Exchange rate regimes and macroeconomic management in Asia 978-9622090538

Citation preview

in Asia

in Asia

To

Max (Cavoli) and Shreyas (R吋 an)

in Asia

Tony Cavoli and Ramkishen S. Rajan

禮品港式學出瓶社

HONG KONG UNIVERSITY PRESS

in Asia

-MK

。、d

ρiu

DA vd3d I eo

瓜

ι 月

曲

的開恥

ng

d

ln

g 叮叮 nkg

。 Tony

WFU

加訕訕

到 wmg

bh 恥伽 tM

HK7AH mMUKm

Cavoli and Ramkishen S. Rajan 2009

ISBN 978-962-209-053-8

All rights reserved. No portion of this publication may be reproduced or transmitted in any form or by any mean日， electronic or mechanic祉， including photocopy, recording , 。r any information storage or retrieval system, without permission in writing from the publisher.

pb一官。

止 τμ

-H -H 一

O-NF c-m I-n-E C-u o-w P-w H-o EM r-1-I

cd-LH J-c71 u-PA

British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Libr征y.

Print巴d

in Asia

and bound by Unit巴d League Graphic & Printing Co. Ltd. , Hong Kong , China

Contents

For巴word

Vll

Preface and Introduction

lX

Acknowledgements

Xlll

Section 1: Exchange Rate Regimes and Monetary Frameworks in Asia 1.

Asian Exchange Rate Regimes a Decade since the 1997-98 Crisis

2.

Open Economy Inflation Targeting Asian Econornies

in the Crisis-hit East

39

3.

Estimating Monetary Policy Rules for Selected East Asian Econornies

51

4.

Estimating the Flexibility of the Indian Rupee

63

5.

Characterizing Singapore's Exchange Rate Policy

75

R巴拉mes

Section 2: International Capital Flows , Financial Integration and Macroeconomic Management in Asia 6.

in Asia

Capital Inflows Surges and Monetary Pre-Crisis

Managem巴nt

in East Asia

3

89

91

Contents

VI

7.

lntemational Capital Flows in East Asia: Trends , Pattems and Policy lssues

107

8.

lnternational Capital Flows and Macroeconomic Policy Options for Small and Open Asian Economies

131

9.

Financial lntegration in East Asia: What Does the Literature Tell Us?

145

10. What Is the Degree of Monetary China and Hong Kong?

Int巴 gration

between Mainland

169

Notes

189

Reference

209

Subject Index

229

in Asia

Foreword

Asian currency and financial developments are no longer relevant just for area specialists. This was dramatically brought home by the global repercussions of the Asian crises in 1997-98. While Asian financial relationships with the rest of the global economy are much quieter now , there are some striking parallels between the situation of the United States today and many of the Asian economies in the years just before the crises hi t. Large current account deficits financed by huge capital inflows and probl巴ms in domestic financial systems that had been allowed to go unchecked for much too long are the most obvious of these parallels. We have also learn巴 d ， if we did not know before , that large private capital inflows are not a sure sign that everything is right with an economy. In contrast to the highly sc巴ptical evaluation of credit risks that we hav巴 traditionally associated with international bankers , we have discovered that they often herd together continuing to give th巴 benefit of the doubt to suspect foreign borrowing as long as their fellow investors are doing the same. This is a world in which market perceptions ofth巴 fundamentals can change quickly , and yester由y's capital inflows can suddenly become today's outflows. Some influ巴ntial economists have pointed to strong similarities between today' s international monet訂Y upturns and the initial postwar international monetary regime negotiated at Bretton Woods. While the early advocates of this optimistic view assumed that the situation today was most like the initial days of Bretton W oods with decades of prosp位 ity ahead, others have pointed to similarities with the last days of Bretton W oods such as the dangerous US fiscal position. These developments and controversies illustrate the strong need to keep an open mind in assessing international financial developments and the dangers of becoming blinded by the rigid adherence to particular th巴oretical models. This volume by Tony Cavoli and Ramkishen S. Rajan provides an excellent example of how open minded car巴ful empirical research and the thoughtful posing of theoretical issues can help the functioning of national policies and their interactions in both the regional and global contex t. The read巴r will find no

in Asia

Foreword

VIII

oversimplified answers to the great questions of international monetary strategy but what will be found is careful study of a number of key aspects of Asian and international monetary and financial issues. Let me off，巳r an exampl巴. The growth of international capital flows has led many market analysts and especially academic economists to assume that except for the low巳 st income countries we live in a world of virtually complet巳 mt巳gratlOn of international money markets. But there is a huge difference between substantial and perf，巴ct capital mobility and Cavoli and Rajan show convincingly that it is the former rather than the latter world in which most Asian economies Ii ve. While international capital mobility is substantial, it is not so high that most Asian countries cannot sterilize the effects of capital inflows on national monetary aggregates in th巴 short and medium terms. Cavoli and Rajan find that this holds both b巳for巴 and after the Asian crises. An important implication is that we should not assume that we have an automatically self correcting international monetary system as is implied by strict adherenc巴 to the monetary approach. Cavoli and Rajan also provide important research on the evolution of Asian exchange rat巴 regim巴 s. While som巴 have argued that in the past crisis period most Asian countries have reverted to a soft dollar standard , Cavoli and Rajan show convincingly that this is not the case. But free floating has not been th巴 choice of the day either. Thus discretionary exchange rate management has become the regime of choice. Recent research on the instability of intermediate exchange rate regimes in a world of high capital mobility finds that countries do not necessarily need to go all the way to one of th巴 corners of hard fixes or free floats to substantially reduce the incidence of currency crises. However , time inconsistency problems make it politically difficult to manage in the middle in a stable manner. A k巴Y to such successful management is to minimize the role of political pressures on exchange rate policies. Cavoli and Rajan' s cogent cas 巴 study of Singapore' s exchange rate policy illustrates both that fairly small open economies can live successfully with a flexible exchange rate and that it is possible to manage such a regime along the lines of a basket based crawling band without creating the types of inconsistencies between exchange rate and monetary policy that invite speculative attacks. These are just a few of the rich menu of topics addressed by Cavoli and Rajan. It is particularly impressive how clearly their analysis is presented even when th巴 underlying issues are quite technical. This makes the volum巴 a valuable contribution for government officials , private sector analysts , and students as well as researchers.

們δ

且

泊

巾。

C

尸lw

FMmTM no m

PTi--inu

O

趴 mdd

)I'

uodv

H

in Asia

，扭扭

wb 間 G

DKm 凹 的 FMm

mωcn

THTC bobt cIEU

Preface and Introduction

With the ris 巴 of China , India and re-emergence of East Asia from the financial crisis of 1997-98 , there has been a significant interest in understanding ongoings in monetary and exchang巴 rate issues in the larger Asian region. 孔10netary issu巴S in Asia have taken on particular prominence as the region holds the largest international reserves in the world and consequently plays a significant role in global macroeconomic imbalances. In addition , there are also a great variety of monetary policy regim的 at play in th巴 region refl巴cting 巴ach country' 日 ne巴ds and policy preferences. This volume consists of t巴n chapters , equally divided into two sections , viz. exchange rate and monetary policy issues in Asia, and international capital flows , financial integration , and macroeconomic management in Asia. Issues explored in this volume include Asia' s exchange rate regim品， including the extent of exchange rate flexibility in Asia , the region's pre- and post-crisis experience with using monet訂y policy to manage capital inflows , the increasing degree of integration of Asia's financial markets. Some of the chapters in this book are considerable updates of journal articles already published by the authors and have thus gon巴 through a process of vigorous refereeing and editing. In addition , a few chapters have been specifically written for the volum巴 to ensure that all relevant issues are systematically covered. The chapters in this book are focussed on import創1t monetary and exchange rate policy issues of contemporary relevance , but are inform巴d by data and rigorous empirical analysis. We have also provided a number of key references in each chapter in order to document the arguments made , but also in case inter巴 sted readers want to follow up on the issues discussed. While each of the chapters can be read as self-contained papers , we have tried to ensure minimal overlap between the chapters and endeavoured to make them as seamless as possible for read巴rs interested in working through the book chapterby-chapter. We summarize each ofthe ten chapters below.

