Economics and Finance in Mauritius : a Modern Perspective 978-3-319-39434-3, 3319394347, 978-3-319-39435-0, 3319394355

Table of contents :
Front Matter ....Pages i-xxii
Front Matter ....Pages 1-1
An Overview of the Mauritian Financial System and Its Economy (Indranarain Ramlall)....Pages 3-24
Front Matter ....Pages 25-25
Banking Sector Analysis (Indranarain Ramlall)....Pages 27-52
Developing a Credit Risk Model for Mauritian Bankers (Indranarain Ramlall)....Pages 53-71
Front Matter ....Pages 73-73
Stock Market Analysis (Indranarain Ramlall)....Pages 75-96
The Behaviour of Foreign Investments in the Stock Exchange of Mauritius (Indranarain Ramlall)....Pages 97-134
The Impact of US Subprime Crisis on SEMDEX (Indranarain Ramlall)....Pages 135-150
Front Matter ....Pages 151-151
Monetary Policy Analysis (Indranarain Ramlall)....Pages 153-166
Money Demand Analysis in Mauritius (Indranarain Ramlall)....Pages 167-193
Impact of Debit and Credit Cards on Currency in Circulation in Mauritius (Indranarain Ramlall)....Pages 195-211
A Posthumous Note on the Lombard Rate in Mauritius (Indranarain Ramlall)....Pages 213-226
Front Matter ....Pages 227-227
What Drives the Capital Structure of Non-Listed Mauritian Firms? (Indranarain Ramlall)....Pages 229-248
Capital Structure Analysis of Exporting and Non-exporting Mauritian Firms: A Pre- and Post-crisis Investigation (Indranarain Ramlall)....Pages 249-274
Front Matter ....Pages 275-275
What Factors Drive Hedging Among Mauritian Firms? (Indranarain Ramlall)....Pages 277-308
Front Matter ....Pages 309-309
Developing a Financial Stability Model for Mauritius (Indranarain Ramlall)....Pages 311-360
Front Matter ....Pages 361-361
Determinants of House Prices in Mauritius: A Hedonic Price Model à la Quantile Bootstrapped Approach (Indranarain Ramlall)....Pages 363-414
Front Matter ....Pages 415-415
Stylized Facts About the Mauritian Economy (Indranarain Ramlall)....Pages 417-422
Author’s Vision for 2030 (Indranarain Ramlall)....Pages 423-430
Back Matter ....Pages 431-468

Citation preview

Economics and Finance in Mauritius A Modern Perspective Dr Indranarain Ramlall

Economics and Finance in Mauritius

Indranarain Ramlall

Economics and Finance in Mauritius A Modern Perspective

Indranarain Ramlall University of Mauritius Moka, Mauritius

ISBN 978-3-319-39434-3    ISBN 978-3-319-39435-0 (eBook) DOI 10.1007/978-3-319-39435-0 Library of Congress Control Number: 2016948430 © The Editor(s) (if applicable) and The Author(s) 2017 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. Cover image © Peter Hermes Furian / Alamy Stock Photo Printed on acid-free paper This Palgrave Macmillan imprint is published by Springer Nature The registered company is Springer International Publishing AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Dedicated to my parents, to all those who are striving hard for a better world and to God. Noble people use their intelligence, knowledge, time and efforts to create things for the betterment of the society and the world. Certain people use their intelligence and knowledge to backstap other people or to create hurdles in the advancing path of the noble people. Just imagine how world would have been if all people were noble! An unknown knowledge seeker.

This book breaks new ground in Mauritius, bears testimony to the author’s dedication to a thorough understanding of the various components of the economic and financial sectors. Academic researchers and students will have ample opportunities to wade through the calculations in support of the findings, whereas policy-makers will have food for thought, and possibly action, in the light of those findings. —Pierre Dinan, Fellow of the Institute of Chartered Accountants in England and Wales.

Preface

This book seeks to provide a comprehensive analysis of the economic and financial conditions of Mauritius. A rigorous analysis of the financial sector of Mauritius is provided, incorporating the banking sector, the stock market, and monetary policy, as well as the capital structure of Mauritian firms. The author not only builds a credit risk model for Mauritian bankers, but also develops a financial stability model to provide the reader with a full account of the Mauritian economy. The hedging practices of Mauritian firms are given due consideration, and the book also includes an in-depth analysis of the housing market. The book builds on the author’s lifetime of experience as an academic and practitioner within the Mauritian economy, within the realm of both the private and public sectors. In a spirit of patriotism the author hopes that this book will offer a valuable contribution to the future development of his country. The author ends with a chapter outlining his hopes for a 2030 vision for Mauritius. He anticipates that this book will provide a new window for researchers, students, policymakers, central bankers and economists. Comments and suggestions about the book can be made to the following email addresses: [email protected] or [email protected]. April 2016

Dr. Indranarain Ramlall ix

Contents

Part I  An Overview of the Mauritian Economy 1

   1

An Overview of the Mauritian Financial System and Its Economy   3

Part II  The Mauritian Banking Sector

  25

2

Banking Sector Analysis  27

3

Developing a Credit Risk Model for Mauritian Bankers  53

Part III  The Stock Market in Mauritius

  73

4

Stock Market Analysis  75

5

The Behaviour of Foreign Investments in the Stock Exchange of Mauritius  97 xi

xii Contents

  6 The Impact of US Subprime Crisis on SEMDEX 135

Part IV  Monetary Policy in Mauritius

 151

  7 Monetary Policy Analysis 153   8 Money Demand Analysis in Mauritius 167   9 Impact of Debit and Credit Cards on  Currency in Circulation in Mauritius 195 10 A Posthumous Note on the Lombard Rate in Mauritius 213

Part V The Business Sector and Its Financing Structure in Mauritius

 227

11 What Drives the Capital Structure of  Non-­Listed Mauritian Firms? 229 12 Capital Structure Analysis of Exporting and  Non-exporting Mauritian Firms: A Pre- and Post-crisis Investigation 249

Part VI  Hedging in Mauritius

 275

13 What Factors Drive Hedging Among Mauritian Firms? 277

 Contents 

Part VII  Financial Stability Risk Assessment in Mauritius

xiii

 309

14 Developing a Financial Stability Model for Mauritius 311

Part VIII  House Price Modelling in Mauritius

 361

15 Determinants of House Prices in Mauritius: A Hedonic Price Model à la Quantile Bootstrapped Approach 363

Part IX  Stylised Facts and Vision for Mauritius

 415

16 Stylized Facts About the Mauritian Economy 417 17 Author’s Vision for 2030 423 Bibliography431 Index451

List of Figures

Fig. 1.1 Sector contribution to GDP  Fig. 1.2 Components of expenditures on Gross Domestic Product  Fig. 1.3 Composition of consumption in Mauritius  Fig. 1.4 Structure of exports and imports under expenditures on Gross Domestic Product  Fig. 1.5 Unemployment rate  Fig. 1.6 Tourist earnings and arrivals  Fig. 1.7 Gross Domestic Product per capita  Fig. 1.8 Headline inflation rate-year end  Fig. 1.9 ICE Brent Crude (US$ per barrel)  Fig. 1.10 Budget deficit to GDP in per cent  Fig. 1.11 Sector-wise distribution of credit as at end December 2015 Fig. 1.12 Evolution of Key Repo Rate  Fig. 1.13 Interest rate spread for the banking sector  Fig. 1.14 Interest rate spread from the World Bank Development Indicators  Fig. 1.15 Evolution of the exchange rates-Clash of the Titans Fig. 1.16 Composition of corporate debt structure in Mauritius  Fig. 1.17 Balance of payments  Fig. 1.18 Public sector debt Fig. 2.1 Banking sector assets to GDP Fig. 2.2 Chernoff faces for the Mauritian banking sector Fig. 2.3 Network analysis for period June 2005 to December 2010

7 8 9 10 11 12 12 13 14 15 16 17 18 18 19 20 21 22 29 48 50 xv

xvi 

List of Figures

Fig. 2.4 Network analysis for period January 2011 to July 2016 Fig. 4.1 Market capitalization on official market as at end December 2015  Fig. 4.2 Bank financing versus equity financing Fig. 4.3 Banking sector assets versus market capitalisation Fig. 4.4 Price-earnings ratio  Fig. 4.5 Foreign investments (official market) on the Stock Exchange of Mauritius  Fig. 4.6 Turnover and volume analysis on official market Fig. 4.7 Performance of indices on the official market  Fig. 4.8 Analysing conventional blue chips stocks in SEM under normalised values Fig. 4.9 Transposing Figure 4.8 into an Elliott Wave analysis Fig. 4.A.1 Diagrammatical view of the interactions among various sectors (Source: Author’s Illustration) Fig. 5.1 An overview of SEM features Fig. 5.2 Foreign transactions in SEM Fig. 7.1 Schema of the interactions among households, banks and public debt in Mauritius Fig. 7.2 Effectiveness of the various channels of transmission mechanism in Mauritius Fig. 8.1 Response to Cholesky One S.D. innovations Fig. 8.2 Variance decomposition Fig. 8.A.1 Inverse roots of AR characteristic polynomial Fig. 9.1 Ratio of debit to credit cards  Fig. 12.1 Theories of capital structure Fig. 14.1 Stress index BoM Z-score Fig. 14.2 Stress index BoM capital ratio Fig. 14.3 Stress index-MCB bad debts Fig. 14.4 Stress index-MCB Z-score Fig. 14.5 MCB deposit gap and loan gap Fig. 14.6 Index-MCB beta Fig. 14.7 Stress index-loan to deposit ratio Fig. 14.8 Domestic credit by banks stress index Fig. 14.9 USD volatility stress index Fig. 14.10 EUR volatility stress index Fig. 14.11 GBP volatility stress index Fig. 14.12 Exchange market pressure index Fig. 14.13 Tourist arrivals stress index

