ickes, Harold, 225 Income taxes. See also Taxes in Great Depression (receipts from), 221–22, 223, 224 Treasury-bond interest exempted from, 71 and Liberty bonds, 71–72, 73–74, 80–82, 83, 89 for WWI financing, 56–58 postwar criticism of, 147 postwar reduction of, 148, 150, 151, 152 through Revenue Act (1918), 59–61 through War Revenue Act (1917), 58–59 Inelastic currency, 21–23 Inflation and interest rate stabilization program, 340, 341 post-WWII, 344 savings bonds as defense against, 341 and Treasury indebtedness, 338 Installment purchase plans, for Liberty bonds, 92–93 Institutional developments, during 1917–1939 period, 4–6. See also Auction sales; Integration of cash management and debt management; Regular and predictable issuance; Statutory control Integration of cash management and debt management, 4, 5, 302 Interest rates on certificates of indebtedness, 117–20 for Liberty bonds as below-market, 102–104 and conversion option, 72–73, 144 low rates as Treasury objective, 186 and short-term debt issuance, 368 and Treasury–Federal Reserve Accord (Meltzer), 348 Interest rate stabilization program (WWII), 339–41 change in, 345 dismantling of, 346 International Manhattan Company; 213 Kang, Sung Won, 102 Kaufman, Henry, 368 Kemmerer, Edwin, 21 Kennedy, David, 52, 102 Kent State University killings, 358 Korean War, 346, 347

La Follette, Robert, 101 Land Bank lending, 229 Leff, Mark, 225–26 Leffingwell, Russell, 82, 189, 191 Lender of last resort, lack of (pre-WWI), 21 and Federal Reserve Act, 24 Lend–Lease Act (1941), 339 Leuchtenburg, William, 225 Liberty bond(s) and bond purchase fund, 164 and certificates of indebtedness, 110, 114–15 complex provisions of, 144 exchange offerings for, 311 1919 amount of, 154 notes secured by, (1919–1921) discount rates on, 190 outstanding (June 30, 1930), 263 and preference rate, 188 question of selling at below-market interest rates, 102–108 redemption of, 217 refinancing of, 172 reverse auction of, 211 secondary market for, 104–107 Liberty Bond (Loan) Acts, 157, 166 First (1917), 69–72, 96, 109, 110 and certificates of indebtedness, 119 subscribers and allotments for, 94 Second (1917), 73–74, 167, 265, 273–74 ceiling on outstanding certificates set by, 110 and certificates of indebtedness, 119 coupon rate of, 139 subscribers and allotments for, 94 Supplement to (1918), 80–81 Third (1918), 75, 77, 78, 167, 273–74 ceiling on outstanding certificates set by, 110 Section 15 of, 164 subscribers and allotments for, 94 Fourth (1918), 79–80, 167, 265 subscribers and allotments for, 94 tax exemption for, 80, 81 Victory (1919), 82, 167 ceiling on outstanding certificates set by, 110 section 6 of, 156 subscribers and allotments for, 94 Liberty Loan army, 95 Liberty Loan Association, 93 Liberty Loans, 5, 7, 64–66, 69, 85 First, 69–73, 80 conversion of, 72–73, 77–78, 85–89 Federal Reserve in financing of, 133–36 and 1930 debt management, 263 redemption options on, 263 refinancing of, 279, 283 secondary market for, 104–106, 107

387

Index

Second, 73–75, 80 conversion of, 75–76, 77–78, 86, 87, 88–89 exchange offerings for, 311 notes in exchange offerings for, 163 redemption of, 161, 180, 261, 263 refinancing of, 279 retirement of, 161, 172–79 secondary market for, 104–106, 107 unexpectedly large payments on, 114 Third, 75–79, 80 exchange offerings for, 311 installment purchases of (Liberty Loan Association), 93 notes in exchange offerings for, 163 refinancing of, 279 retirement of, 161, 162, 172–79 secondary market for, 106–107 Fourth, 79–82 installment purchases of (Liberty Loan Association), 93 and lending by national banks, 140 management of maturity of, 279 and 1930 debt management, 263 redemption of, 283 redemption options on, 263 refinancing of, 283–85, 314–15 refunding of, 271 secondary market for, 106–107 Victory, 82–84, 140, 162, 163 exchange offerings for, 311 notes in exchange offerings for, 163 redemption options on, 263 refinancing of, 279 as tax-exempt, 83, 273 Liberty Loans, marketing of, 91, 93–96, 143 as “capitalizing patriotism,” 102, 108 coercive aspect of, 100–102, 108 and discount rate policy, 137, 139–40, 142 first campaign, 96–97 as fixed-price offerings, 91–92 fourth campaign, 98 installment purchase plans, 92–93 and oversubscription, 96–97 preferential allotments for small investors, 93 and prompt delivery, 93 second campaign, 97–98 third campaign, 98 Victory campaign, 99–100 Loan rates, 15-day by Federal Reserve Bank of New York, 138, 186–89 Longworth, Nicholas, 62 Marketing of Liberty Loans, See Liberty Loans, marketing of Market for Treasury securities. See Primary market; Secondary market; Treasury market Martin, William McChesney, 348