in Asia

x

Pr~fáce

and 1ntroduclio l1

Chapter 1 is entitl巴d “Asian Exchang巴 Rate Regimes a Decade since the 1997-98 Crisis". An immediate lesson that many observers appear to have drawn from recent financial crises in emerging market economies in the 1990s is that the only viable exchange rate option boils down to one between flexibili旬， on the one hand , and “ credible pegging" , on the other. According to this view (which was dominant in the late 1990s and early 2000s but still has a number of followers) , emerging econorni巴 s have to gravitate to one of these two extremes. Any currency arrangements that lie in between these polar extremes or corners (i.e. those in the “ middle") are viewed as being inh巴rently unstable and crisis-prone. This chapter compiles and discusses the de jure or official exchange rate regimes in the various Asian economies. Recognizing that countries do not always follow their policy pronouncements , the chapter also reviews the evidence regarding the de 戶 cto or actual exchange rate regimes in s巴l巴cted Asian countries , particularly Indonesia, Thailand , Korea and the Philippines. Chapter 2 is entitled “ Op巴n Economy Inflation Targeting Regimes in the Crisis-hit East Asian Economies". Since the Asian crisis of 1997-98 it has been recognized that exchange rate and monetary policy strategies must involve a “ fairly high" element of flexibility rather than a single-minded defense of a p副ticul紅 rate One way thi 日 flexibility might be introduced is by a country adopting an open economy inflation targeting arrangemen t. This particular policy regime has been implem巴 nted in s 巴 veral Asian countries in recent y 巴 ars ， but the normative implications of inflation targeting appear at times to be at odds with the requirements I巴garding exchange rate flexibility. This chapter pr巴 sents an analysis of som巴 of the issues relevant to central banks in small and open economies implementing an inflation targeting arrangement with specific focus on the role of the exchange rate Chapter 3 is entitled “ Estimating Monetary Policy Rules for Selected East Asian Econornies". Monetary policy rules have become an accepted way to ascertain the revealed preferences of central banks' monetary policy regimes by estimating how the instrument of monetary policy reacts to key economic variables. In the recent literature , monetary policy rules where the nominal interest rate is the policy instrum巴nt is closely associated with inflation targeting as th

in Asia

Pref，μ

and

lntroduc lÌofl

XI

range of measures so that as many of the salient characteristics are captured , as well as to ensure the robustness of the results. While the Reserve Bank of India (RB I) is comrnonly beli巴ved to target the real effective exchange rate (REER) , the results in this chapter indicate that the Indian rupee is predominantly influenced by the US dollar, with th巴 euro slowly gaining in significance as wel l. Chapter 5 is entitled “ Characterizing Singapore' s Exchange Rate Policy". Chapters 1 and 2 stress that there can be a significant divergence in the de facto versus de jure exchange rate regimes operated by economies. While much of the recent literature including chapter 2 has focussed on the crisis-hit 巴 conomies ， Korea and Thailand in particul缸， scant attention has been paid to Singapore , which officially targets its nominal effective exchange rate or NEER (around a band). This chapter examines the degree of exchange rate intervention for Singapore using various methods of assessing defacto exchange rate regimes. In the main we show that while the Singapore dollar is primarily influenced by the US dollar , in keeping with its de jure classification of a basket pegged regime , other major currencies such as the yen and the euro also impact the Singapore dollar. There is also evid巴nce to indicate that Singapore uses the NEER strategically as a policy instru l11ent to satisfy dOl11estic inflation objectives. Chapter 6 i日 entitled “ Capital Inflows Surges and Monetary Management in East Asia Pre-Crisis". This chapter develops a si l11ple 1110del to exal11ine the reasons behind th巴 capital inflow su您的 into selected Asian economi巴 s in the 1990s prior to the financial crisis of 1997-98. The analytical l110del shows that persistent uncovered int巴rest differentials and consequent capital inflows l11ay be a result of cOl11plete 1110netary sterilization, perfect capital l11obility, sluggish response of interest rates to domestic monetary disequilibrium, or some cOl11bination of all three. Using the 1110del as an organizing fra l11ework, the chapter undertakes a series of related sil11ple el11pirical tests of the dynal11ic links betwe巴n international capital flows , the extent to which they are sterilized and uncovered interest rate differentials in the five crisis-hit economies (Indonesia, Korea , Malaysia, the Philippines and Thailand). Chapter 7 is entitled “ International Capital Flows in East Asia: Trends , Patterns and Policy Issues". While th

in Asia

XII

Pr~fáce

and 1ntroduclio l1

into and out of the region. The chapter also examines the 巴 xtent of monetary sterilization and the degree of capital mobility in selected emerging Asian econoruies. Chapter 8 is entitled “ International Capital Flows and Macroeconoruic Policy Options for Small and Open Asian Economies". Asia has benefited greatly by being relati、rely open to global trade and financial flows. However, this op巴nness also leaves the region somewhat vulnerable to a variety of external shocks that could spillover to the domestic economy. How should monetary authoriti巴 s respond to various exogenous shocks , and what does the response depend on? This is the S巴t of questions 巴xplored in this chapter within the context of a simple analytical framework. The framework used is a modified version of the age-old Swan diagram. Th巴 merit ofth巴 Swan diagram is that it helps us think through various combinations of policies needed for an economy to maintain or regain internal and external balance simultaneously in the face of various shocks. Chapter 9 is entitled “ Financial Integration in East Asia: What Does the Literature T巴11 U s?" This chapt巴r offers a s巴lective survey of the recent empirical literature on financial integration in East Asia , the focus being on alternative definitions of financial integration and measurement issues and results. In particul缸， this paper concentrates on the ASEAN-5 plus 3 or the APT economic group (i.e Indonesia, Thailand , Malaysia, Philippines , Singapore , Korea, China and Japan) as well as Hong Kong and Chinese Taipei. These are the economies that have consciously att巴mpted to intensify intraregional mon巴tary and financial cooperation in the last few years , particularly since the East Asian crisis of 1997-98. While ther巴 is a large body of literature on the measurement of financial integration , we exarnine the most widely used measures and place them into three broad categories. The first category refers to the price based conditions involving mainly debt flows These are largely embodied in the interest parity conditions. The second category involves quantity based measures such as savings-investment corr巴 lations and consumption correlations , current account dynamics and gross capital flows. The third category can be broadly classified as regulatory or institutional factors (such as capital controls and prudential regulations) , as well as non-debt flows such as the co-mov巴ment of stock

in Asia

Acknowledgements

Much of the work contained in this book was undertaken while the first author was at the School of Economics and Finance , Queensland University of Technology , Brisbane , Australia. As such , he would like to acknowledge the support of his colleagues and the generous provision of resources by the School during that time. He would also like to thank his cu汀巴nt plac巴 of employment, School of Commerce and Centre for Regulation and Market Analysis at the University of South Australia in Adelaide , for offering the n巴cessary support that helped in the timely completion of this projec t. Th巴 second author would like to acknowledge th巴 support of his colleagues and resources provided by his current place of employment, the School of Public Policy at George Mason University (SPP-GMU) in Virginia , USA as well as the School of Economics , University of Adelaide where he holds a Visiting position Both authors would also like to acknowledge th巴 valuable research students who worked on various chapters: Jeff Kim , Surabhi Jain , Alice Ouyang , Sadhana Srivastava , and Nicola Virgil l. Pritam Banerje巴 helped with the final editing of the book. The support extended to us by Colin Day of Hong Kong University Press in over-seeing this project from the proposal stage to final publication is also much appreciated. Lastly , but most importantly , our family members have remained unstinting in their support of our respective careers and have provided us the stability necessary to remain focused on our writings. Tony Cavoli Ramkishen S. Rajan March 2009

in Asia

in Asia

三

in Asia

Exchange

Rate Regimes and Monetary Frameworks in Asia

in Asia

Asian Exchange Rate Regimes A Decade since the 1997-98 Crisis1

1. Introduction An immediate lesson that many observers appear to have drawn from the financial cnses m em巳rging market economies in th巴 1990s i日 that th巳 only viable 巴xchange rale oplion boils down lo one belween Oexibilily on lhe one hand , and “ credible pegging" on th巳 other. According to this view (which was dominant in the late 1990s and early 2000s but still has a number of followers) , emerging economies have to gravitat巳 to one of these two 巳xtremes. Any currency arrangements that lie in between these polar extremes or comers (i.e. , those in the "middle") are viewed as being inherently unstable and crisis-pron巳. It used to be commonly believed that this so-called “ bipolar view" drew analytical support from the "Impossible Trinity" which 巴 ssentially states that a country with an open capital account cannot simultaneously conduct a completely ind巴pendent monetary policy and pursue a completely rigid or fixed exchange regime. However , the Impossible Trinity does not on its own imply that in an mcr巳 asingly globalized world economy an interm巴diat巴 regime is unviabl巴， or that countries will be compelled to abandon the middle ground. 2 For instance , a country could choose to maintain an interm巳diat巳 exchange rat巳 regime while forsaking a degree of monetary policy autonomy. In other words , the analytical basis in support of the bipolar view is rather weak (particularly since some dev巳loping countries still maintain capital controls that are not entirely porous). Indeed, the only analytical support offered against intermediate regim巳 s is their lack of v巳rifiability or transparency; simple regimes are more verifiable by market participants than complicated ones (Frankel , Schmukl巳r and S巳rv巳n ， 2000). The other commonly repeated weakness of intermediate regimes is that they are more crisis-prone (Bubula and Otker-Robe , 2003). How叮叮， a more careful 巳xamination of the links betwe巳n de facto exchange rate regimes and cu汀ency crises suggests that there is no evidence that either of the two comers is n巳c巳 ssarily less crisis-pron巳 than interm巴 diate regimes in genera1. 3

in Asia

4

Eλ change

Rate Regimes

Theremaind巴r of this chapter in organized as follows. The next section compiles and discusses the de jure or official exchange rate regimes in various Asian economies. 4 R巴 cognizing that countri 巴 s do not always follow their policy pronouncements , section 3 presents some simple de facto exchange rate regime measures for selected Asian countries. Since different measures inevitably capture different characteristics of any regime , it is critical to use a number of methodologies as an attempt to provide some robustness to the results. To preview the main conclusions from sections 2 and 3, it is evident that Asia is home to a wide array of exchange rat巴 reglm肘， though there are signs of gradual movement towards somewhat greater exchange rate flexibility in many Asian countries. However, the propensity for foreign exchange intervention and exchange rate management among regional central banks remains fairly high in many instances. Section 4 discusses th巴 analytical reasons that have motivated th巴 mov巴 towards greater exchange rate flexibility in Asia in general , but also the reasons why many Asian economies continue to remain circumspect about adopting a flexible regime per se. Section 5 concludes the chapter.