51 78 80 82 83 84 85 86 88 89 91 103 104 158 160 185 186 189 200 252 336 336 337 337 338 338 339 339 340 340 341 341 342

  List of Figures 

Fig. 14.14 Fig. 14.15 Fig. 14.16 Fig. 14.17 Fig. 14.18 Fig. 14.19 Fig. 14.20 Fig. 14.21 Fig. 14.22 Fig. 14.23 Fig. 14.24 Fig. 14.25 Fig. 14.26 Fig. 14.27 Fig. 14.28 Fig. 14.29 Fig. 14.30 Fig. 14.31 Fig. 14.32 Fig. 14.33 Fig. 14.34 Fig. 14.35 Fig. 15.1 Fig. 17.1 Fig. 17.2

Tourist earnings stress index Cheques stress index MACSS stress index Balance of payments stress index Government debt stress index-cyclical components Cyclical components of GDP and public debt Public debt sustainability metric index Stress-interbank market Bid-cover stress index TED spread stress index Monetary condition stress index Currency in circulation stress index Broad money multiplier stress index Excess cash holdings stress index Net international reserves stress index Import cover stress index Oil stress index Trade finance stress index SEMDEX stress index Net foreign investments Cyclical components of GDP Mauritius financial system strength index Concave and convex Static economic progress Real effective economic progress

xvii

342 343 343 344 344 345 345 346 346 347 347 348 348 349 349 350 350 351 351 352 352 353 380 424 425

List of Tables

Table 2.1 Risk-weighted capital adequacy ratio Table 2.2 ABC Banking Corporation Limited Table 2.3 AfrAsia Bank Limited Table 2.4 Bank of Baroda Limited Table 2.5 Bank One Limited Table 2.6 Banque des Mascareignes Limited Table 2.7 The Mauritius Commercial Bank Limited Table 2.8 The State Bank of Mauritius Limited Table 2.9 Sectors considered in analysing the sector-wise distribution of credit in Mauritius Table 3.1 Econometric results under OLS approach Table 3.2 Econometric results for limited dependent models Table 3.A.1 List of variables Table 3.A.2 Summary statistics Table 3.A.3 Correlation matrix of independent variables used in the model Table 4.A.1 List of ATS operators (January 2016) Table 4.A.2 List of issuers on official market Table 5.1 Definition of the variables Table 5.2 Summary statistics for stock market returns Table 5.3 ADF unit root tests Table 5.4 Results for cointegration between purchases and sales for the post-­crisis period

28 30 32 35 38 40 42 45 49 59 66 69 69 70 92 93 106 110 111 113 xix

xx 

List of Tables

Table 5.5 Results for cointegration between bank rate and NetPurchases: pre crisis period Table 5.6 Results for cointegration between bank rate and purchases: pre crisis period Table 5.7 Results for cointegration between bank rate and sales: pre crisis period Table 5.8 Results for NetPurchases and USDOLLAR: pre crisis Table 5.9 Results for PURCHASES and USDOLLAR: pre crisis Table 5.10 Results for SALES and USDOLLAR: pre crisis Table 5.11 Dependent variable: net purchases under SEMDEX pre crisis and post crisis results Table 5.12 Dependent variable: net purchases under MCB pre crisis and post crisis results Table 5.13 Dependent variable: net purchases under SBM pre crisis and post crisis results Table 5.14 Dependent variable: net purchases under SUN pre crisis and post crisis results Table 5.15 Dependent variable: net purchases under NMH pre crisis and post crisis results Table 5.16 Dependent variable: net purchases under ROGERS pre crisis and post crisis results Table 5.17 Dependent variable: net purchases under IBL pre crisis and post crisis results Table 5.18 Dependent variable: D(SEM) under SEM pre crisis and post crisis results Table 5.19 Dependent cariable: D(MCB) under SEM pre crisis and post crisis results Table 5.20 Dependent variable: D(SBM) under SEM pre crisis and post crisis results Table 5.21 Dependent variable: D(SUN) under SEM pre crisis and post crisis results Table 5.22 Dependent variable: D(NMH) under SEM pre crisis and post crisis results Table 5.23 Dependent variable: D(ROGERS) under SEM pre crisis and post crisis results Table 5.24 Dependent variable: D(IBL) under SEM pre crisis and post crisis results Table 6.1 Summary statistics for stock market returns

114 116 117 119 120 121 122 122 123 124 124 125 125 126 127 128 128 129 130 131 138

  List of Tables 

Table 6.2 Results for granger causality Table 6.3 Unit root tests Table 6.4 Cointegration results under monthly horizon analysis Table 6.5 Results for Group 1 (FTSE, DAX and CAC-40) Table 6.6 Results for Group 2 (DJIA and BSE) Table 6.7 Results for Group 3 (NIKKEI 225 and JSE) Table 8.1 Definition of variables Table 8.2 Descriptive statistics Table 8.3 Correlation coefficients Table 8.4 Unit root tests Table 8.5 Empirical results for Mauritius Table 8.A.1 VEC Residual serial correlation LM tests Table 8.A.2 VEC Residual portmanteau tests for autocorrelations Table 8.A.3 VEC residual heteroskedasticity tests: no cross terms (only levels and squares) Table 8.A.4 VEC Residual heteroskedasticity tests: includes cross terms Table 9.1 Using number of credit, debit cards and ATMs as independent variables Table 9.2 Using a proxy for debit card usage as independent variable Table 9.A.1 Correlation coefficients for the variables Table 10.1 Summary statistics for Lombard Rate and yields on different maturities Table 10.2 Unit root tests Table 10.3 Results of the cointegration test under weekly horizon Table 10.4 Results of the cointegration test under weekly horizon Table 10.A.1 Results for 91 days Table 10.A.2 Results for 182 days Table 10.A.3 Results for 364 days Table 10.A.4 Results for bank rate Table 10.A.5 Results for 91 days and 182 days Table 10.A.6 Results for 91 days and 364 days Table 10.A.7 Results for 182 days and 364 days Table 11.1 Definition of variables Table 11.2 Summary statistics Table 11.3 Correlation coefficients Table 11.4 Correlation coefficients

xxi

139 139 142 143 146 148 176 178 179 180 181 189 190 190 191 204 206 210 216 216 217 219 222 222 223 223 224 224 225 235 235 239 240

xxii 

List of Tables

Table 11.5 Results of regression Table 12.1 Definitions of variables Table 12.2 Summary statistics for exporting and non-exporting firms in the pre crisis period Table 12.3 Summary statistics for exporting and non-exporting firms in the pre crisis period Table 12.4 Results under short-term leverage Table 12.5 Results under long-term leverage Table 13.1 Pearson correlation coefficients Table 13.2 Determinants of the decision to hedge under probit and logit Table 14.1 Components of the Mauritian financial system stress index Table 14.2 Composition of MFSSI Table 14.A.1 Vector error correction estimates for MCB loans and deposits Table 14.A.2 Vector error correction estimates for current account and external debt Table 15.1 Components of characteristics scrutinized in the model Table 15.2 Weight based on average values to sieve out core determinants Table 15.A.1 Quantile regressions Table 15.A.2 Quantile regression based on bootstrapped standard errors Table 15.A.3 Semi-log regression under bootstrapped approach Table 15.A.4 Summary of main findings Table 15.A.5 Summary statistics Table 15.A.6 Correlation coefficients

241 259 260 260 265 266 298 299 316 331 353 354 372 392 398 400 403 406 407 408

Part I An Overview of the Mauritian Economy

1 An Overview of the Mauritian Financial System and Its Economy

1

Introduction

The aim of this chapter is to introduce the main features and components which characterize the Mauritian economy and its financial system. The first part of this chapter focuses on the various components which make up the Mauritian financial system. The second part deals with the various features of the Mauritian economy not only descriptively but also under a critical lens of analysis. For instance, it is found that the interest rate spread data posted under the World Bank development indicators are not compatible with the reality of the Mauritian economy or with the data disseminated in the Bank of Mauritius monthly statistical bulletins. In a nutshell, the overall aim of this chapter is to make the readers conversant with the Mauritian economy.

© The Author(s) 2017 I. Ramlall, Economics and Finance in Mauritius, DOI 10.1007/978-3-319-39435-0_1

3

4 

Economics and Finance in Mauritius

2

 n Overview of the Mauritian Financial A System

Various financial services prevail in Mauritius such as custodial services, payment services, credit services, asset accumulation, reinsurance, consultancy services and risk assessment, amongst others. The Mauritius financial system consists of the following main components:

2.1

 egulators such as the Bank of Mauritius R and the Financial Services Commission

While the Bank of Mauritius is responsible for regulating and supervising the banks, the Financial Services Commission oversees all the rest of the non-banks, including the stock market, global businesses, leasing companies, insurance companies, fund management companies and one derivative company. Established in 2001, the Financial Services Commission operates under the Financial Services Act 2007, the Securities Act 2005, the Insurance Act 2005 and the Private Pension Scheme Act 2012. The Financial Services Commission regulates two important licenses securities exchanges in Mauritius, namely, the Stock Exchange of Mauritius and The Bourse Africa Limited, previously known as the Global Board of Trade.