Maturity extension in 1933–1934 period, 281–82 in 1935 and beyond, 293 McAdoo, William, 2, 2n, 82, 179, 217 and Aldrich–Vreeland Act, 24 and bond purchase fund, 164 and certificates of indebtedness, 111, 117–18, 120 and Federal Reserve Banks as fiscal agents, 68 and Federal Reserve Board, 132 and interest rate on federal deposits, 33 and Liberty Loans, 64, 73, 75, 77, 78, 79, 108, 133, 137, 139 conversion option for, 72, 74, 77, 85 and falling prices, 76, 76n marketing of, 91, 95–96, 96, 97, 99, 101, 102, 103 maturity of, 70 taxation of, 71, 80, 81 for short-term financing, 82 and uncertainty in bond yields (WWI), 340 and WWI financing, 54 and WWI foreign loans, 53 WWI tax requests from, 60–61 and WWI war profits tax, 63 McKinley, William, 42 Mellon, Andrew, 2, 2n, 150, 154, 217, 333 and Adjusted Service Certificate Fund, 250, 250–51 as Ambassador to Great Britain, 261 on bond distribution, 202 broadened authority requested by, 314–17 and congressional authority, 317 and debt management in 1930s, 261, 264–65 and excess profits tax, 150 and exchange offers, 311 in Great Contraction, 279 and extending of maturity, 279 and national bank notes, 323 refinancing program of, 261, 278 regular schedules introduced by, 288, 302 and Revenue Act (1932), 223 and short-term debt, 162, 163, 180 on taxation, 150–51, 152, 153, 154 and tax date financing, 180 on veterans’ certificates payoff, 266 on zero coupon bills, 212 “Mellons” (bonds), 180–83, 263, 265 Meltzer, Allan, 230, 326, 348 Mills, Ogden, 2, 175, 217, 232n, 261, 333 and anti-hoarding campaign, 324–25, 325 and cash management through Treasury bills, 295 and debt extension, 271, 272, 278, 279 and distortion from War Loan Deposit Account rates, 208–209 on oversubscriptions, 306 and regular and predictable bill offerings, 268

388

Index

Mills, Ogden (cont.) and Treasury bills, 333 and weekly bill offerings, 302 Monetary expansion, Federal Reserve program of, 275 Monetizing minimizing of as goal (1951), 348 of WWII debt, 342, 344 Money. See also Federal Reserve bank notes; Federal Reserve notes bank money, 16–21 Treasury money, 14–16 Money supply, contraction of (1929–1933), 322 Morgan, J. P., 36, 37–38, 41, 41n, 54, 64 Morgan Stanley & Co., 366 Morgenthau, Henry, 2, 240–41n, 279, 333 and “as needed” approach, 279, 288–89, 293, 333 and bond auctions, 287, 288, 290, 292, 309, 354 and congressional authority, 317 and exchange offers, 311 and investor choice, 350 and limits on bank subscriptions, 309, 310–11 and return to fixed-price bond offerings, 291 and savings bonds, 256–57 and 13-week bills, 301 and WWII financing, 341 Multi-option exchange offerings, 350 Gaines’s criticism of, 355 in refundings, 357, 358, 359 Multiple-price auction format, 362, 369, 370 National Bank Act, 16n, 17n, 20n, 26 amendment of (1916), 27n Section 18 of, 27–28 National Banking System, 16n National bank notes, 17 extinction of, 328–31 and Federal Reserve notes, 25, 27–28, 319 during Great Contraction, 321–24 and anti-hoarding campaign, 324–25 and Federal Reserve open market purchase program (1932), 325–27 and Glass–Borah amendment, 327–28, 329, 330 during Great Depression, 321 Treasury bonds as collateral for, 17, 30, 31, 45 between WWI and Great Depression, 319–21 National banks pre-WWI debt held by, 1 reserve requirements for (pre-WWI), 20 Treasury deposits in, 32–33 National Credit Corporation, 227–28, 275 National Currency Act (1863), 16n