2. Official Exchange Rate Regimes in Asia Until 1998 it was fairly easy to obtain de jure exchange rate classifications as this data was compiled from national sources by the lMF. Specifically , between 1975 and 1998 th巳 IMF' s Annual Report on Exchange Arrangements and Exchange Restrictions was based on self-reporting of national policies by various govemments with revisions in 1977 and 1982. Since 1998 - and in response to criticisms that there can be significant divergences between de facto and de jure policies - the IMF' s exchange rat巴 classification methodology has shifted to compiling unofficial policies of countries as determined by Fund staff. 5 While the change in lMF exchange rate coding is welcome for many reasons (including the fact that th巴 new set of categories is more detailed than the older one) , the lMF is no longer compiling the dejure r巳gimes. The only way this can be done is by r巴ferring to the website of each central bank or other national sources individually and wading through relevant materials. Th巴 results of this det巳ctive work are summarized in Table l. 6 As is appare訓， the de jure exchange rate regimes in Asia span a wide spectrum. A number of smaller Asian economi巳 s app巳征 to prefer som巴 form of single currency pegs. This is true of the Hong Kong SAR (whose currency board arrangement is pegged to the US dollar) , Brun巴i(pegg巴d to th巳 Singapore dollar) and Bhutan and Nepal (pegged to the lndian rupee). ln contrast , Bangladesh , Sri Lanka and the crisis-hit economi巴 s of Indonesia, Korea , the Philippines and Thailand officially operate flexible exchange rate regimes. Th e flexible exchange rates in the four East Asian countri的 are accompanied by inflation targeti月 frameworks (s自 chapter 2). A number of other Asian countries have adopted a variety of intermediate regimes (currency baskets , crawling bands , adjustable pegs and such). For instance ,

in Asia

Asian Exchange Rale Regimes a Decade since the 1997-98 Crisis

5

Table 1 De jure Exchange Rate Regimes in Asia (As per Country Central Bank Websites unless otherwise stated) Country

Official Policy Pronouncements (direct quotes)

Bangladesh

The exchange rates of the taka for inter-bank and customer transactions are set by the dealer b臼1ks themselv郎， based on demand-supply interaction The Bangladesh Bank i日 not present in the market on a day-to-day basi日 and undertakes purchas巳 or sale transactions with th巳 dealer banks only as needed to maintain orderly market conditions

Bhutan

Except for the lndian rup∞ to which the ngultrum is pegged at parity , and which circulates freely in Bhutan , paying or receiving payments in any other foreign currency for 仕ansactions in Bhutan is illegal The Government may , by order, at any time , on the recommendation of the Board , declare an external value for the ngultrum , having due regard for the obligations which Bhutan has assumed in accordance with th巴 provisions of any international monetary agreement to which it is a party , or to which it has adhered

BruneiA currency interchangeability agrl∞ment was established between Singapore Darussalam 1 and Brunei Darussalam, which remains in effect till today and continues to play a central role in relations between the two countries. This agreement allows both countries to interchange their currencies at p位 without either country running the risk of currency exchange rate fluctuations which thus further facilitates trade and commerce between th巳 two countries. The individual currencies are acceptable as customary tender when circulating in the country in which they are not legal tender. Cambodia

N.A

China

China announced on July 21 , 2005 the adoption of a managed floating exchange rate regime based on market supply and demand with reference to a basket of currencies. Since then , the new exchange rate system has operated stably, and the RMB exch但1ge rate has been kept basically stable at an adaptive and equilibrium leve l. The exchange rate of th巳 RMB against the US dollar has been moving both upward and downward with greater t1 exibility.

HongKong SAR

Since 1983 the Hong Kong dollar has been linked to the US dollar at the rate of HK$7.8 to US$ 1. The link is maintained through the operation of a strict and robust Currency Board system which requires both the stock and the t1 0w of the Monetary Base to be fully backed by foreign reserves. Any change in the size of the 孔10netary Base has to be fully matched by a corresponding change in the foreign reserves.

India

The exchange rate policy in recent years has been guided by the broad principles of careful monitoring and management of exchange rates with t1exibili句， without a fixed target or a pre-announced target or a band , coupled with the ability to intervene if and when necessary.

in Asia

Eλ change

6

Indonesia

Rate Regimes

In July 2005 , Bank Indon巳 sia launched a new monetary policy framework known as the Inflation Targeting Framework , which has four basic elements 的 follows: (l) use of 出e BI rate as a reference rate in monetary control in replacement of the base money operational target , (2) forward looking monetary policymaking process , (3) more transparent communications strategy , and (4) strengthening of policy coordination with the Government The rupiah exchange rate is determined wholly by market supply and demand , However, Bank lndonesia is able to take some actions to keep the rupiah from undergoing excessive fluctuation ,

Korea

Inflation targeting is an operating framework of monetary policy in which the central bank announces an explicit inflation target and achieves its target directly , This is based on the recognition that to achieve sustainable economic growth , it is important above all else that inflation expectations , which have a great effect on wage and price deci日lOn日， should be stabilized. In this regard , inflation targeting places great emphasis on inducing inflation 巴xpectations to conv巳rge on the central bank' s inflation target level by the prior public announcement and successful attainment of that target level The exchange rate is , in principl巴， decided by the interplay of supply and demand in the foreign exchange markets. However , the Bank of Korea implements smoothing operations to deal with abrupt swings in 由e exchange rate caused by temporary imbalances between supply and demand , or radical changes in market sentiment

Laos

The Bank of Laos announces the exchange rate derived from the market and officially adjusted , bas巳d on the daily average trading rate of th巳 int巳rbank market to the commercial banks and the foreign exchange bureaus as a reference to determ the Bank of Laos determines the exchange rate on its own for the commercial banks and foreign exchange bureaus for implementation.

Malaysia

On July 訓， 2005 , Malaysia shifted from a fixed exchange rate regime of USDl =R孔13.80 to a managed float against a basket of currencies. Under the managed float system, the ringgit exchange rate is largely determined by ringgit demand and supply in the foreign exchange market. Th巳 Cen仕al Bank does not actively manage or maintain the exchange rate at any particular level 一- economic fundamentals and market conditions are the primary determinants of the level of the ringgit exchange rate. In this regard , the Central Bank intervenes only to minimiz巳 volatility ， and to ensure that the exchange rate does not become fundamentally misaligned

Myanmar

N.A.

Nepal

In the review year , the exchange rate of the Nepalese rupee vis-à-vi日 the Indian rupee remained constant, and NRB intervened 44 times in the foreign exchange market. Currently , Nepal is adopting a dual exchange rate 缸Tangement. It i日 dual because the Nepali currency is pegged to the Indian currency (IC), whereas it floats with the convertible currencies. This system 。f exchange rate was introduc巴d on February 12 , 1993.

in Asia

Asian Exchange Rale Regimes a Decade since the 1997-98 Crisis

7

Pakistan2

Pakistan has adopted the floating inter-bank exchange rate as the preferred option since 2001. State Bank of Pakistan has attempted to maintain real effective exchange rate at a level that keeps the competitiveness of Pakistani exports intac t. But, like other Central B ank:s, it does intervene from time to time to keep stability in the market and smooth excessive fluctuations. The current framework of monetary-cum-exchange rate policies and the underlying economic analysis in Pakistan can , thus , be broadly ch訂acterized as judgment - and discretion-based rather than model - or rule-based.

Philippines

The primary 0吋叫ive of Bangko Se凶al ng Pilipinas' monet叮 policy is to promote a low and stable inflation conducive to a balanced and sustainable economic growth. The adoption of inflation targeting framework for monetary policy in J但lU但-y 2002 is aimed at achieving tbi日 objective The Monetary Board detennines the exchange rate policy of the country , determines the rates at which the Bangko Sentral buys and sells spot exchange , and establishes deviation limits from the effective exchange rate or rates as it deems proper.

Singapore

Since 1981 , monetary policy in Singapore has been centred on the management of the exchange rate. (1) The Singapore dollar is managed against a ha只ket of cu付'enCle肉。fits m 貝jor trading pa叫 ners and competitors (2) The Mo閥的 Authority of Si月 apore operates a managed float regime for the Singapore dollar. The trade-weighted exchange rate i日 allowed to fluctuate within an undisclosed policy band , rath巳r than kept to a fixed value. (3) The exchange rate policy band is periodically reviewed to ensure that it remains consistent with the underlying fundamentals of the economy (4) The choice of the exchange rate as the intermediate target of monetary policy implies that MAS gives up control over domestic in仗時 st rates (and money supply)

SriLanka

The Central Bank continues to conduct its monetary policy under an independently floating exchange rate regime within a framework of targeti月 monetary aggregates with reserve money (i 已， high powered money) as the operating target and broad money (M2b) 的 the intermediate target

Taiwan

Prior to February 1979 , management of foreign exchange in Taiwan was characterized by a central clearing and settlement system. Following the establishment of the Taipei Foreign Exchange Market in February 1979 , a flexible exchange rate system was formally implemented. Since then , the NT dollar exchange rate has been determined by the market. However , when the market is disrupted by seasonal or irregular facto郎， the Bank will st巳pm.