2.2

Banks

Mauritius is basically a bank-based financial system so that the banks account for a large chunk of the total assets of the financial system. There are four major banks – two domestic banks (the Mauritius Commercial Bank Limited and the State Bank of Mauritius Limited) and two foreign banks (Barclays Bank Limited and the HSBC Bank (Mauritius) Limited). The full list of institutions regulated by the Bank of Mauritius is found in the appendix section of this chapter.

1  An Overview of the Mauritian Financial System and Its Economy 

2.3

5

Insurance Companies

It can be said that the banking sector and the insurance sector move hand in hand in terms of business activities chiefly when loans taken require insurance cover. Insurance companies thereby play an important role in the local financial system.

2.4

Leasing Companies

Leasing companies are now considered as an important part of the financial system in Mauritius as they enable the use of equipment at cheaper cost for those who are not financially very healthy.

2.5

Derivative Company

The Bourse Africa is the only derivatives company established in Mauritius and it has cross-border operations with futures being the main financial instrument in use. The company focuses on three core activity types, namely, commodity derivatives, currency derivatives and equity.

2.6

Fund Management Companies

Wealth management for both individuals and corporates is effected via fund management companies. Some of these companies automatically emanate as a subsidiary company from a bank’s mainstream activities. Insurance companies also have them as separate companies to offer a ‘one-­ stop’ financial services and products to potential clients. For instance, the Anglo-Mauritius Financial Services Limited provides a whole set of personal and corporate wealth management services along with advisory services to distinct clients.

6 

Economics and Finance in Mauritius

2.7

Other Companies

There are also other types of companies, including pension funds, the Mauritius Housing Corporation and the Mutual Aid Association. There are also a number of offshore companies – namely, the Global Business License category 1 and Global Business License category 2. Global Business License category 1 is used to structure investments projects with Mauritian double tax avoidance treaty countries such as Singapore, India and South Africa to avail themselves of various benefits such as no capital gains tax being paid, and no tax being imposed on dividends. Global Business License category 2 is used for conducting businesses outside Mauritius and is used chiefly to hold or own investments and assets. Several benefits accrue to such type of license holders such as no capital gains tax being imposed, exemption from taxation, among others.

3

 Critical Assessment of the Mauritian A Economy

The objective of this section is to discuss the different aspects of the Mauritian economy, including public debt, the level of banks’ non-­ performing loans, the balance of payments situation, tourist earnings, and the sector contribution to GDP, amongst others. In addition, a critical assessment is undertaken in the discussions to unleash an enhanced state of analysis. Focus is also laid on key macroeconomic variables such as inflation rate, interest rates and unemployment. It is vital to gain insight about which sectors contribute significantly to the gross domestic product. Thus, an analysis of the sector contribution to GDP becomes warranted. Using a cut-off value of a 4.5% contribution to GDP in 2014, it becomes clear that eight sectors represent around 61% of total contribution to GDP in Mauritius. These sectors are depicted in the Figure  1.1. Probing deeper, it transpires that the manufacturing, wholesale and retail trade and financial and insurance activities sectors account for around 35% of the total sector contribution to GDP.  Thus, policy makers should be cautious of these sectors

7

1  An Overview of the Mauritian Financial System and Its Economy 

as ­deterioration in any one of these three sectors will definitely lead to bearish impacts on the general economic performance of Mauritius. The authorities should therefore always develop better policies which would create a conducive business atmosphere for these three key sectors in view of leveraging economic growth. The analysis clearly showed that the financial sector does play a preponderant role in the level of economic growth in Mauritius (Fig. 1.1). It is widely known that aggregate demand constitutes a combination of consumption, investment, stocks, imports and exports. In that respect, an analysis is undertaken with respect to the components of expenditures on GDP.  Consumption is found to represent the largest chunk of expenditures, with a steady increase noted throughout the years as to attain 88% in 2014. On the other hand, investment is found to

Manufacturing Wholesale and retail trade

Sectors

Financial and insurance activities Public administration and defence Accommodation and food service activities Transportation and storage Real estate activities Construction

0.00%

5.00%

10.00%

15.00%

20.00%

Percentage contribution 2006

2007

2008

2009

2010

2011

2012

2013

2014

Fig. 1.1  Sector contribution to GDP

(Source: Author’s illustration based on data gleaned from the Central Statistics Office, Mauritius)

8 

Economics and Finance in Mauritius

hover around 20% and to be experiencing a gradual decline over the years. Trade ­balance unleashes a general bearish effect on aggregate expenditures based on the value of imports systematically exceeding that of exports. Since higher consumption entails restraint in terms of higher future output, this implies that there is a need to re-engineer the Mauritian economy in such a way as to bolster investment and induce a positive trade balance in order to scale down the significance of consumption in aggregate expenditures (Fig. 1.2). To gain deeper insights about the level of consumption, total consumption is split into household and government consumption. Household consumption is found to account for more than 80% of total consumption in Mauritius  – a level that has remained constant over the years. Thus, households drive consumption in Mauritius to such an extent that 100.00%

80.00%

60.00%

40.00%

20.00%

0.00%

–20.00%

2006

2007

2008

2009

2010

2011

2012

2013

2014

Final consumption expenditure

Gross domestic fixed capital formation

Increase in inventories

Exports of goods and services

Less Imports of goods and services

Fig. 1.2  Components of expenditures on Gross Domestic Product

(Source: Author’s illustration based on data gleaned from the Central Statistics Office, Mauritius)

9

1  An Overview of the Mauritian Financial System and Its Economy  100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

2006

2007

2008

Households

2009

2010

2011

2012

2013

2014

General Government

Fig. 1.3  Composition of consumption in Mauritius

(Source: Author’s illustration based on data gleaned from the Central Statistics Office, Mauritius)

any shock to a household’s income, such as a hike in the income tax rate, will automatically have a detrimental effect on the economy. It can thus be concluded that the consumption structure in Mauritius has not altered much over the last decade (Fig. 1.3). It is also a long-standing feature of the Mauritian economy that the country imports more than it exports. To bolster economic power, a country try to reverse this pattern as this allows it to accumulate reserves, which can be used for productive uses and thereby leverage the country’s economic growth. A decomposed assessment is made with respect to the imports and exports of goods and services, respectively, as depicted in (Fig. 1.4). Over the years 2013 and 2014, exports of goods and exports services, contributed almost equally to the total exports of goods and services. In the case of total imports, the imports of goods always represented a major component under more than 65% of the total value of imports.

10 

Economics and Finance in Mauritius

75.00% 65.00% 55.00% 45.00% 35.00% 25.00% 15.00% 5.00% –5.00%

2006 2 006

2007 2007

2008 2008

2009 2009

2010 2010

2011 2011

2012 2012

2013 2013

2014 2014

Exports Goods over Total Exports

Exports Services over Total Exports

Imports Goods over Total Imports

Imports Services over Total Imports

Fig. 1.4  Structure of exports and imports under expenditures on Gross Domestic Product

(Source: Author’s illustration based on data gleaned from the Central Statistics Office, Mauritius)

Unemployment is regarded as a vital macroeconomic evil as it trails behind undesired effects at both micro and macro levels. At the micro level, the unemployed persons feel depressed and rejected by the society, and, in extreme cases, this can occasion violence and even suicide. At the macro level, high unemployment is synonymous with the loss of potential output since production will be below the natural rate of economic activities. From 2008 onwards, unemployment adopted something of an upward trend, likely due to the impact of the global financial crisis. As at the end of 2015, unemployment stood around 8%. With more graduates in the market, such a figure is expected to be lower. Although unemployment is a significant macroeconomic evil, yet, underemployment also constitutes another aspect of the labour force, which unfortunately, at the present time, has not been well studied and assessed via metrics. The chief drawback of underemployment is that skills and knowledge gained at tertiary or higher education levels are not matched to the job requirements to such an extent that there is an inherent loss of time and resources which could have been employed in a more judicious manner (Fig. 1.5).