National debt. See also Budget deficit permanence of established, 278 at WWI end, 7, 147 National Industrial Recovery Act (1933), Title II of, 236 New Deal, 235–36 debt management during, 279–81, 302 on “as needed ” basis, 279, 281, 289, 292, 293, 296 maturity extension (1933–1934), 281–82 maturity extension (1935 and following), 293 and 1935 bond auctions, 286–93 refinancing of Liberty Loans, 283–85 reintroduction of bond auctions (1935), 287 regular weekly bills, 296–302 and tax date bills, 279, 293–96, 302 oversubscriptions during, 307–10 revenue acts during, 223, 225–26 “New York exchange,” 19 New York Stock Exchange failure of to list certificates and notes, 197 Liberty bonds sold to Treasury on, 198 revival of bond trading on, 65 secondary market on, 185 trading in government securities on, 186 and Treasury bond market, 34, 34n, 196–98 Nixon, Richard, 358 Nonmarketable debt, 247, 248, 259 for government trust funds, 247, 256 and Adjusted Compensation Payment Act (1936), 252–53 Adjusted Service Certificate Fund, 248–51, 267 and Emergency Adjusted Compensation Act (1931), 251–52 New Deal issues of, 281 for Old-Age Reserve Account, 254–56 savings bonds, 247, 256–58, 259, 296, 341, 342 for Social Security Trust Funds, 253 for Unemployment Trust Fund, 254 Notes. See Federal Reserve notes; National bank notes; Treasury notes Old-Age Reserve Account, 247, 253, 254–56, 259 Over-the-counter-market (open market), 185, 198 for certificates of indebtedness, 186–92, 196 and repurchase agreements, 192 and wire transfers, 195 for Treasury bonds, 196–98 Oversubscriptions and allocation problem, 201–202 insistence on, 101 and interest rate disparity, 208

389

Index

of Liberty bonds, 143 as symbol of national unity, 91, 92, 96, 97, 108, 108n in 1930s, 303 exchange offerings as limiting, 312 during Great Contraction, 303–307 during New Deal, 307–10 1935 auction process as avoiding, 290 of 1920s, 205–207, 207n and padding, 304, 306, 307, 310, 312 and preferential allotments for exchange of maturing securities, 307, 308, 311 risk of in selling long-term debt, 350 Padding of fixed-price subscriptions, 304, 306, 307, 310 exchange offerings as limiting, 312 precautions taken against, 310 Panama Canal Bonds, 30, 31, 45–46, 286, 328, 337 and currency privilege, 329 redemption of, 330, 331 tax exemption for, 71 Patriotism and certificate-of-indebtedness campaign, 118, 120, 120–21n money payment as antithetical to (Coolidge), 248n and Spanish–American War bonds, 42, 44 in WWI marketing campaigns, 199 Liberty Loan campaigns, 55, 75–76, 88n, 92, 95, 96, 101, 102, 103, 104, 107, 108, 134n “Patriots for publication,” 75–76 Payments from foreign governments (1920s), 154, 157–58 Payne–Aldrich Tariff Act (1909), 30, 45n, 61 Popular loan 1898 bond offering as, 41–42, 43 Liberty bonds as, 91–92, 143 First Liberty Loan, 70, 71–72 savings bonds as, 256–57, 258 Preferential allotments, 8 for Liberty bonds, 93 for tender of maturing securities (exchange offerings), 8, 307, 308, 311 Preferential rates, for loans and discounts secured by Treasury securities, 187, 188 Primary market. See also Auction sales; Fixed-price offerings evolution in marketing for, 199, 216 and allocation problem, 200–207 Treasury bills introduced, 209–15 and War Loan Deposit Accounts, 207–209 during Great Depression, 303 and exchange offerings for maturing securities, 307, 310–12 and oversubscription during Great Contraction, 303–306