Thailand

Since July 2, 1997 , Thailand has adopted the managed-float exchange rate in which the value of the baht i 日 determined by mar1主et forc郎， namely demand and supply in both on-shore and off-shore foreign exchange market, to let the currency move in line with economic fundamentals. The Bank of Thailand will intervene in the market only when necessary , in order to prevent excessive volatilities and achieve economic policy targets repm巴，

Under the in目前的n targeting framework , the Bank of Thailand implements its monetary policy by influencing short-term money market rates via the selected key policy rate , currently set at the 14-day repurchase rate

in Asia

Eλ change

8

Vietnam

Rate Regimes

Vietnam has adopted a crawling peg with the US dollar for its exchange rate. The State Bank of Vietnam sets the official exchange rate daily , and commercial banks set their dealing rate within a trading band of plus or minus 0 之5 percent. The State Bank of Vietnam tends to keep the dong depreciated against the US dollar by keeping the exchange rate on an upward trend.

1) Based on information available from Brunei Ministry of Finance. http://www .tïnance gov.bnlbcb/b cb_index.htm.

2) Bas巳d on speech by former Pakistan central bank Governor (Husain , 2005) Source: Compiled by author with assistance of Nicola Virg i11 from websites from various central banks and other official sources with minor modifications. Central Bank websites available here: http://www.bis.org/c banks.htm.

according to the Reserve Bank of India (RBI) , India “ rnonitors and manages the exchang巳 rates with flexibility without a fixed target or a pre-announc巳d target or a band , coupled with the ability to intervene if and when necessary".7 Vietnam officially maintains a crawling peg and band around the US dollar. Singapore officially manages its currency against a basket of currencies , with the tradew巳ight巴d exchang巳 rate us巴d as an interm巳diat巳 target to ensure that the inflation target is attained. 8 While Singapore' s currency basket regime follows a more strategic orientation , both China and 弘1alaysia in July 2005 officially shifted to what may be best referred to as a more mechanical version of a currency basket r巳gim巳 (i.e. ， k巴巴ping th巳 trade-weighted 巳xchang巳 rate within a certain band as a goal in and of itsel f). The remaining Asian economi凹， viz. , Taiwan , Pakistan and Laos , seem to op巳rate rather ad hoc manag巳d floats or adjustable pegs. Overall therefore , it is readily apparent that “ one-size does not necessarily fit all" when it comes to the choic巳 of 巳xchang巳 rate regim巳 s in Asia.

3. De Facto Exchange Rate Regimes in Crisis-hit Asia 3.1 Existing

Class{βcations

As noted , the IMF has replaced its compilation of the de jure exchange rate regimes with the behavioural classification of exchange rates. Th巴 new IMF coding is based on various sources , including information from IMF staff, press reports , other relevant papers , as w巴11 as the behaviour of bilateral nominal exchange rates and reserves. 9 Table 2a summarizes the definitions of various IMF exchange rate classifications. As is apparent , the 1孔1F has eight exchange rat巴 categories. Table 2b reclassifies the original IMF definitions into three broad categories , v妞，“hard peg" ,“ soft peg" and “ floating regimes". lO Table 3 categorizes Asian exchange rates based on the new IMF classifications as of July 2006.

in Asia

Asian Exchange Rale Regimes a Decade since the 1997-98 Crisis

Table 2a IMF Descriptions of Exchange

9

Rat巴 R巴gimes

Type

Description

Exchange Arrangements with No Separate Legal Tender

Th巳 currency

Currency Board Arrangements

A monetary regime based on an explicit legislative commitment to exchange domestic curr官ncy for a specified foreign currency at a fixed exchange rate , combined with restrictions on the issuing authority to ensure the fulfilment of its legal obligation. This implies that the domestic currency will be issued only against foreign exchange and that it remains fully back，巴d by foreign assets , eliminating traditional central bank functions , such as monetary control and lender-of-last-resort , and leaving little scope for di 日 cretionary monetary policy. Some flexibility may still be afforded , depending on how strict the banking rules of the currency board arrangement are.

Other Conventional Fixed Peg Arrangements

The country (forrnally or de facto) pegs its currency at a fixed rate to another currency or a basket of currencies , where the basket is formed from the Cl凹encies ofm吋 or tradi月 or financial partners and weights reflect the geographical distribution of trad巴， serVlces , or capital flows. The cum巴ncy composites can also be stand征dized， as in the case of the SDR. There is no commitment to keep the parity irrevocably.τne exchange rate may fluctuate wi也mnarrow margins of less than :!: 1 percent around a central rate , or the maximum and minimum value of the exchange rate may remain within a narr‘ow margin of 2 percent for at least three months. The monet訂y authority stands ready to m剖nt剖且由e íïxed parity through direct intervention (i.巴， via sale/purchase of foreign exchange in the mark巳t) or indirect interv巳ntion (e.g. , via aggressive use of interest rate policy , imposition of foreign exchange regulations , exercise of moral suasion that cons仕ains foreign exchange activ旬， or through intervention by other public institutions). Flexibility of monetary policy , though limited , is greater than in the case of exchange arrangements with no separate legal tender and currency boards because traditional central banking functions are still possible , and the monet征 y authority can adjust the level of the exchange rat巴， although r巳latively infrequently

Pegged Exchange Rates within Horizontal Bands

The value of the cu虹ency is maintained within certain m旺gins of fluctuation of at least :!: 1 percent around a fixed central rate or the margin between the maximum and minimum value ofthe exchange rate exceeds 2 percent. It also includes arrangements of countries in the exchange rate mechanism (ERM) of the European Monet訂y System (EMS) that was replaced with the ERM Il on January 1, 1999. There is a limited degree of monetary policy discretion , depending on the bandwidth

in Asia

of another coun住y circulates as the sol巳 legal tender or the member belongs to a monetary or currency union in which the same legal tender is shared by memb巴rs of the union. Adopting such regimes implies the complete surr巴nder of the monetary authoriti巴 s' independent control over domestic monetary policy (仔 f凹 o rn 官m ‘1 油泊叫 叫1 do a 叫lla 削r‘泣a 抓t1昀 on 的 )，

10

Eλ change

Rate Regimes

Crawling Pegs

The currency is a甸的ted periodically in small amounts at a fixed rate or in response to changes in selective quantitative indicators , such as past inflation differentials vis-à-vi日 m呵。r trading partne肘， differentials between the in f1 ation target and expected in f1 ation in m吋 or tradi月 P訂tners ， and so forth. The rate of crawl can be set to generate inflati Ol (backward looking) , or set at a pre-announced fixed rate and/or below the projected in f1 ation differentials (forward looking). Maint組ning a crawling peg imposes constraints on monetary policy in a manner similar to a fixed peg system

Exchange Rates within Crawling Bands

The currency is maintained within certain f1 uctuation margins of at least :t 1 percent around a central rate - or the margin between themaximum 阻ld minimum value of the exchange rate exceeds 2 percent - and 自己 C巳ntral rate or margins a臼 a你的ted periodically at a fixed rate or in response to changes in selective quantitative indicators. The degree of exchange rate flexibility is a function of the bandwidth. Bands are either symmetric around a crawling central parity or widen gradually with an asymmetric choice of the crawl of upper and lower bands (in the latter case , there may be no pre-announced central rate). The commitment to maintain the exchange rate within the band imposes constraints on monet征y policy , with the degree of policy independ巳nce being a function of the bandwidth

Managed Floating

The monetary authority attempts to influ巳nce the exchange rat巴 without having a specific exchange rate path or targe t. Indicators for managing the rate are broadly judgmental (e.g. , balance of payments position , international reserves , parallel market developments) , and adjustm凹的 may not be automatic. Intervention may be direct or indirec t.

、廿thNo

Predetermined Path forthe Exchange Rate

Independently Floating

The exchange rate is market-determin叫， with any official foreign exchange market intervention aimed at moderating the rate of change and preventing undue tluctuations in the exchange rate , rather than at establishing a level for it.

Source: Taken direc t1y from IMF website on Classijïcation of Exchange Rate Arrangements andMonetaη Frameworks. http://www.imf.org/extem叫Jnp/mfd/erl2006/eng/0706.htm

Table 2b Broad Categorizations of Exchange Rate Regimes Hard Pegs

80ft Pegs

Exchange Arrangem巴nts with No Separate Legal Tender (includes Do l1arization and Currency Union)

Conventional fixed pegs

Currency Board

Crawling band

Horizontal band

Floating Regimes Ind巴pendent

floats

Managed tloats

Crawling peg

Source: Using IMF de facto classifications (see Table 2a) recategorized based on Bleaney and Francisco (2005). in Asia

Asian Exchange Rale Regimes a Decade since the 1997-98 Crisis

Table 3

De 戶cto

11

IMF Exchange Rate Classifications as of July 2006

Country

As of July 2006

Bangladesh

Managed floating with no pred巳t巳rmined path.

Bhutan

Other conventional fixed peg arrangement (against a single currency).

Brunei Darussal副n Currency board arrangement Cambodia

Managed floating with no predetermined path.

China

Other conventional fixed peg arrangements

Hong Kong SAR

Currency board arrangement

India

Managed floating with no predetermined path

Indonesi且

Independently floating.

Japan

Independently floating.

Korea

Independently floating.

Laos

Managed floating with no predetermined path

Malaysia

Managed floating with no predetermined path.

扎1yanmar

Managed floating with no pred巳t巳rmined path.

Nepal

Conventional pegged arrangement (against a single currency).

Pakistan

Other conventional fixed peg arrangements

Philippines

Independently floating.

Singapore

Managed floating with no pred巳t巳rmined path.

Sri Lanka

Managed floating with no pre-determined path.