11

1  An Overview of the Mauritian Financial System and Its Economy 

9.6 9.0 8.5

7.6 7.2

2005

2006

2007

2008

7.8

8.0

8.0

2012

2013

7.8

8.0

7.3

2009

2010

2011

2014

2015

Fig. 1.5  Unemployment rate

(Source: Author’s illustration based on data gleaned from Bank of Mauritius)

The impact of the US subprime crisis is clearly visible through the decline in tourist arrivals for the year 2009. With the vigorous efforts done on the part of the government, this pattern was reversed and an upward trend was resumed, with 1.2 million being now a reasonable target for hotels and tour operators. Interestingly, in spite of the upward trend noted for tourist arrivals, tourist earnings do not yet appear to have kept a perfect pace with tourist arrivals based on a decline in earnings in 2013. Fortunately, things appear to have improved for the years 2014 onwards. More global roadshows and campaigns should be organized to further buttress the tourism sector as it constitutes a key component of the exporting arm of Mauritius (Fig. 1.6). From 2005 to 2015, GDP per capita (in Rs) rose by a hefty 107% while GDP per capita (in US dollars) increased by just 74%. As a matter of fact, as demonstrated (Fig. 1.7), the exchange rate effects can play against Mauritius. For instance, despite a hike in GDP per capita (in Rs) from Rs 306,620 in 2014 to Rs 322, 061 in 2015, yet, GDP per capita (in US dollars) fell from $9,553 to $8,816. Should the rupee depreciate further against the US dollar, this will reduce the GDP per capita, which is denominated in US dollars. Nonetheless, the Figure 1.7 clearly shows a sustained level of economic performance being accomplished by Mauritius over the last decade. However, more remains to be done in

12 

Economics and Finance in Mauritius 60,000

14,00,000 12,00,000

50,000

10,00,000 8,00,000

Number

In Millions

40,000 30,000

6,00,000 20,000

4,00,000

10,000 0

2,00,000

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

0

Year

Tourism Earnings

Tourist Arrivals

Fig. 1.6  Tourist earnings and arrivals

(Source: Author’s illustration based on data gleaned from Bank of Mauritius)

3,50,000

12,000 3,22,061.14

3,06,619.54 2,91,044.47 9,512.92

In Rupees

2,50,000

8,816.06

2,39,259.44 7,675.61 2,26,348.75 7,344.45 2,20,489.81 6,775.63 6,795.53 1,96,831.31

2,00,000

1,50,000

9,552.70

2,73,864.90 8,584.42 8,828.78 2,57,912.78

10,000

8,000

6,000

1,72,969.77 5,061.18 5,174.92 1,55,825.26

4,000 1,00,000 2,000

50,000

0

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

Fig. 1.7  Gross Domestic Product per capita

(Source: Author’s illustration based on data gleaned from Bank of Mauritius)

In US Dollars

3,00,000

13

1  An Overview of the Mauritian Financial System and Its Economy 

terms of creating the best human capital and increasing the level of diversification of the economy such as moving towards other exporting sectors rather than just relying on the tourism sector (Fig. 1.8). Inflation is widely perceived as bad because it eats up the purchasing power of money. However, there is also a mistaken perception that inflation is always bad. In fact, a mild dose of inflation is considered as beneficial as it assists in “greasing the economic wheels” of a country by inciting production to avail itself of stock appreciation. Bearish demand at the global level, with China being subject to deflationary pressures, constituted the main driver behind the recent decline in oil prices. Such a fall lowered the level of inflation in Mauritius so that it stood at around 1.7% as at the end of 2015. Ironically, such a low inflation rate signifies that it could be a propitious time for the Bank of Mauritius to embark on the formal introduction of an inflation-targeting framework. It is also vital to note that the global fall in demand incited by China unleashed a decline in global oil prices which then ricocheted into a lower inflation rate for Mauritius (Fig. 1.8). The price of oil conveys two important items of information. First, it tends to signal the potential inflation rate. Second, it also provides information about the future global economic state. Based on a s­ustained

10.7

8.8

6.9 5.6

5.1

5.1

5.1 3.6

4.0

1.7

2005

2006

2007

2008

2009

2010

1.7

2011

2012

2013

2014

2015

Fig. 1.8  Headline inflation rate-year end

(Source: Author’s illustration based on data gleaned from Bank of Mauritius)

14 

Economics and Finance in Mauritius

declining trend noted from 2013 onwards, a 65% record decline is noted from end 2013 to end 2015. Such a state of affairs signifies that the global economy may be trapped into a state of a depressed level of economic activity. In that respect, this calls for a re-engineering of the aggregate demand function in order to rely more on internal demand. Being a small open economy, it is undeniable that adverse effects will be felt. However, the adoption of cushioning mechanisms such as a sound diversification of the activities base, coupled with innovative entrepreneurship which will create positive foreign demand, will be best able to offset any adverse repercussions (Fig. 1.9). Throughout the last decade, Mauritius has been able to maintain a not so healthy position in its budget deficit which hovers at less than 6% over GDP. From 2008 onwards, the deficit was squeezed further, ending at 3.2% of GDP as at end 2015. It is important to analyse the budget deficit to GDP ratio as it sends strong signals of the extent of real economic advancement being accomplished. For instance, if, in a specific year, the budget deficit to GDP happens to be 3.5%, then, to achieve real economic progress, the economic growth level has to be relative much higher than 3.5% (Fig. 1.10).

107.7

109.2

110.7

92.3

91.6

75.2

63.3

62.3 57.6

43 38.9

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

Fig. 1.9  ICE Brent Crude (US$ per barrel)

(Source: Author’s illustration based on data gleaned from Bank of Mauritius)

15

1  An Overview of the Mauritian Financial System and Its Economy 

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

-1.8

-2.7 -3.0

-3.2

-3.2

-3.2 -3.5

-4.3 -5.0 -5.3

Fig. 1.10  Budget deficit to GDP in per cent

(Source: Author’s illustration based on data gleaned from Bank of Mauritius)

Seven key sectors account for the lion’s share of credit provided by the banking sector in Mauritius. More specifically, 92.38% as at end December 2015 was given to the construction, tourism, traders, personal, financial and business services, agriculture and fishing, and ­manufacturing sectors in Mauritius. Excessive credit risk concentration is found in the construction sector. Other sectors make less use of credit. It can be posited that some potential crowding-out effects prevail as to some sectors availing of easy access to credit while other sectors are being restricted in funding. Nonetheless, it is clear that banks should now practise a more inclusive lending behaviour by shifting to other sectors in order to secure a well-diversified credit portfolio. Above all, since many construction and real estate builders are already subject to high debt levels, such a sector is expected to be already creditsaturated on the back of there being no viable fresh projects in the pipeline, perhaps with the exception of the newly contemplated government projects such as smart cities and the new government office building at Ebene. As from January 2014, fortunately, the Bank of Mauritius has undertaken some measures to bolster the economy, including the introduction of credit limits via loan-to-value and debt-to-income ratios in an attempt to contain the risks in the construction sector (Fig. 1.11).

16 

Economics and Finance in Mauritius

Construction

30.09%

Tourism

16.85%

Traders

11.11%

Personal

10.16%

Financial and Business Services

9.31%

Agriculture & Fishing

7.48%

Manufacturing

7.37%

Other

1.67%

Infrastructure

1.57%

Transport

1.56%

Information Communication and Technology

0.65%

Public Nonfinancial Corporations

0.62%

Education

0.50%

Professional

0.44%

Media, Entertainment and Recreational Activities

0.31%

Health Development Certificate Holders

0.17%

Freeport Enterprise Certificate Holders

0.13%

Human Resource Development Certificate Holders

0.00%

0%

5%

10%

15%

20%

25%

30%

35%

Fig. 1.11  Sector-wise distribution of credit as at end December 2015

(Source: Author’s illustration based on data gleaned from Bank of Mauritius)

The Key Repo Rate replaced the Lombard Rate in 2006 as the new interest rate to indicate the monetary policy stance to be adopted by the Bank of Mauritius. Since the onset of the crisis, the Key Repo Rate has maintained a falling trend based on the adoption of a doveish monetary policy stance adopted by the Bank of Mauritius. Such a strategy is considered to be aligned to the interest rate-setting paths adhered to by major foreign economies. Indeed, by virtue of Mauritius being a small open economy, its monetary policy strategy usually constitutes a function of both local and international developments, and, in recent years, the latter have weighed more on the scale due to the US subprime ­crisis. An important point to note, and one which is often not cited by policy makers, is that as a small economy, Mauritius automatically bears external shocks such as oil price hikes, which eventually generate higher inflation, and thus higher interest rates, compared to advanced economies who may have more power and resources to alleviate the impact of oil price changes on their inflation rates and on their interest rates. However, this is beneficial to Mauritius in terms of higher interest rates

1  An Overview of the Mauritian Financial System and Its Economy 

17

10 9 8 7 6 5

Dec-06 Mar-07 Jun-07 Sep-07 Dec-07 Mar-08 Jun-08 Sep-08 Dec-08 Mar-09 Jun-09 Sep-09 Dec-09 Mar-10 Jun-10 Sep-10 Dec-10 Mar-11 Jun-11 Sep-11 Dec-11 Mar-12 June-12 Sep-12 Dec-12 Mar-13 June-13 Sep-13 Dec-13 Mar-14 June-14 Sept-14 Dec-14 Mar-15 June-15 Sept-15 Dec-15

4

Fig. 1.12  Evolution of Key Repo Rate

(Source: Author’s illustration based on data gleaned from Bank of Mauritius)

reflecting a larger scope of manouevre of its monetary policy compared with developed economies which then get entangled in the liquidity trap during crisis conditions. Thus, when all the power of interest rate policy is used up by the liquidity trap, developed economies have had to have recourse to unconventional monetary policy such as quantitative easing while Mauritius has stuck to interest rate adjustments (Fig. 1.12). The interest rate spread is computed as the mid-values of the prime lending rate and the savings rate. Since the prime lending rate pertains to the lending rate made to the best borrowers, the above interest rate spread can be deemed as the minimum level of interest rate spread. Indeed, Ramlall (2015) computed the interest rate spread for the banking sector by using a change in the equity base as a proxy for profitability to end up with a spread level of 7%. In that respect, it can be posited that the interest rate spread should hover in the range of 6–8% in Mauritius. However, the World Bank Development Indicators point to an interest rate spread of much lower values, not compatible with the data from the Bank of Mauritius and thus not really reflecting the reality of the Mauritian financial system. As depicted in the Figure 1.14, the interest rate spread from the World Bank data shows a range of 0.5–2.4% for the years spanning from 2008 to 2014, in stark contrast to the findings shown in the first figure of interest rate spread (Figs. 1.13 and 1.14).