and oversubscription during New Deal, 307–10 and Mellon, 217 for Treasury bonds (pre-WWI), 13, 34, 46 Project finance approach to Treasury debt, 5, 6, 29, 46, 109, 313, 337 end of, 315 Mellon’s move away from, 217 and Morgenthau’s “as needed” approach, 288–89 Prompt delivery sales, for Liberty bonds, 93 Public notification campaign, in call for Second Liberty bonds, 176 Public protest, against 1895 private bond sale, 38–39, 40 Public works in Hoover’s program, 226 “self-liquidating,” 229 under New Deal, 236 Put option, on certificates of indebtedness, 114, 116 Quincey, Charles E., 192, 197 Recession (1920–1921), 152, 162 Recession (1948–1949), and long-term bonds, 349 Recession (late 1950s), 351 Reconstruction Finance Corporation (RFC), 228–30, 323–24 financing of, 275–78, 278 gold purchases of, 239–42, 282 and Mills, 261 preferred bank stock purchases by, 234, 235 Treasury expenditure to, 236 Reconstruction Finance Corporation Act (1932), 228, 275 Redemption options, 144. See also Conversion option Refinancing of debt of certificates of indebtedness (1920s), 169 vs. financing of new projects, 5 during Great Contraction, 278 of Treasury bills, 261, 262 of Liberty Loans, 172, 279, 283–85, 314–15 Victory notes, 168, 169 of short-term debt (1921–1923), 161–63, 217 and WWI debt, 7 Reflation (1930s), 323, 327 of commodity prices, 238–39 and gold bullion standard, 242 and gold purchase program, 240 Refunding (July 1958), 351–52 Friedman’s critique of, 353 Refunding (February 1969), 357 Refunding (May 1970), 357–59 Registered bonds, 29–31 Regularization of Treasury operations, 168, 168n, 169–70, 180, 210 Volcker on, 365

390

Index

Regular and predictable issuance, ix, 4, 6, 333 of bills, 268–71, 278, 279, 296–98, 368 13-week bills, 269, 270, 278, 279, 297, 298–302, 333 and bond auctions, 293, 356–57, 360–64, 365, 367–68 early version of, 7 Friedman in favor of, 353, 356 Gaines in favor of, 355, 356 of notes and bonds (1970s), 356–57 advent of auction offerings, 357–63 advent of regular offerings, 363–65 embraced as debt management strategy, 365–67 enduring value of, 367–68 permanent debt as condition for, 278 Repurchase agreements (“repos”), 192 Reserve requirements, 20–21 Revenue Act (1916), 58, 59, 61, 71, 78 Revenue Act (1917), 62, 78 Revenue Act (1918), 59, 59–61, 61, 63-64, 79, 151, 223, 224 Revenue Act (1921), 151 Revenue Act (1924), 151, 152, 249 Revenue Act (1926), 151, 152–53, 224 Revenue Act (1928), 153 Revenue Act (1932), 221–22, 223, 224 Revenue Act (1935), 225, 226 Revenue Act (1936), 225–26 Revenue Act (1938), 226 Revenue Act (1939), 226 Reverse auctions, 210, 211 RFC. See Reconstruction Finance Corporation Rockoff, Hugh, 102 Roosa, Robert, 350 Roosevelt, Franklin, 230, 281 banking holiday proclaimed by, 231 and early payment of veterans’ bonuses, 252 and gold bullion standard, 237–44 and gold drain, 233 and New Deal, 235–36 and reflation, 238–39, 240, 242 and reopening of banks, 232, 235 Roosevelt administration, and fixed-price subscription offerings, 307 Rothschild, N. M., 37 St. Clair, Labert, 96, 98 Salomon Brothers, 370 Salomon Brothers & Hutzler, 192, 197, 213, 299 Savings bonds, 247, 256–58, 259, 341 sales of, 296, 342 Savings stamps and certificates. See War-savings stamps and certificates Scaled allotments, 199, 202–205 Schiff, Mortimer, 76, 78 Schwartz, Anna, 24, 322