Thailand

Managed floating with no predetermined path

Vietn位n

Other conventional fixed peg

a訂angements

(again日t

a single currency)

(against a single currency)

Source: IMF data on Cla爪sification of Exchange Rate Arrangements and Monetary Frameworks. http://www .imf.org/externaνnp/mfd/er/2006/eng/0706.htm

As is

appar巳nt

from a comparison of Tables 1 and 3 ,

th巴r巳 is

no

discr巴pancy

between the de jure and de facto regimes of Bhutan , Brunei , Hong Kong SAR and Nepal , all of which op巳rat巳 fix巳d 巳xchange rat巴 s to a single currency. Similarly , India , Lao PDR and Singapore are categorized as managed floaters , broadly consist巴nt

with their official

pronouncem巳nts.

Vietnam, which

us巴d

to be in this

category , has more recently been classified as having a conventional fixed peg I 巴gime ，

in Asia

in contrast to its official

pronouncem巴nt

of maintaining a crawling

p巴g

and

12

Eλ change

Rate Regimes

band around th巴 US dollar. Bangladesh , Sri Lanka and Thailand have been characterized as managed floaters (with no predetermined exchange rate path) d巴 spite their official declarations of being independent floaters. Pakistan is defined as oper瓜1月 conventional fixed peg arrangem凹的 (against a single currency) despite proclaiming to be an independent floate r. Japan , Korea and the Philippin巴 sare characterized as independent floaters , consistent with their official assertions. 11 Contrary to the public pronouncement of the Chinese authorities that th巴 cu訂ency is a based on a currency basket , recent empirical studies suggest the de facto regime appears to be a soft p巴g to th巴 US dollar with th巴 IMF classifying China under “ other conventional fixed peg arrangemen的，\12 The 孔1alaysian ringgit since its official depegging is defined as being a managed floater with no predetermined path. This is consistent with empirical analysis which suggests that the ringgit closely tracks a trade-weighted basket since its depegging in July 2005 , not unlike the Singapore dollar. In their seminal paper, Reinhart and Rogoff (2004) develop a so-called “ natural classification" based on market information such as black mar1主et or parallel rates (rather than official rate) , the stati 日 tical behaviour of exchange rate , reserv巴 s and interest rates as well as country chronologies using a five-year window (to prevent sporadic exchange rate changes). The authors apply the methodology to 153 countries from 1946 to 2001 and find , among other thin軒， that nearly half of the “ official pegs" are bet1er characterized as managed or freely floating arrangements or limited flexibility.13 More generally , once one uses de 戶 cto classifications , the bipolar view on exchang巴 rat巴 regimes which was bas巴d largely on de jure exchange rate classification is no longer obvious. This is also borne out in the case of de facto IMF coding for Asia. Referring to Figure 1, while there has been a discernible trend towards greater exchange rate flexibility from 1998 to 2003 , there was a slight reversion to soft dollar pegs in the last two years. 14 Unlike the new IMF classification , Reinhart and Rogoff are careful to distinguish between a flexible exchange rate regime and on巴 that is fre巴ly falling rate per se. They define the latter as episodes in which the 12-month rate of inflation equals or 巴xceeds 40 percent unless ther巴 is some type of pre-announced or narrow band. The authors also define the six-month period immediately after a crisis as being fre巴ly falling if there is a sudden transition from a fixed or quasifixed to more flexible exchange rate regime. Thus , in 1998 , while the IMF codes Indonesia, Korea and Thailand as “independ巴ntly floating" , Reinhart -Rogoff more accurately characterize them as “ freely falling弋 15 Notwithstanding this difference , by and large , the 1孔1F and Reinhart-Rogoff reach th巴 same conclusion regarding the Asian currency arrangements. While the more detailed classifications of Reinhart Rogoff make it preferable to the IMF coding , th巴 latt巴r is far more frequently updated than the former.的 Somewhat surprisingly , both the IMF and ReinhartRogoff coding characterize Japan and Korea as independ巴ntly floating despite the sharp reserve build-up in both countries (Willett, Ki m and Nitithanprap訟， 2005)

in Asia

Asian Exchange Rale Regimes a Decade since the 1997-98 Crisis

13

1(;

14

12

10

6,

倉

1998

1999

企

金

企

倉

倉

企

2 渦。

200 1

2001

2 間可

2004

200多

---A- h 0) is referr巳d to as “flexibl巴" inflation targeting. Output is secondary to the inflation target, but the weight on the output objective prevents excessive volatility in output and d巳lays th巳 attainment of the inflation targe t. Ifλ1 > 0 and ~ = 0 , it is referred to as strict inflation targeting where the inflation target is achieved at any cos t. We r 巳 examm巴 the issue of strict versus fl巳xible inflation targ巳ting more broadly in Section 5.

3. The Role of Exchange Rate under Inflation Targeting 3.1 A Simple Model For much of this last decade , the literature on MPRs has developed in a closed economy context (Ball , 1997 and Svensson, 1997). It is only recently , when inflation targeting has been suggest巴d as a s巴rious policy option for small and open emerging economies , that research has begun to focus on rules in open economy models and , cons巴quently ， the role of the exchange rate. For instance , Fischer (200 1) notes that “ in most countries , even those with floating exchange rate regimes , monetary policy is likely to respond to som巴 extent to movements in the exchange rate" (p. l3)

in Asia

Open Ecof! omy Injiation Tärgeling Regirnes in the CrisÎs-hil

E，此st

Asian Economies

43

Conventionally an inflation targeting arrangement ought to be accompanied by a flexible exchange rate , with the interest rate used as the monetary policy instrument. It is generally recognized that for small and open economies in Asia and elsewhere , fluctuations in th巴巴 xchang巳 rate can hav巳 significant and direct impact on the domestic economy. In particular , assuming a significant degree of pass-through from 巴xchange rate changes to domestic inflation , it has sometimes b巴巴 n argued that exchange rate f1 uctuations ought to be explicitly incorporated in any MPR.9 We can evaluate this and oth巳r lSSU巳 s with the aid of the following simple modepo }'t+1

= 戶lYt

-

ß2 rt

-

ß3 e t +t f+

π叫 =π ， +αIY r 一 α ， (e， -et-l)+1J,+

er

=er, +V,

(2)

(3) (4)

wher巴 r is th巳 real interest rate , e is the real exchange rate (increas巳 implies appreciation of domestic currency), E， μand v represent demand, inflation and (rcal) cxchangc ratc shocks rcspcctivcly. (2) is an opcn cconomy IS curvc whcrc output is determined by its own lag, the real interest rate (the traditional transmitter of policy in a closed 巳conomy inflation targ巴ting system) and the real exchange rate. (3) is a conventional Phillips relation exhibiting inflation persistence and where output and the real exchange rate 巴xplain the rate of inflation next period. To be more specific about the exchange rate transmission channel in (2) and (3), a fall in e (real depr巳ciation) leads to a higher inflation dom巴 stically (pass-through) as well as boosts net exports and thus outputY (4) is a reduced form relationship between the real exchang巳 rate and the real int巴rest rate. We can specify a loss function similar to (1), where the central bank positions its policy instrument to minimize inf1 ation and output d巴viations n巴xt period. 12 Given the quadratic nature of (1) and the linear constraints (2) to (4) ，的 is convention , we can expect the optimal MPR to also be linear. By substituting (4) into (2) and (3) , re-substituting the resulting equations into (1) and differentiating with respect to r" w巴 can derive the values ofl"， 元，此 andfe in (5) 的 follows:

們 =jππ，+ fvY'

λV， + λ e，_↓

(5)13

where

j~ = π

九αzθ

九 α;θ2 +λ， A'

r 一 λα1α2θ+λ， A戶1 J

in Asia

rλ1α;θ'+λ， A'

(6)

(7)

44

Eλ change

Rate Regimes

4-2

之 -J

-一勻心

明 -A

λ 一十

α-α

仇。 -AO

吋缸司三門Jhn1】

(9)

一­

川九 αiθ2 +λ2AZ

f

「 人 7A

(8)

Jj

3-2

r 一九 αjθ

wh巳r巳 .A= 戶 2 +扎 θ

3.2 Fear 01 Floating S 巳veral observations can be made about the role of th巴 exchange rat巳. As long as the central bank is committed to pursuing monetary policy as specified by the 10ss function in (1) (i.e. and ~ > 0) the optimal rule will suggest that the instrument of policy will a1ways react , in some way , to the rea1 exchange rate even if the central bank is a strict output targ巳ter. Th巳 key point to recognize h巳r巳 1日 that it is optima1 for the centra1 bank to respond to exchange rate movements insofar as any exchange rate shock aff，巳cts its ability to reach its targe t. Henc巴， for an inflation targ巴t巴r the centra1 bank will react to exchange rate shocks in the process of achieving the inflation targe t. This particular behavior by the c巴ntral bank is misinterpreted as exhibiting fear of floating (Eichengreen , 2001). Nonethe1ess , by observing the composition offv by (8) , it i日 clear that the reaction of the instrument to the exchange rate is not re1ated to fear of floating. Why? First , if a c巴ntral bank were to exhibit a fear of floating , it would follow that the exchange rate (rea1 or nomina1) wou1d appear in the 10ss function. We know from (1) that it do巳 s no t. S巳cond ， given that ther巴 is no fear of floating coeffici巳nt in the 10ss function, obvious1y, one will not appear in the optima1 ru1e. From (8) we see thatfv is made up of policy pr巳ferenc巳 s governing inflation and/or output and any feedback from the structura1 mode1; there is no fear of floating per se. Howev巳r ， in op巴n econOlll1巳 s wher巳 the proportion of trad巳d to non-traded goods is re1ative1y high , the use of domestic inflation may not sufficiently represent price changes of the consumption baske t. Inst巳叫， th巴 c 巳ntral bank might choose to undertake targeting CPI inflation. Consider the following simple d巳finition of CPI inflation