18 

Economics and Finance in Mauritius 4.60 4.40

Percent

4.20 4.00 3.80 3.60

Dec-06 Mar-07 Jun-07 Sep-07 Dec-07 Mar-08 Jun-08 Sep-08 Dec-08 Mar-09 Jun-09 Sep-09 Dec-09 Mar-10 Jun-10 Sep-10 Dec-10 Mar-11 Jun-11 Sep-11 Dec-11 Mar-12 June-12 Sep-12 Dec-12 Mar-13 June-13 Sep-13 Dec-13 Mar-14 June-14 Sept-14 Dec-14 Mar-15 June-15 Sept-15 Dec-15

3.40

Fig. 1.13  Interest rate spread for the banking sector

(Source: Author’s illustration based on data gleaned from Bank of Mauritius) 2.5

2

1.5

1

0.5

0

2008

2009

2010

2011

2012

2013

2014

Fig. 1.14  Interest rate spread from the World Bank Development Indicators

(Source: Author’s illustration based on data gleaned from World Bank Development Indicators [The data was gleaned on 20 March 2016])

Prior to the onset of the crisis, the Mauritian rupee depreciated vis-à-­ vis the major currencies such as the US dollar, the pound sterling and the euro. However, after the crisis, things veered course with appreciation of the rupee being more in the scene. Interestingly, when there was a change

19

1  An Overview of the Mauritian Financial System and Its Economy 

in the governor of the central bank of Mauritius, a depreciating trend manifested with vigour as depicted with the vertical line stretching at the period January 2015. The latter gave the author an idea as to whether there is a feasible central bank’s governor bias onto the exchange rates in Mauritius – akin to clash of two opposite strategies, here, labelled as the ‘Clash of the Titans’. Consequently, the author drew the below figure with change in governor tenure being captured by the vertical dotted lines. Interestingly, it appeared as if Governor A adhered to a more bearish stance to exchange rates (meaning to a depreciating rupee) which benefitted the tourism sector in terms of easy wealth accumulation via the euro and pound sterling deposits held in banks. In contrast to Governor A, Governor B clung to a more hawkish stance to exchange rates (an appreciating rupee) which benefitted the importing arm of Mauritius and also alleviated the cost of external public debt. It might be concluded, therefore, that there are strong indications that one governor tends to favour a depreciating rupee as he considers exports to reflect the main driver of economic growth while the other governor prefers an appreciating rupee as he believes that lower import costs induce higher economic growth (Fig. 1.15). 69

Governor B-Feb 2007

64

Pound Sterling

Governor A-30 Dec 14

Governor A

59 54 49

Euro

44 39

US dollar

34

Jul-15

Fig. 1.15  Evolution of the exchange rates-Clash of the Titans

(Source: Author’s illustration based on data gleaned from Bank of Mauritius) Note: Names of the governors have been intentionally removed to avoid any direct targeting

Jan-16

Jul-14

Jan-15

Jul-13

Jan-14

Jul-12

Jan-13

Jul-11

Jan-12

Jul-10

Jan-11

Jul-09

Jan-10

Jul-08

Jan-09

Jul-07

Jan-08

Jul-06

Jan-07

Jul-05

Jan-06

Jul-04

Jan-05

Jul-03

Jan-04

Jul-02

Jan-03

24

Jan-02

29

20 

Economics and Finance in Mauritius

Since most of the corporate debt is tilted towards local debt, this implies a low level of currency risk. In that respect, currency shocks are less likely to generate direct detrimental impacts on the leverage structure of Mauritian corporates. The high proportion of local debt signifies a large degree of home bias. From a financial stability perspective, a depreciating rupee would inflate the level of foreign debt and its repayment costs so that the probability of the default rates of companies, imbued with high proportion of foreign debt, will automatically scale up (Fig. 1.16). Prior to undertaking the balance of payments analysis, it is deemed considerate to fully understand the various components of the balance of payments. At a broader level, the capital and financial account plus current account plus net errors and omissions must sum to zero. Net errors and omissions do not constitute an accommodating item, but simply manifest from imperfections in the source of the data along with feasible compilation errors. Financial account captures direct investment plus portfolio investment plus other investment plus reserve assets (accommodating item to induce balancing). Except for reserve assets, all items are considered as autonomous items. In essence, a balance of payment surplus signifies an accumulation of reserves while a balance of payment 100.00% 90.00% 80.00% 70.00% 60.00% 50.00% 40.00% 30.00% 20.00% 10.00% 0.00%

2011

2012 External debt

2013

4Q-2014

1Q-2015

Domestic debt

Fig. 1.16  Composition of corporate debt structure in Mauritius

(Source: Author’s illustration based on data gleaned from Bank of Mauritius Financial Stability Report 2015)

21

1  An Overview of the Mauritian Financial System and Its Economy 

deficit means a drawing down of reserves. More specifically, a balance of payments deficit implies more debits (payments) than credits (receipts). For instance, imports have a negative sign while exports have a positive sign. Similarly, direct investments abroad are considered with a negative sign. Thus, negative signs do not always signal negative messages. For instance, portfolio investments abroad are considered with a negative sign and are found under assets in the balance of payments. The key item to gauge on the balance of payments relates to the reserve assets position. In fact, since the balance of payments must, by definition, always balance, the official reserve item reflects the level of transfer of international reserves which must move in or out. A positive sign for reserve assets implies bringing in additional reserves to fund a balance of payments deficit. Conversely, a negative sign under the balance of payment implies removing excess reserves due to a balance of payments surplus. In a nutshell, a balance of payments deficit is associated with a loss of reserves while a balance of ­payments surplus is linked to a gain in reserves. From 2011 onwards, Mauritius posted a surplus in its balance of payments. However, the quality of the balance of payments sustainability is not sound as a sudden fall in capital inflows can lead to problems in 50,000

0

40,000

IN MILLIONS

20,000

-5,000

-5246.85

-6041 37,916

34,477

33,172

10,000 -

3,061 2012 (1,651)

2011

2013

(10,000) (20,000)

-10,000

18,586 8,397

3,238 2014

-15,000

(21,824) (42,874)

(36,265)

(36,234)

-16,580

(30,000) (40,000) (50,000)

IN MILLIONS

30,000

-20,000

CURRENT ACCOUNT

CAPITAL AND FINANCIAL ACCOUNT

NET ERRORS AND OMISSIONS

RESERVE ASSETS

-23019 -25,000

Fig. 1.17  Balance of payments

(Source: Author’s illustration based on data gleaned from Bank of Mauritius)

22 

Economics and Finance in Mauritius 64%

100% 90%

63%

80% 70%

62%

60% 61%

50% 40%

60%

30% 20%

59%

10% 0%

58%

External debt

Domestic debt

Total public sector debt to GDP

Fig. 1.18  Public sector debt

(Source: Author’s illustration based on data gleaned from Bank of Mauritius Annual Report 2015)

funding the current account deficit, though the latter has subsided in magnitude in 2014 relative to 2013 (Fig. 1.17). Throughout the recent years, public debt rose at both the domestic and external levels, reaching a rate of 63% to GDP in June 2015 (Fig. 1.18). In that respect, efforts should be made at the government level to curtail unnecessary expenditures since it is widely believed that real winners in the years to come will comprise of economies which have a satisfactory level of economic growth and sustainable public debt levels.

Appendix List of banks, non-bank deposit taking institutions, money-changers and foreign exchange dealers licensed by the Bank of Mauritius, as reported as at end January 2016 by the Bank of Mauritius.

1  An Overview of the Mauritian Financial System and Its Economy 

Banks 1. ABC Banking Corporation Ltd 2. AfrAsia Bank Limited 3. Bank One Limited 4. Bank of Baroda 5. Banque des Mascareignes Ltée 6. Banque Privée de Fleury Limited 7. BanyanTree Bank Limited 8. Barclays Bank Mauritius Limited 9. Century Banking Corporation Ltd 10. Deutsche Bank (Mauritius) Limited 11. Habib Bank Limited 12. HSBC Bank (Mauritius) Limited 13. Investec Bank (Mauritius) Limited 14. MauBank Ltd 15. P.T Bank Internasional Indonesia 16. SBI (Mauritius) Ltd 17. Standard Bank (Mauritius) Limited 18. Standard Chartered Bank (Mauritius) Limited 19. SBM Bank (Mauritius) Ltd 20. The Hongkong and Shanghai Banking Corporation Limited 21. The Mauritius Commercial Bank Limited 22. Warwyck Private Bank Ltd

Non-Bank Deposit Taking Institutions 1 . AXYS Leasing Ltd 2. Cim Finance Ltd 3. Finlease Company Limited 4. La Prudence Leasing Finance Co. Ltd 5. Mauritius Housing Company Ltd 6. Mauritian Eagle Leasing Company Limited 7. SICOM Financial Services Ltd 8. The Mauritius Civil Service Mutual Aid Association Ltd

23

24 

Economics and Finance in Mauritius

Money-Changers (Bureaux de Change) 1 . Abbey Royal Finance Ltd 2. Change Express Ltd 3. Easy Change (Mauritius) Co. Ltd 4. EFK Ltd 5. Iron Eagle Ltd 6. Moneytime Co. Ltd 7. Unit E Co Ltd 8. Viaggi Finance Ltd 9. Vish Exchange Ltd

Foreign Exchange Dealers 1 . British American Exchange Co. Ltd 2. Cim Forex Ltd 3. Island Premier Foreign Exchange Ltd 4. Shibani Finance Co. Ltd 5. Thomas Cook (Mauritius) Operations Company Limited

Bibliography Bank of Mauritius, Annual Reports, various issues. Bank of Mauritius, Monthly Statistical Bulletins, various issues. Bank of Mauritius Financial Stability Report, 2015. Central Statistics Office, Mauritius, database. World Bank, World Bank Development Indicators online database.