Seasonal variation and discount privilege, 27 and inelastic currency (pre-WWI), 21–22 Secondary market, 185 and bond purchase fund, 79 for certificates of indebtedness, 185, 191–92 and wire transfers, 195 and exchange offers, 312 growth of (1920s), 217 for Liberty bonds, 82, 104–107, 185 and oversubscription problem, 208–209, 310 in postwar period, 198 for Treasury bonds, 31–34, 46 Second Liberty Bond Act (1917), 73–74, 167, 265, 273–74 ceiling on outstanding certificates set for, 110 and certificates of indebtedness, 119 coupon rate of, 139 Supplement to (1918), 80–81 Second Liberty Loan. See under Liberty Loans Second war loan (WWII), 343 Secretaries of the Treasury (World War I to Great Depression), 1, 2. See also individual secretaries Securities, Treasury. See Treasury securities; specific types of security “Sedalia, Missouri, plan,” 19n Seventh war loan (WWII), 343 Shaw, Leslie, 23 Sherman Silver Purchase Act (1890), 13n, 15–16, 16n Silver certificates, 14, 15–16, 16n, 21, 319, 331 Silver dollars, 14, 15–16, 16n Simmons, Furnifold, 53 Simon, William, 369 Single-price auction format, 361–62, 368–70 Sinking fund (1920s), 154, 156–57, 161, 163, 180, 197–98 Sixth war loan (WWII), 343 Smoot, Reed, 210, 212 Snyder, John, 346, 348 Social Security Act (1935), 223, 247, 253, 255–56 Social Security Act Amendments (1939), 255nn Social Security Trust Funds, 247, 253, 312 Old-Age Reserve Account, 247, 253, 254–56, 259 and question of sufficient resources, 259 Unemployment Trust Fund, 247, 253, 254, 259 Social security wage taxes, 296 Spanish–American War Bonds, 29, 30, 46, 69, 313, 337 and below-market pricing, 108 blowout sale of, 133 in exchange offering, 44

391

Index

and fixed-price offerings, 91 individual investors for, 91 interest rates of, 143 and preferential allotments, 93 subscription offering of (1898), 41–44 tax exemption for, 71 Spooner Act (1902), 30, 45 Sprague, O. M. W., 54 Sproul, Allan, 345 Stalnecker, Mark, 368 Stamp tax, on Federal Reserve promissory notes, 193 State unemployment funds, and Social Security Trust Funds, 253, 254, 254n, 259 Statutory control of Treasury debt management, 313 before First World War, 29 before Great Depression, 5, 166–67, 313–14 managerial authority expanded (Depression years), 5–6, 314–17, 337 Stevenson, David, 51, 52 Stewart, John, 36 Strong, Benjamin, 68n, 100, 104, 141n, 194, 195 Surplus receipts (1920s), 158–59, 161 Taussig, Frank, 58 Taxes. See also Income taxes in Great Depression, 221 and Britain’s departure from gold standard, 222–23 and New Deal revenue acts, 223, 225–26 and 1932 Revenue Act, 221–22, 223 high rates of (WWI end), 147 Mellon on economy stimulated by lowering rates on, 151, 152 and “regularization” of financing operations, 7–8, 168, 168n, 169, 180, 210, 279, 293 tax policy of 1920s, 150–54 and sinking fund, 156–57 for Social Security, 255n, 296 and veterans’ WWI bonuses, 248, 248n, 249 Victory notes exemptions from, 83, 273 for WWI financing, 55–56 certificates of indebtedness issued in anticipation of receipts from, 115–17, 121 through corporate income taxes, 61 through excess profit and war profit taxes, 61–64 through individual income taxes, 56–61 Third-degree committees, in selling of Liberty bonds, 101n Third Liberty Bond Act (1918), 75, 77, 78, 167, 273–74 ceiling on outstanding certificates set for, 110 Section 15 of, 164 Third Liberty Loan. See under Liberty Loans Third war loan (WWII) 343 Thomas amendment, 237, 242