^'l

wh巳reπC

慨:f+(1 一 ω)πt

咀EA JSR、

π

)

AHV

is CPI inflation , l[ fr is trad巳d goods inflation and πis dom的 ticinflation as in (3) above. The parameterωrepresents the degree of trade openness by reflecting th巳 proportion of traded goods in the domestic economy consumption basket. Traded goods prices are high1y dependent on changes in the exchange rate. 14

in Asia

Open Ecof! omy Injiation Tärgeling Regirnes in the CrisÎs-hil

E，此st

Asian Economies EA 咀

、‘，/

(l

π;' =-pl!,.e,

45

where ρcaptures the degree of pass-through to traded goods pric巴 s. If the central bank targets domestic inflation then the loss function would be as given by (1). However, if th巴 central bank wishes to target CPI inflation the loss function becomes: 人= K， δ(丸 (π!C~1)2 +λ2Yt~1}

(1 2) 2

= K， O{丸 [ωp拉什， +(1ω)π仆，]' +λ2Yt +1 }

where λl' represents the central bank preference for targeting CPI inflation. From (12) , we can see that a CPI targeting central bank can exhibit some additional policy pr巴ferenc 巴 to managing exchang巴 rate volatility and that the parameters representing openness (ω1) and pass-through (p) feature significantly in th巴 loss function. By making the requisite substitutions and repeating the process used to derive (6) to (9) , one can derive the optimal rnle for CPI inflation targeting as follows: 1"' 一九鼎M(lω) JIT

γ

Qμ'4 月一

弓之

+一

--J

λTr

二 -h

l-*

ω 一;

*-

M 一+

A 一

v

叭叭一九

fFJc

(1 3)

ι1\1' +λ7 A '

l-

j , = λ1，M(ωIp +(1一 ω)α0+λ2Aß3 C

2

叩門J

λl'M + λoA'

(1 4)

(1 5)

WP-A

(1 6)

where M =(ωpθ+ (1一 ω)α2 ) The optimal rnle co巴fficients (jC) for CPI inflation targeting are given by (13) to (1 6). Notice that the extent with which exchange rate movements permeate through the domestic economy (as captured by ωand ρ) is now explicitly captured as part of a central bank's optimal monetary policy.

4. Concerns with Fear of Floating Behaviour Despit巳 the

ways that th巴巴xchang巴 rate can be incorporat巳d into an open 巳conomy inflation targeting a叮angeme前， there remain significant concerns about doing so. First, if one attempts to control the inflationary 巳ffects of exchange rat巴 changes ， it effectively implies raising interest rates during periods of exchange rate weakness

in Asia

46

Eλ change

Rate Regimes

to and vice versa during periods of exchange rate str巴ngth. The concern is that responding too heavily and frequently to currency movements in the short-term could risk transforming the flexible inflation target to a de facto soft currency p巴g which in turn tends to be crisis-prone. This observation may be especially pertinent to som巴 Asian economies where there are concerns of a reversion to exchange rate based monetary policy regime (Cavoli and R吋姐， 2006a). Second , insofar as interest rate changes have a lagged effect on the economy on the one hand , and passthrough from exchange rates tends to be fairly immediate on the other, the central bank will have to forecast short-term exchange rate movements. This is near impossible to do on a consistent basis. One way to partially overcome the problem of exchange rate fluctuations on inflation is for the central bank to focus on “ core" rather than “ headline" inflation (the former being h巴adline inflation minus food and energy prices).15 Referring to Table 1, one sees that a number of the Asian central banks pursuing inflation targeting arrangements are in fact targ巴ting core inflation. The benefit of doing so is that any exchange rate fluctuations that directly impact the imported price of foodstuffs and energy will be excluded. While targ巴ting core inflation does not completely offset the impact of exchange rate fluctuations on all domestic prices (as a country could be importing other goods and ther巴 could b巴 a seeping through of non-core price inflation into overall inflation) , it has been seen as a way of addressing the 巴xchange rate debate for small and open economies. 16 While targeting core inflation helps to loosen the tie between exchange rates and domestic monetary policy , there is a more basic concern with exchange rate movements on the monetary transmission mechanism, viz. what if exchange passthrough is incomplete such that nominal exchange rate changes do not immediately translate into real exchange rate changes? If this happens , it implies that the real exchange rat巴 will not rev巴rt to its original value (i.e. purchasing power parity will not hold) , which in turn could impact domestic output , growth and inflation over tim巴 In other words , a fl 巴xible exchange rate could lead to persistent exchange rate misalignment which could be sustained over prolonged periods. Insofar as these exchange rate misalignments have a sustain巴 d impact on the r

in Asia

Open Ecof! omy Injiation Tärgeling Regirnes in the CrisÎs-hil

E，此st

Asian Economies

47

5. Strict versus Flexible Targeting There is a school of thought that argues that as long as the country' s inflation outlook r巳mains consistent with the m巳dium t巳rm inflation target range (i.e. the policy reference period) , the central bank has space to use its judgment to judiciously to me巳t oth巴r objectives and respond 巳ff，巳ctiv巳ly to various shocks and “ obvious" asset price misalignments in the interim. 19 This suggests a degree of discretion in being abl巴 to prick “ ass巴t pnce “bubbl巳 s" induding 巴xchange rate and housing ones (or better still , be preemptive so as to prevent bubbles from forming in the first instance). However, multiple targeting (ov巳r and above inflation and output) 的 not without its drawbacks. On巴， multiplicity of objectiv巳 slflexibility in impl巴menting the inflation target invariably complicates the communication strategy of the central bank's monetary policy. As Mishkin (2002) notes: The KISS principle (“ Keep It Simple Stupid") suggests that monetary policy should be articulated in as simple way as possible. The beauty of inflation target regimes is that by fOCllSi月 on one objective - inflation - commllnication is fairly straightforward (p 14) Two , when monetary authorities explain their monetary policy actions by to th巴 need to ensure output or exchange rate stability ,“the political debate about monetary policy is likely to focus on short-run issues" , (Mishkin , 2002 , p. 11) be it job creation , exchange rate stability of even asset price stability. This in turn may “ obscure the transparency of monetary policy and make it less likely that the public will support a monetary policy that focuses on long-run considerations" (Mishkin 2002 , p. 14) and may worsen the output-inflation tradeoff. To be sure , there is a significant difference between keeping an eye on asset price changes as offeri月 information on underlying economy compared to explicitly targeting them. The former is rather uncontroversial; the latter is no t. 20 There is a concern that central banks are not able to estimate bubbles or misalignments (wouldn't they be rich if they could?) and there could also be instances where various asset prices give conflicting signals. 21 Ben Bernanke of the Federal Reserve has argued strongly against th巴 central bank att巴mpting to respond to asset price bubbles. As he notes (Bernank巴， 2002)

r巴ferring

If we cOllld accurately and painlessly rid asset markets of bllbbl郎， of co盯se we would want to do so. Bllt as a practical matter , this is easier said than done , particularly if we intend to use monetary policy as the instrument, for two main reasons. First, the Fed cannot reliably identify bubbles in asset prices. Second , even if it cOllld identify bubbles , monetary policy is far too blllnt a tool for effective llse against them... . (A)s a society , we wOllld like to find ways to mitigat巴 the potential instabilities associated with asset-pric巴 booms and bllStS.

in Asia

48

Eλ change

Rate Regimes

Monetary policy is not a useful tool for achi巴ving this objective , however. Even putting asid巴 the great difficuIty of identifying bubbles in asset prices , monetary policy cannot be directed finely enough to guide asset prices without risking sev巳E巳 collateral damag巴 to th巳 economy. A far better approach , 1 believ巴， is to use micro-l巳V巳1 policies to reduc巳 the incid巴nce of bubbles and to protect the financial system against th巳ir effects. 1 hav巳 alr巳 ady m巳ntioned a vari巳ty of possible measur巴日， including supervisory action to ensure capital adequacy in th巴 banking syst巳m ， stress testing of portfolio 日， mcreas 巴d transparency in accounting and disclosure practices , improved financialliteracy , greater care in the process of financial liberalization , and a willingness to play the role of lender of last resort when needed. (pp. 3 and 8)22 Even if there is a case for the central bank to respond to signs of “ obvious bubbles" , it probably cannot be incorporated in an explicit rule. If the monetary authority chooses to respond to such misalignments infrequently they should do so on a discretionary basis. This leads us to the n巳xt issu巳 as to whether an inflation targeting arrangement errs on the side of policy rigidity and discipline or discretion and fl巳xibility? While th巴巴xact balance betwe巳n flexibility and rigidity wiU no doubt vary between countries (and possibly over time within a country) , broad rul巴s of thumb sugg巳 st: (a) th巴 l巴ss credible th巳 C巳ntral bank (i.e. poorer its inflationfighting track record); (b) the less its technical ability; and (c) the lower its political indep巳ndenc巴， th巳 more advisable it is to pre-commit to a “ strict" or “ hard" inflation target (i.e. preference of a rule over discretion).