Part II The Mauritian Banking Sector

2 Banking Sector Analysis

1

Introduction

This chapter focuses on the Mauritian banking sector. Given that Mauritius is endowed with a bank-based financial system, it becomes vital to undertake some specific bank analysis to identify some of its most important aspects. However, prior to embarking on bank-specific analysis, the chapter offers a general analysis of the background of the banking sector in Mauritius.

2

 eneral Analysis of the Banking Sector G in Mauritius

As shown in Table 2.1, Mauritius is endowed with a solid banking sector with the capital adequacy ratio remaining above 17% as at the end of June 2015. The chief implication of the solid capital base is that the local banking sector is ready to withstand any adverse shocks that may undermine any upside risk to financial stability. Interestingly, such a high capital adequacy ratio corroborates the fact that Mauritius is subject to © The Author(s) 2017 I. Ramlall, Economics and Finance in Mauritius, DOI 10.1007/978-3-319-39435-0_2

27

28 

Economics and Finance in Mauritius

Table 2.1  Risk-weighted capital adequacy ratio Period

Jun-14

Sep-14

Dec-14

Mar-15

Jun-15

Tier 1 capital Tier 2 capital Capital Base (A) Total Risk-Weighted Assets (B) Total on-balance sheet risk-weighted credit exposures Total non-marketrelated off-balance sheet risk-weighted credit exposures Total market-related off-balance sheet risk-weighted credit exposures Total risk-weighted assets for operational risk Total foreign currency exposures Capital charge for trading book position exceeding 5% or more of its total assets Capital Adequacy Ratio (A/B)

91,767 15,683 107,450 621,301

95,109 15,919 111,028 653,417

98,372 13,466 111,838 676,875

106,728 13,991 120,719 728,884

101,276 16,641 117,917 686,826

521,327

556,284

573,592

621,704

591,678

52,055

49,841

54,422

52,406

43,251

2,233

2,127

2,170

4,281

2,950

42,864

43,248

43,739

46,215

45,844

2,823

1,918

2,565

4,081

3,071

387

197

32

17.3%

17.0%

16.5%

16.6%

17.2%

Source: Bank of Mauritius, Annual Report 2015a

high interest rate spreads which benefit the banks in form of high profits which eventually feed into the stronger capital base. As per the 2015 financial stability report issued by the Bank of Mauritius, the percentage of non-performing loans to total loans stood around 5.1% as at end March 2015, a figure that is indicative of the low loan risk present in today’s Mauritian banking sector. The significance of the banking sector cannot be ignored in Mauritius. A quick glance at Fig. 2.1 clearly demonstrates an upward trend for both the total assets of the banking sector and GDP at market prices. As at the end of 2015, the total consolidated assets of the banking sector over GDP

2  Banking Sector Analysis 

29

Millions

Banking assets to GDP 14,00,000

305% 300%

12,00,000

295%

10,00,000

290% 285%

8,00,000

280% 6,00,000

275% 270%

4,00,000

265%

2,00,000 -

260% 2010

2011

Consolidated assets of banking sector

2012

2013

GDP at market prices

2014

2015

255%

Banking sector assets to GDP

Fig. 2.1  Banking sector assets to GDP (Source: Author’s illustration based on data gleaned from Bank of Mauritius Monthly Statistical Bulletins, various issues and GDP data from Central Statistics Office, Mauritius)

hovered at around 293%.1 Thus, the banking sector constitutes a core component of the financial system in Mauritius. It is important, therefore, that the authorities put in place sound regulatory and supervision measures.

3

 pecific Analysis of the Banking Sector S in Mauritius

ABC Banking Corporation Limited began its operations in December 2010 (Table 2.2). From the calculation of the various ratios given here, it is clear that the bank is still in an infant stage as the level of loans is not yet aligned to the level necessary to exploit the full potential of their deposits. Ironically, the total loans over deposits from customers’ ratio has undergone a steady decline over the years. Such a finding clearly shows the extent to which the market is being controlled by two major banks in Mauritius  Estimate for GDP for year 2015 was used based on data from the Central Statistics Office.

1

30 

Economics and Finance in Mauritius

Table 2.2  ABC Banking Corporation Limited Year

2010

Loans to deposits assessment Loans and advances 0.94 to customers/ Deposits from customers Total loans/Deposits 1.02 from customers Costs assessment Personnel expenses/ 0.78% Total assets Depreciation and 0.07% amortization/Total assets Provision for credit 0.85% impairment on financial assets/ Total assets Asset structure composition Investments/Total 0.50% assets Total loans/Total 81.94% assets Property, Plant and 0.30% equipment/Total assets Cash and cash 9.72% equivalents/Total assets Intangible assets/ 0.01% Total assets Profitability assessment Net interest income/ 3.57 Net fee and commission income Profits/Total assets 0.61% Net interest income/ 2.36% Total assets Net fee and 0.66% commission income/Total assets Equity assessment Capital and reserves/ 9.25% Total assets

2011

2012

2013

2014

2015

0.76

0.60

0.37

0.32

0.27

0.82

0.63

0.52

0.37

0.38

1.34%

1.02%

0.72%

0.72%

0.71%

0.20%

0.15%

0.12%

0.12%

0.13%

0.97%

0.47%

0.17%

0.69%

0.58%

2.37%

27.86%

29.60%

28.23%

24.52%

67.29%

55.09%

46.81%

33.45%

35.00%

0.87%

0.67%

1.63%

1.56%

1.81%

20.79%

8.45%

16.13%

31.06%

34.14%

0.50%

0.34%

0.21%

0.15%

0.10%

2.61

3.08

2.34

2.75

4.68

−0.82% 1.45%

−0.21% 1.45%

0.08% 1.03%

0.17% 1.27%

0.55% 1.42%

0.56%

0.47%

0.44%

0.46%

0.30%

7.27%

5.47%

4.38%

4.31%

4.93%

Source: Author’s computations based on data gleaned from Mauritius Bankers Association, various issues

2  Banking Sector Analysis 

31

such that market penetration is highly difficult for new entrants. In that respect, the bank had no choice but to channel most of its assets to other channels, namely, cash and investments, to ensure minimum returns over rather stable costs incurred such as staff costs and depreciation. The interesting feature of the bank is that it managed to generate profits after two years of losses. Such a result substantiates the notion that the banking business in Mauritius is definitely viable and profitable as it avails itself of strong deposit funding from Mauritians who are highly risk-averse and tend ot avoid investment in sophisticated products so that they simply channel most of their income to banks as deposits. Alternatively stated, the Mauritian banking business culture might be seen to be characterized by so-called ‘lazy’ banking. Coupled with the latter, the high interest rate spread reflects an opportune avenue for harnessing profits to local banks. In terms of earning capacity, it is clear that interest income reflects the core area of profit-earning capacity for Mauritian banks, based on the high interest rate spread compared to the structure of fees and commissions. In spite of the fact that the ABC Banking Corporation Limited managed to register positive profits, in terms of total assets, it is evident that the bank is not fully optimizing on its resources. As a matter of fact, over time, the bank seems to have more of its assets tilted towards cash and investments in lieu of loans. If such a state of affairs prevails in future, it is most likely that the bank will be unable to sustain its assets to thereby be subject to undermined profitability that would in turn influence its operating activities. The bank may try to tap into new and innovative products, let alone scaling up its loan market penetration to give a real boost to its profits. The low loans to total assets ratio again buttressed the notion of the Mauritian banking sector being controlled by MCB Ltd. and SBM Ltd. Nonetheless, the high portion of assets being skewed towards investments may signify greater focus towards investment banking activities to partly neutralize the dominance of the two big banks in the local market (Table 2.3). Despite being incorporated in 2007, AfrAsia Bank Limited managed to have a large portion of its assets tilted towards loans. Above all, the bank obtained many awards such as the Best Local Private Bank in Mauritius in 2012 and 2013 by Euromoney with its activities being split into three main

2008

Loans to deposits assessment 0.39 Loans and advances to customers/Deposits from customers Total loans/Deposits 0.88 from customers Costs assessment Personnel expenses/ 1.78% Total assets 0.14% Depreciation and amortization/Total assets 0.29% Net Allowance for credit impairment/ Total assets Asset structure composition Investments/Total 12.53% assets Total loans/Total assets 72.32% 0.43% Property, Plant and equipment/Total assets 11.31% Cash and cash equivalents/Total assets Intangible assets/Total 0.62% assets