Timberlake, Richard, 27 TIPS (Treasury Inflation-Protected Securities), 2, 337n Trading with the Enemy Act (1917), 231 Treasury bills, 2 auction sales of, 8 vs. auctioning of bonds, 292 expansion of (1959), 353 for debt management, 8 during Great Depression, 281 outstanding amounts of, 297 introduction of (1929), 5, 199, 209–12, 216, 217, 314 first auction, 212–13 and illiquidity due to bookkeeping requirement, 214–15 subsequent auctions, 213–14 post-WWII, 345 refinancing of, 261 regular and predictable issuance of, 268–71, 278, 279, 296–98, 368 13-week bills, 8, 269, 270, 278, 279, 297, 298–302, 333 tax date, 279, 293–96, 302, 333 term to maturity of, 295 in World War II, 341, 342 Treasury bonds, 2 as amount of debt, 217 as collateral for bank notes and public deposits, 34 for currency issues and public deposits, 46 for national bank notes, 17, 31, 45, 319, 320, 327 Congressional authorization for sale of (Liberty Bond Acts), 167 decline in price of (1931), 266 disadvantages in auctioning of, 292–93 and flight from dollar (1933), 241 in Great Depression (rise in yields of, 1931–1932), 223 “Mellons,” 180–83, 263, 265 outstanding (June 30, 1914), 30, 31 outstanding (June 30, 1930), 263 over-the-counter market for, 196–98 paying off, 168–72 predictable maturities lacking for (New Deal years), 302 pre-WWI sales of, 34, 46 1894 auction sales, 34–36 1895 private sale, 35, 37–39 1896 auction sale, 39–41 1898 subscription offering of Spanish– American War bonds, 41–44, 46 1900 exchange offering, 44 Panama Canal Bond auctions, 45–46 regular and predictable auctions of, 293, 356–57, 360–64, 365, 367–68 on model of bills, 333 scaled allotments for, 202–203 secondary market for (pre-WWI), 31–34, 46

392

Index

Treasury bonds (cont.) statutory limitations on, 167 tax-date sales of, 337 tax exemption for, 71 and veterans’ certificates financing, 267–68 in World War II, 341, 342 yields on, 264 Treasury cash management. See Cash management Treasury debt. See Treasury indebtedness Treasury debt management. See Debt management Treasury deposits, in national banks, 32–33 Treasury expenditures in Great Depression, 226–27 (see also Budget deficit) and Reconstruction Finance Corporation, 228–30 in 1920s, 159 in 1927–1939 period, 236, 237 reduction of (1920s), 148–50 Treasury–Federal Reserve Accord, 348 Treasury financing during Great Depression, 221 (see also Great Depression) with nonmarketable debt, 247, 248 for government trust funds, 247–53 for Social Security Trust Funds, 253 and tax reduction vs. debt reduction (1920s), 147, 153–54 debt-reduction measures, 154–59 tax policy, 150–54 of World War I effort, 53–55 through debt financing, 54–55 through Liberty Loans, 5, 7, 64–66, 69, 85 (see also Liberty Loans) through tax financing, 55–64 through war-savings stamps and certificates, 66–67 of World War II, 341, 342, 343 Treasury indebtedness (debt), 1, 29–31 decrease in (1920s), 148, 154–55, 217 explosion of (1916–1919), 143 during Great Depression, 236–37, 238 composition of (New Deal), 279–81 intermediate-term, 180, 199, 217 long-term, 182, 183 (see also Maturity extension) as economic conundrum, 349 for New Deal financing, 281–82 timing of sale of, 350 long-term vs. short-term (prior to WWI), 313 1919 amount of, 161 in post-1939 period, 337–38 pre-war vs. postwar, 198 short-term and banking system collapse, 276 and gold crisis, 273, 274 and interest rates, 368

and Liberty Loan campaigns, 7 for 1933 financing, 281 refinancing of (1921–1923), 161–63, 217 for veterans’ loans, 271 statutory control of, 313 before Great Depression, 5, 166–67, 313–14 managerial authority expanded (Depression years), 5–6, 314–17, 337 term to maturity of (New Deal years), 294 from World War I to eve of World War II, 1, 3–4 Treasury Inflation-Protected Securities (TIPS), 2, 337n Treasury market, 1 Federal Reserve support of, 131, 140–42 and ”borrow and buy” program, 98, 136–37 first financing, 131–32 and First Liberty Loan, 133–36 and wartime discount rate policy, 137–40 four pillars of, ix Mellon’s transformation of, 217 pre-World War I, 1 WWI transformation of, 49 Treasury money (pre-WWII), 14–16 and bank money, 16 Treasury notes, 2, 167 as amount of debt, 217 during Great Contraction (and RFC financing), 275–76, 278 1920s sales of, 163 outstanding (June 30, 1930), 263 over-the-counter market for, 198 regular and predictable auctions of (1970s), 278, 356–57, 359, 365 advent of regular offerings, 363–65 and February 1969 refunding, 357 and first note auctions, 359–60 and May 1970 refunding, 357–59 on model of bills, 333 question of reason for success of, 362–63 subsequent auctions, 360–62 scaled allotments for, 203 statutory limitations on, 167 tax-date sales of, 337 in World War II, 341, 342 Treasury policy, finance vs. monetary, 141n Treasury receipts and expenditures (1920s), 159 Treasury receipts and expenditures (1927– 1939), 236, 237 Treasury securities. See also specific types of security market for, 1 overpricing of, 205, 206 regular and predictable basis for, ix, 4, 6, 278, 333 (see also Regular and predictable issuance) sales volumes of, 199, 200