6. Conclusion In th巴 final analysis , regardl巴 ss of the extent of flexibility or discretion that is pursued , it is imperative that the central bank operating a flexible inflation targeting arrang巳m巴nt communicate effectively to the public the lexicographic ordering of its objectives and the time frame over which the central bank is committed to r巴turning inflation to targe t. The central bank n巴巴 ds to be publicly committed to relinquish all other goals in order to meet the inflation target. The new inflation targeters in Asia hav巳 thus far not faced significant trad巳offs between inflation and other objectives in view of the fact that the global economic environment has , until r巴c巴ntly ， b巴en non-inflationary. In other words , given that inflation has never really threatened to overshoot its predetermined band , many of the Asian central banks have b巴巴n largely free to use monetary policy to a1tain other goals such as smoothing exchange rate changes. Put another way , the credibility of the system has , to dat巴， not been seriously challenged. Many Asian inflation targeting central banks appear to take into account exchange rate movements 一­ whether by targ巴ting CPI inflation or by possessing som巴 exchang巴 rate objective over and above that implicitly given by CPI inflation targeting as described in section 3. 23

in Asia

Open Ecof! omy Injiation Tärgeling Regirnes in the CrisÎs-hil

E，此st

Asian Economies

49

It would appear though , that there is an asymmetry in the way that central banks

treat exchange rate movements. Specifically , they do not always alter interest rates in response to upward (buying) pressure on their currencies , preferring to intervene in the foreign exchange market and accumulate reserves , but they are more willing to hike interest rates (or use other measures such as tightening capital controls) in the midst of downward (selling) pressure on their cnrrencies. This in turn inevitably has led to a rapid stockpiling of int巴rnational reserves which have had to be sterilized so as to prevent a domestic monet訂y overhang. 24

in Asia

in Asia

Estimating Monetary Policy Rules for Selected Asian Economies

1. Introduction One of the dominant issues surrounding the choice of monetary and exchange rate policy in Asia is that of the role of th巴巴xchang巳 rate and what policy the crisisaffected countries , viz. Korea , Thailand , Indones間， Malaysia and the Philippines , actually impl巳ment巴d aft巳r the crisis. Th巴 conv巴ntional vi巳w is that th巴 se countries made their policy choices on the basis of the “ bipolar view" , that 芯， intermediate I 巴 gimes ar巳 inherently crisis-pron巳 and that th巳 best choice is to head for the extremes. In terms of de jure monetary policy regimes , Korea , Thailand, Indonesia and the Philippines chos 巳 the floating corner by electing to institut巴 inflation targeting systems. Malaysia chose the opposite corner and implemented a rigid fix of the ringgit to the US dollar which it forsaked in July 2005 (s巴巴 chap胞的 1 and 2). However, there is some doubt about whether the floating group actually gave up the desire to fix their r巳 spective currencies (especially to th巴 US dollar). Th巳re is evidence to suggest that these countries still maintained some degree of fixity to th巴 US dollar and the yen aft巳r th巴巴叮叮ts of the crisis had subsided (see chapt臼 1). This chapter expands on and contributes to the literature on this topic by investigating the behaviour of th巳 policy variables that matter for policy making. It does this by estimating a Taylor type monetary policy rule (MPR) for Korea , Thailand , Indonesia and the Philippines. The recent literature contains research on MPRs where the nominal interest rate and, indeed , th巴 norninal 巳xchang巳 rat巳 ar巳 the stated instruments of monetary policy, and the MPRs assess how each respective instrument reacts to inflation, output and the exchange rat巳. Our objective is to investigate the appropriateness of an interest rate MPR for each country. Naturally , we have a priori expectations about how well th巳 MPRs might work on the basis of the stylized facts regarding pre and post crisis policy regime. As mentioned , there have been suggestions that som巴 Asian central banks ar巳 reverting to US dollar or basket peg arrangem凹的 m recent times. Our interest is whether this is reflected in the MPRs despite an

in Asia

52

Eλ change

Rate Regimes

interest rate based inflation targeting bas巴d policy for many Asian countries. It is possible that the strength of the relationship between the exchange rate and other important macroeconomic variables is such that the exchang巴 rate is a significant factor in determining the effectiveness of monetary policy and , consequently , the type of MPR employed. The remainder of the chapter is divided into two sections. Section 2 presents the models and 巴 stimation procedures. Section 3 discusses the results and Section 4 offers some concluding remarks. To preview the main conclusions , interest rate MPRs tend to contain stronger and more stable reactions to current and expect巴d inflatÌ on than to exchange rates and the interest rate rules' reaction to the exchange rate is weaker than anticipated given the desire for Asian central banks to engage in some form of exchange rate management.

2. Taylor-Type Forward Looking MPRs The MPR that will be the subject of our analysis is similar in sp巴cificatÌon to those found in Clarida , Gali and Gertler (1 998) which has spawned many subsequent works in this ar巳a. 1 Consider the following expression for the short-term nominal market inter，的t rate: ρ )r; + ρ'1;-1 +V ,

I

JSZ\

。三~p三 l

、-Y

代=(1

where r, is the short-term nominal interest rate at t , r,* is central bank's target for the short-term nominal inter巳 st rate - put another way , its policy instrument. The parameter , p , captures the persistence of the market interest rate. The policy instrument is spωifi巳d as follows

可 =r+ 戶(耳πt十pln， l 一的 +γ(E[Yl+qln， l- y*)+o(E[心 In， l-s*)

(2)

is the equilibrium nominal interest rate , 7f, +p is the rate of inflation betw巳en periods t and t+p , 7f* is an inflation target , Y'+n is real output growth over q periods , t+ q y' is the long run or steady state real output growth , s, is the nominal exchange rate, and s* is the steady state value. 2 Substituting (2) into (1) obtains: wh巳r巴 r

代= (1- p){α + ßÆ'[π'+1' I!:~i 1+ yÆ' [x ,+q In ,l + oÆ'[ 寄，:" In ,]} + pr'-l + V where α ==r-ßπ忙 ， X， == errors as follows:

t

(3)

y, - y' and sl == s, - s*. If we group th巳 unobs巳rv巳d forecast

St 三一(1- p){ß(π卅一 E[ 7r l+ p Int]) + y(x卅一耳X t+ q Int])+δ (s;~" -E[比 In， ]}+v，

in Asia

Estùnating 九10netary

Policy Rules!ór SeZected Asian Economies

53

we can then 巴xpress (3) in terms of the realized variables: 代= (1

ρ)α+ (1

ρ )ßπ 卅+(1

ρ)戶卅+ (1

ρ )08，+n + ρr'_l +e ,

(4)

By estimating an interest rate MPR we seek information about the effectiveness of each MPR over the sample. A priori , we would expect to find that for the crisisaffected countri巴 s of Korea, Thailand, Indonesia and the Philippin巴 s which have adopted inflation targeting regimes with interest rate as the instrume帥， the interest rate rule reacts quite strongly to inflation especially after the crisis.

3. Estimation and Results Equation (4) forms the basis of our empirical testing. The procedure to be used estimate [αβ 訴 ρ，割的 GMM. GMM is an appropriat巴巴 stimator 汀， for each MPR given by equation (4) , a vector of instruments within the policymakers information set ， μfεQ" can b巳 found such that the exogeneity condition , E[ e， lμJ is satisfied. We separately estimate MPRs for the four official inflation targeters noted above Data is from IFS except for Thailand's Industrial Production which is sourced from th巴 Bank of Thailand (BOT) website. The instruments (/1,) for the GMM estimates are taken from 1-6 , 9 and 12 monthly lagged values of the regressors. Th巳 sample used is 1985Ml to 2006M12 unl巴 ss otherwis巳 specified. This sample period spans two quite distinct policy regim 肘， punctuated by the Asian crisis between mid 1997 and 1998. Tables 1 to 4 report GMM estimates for the MPRs. We test for the reaction of current inflation (p=O) , expected inflatÌ on (p=12) and current and expected exchange rate gap (m=O and 12 respectively).3 Each table has two panels - panel a presents th巴 results wh巳re the nominal exchange rate per US dollar was employed as a regressor while panel b presents those results where the nominal effective exchange rate (NEER) is a r巴gressor. The reaction to inflation is generally quite strong for Korea and Thailand. The co巴fficients are well abov巴 1 and up to about 2.8 for Korea and up to about 2.6 for Thailand. In other words , a 1% increase in either current or expected inflation brings about a rise in interest rates of up to 2.8% for Korea or up to a 2.6% ris巳 for Thailand. The exception is the Philippines where the reaction is low but still statistically significan t. The r巳action of the interest rate to inflation is variable for Indonesia. More specifically, the reaction for current inflation is more statistically significant than for expected inflation but , especially for Korea and Thailand , the actual coefficient value for expected inflation is higher. The reaction to the exchang巴 rate is mixed - the sign is incorrect (from the perspective of a policy reaction) and also statistically insignificant in many cases especially so for Indonesia and the Philippin巴 S.4 As in the case of inflation , the

h巴r巴 to

in Asia

54

Eλ change

Rate Regimes

Table 1 GMM Estimates for

Kor拭 Dependent

variable , r,

a. s = nominal exchange rate per US dollar Cm的t

π，

0.04 (0.69) 1. 21 (2.63)

,

sg

0.05 (0.90)

0.86 (1.3 7)

-0.58 (-1. 87)

。 .10

(-0.96)

0.97 (1. 93) 1. 99 (2.75)

π1+12

y',

。 10 (-0.77)

。 45 (1.1 8)

1. 97 (3 0)

.4

1. 29 (1. 92)

(1.1 8)

0.02 (0.09)

0.01 (0.06)

。.46

0.01 (0.03)

Sgt+12

(99.28)

。 97

0.95 (8 1.3 1)

0.97 (93.89)