Year

Table 2.3  AfrAsia Bank Limited

0.61

0.83

1.00% 0.07%

0.35%

13.75% 73.87% 0.15%

11.08%

0.17%

0.95

1.20% 0.09%

0.18%

12.24% 79.47% 0.19%

3.43%

0.22%

2010

0.42

2009

0.08%

6.47%

74.54% 0.08%

18.09%

0.15%

0.04%

0.76%

0.84

0.57

2011

0.08%

3.05%

73.44% 0.06%

22.65%

0.11%

0.03%

0.78%

0.83

0.65

2012

0.05%

5.08%

73.17% 0.09%

21.37%

0.00%

0.03%

0.63%

0.85

0.51

2013

0.08%

4.28%

75.54% 0.14%

19.41%

0.37%

0.03%

0.61%

0.87

0.42

2014

0.07%

3.42%

86.02% 0.13%

9.85%

0.68%

0.03%

0.40%

0.94

0.32

2015

32  Economics and Finance in Mauritius

2.43

0.41% 1.19% 0.49%

7.56%

4.25

0.34% 1.00% 0.24%

10.54%

6.32%

0.26%

0.69% 1.30%

5.11

7.57%

0.46%

0.81% 1.45%

3.18

6.98%

0.50%

0.96% 1.16%

2.35

7.06%

0.34%

0.47% 1.40%

4.15

Source: Author’s computations based on data gleaned from Mauritius Bankers Association, various issues

Profitability assessment 2.40 Net interest income/ Net fee and commission income Profits/Total assets −1.35% Net interest income/ 0.65% Total assets 0.27% Net fee and commission income/ Total assets Equity assessment Capital and reserves/ 7.04% Total assets 6.16%

0.29%

0.24% 1.17%

4.04

2  Banking Sector Analysis 

33

34 

Economics and Finance in Mauritius

areas: private banking and wealth management; corporate and investment banking; and international banking solutions. It can be thereby concluded that AfrAsia Bank managed to consolidate its position as a reputed bank over a very short period of time. The low loans to customers over deposits from customers ratio clearly shows the international intent of the bank, let alone tapping on over activities as encapsulated under its three-pronged focus in banking activities. More precisely, the bank lends heavily to other banks under total loans in view of harnessing higher profitability levels. This can also be explained by the international links reflected by the bank in many countries such as Singapore, South Africa and Zimbabwe. In terms of profitability assets, as is common to all other Mauritian banks, net interest income predominates over net fees and commissions in spite of some conspicuous increases noted in the latter for the years 2010, 2012 and 2013. Compared to the experience of mature banks such as SBM Ltd. and MCB Ltd., the AfrAsia Bank currently has a relatively low equity base. The latter will require time to be bolstered chiefly with profits consolidating on the capital base of the bank to withstand any feasible adverse shock. Bank of Baroda has been active in Mauritius since 1962. However, its loans to customers are currently funded more by banks’ deposits in lieu of customers’ deposits (Table 2.4). Consequently, in the event that the banks which fund Bank of Baroda are subject to funding pressures, this would automatically impound on Bank of Baroda’s funding structure. Although such a risk is only latent at present, it could possibly arise in future. In fact, the bank is already subject to negative pressures, as evidenced by losses in 2015 and dwindling equity levels over recent years. In addition, it is clear that the bank does not avail itself of a robust market structure as depicted by most of its assets being tilted towards investments throughout the different years. It appears that the strategy deployed by the bank has remained static throughout the different years with less than 50% of the total assets being channeled towards loans. Such a statement is bolstered by its steadily falling capital and reserves over recent years despite it having operated in Mauritius for many decades. In that respect, it is expected that products offered

Year 2007 Loans to deposits assessment NA Loans and advances to customers/ Deposits from customers 0.26 Total loans/Total deposits from customers and banks Costs assessment Personnel expenses/ 0.21% Total assets 0.03% Depreciation and amortization/Total assets Asset structure composition Investments/Total 71.94% assets Total loans/Total 22.81% assets

Table 2.4  Bank of Baroda Limited

0.51

0.22% 0.03%

49.46% 43.18%

0.21% 0.04%

67.30% 26.57%

0.64

NA

0.35

2009

2008

34.70%

61.77%

0.01%

0.16%

0.45

0.51

2010

38.18%

58.49%

0.02%

0.15%

0.47

0.69

2011

40.94%

54.83%

0.01%

0.11%

0.47

0.74

2012

41.45%

55.29%

0.01%

0.10%

0.47

1.03

2013

44.15%

43.33%

0.02%

0.16%

0.55

1.26

2015

(continued)

38.90%

56.19%

0.01%

0.10%

0.52

1.11

2014

2  Banking Sector Analysis 

35

2007

1.47% 1.76%

11.79%

8.53%

4.81%

1.05%

2009

1.07% 1.33%

3.09%

NA

2008

5.45%

0.71% 1.04%

2.24%

0.46%

2010

4.47%

0.63% 0.92%

2.27%

0.40%

2011

3.85%

0.53% 0.79%

3.18%

0.42%

2012

3.39%

0.52% 0.63%

2.40%

0.36%

2013

3.06%

0.25% 0.47%

3.92%

0.29%

2014

Source: Author’s computations based on data gleaned from Mauritius Bankers Association, various issues.

NA Property, Plant and equipment/Total assets 1.39% Cash and cash equivalents/Total assets Profitability assessment Profits/Total assets 0.97% Net interest income/ 1.28% Total assets Equity assessment Capital and reserves/ 8.51% Total assets

Year

Table 2.4 (continued)

4.93%

−0.20% 0.28%

11.47%

0.44%

2015

36  Economics and Finance in Mauritius

2  Banking Sector Analysis 

37

by Bank of Baroda are not likely to be as competitive as those offered by some of its competitors. It is interesting to note that since Bank of Baroda constitutes an international bank and is one of the largest banks in India, it could be that the parent bank is simply trying to harness some bouts of benefits via from its foreign branch chiefly for Indian expatriates who work in Mauritius. Based on such an analysis, it becomes clear that the bank is less likely to undertake vehement measures to substantially increase its market share, trailing behind a state of lazy banking. The steadily diminishing equity base of the bank could herald potential problems looming large on its existence should a major shock wipe out an important chunk of its capital base. Bank One Limited is shown to be benefiting from a sound client base, represented by a relatively satisfactory level of the loans to deposit ratio (Table 2.5). This is also being demonstrated by its asset structure, whereby the lion’s component emanates from loans. Such a finding endorses the notion of conventional banking being practised by the bank. Nearly a quarter of the total assets of Bank One Limited is skewed towards cash which is considered to be an asset based on low returns. However, the burgeoning level of cash could be synonymous to some important investment which is in the pipeline. In that respect, the high level of cash projects a need to undertake a major investment project without resorting to external borrowing in order to concentrate the gains of such investment to the bank only. As is the case with any other bank, the ratio of net interest income over net fee and commission income has been increasing over time, clearly highlighting the significance of net interest income as the crude form of earning capacity for the bank. Over the years 2009 to 2014, net fee and commission income to total assets has not changed much in spite of an upward trend registered by total assets. Subsequently, it can be deduced that Bank One Limited adopted a strategy of low fee and commission in view of consolidating its profitability level. The capital base of the bank fell gradually between 2008 and 2012 before picking up sharply for the last two years. Overall, a rather satisfactory performance is noted for Bank One Limited. Bank des Mascareignes constitutes a wholly-owned subsidiary of Groupe BPCE, France and thus benefits from the advantages of strong

38 

Economics and Finance in Mauritius

Table 2.5  Bank One Limited Year

2008

2009

Loans to deposits assessment Loans and 0.64 0.59 advances to customers/ Deposits from customers Total loans/ 0.71 0.61 Deposits from customers Costs assessment Personnel 1.84% 1.37% expenses/Total assets Depreciation and 0.52% 0.29% amortization/ Total assets Asset structure composition Investments/Total 14.04% 10.30% assets Total loans/Total 63.56% 54.82% assets Property, Plant 4.68% 3.12% and equipment/ Total assets Cash and cash 7.76% 23.95% equivalents/ Total assets Intangible assets/ 0.38% 0.45% Total assets Profitability assessment Net interest 1.24 2.09 income/Net fee and commission income Profits/Total −1.77% 0.37% assets Net interest 0.43% 1.26% income/Total assets Net fee and 0.35% 0.60% commission income/Total assets Equity assessment Capital and 7.87% 6.10% reserves/Total assets

2010

2011

2012

2013

2014

0.63

0.64

0.69

0.76

0.68

0.64

0.65

0.69

0.77

0.70

1.24%

1.30%

1.20%

1.27%

1.41%

0.25%

0.22%

0.18%

0.20%

0.21%

10.15%

11.57%

8.08%

8.41%

7.17%

57.83%

58.72%

61.57%

66.12%

60.14%

2.31%

2.04%

1.70%

1.75%

1.82%

23.94%

20.33%

22.24%

18.46%

26.17%

0.33%

0.33%

0.29%

0.35%

0.21%

3.09

3.43

3.70

4.10

5.43

1.21%

1.12%

1.05%

0.30%

0.65%

2.09%

2.27%

2.04%

2.75%

3.13%

0.68%

0.66%

0.55%

0.67%

0.58%

5.78%

6.06%

5.82%

7.62%

8.16%

Source: Author’s computations based on data gleaned from Mauritius Bankers Association, various issues