393

Index

secondary market for, 185 (see also Secondary market) types of, 2 in post-WWI period, 7–8 yields on, 264, 280 Treasury Tax and Loan system, ix War Loan Deposit Account system as forerunner of, 5 Treasury War Loan Deposit Accounts, 5, 7, 8, 109, 111, 199, 204, 207–209, 210, 213, 216, 307, 308, 324–25, 330 Truman, Harry, 348 Trust funds, government, 247–53 Tugwell, Rexford 226 Underpricing allocation problem with, 8, 200–207 and exchange offers, 312 of New Deal offerings, 308 and oversubscriptions, 310 (see also Oversubscriptions) Underwood–Simmons Tariff Act (1913), 57, 71 Unemployment Trust Fund, 247, 253, 254, 259 United States Government Life Insurance (USGLI) Fund, 247, 250, 251n U.S. notes (“greenbacks”), 14, 15, 21, 319, 331 U.S. Treasury securities. See Treasury securities Vandenberg, Arthur, 265–66 Vanderlip, Frank, 42 Veterans’ bonuses (WWI), 236, 247, 248–49, 265 loan program from, 251–52, 265–68 Veterans’ Bureau, service certificates issued by, 265 Victory Liberty Loan. See under Liberty Loans Victory Liberty Loan Act (1919), 82, 167 ceiling on outstanding certificates set in, 110 section 6 of, 156 Victory Loan drive (WWII), 342, 343, 349 Victory notes, 161, 162, 167, 174 and bond purchase fund, 164 and exchange offers, 166, 169, 182 and preference rate, 188 redemption of, 161, 162–63, 165, 172 refinancing of, 168, 169 retirement of, 165–66 and taxes, 82–83 Vietnam war, 358 Volcker, Paul, 293, 357, 359–60, 361, 365

Walsh, David, 323 War bonds, for Spanish–American War, 41–44 War Loan Deposit Account, 5, 7, 8, 109, 111, 199, 204, 207–209, 210, 213, 216, 307, 308, 324–25, 330 War Loan drives (WWII), 342, 343 War profit taxes, for WWI financing, 60, 61–64 War Revenue Act (1898), 29, 313 Section 33 of, 41n, 42n War Revenue Act (1917), 58–59, 60, 61, 62, 78 Title VIII of, 193 War-savings stamps and certificates (World War I), 66–67, 154 Wartime discount rate policy, 137–40 When-issued transactions, 38 Wilson, Woodrow and Liberty Loans, 69–70, 96, 108 on post-WWI tax reduction, 147 and World War I, 50 for WWI taxes, 60, 63, 79–80 Winston, Garrard, 174, 175, 217 Wire transfers, 195 Wolcott, Edward, 42 Woodin, William, 2, 232, 240n, 279, 281, 307, 333 World War Adjusted Compensation Act (1924), 248–49, 251n World War I, 7, 49–51 American participation in, 50–51, 52–53 financing of, ix, 4, 5, 7, 53–67 (see also Liberty Loans) bonuses promised to veterans of, 236, 247, 248–49, 265 costs of, 51–53 debt management during, 143–44, 337 certificates of indebtedness for, 114–15 and statutory limits, 313–14 German 1918 offensive in, 60 national debt at end of, 7, 147 unexpectedly sudden ending of, 72, 85 World War II, 338–39, 344 financing of, 337, 341–42, 343 and interest rate stabilization program 339–41 Wyatt, Walter, 194, 232n Yeo, Edwin, 367 Young, Owen, 266–67 Zero-coupon bills, 212 Zero-coupon service certificates, from Veterans’ Bureau, 265 Zero-coupon Treasury bonds, and World War Adjusted Compensation Act, 249