。 .95 (1 37.12)

R 2 adj

0.95

0.95

0.95

0.95

DW

1. 59

1. 60

1. 60

1. 60

Obs

237

227

227

227

-0.33 (-2.15)

0.001 (0.01)

-0.85 (-3.37)

7'_1

b. s = nominal effective exchange rate Cm的t

-0.05 (-0.90)

π，

2.39 (2.57)

.4

1 8 (2.16) 2.23 (7.4 3)

π1+12

y',

1.5 1 (1. 63)

0.07 (0.34)

s弓

1.1 2 (1. 5 1)

(1. 60)

2.88 (8.98) 1. 44 (1. 56)

0.03 (0.13)

0.04 (-0.07)

(3.24)

0.26

Sgt+12

。.45

0.98 (96.69)

0.89 (47.37)

0.98 (89.22)

0.84 (3 1. 22)

R 2 adj

0.95

0.93

0.95

0.92

DW

1. 61

1. 22

1. 62

1. 02

Obs

237

227

228

227

rt_1

Notes: lagged values (1 to 6, 9 and 12 months) of all variables used as instruments. Full

in Asia

Estùnating 九10netary

Policy Rules!ór SeZected Asian Economies

Table 2 GMM Estimates for Thailand , Dependant

55

variable ，科

a. S = nominal exchange rate per US dollar Const

7f,

-0.07 (-0.66)

-0.03 (-0.23)

1. 82 (1. 50)

1rt+12

0.13 (0.34) 0.36

SEt

(1. 50)

-0.27 (-1. 28)

1. 98 (5.58)

2.10 (3.74)

yg,

-0.03 (-0.13)

。 43

(0.65)

1. 80 (5.20)

(0.82)

。 20

。 23 (1. 44)

-0.99 (-2 .48)

-0.55 (-3.52)

。 84

1.39 (2.56)

Sgt+12

0.93 (35.02)

0.96 (6 1. 84)

(1 9 .45)

0.84 (35.85)

R 2 adj

0.83

0.83

。 78

0.83

Dl平

2.53

2.60

1. 79

2.32

Obs

217

213

215

213

-0.12 (-0.82)

-0.13 (-1. 31)

-0.34 (-2.20)

rt_l

b.

S

=nominal effective exchange rate

Const

7f,

-0.15 (-1 .4 8) 2.36 (4.77)

2.27 (4.5 1) 2.14

1rt+12

2.60 (4.20)

(1. 92)

yg, s',

-0.24 (-0.80)

-0.18 (-0.34)

0.31

-1.54 (-2 .4 8)

(1. 20)

sgt+ J2

-0.28 (-0.84)

-0.06 (-0.2 1)

-0.07 (-0.26)

(2 .4 8)

。 83

0.92 (35 .42)

0.96 (60.08)

(35.12)

。 93

0.93 (56.93)

R 2 adj

0.83

0.83

。 83

0.84

Dl干

2.54

2.59

2.52

2.62

Obs

217

213

216

213

T t_l

Notes: lagged values (l to 6 , 9 and 12 months) of all variables used as instruments. Full sample 1985.1 to 2006.12 in Asia

56

Eλ change

Rate Regimes

Table 3 GMM Estimates for the Philippine日， Dependant variable , r, a. s = nominal exchange rate per US do l1 ar Cm的t

。 88

(2 .47) π，

。 70 (1. 69)

0.72

。 .97

(1. 28)

(2.22)

1.36 (4.23)

0.66 (2.92)

-0 .44 (-0.38)

0.68

π1+12

(1. 43) y馬

,

sg

-0.05 (-0.94) 。 18

(0.98)

0.01 (0.13)

-0.11

0.13 (-2.05)

(-1. 09)

1. 04 (2 .44)

-0.56 (0.69)

0.31 (1.1 7)

Sgt+12

(20.76)

。 86

0.89 (22 .1 3)

(1 2.74)

(22.09)

R 2 adj

。 23

0.19

0.19

0.16

DW

2.69

2.74

2.53

2.72

Obs

237

227

227

227

0.97 (2.23)

0.57 (1. 28)

1. 22 (2.76)

7'_1

0.80

。 .92

b. s = nominal effective exchange rate Cm的t

。 79

(2.16) π，

s弓

-1.1 3 (-0.67)

-2.70 (-0.60)

π1+12

y',

1. 25 (3.64)

0.66 (2.77)

0.10 (-0 .4 4)

-0.11

-0.11

(-1.1 3)

(-1. 84)

(-1. 06)

-0.18 (-0.7 1)

-1. 26 (-0.84)

(-1. 91)

0.06

-0.91 。 .70

Sgt+12

(1.1 2)

0.87 (20.10)

0.96 (2 1.70)

0.85 (16.67)

0.92 (19.58)

R 2 adj

。 22

。 12

0.19

0.14

DW

2.70

2.69

2.63

2.66

Obs

237

227

228

227

r'_l

Notes: lagged values (1 to 6, 9 and 12 months) of all variables used as instruments. Full

in Asia

Estùnating 九10netary

Policy Rules!ór SeZected Asian Economies

57

Table 4 GMM Estimates for Indonesia, D巴pendant variabl巴 ， r, a. s = nominal exchange rate per US dollar Const

0.08 (0.50)

1[,

1. 03 (1.1 6)

1.4 1 (6.14)

0.68

1rt+12

(1 2.52) yg,

-0.60 (-0 .46)

(1.3 9)

0.20 (0.27)

0.17 (4.07)

0.98 (53.61)

0.77 (56.65)

R 2 adj

0.85

0.86

Dl平

2.36

2.00

Obs

155

143

s',

0.14

Sgt+12 rt_l

Notes: lagged values (1 to 6, 9 and 12 months) of all variables used as instruments. Full sample 1985.1 to 2006.12. reaction to the exchange rate was quite robust to whether the current gap or exp巴cted gap is used. In th巴 case of Korea , there appears to be a stronger r巴 action to the NEER than to the bilateral exchange rate per USD. For Thailand , the reaction to the expected exchang巴 rate gap is more significant when interacted with CUIT巴nt and not expected inflation. Figures 2-5 show the predicted values of the interest rate against the actual values. These are generated from GMM estimates of the MPR where current inflation and the current output and exchange rate gaps were used as regressors (i.e. , p=q=n=O ).5 There are two panels to each figure. Panel a presents the values for thos 巴 MPRs estimated using th巴 nominal exchange rat巴 as a regressor and panel b contains the results for those estimates where the nominal effective exchange rate (NEER) was used as a regr巴 ssor. Eyeballing each panel shows little difference between the two sets of results. It can b巴 seen that th巴 predicted int巴rest rate , , follows the actual T , quite closely for the entire sample. However, the model seems to fit the data best for the post-crisis sample , 1999.1 to 2006 .1 2. This is an expected result for Korea, Thailand, Indonesia and the Philippines as there is evidence to suggest that central banks of those countries pursued a Taylor rul巴 type monetary policy. The 巴xception is Indonesia where the residuals are higher for the post-crisis period.

r

in Asia

Eλ change

58

Rate Regimes

4. Conclusion The estimations undertaken in this chapter suggest quite clearly a shift in monetary policy regime in that the interest rate 孔1PRs are a better fit after th巴 crisis than befor巴， and th巴巴xchang巴 rate rules are a b巳tter fit befor巳 the crisis. The MPRs generally react quite strongly to inflation , but w巳 akly to outpu t. B 巳yond this , two results warrant emphasizing. First, the relationship between the interest rate and inflation is stronger than that between the exchange rate (bilateral or effective) and inflation. This suggests that the int巴rest rate may be a more suitable instrument for an inflation targeting framework. The results ar'0

-ι，.4

1988

199 日

1 9 9 4-

1992

主之一eC "L1rsive

1996

1998

2000

2002

2 0 0 4-

c:> u t p "L1r coefficienTs

='二三三巴基王三

Figure 4 Recursive OLS Estimates ofBaseline MPR (Dep

Variabl巴，逸的，

1988-2004

1 _~古

1ζ3

。 干事

。。 0.5

-1. 0 198 日

1990

1 9 9 4-

1992

王之 ecurS1V 且

='c

二

1 笠，芒，正三

1998

iu 1"l且 ti..:.>I J.

2000

20日 2

2 0 0 4-

C 口已且.~ricieI1 ts

s 一主二三

1.2 10 0

日

0.6 0. 4-

。 2 《了)0

。 2 。 4

1990

1992

1994

1996

199 日

2000

2002

2 0 0 4-

Figure 5 Recursive OLS Estimates of Baseline MPR (Dep Variable , ERgap) , 1988-2004

in Asia

Characterizin;穹 Singαpore's E.芝change

Rate Policy

85

In keeping with the literature on MPRs we estimate Eq , 10 using leading values for inflation as regressors. We are inter巴 st巳 d in 巴 stimating whether the instrument reacts to what might be considered future values of inflation. To control for the possible endogeneity of inflation we use TSLS. The results ar巴 also pr巳sented in Table 4 where estimates are presented for values of 1r'+k where k = 1 to 4 (quarters). l1 ,12 R巳 assuringly ， the r巳 sults are not dissimilar to those pr巳 sented in Table 3 for the baseline model. For the model using MR , the coefficient values for inflation are significant and se巳 m quit巳 robust to the value of k. As with th巳 baseline estimates , the coefficient values are smaller for the model where ERgap is used and appear to increas巳 as k incr巳ases. Table 4

Mon巳tary

Policy Rul巳 for Singapor