2  Banking Sector Analysis 

39

group networking and a high degree of international cooperation. Such international assistance explains why the loans to deposits ratio exceed one, i.e., the bank is giving more loans than its deposits can support itself. On further investigation it transpires that the bank uses borrowed funds which could act as additional funding source to support the higher levels of loans given to customers. The drawback of such a strategy is that should the parent company in France be subject to funding problems, which may trigger a call-back of funds to jeopardize the loans sustainability structure of the bank. In that respect, external funding risk represents the chief source of risk for the bank. One interesting element to note about the bank is that, in spite of the high levels of loans, the bank channels around 10% of its assets into cash and investments (Table 2.6). In that respect, the bank channels most of its efforts towards loans, with some diversifications into investments and cash to harness the maximum returns possible. The bank managed to build up positive capital levels over time as witnessed by steadily rising equity levels over the years. Like other banks, net interest income acts as the core earning component of the bank. The Mauritius Commercial Bank Limited constitutes one of the two banks which controls a large part of the market share in terms of loans and deposits. From its asset structure, it is clear that the bank focuses most of its activities on loans with some bouts of diversification effects in investments and cash (Table 2.7). An interesting finding to note is that the level of loans to total assets has shrunk steadily since 2012 onwards. Such a state of affairs is also reflected in the total loans to deposits ratio, which fell from 0.93 in 2012 to 0.78 in 2015. Probing deeper, it was found that the value of total loans rose over the whole period, but the increase in total assets and total deposits were higher than that of the loans. Profitability assessments show that net interest income in lieu of net fee and commission income accounts for a greater level of profits throughout the period under analysis. However, in relative terms and over time, the net interest income over net fee and commission income ratio has maintained a gradual declining trend over time since 2009. Findings further show that, over time, the equity base of the bank rose steadily, consolidating its ability to absorb any likely future shocks.

2007

Loans to deposits assessment 1.32 Loans and advances to customers/Deposits from customers Total loans/Deposits 1.32 from customers Total loans/Total 1.29 deposits Costs assessment Personnel expenses/ 0.58% Total assets 0.23% Depreciation and amortization/Total assets Asset structure composition Investments/Total 39.61% assets Total loans/Total assets 47.01% 0.56% Property, Plant and equipment/Total assets 9.72% Cash and cash equivalents/Total assets Intangible assets/Total 0.42% assets

Year 1.05

1.54 1.54

1.20% 0.09%

8.55% 83.86% 0.46%

4.82%

0.14%

2.07 2.01

1.78% 0.14%

5.64% 82.55% 0.46%

8.91%

0.25%

2009

1.25

2008

Table 2.6  Banque des Mascareignes Limited

0.08%

5.61%

83.47% 0.35%

7.45%

0.07%

1.00%

1.41

1.41

0.93

2010

0.03%

11.78%

75.98% 0.30%

8.27%

0.04%

0.76%

1.35

1.35

1.11

2011

0.01%

9.93%

77.77% 0.24%

7.94%

0.03%

0.78%

1.14

1.25

1.12

2012

0.01%

9.07%

75.56% 0.21%

10.59%

0.03%

0.63%

1.2

1.33

1.29

2013

0.04%

6.92%

77.67% 0.19%

10.41%

0.03%

0.61%

1.09

1.19

1.15

2014

40  Economics and Finance in Mauritius

5.004

0.79% 2.27% 0.45%

6.37%

5.484

0.33% 1.32% 0.24%

4.66%

6.26%

0.57%

1.09% 2.32%

4.057

6.86%

0.47%

0.68% 2.38%

5.039

6.15%

0.50%

−3.28% 2.24%

4.484

9.04%

0.53%

0.61% 2.25%

4.204

Source: Author’s computations based on data gleaned from Mauritius Bankers Association, various issues

Profitability assessment 14.44 Net interest income/ Net fee and commission income Profits/Total assets 0.02% Net interest income/ 1.27% Total assets 0.09% Net fee and commission income/ Total assets Equity assessment Capital and reserves/ 5.28% Total assets 8.46%

0.54%

0.39% 2.16%

4.024

2  Banking Sector Analysis 

41

2007

Loans to deposits assessment 0.81 Loans and advances to customers/Deposits from customers Total loans/Deposits 0.82 from customers Total loans/Total 0.81 deposits Costs assessment Personnel expenses/ 1.23% Total assets 0.34% Depreciation and amortization/Total assets Asset structure composition Investments/Total 15.77% assets Total loans/Total assets 64.47% 2.59% Property, Plant and equipment/Total assets 15.28% Cash and cash equivalents/Total assets Intangible assets/Total 0.24% assets

Year 0.8

0.82 0.8

1.15% 0.22%

13.27% 67.46% 2.22%

12.34%

0.20%

0.76 0.74

1.23% 0.37%

21.60% 61.26% 2.09%

13.37%

0.17%

2009

0.74

2008

Table 2.7  The Mauritius Commercial Bank Limited

0.41%

9.52%

70.36% 2.67%

13.05%

0.22%

1.05%

0.83

0.85

0.83

2010

0.59%

5.32%

73.80% 3.31%

11.55%

0.31%

1.10%

0.9

0.92

0.9

2011

0.47%

5.56%

75.32% 3.20%

9.68%

0.34%

1.08%

0.93

0.95

0.92

2012

0.35%

7.45%

72.70% 2.77%

11.13%

0.31%

1.03%

0.92

0.95

0.92

2013

0.24%

8.17%

67.81% 2.40%

14.13%

0.29%

0.99%

0.84

0.86

0.82

2014

42  Economics and Finance in Mauritius

3.58

2.47% 3.12% 0.87%

9.67%

3.59

2.03% 3.19%

0.89%

9.83%

9.79%

0.88%

2.40% 3.36%

3.81

10.29%

0.91%

2.11% 3.18%

3.5

11.48%

1.02%

2.51% 3.45%

3.37

11.70%

1.07%

2.07% 3.31%

3.09

11.57%

1.12%

1.89% 3.23%

2.89

Source: Author’s computations based on data gleaned from Mauritius Bankers Association, various issues

Profitability assessment Net interest income/ Net fee and commission income Profits/Total assets Net interest income/ Total assets Net fee and commission income/ Total assets Equity assessment Capital and reserves/ Total assets 10.20%

1.11%

1.54% 3.02%

2.73

2  Banking Sector Analysis 

43

44 

Economics and Finance in Mauritius

Like the Mauritius Commercial Bank Limited, the State Bank of Mauritius Limited represents another major bank which controls a large chunk of the total loans, deposits and assets of the banking sector (Table 2.8). Approximately 60% of the total assets of the bank is tilted towards loans while 20% make up its investment component. Like any other bank, the main income-earning capacity of the bank emanates from net interest income in lieu of net fee and commission income. Moreover, over time, the bank managed to bolster its equity capital base which is vital to withstand any future adverse shocks. The benefit which the State Bank of Mauritius limited has compared to Mauritius Commercial Bank limited is that government can always intervene to rescue the bank since it is par-owned by the government. In that respect, the bank carries implicit contingent liability to public debt should things turn sour. The robust level of equity capital base accumulated over the years will undeniably help to easily absorb such contingent liability. As expected, the bank also exploits a strong loans to deposits ratio, clearly showing its power in gaining and maintaining a high level of market share over time. Since the lion’s share of total deposits emanates from customers’ deposits, this implies that the bank avails itself of a particularly strong level of deposits base and thus a strong funding structure. Inquisitive readers can resort to further analysis of the remaining banks which fall under the purview of the Mauritian banking sector based on the same approach, i.e., loans to deposits, costs, asset structure, profitability and equity assessments, respectively.

4

 Holistic Assessment of the Mauritian A Banking Sector

To undertake a holistic assessment of the Mauritian banking sector, recourse is made towards high-quality visual tools, in particular, chernoff faces. The main benefit of chernoff faces is that they are able to cater for different features under scrutiny for a large number of units under focus so that a single visual analysis can disclose important information which is often unseen unobserved by quantitative analysis. Focus is laid on ten key financial metrics – net interest income, net fees and c­ ommission, salaries,

2007

Loans to deposits assessment 0.62 Loans and advances to customers/Deposits from customers Total loans/Deposits 0.62 from customers Total loans/Total 0.62 deposits Costs assessment Personnel expenses/ 0.76% Total assets 0.49% Depreciation and amortization/Total assets Asset structure composition Investments/Total 16.92% assets Total loans/Total assets 51.56% 4.37% Property, Plant and equipment/Total assets 21.82% Cash and cash equivalents/Total assets Intangible assets/Total 0.57% assets

Year 0.61

0.66 0.66

0.91% 0.39%

23.22% 54.27% 2.99%

15.26%

0.12%

0.66 0.66

0.94% 0.46%

21.13% 54.71% 3.68%

16.08%

0.32%

2009

0.65

2008

Table 2.8  The State Bank of Mauritius Limited

0.11%

6.76%

56.04% 3.70%

29.32%

0.22%

0.99%

0.70

0.71

0.71

2010

0.06%

6.99%

61.30% 2.99%

23.81%

0.20%

0.95%

0.81

0.81

0.80

2011

0.09%

7.39%

67.97% 2.83%

15.96%

0.19%

1.02%

0.85

0.85

0.83

2012

0.60%

6.13%

66.85% 2.44%

16.61%

0.26%

1.58%

0.87

0.87

0.86

2013

(continued)

0.93%

6.66%

59.47% 2.36%

22%

0.14%

0.93%

0.75

0.75

0.75

2014

2  Banking Sector Analysis 

45

2007

0.92%

13.75%

13.94%

2.34% 2.97%

3.17% 2.97% 0.92%

3.23

2009

3.21

2008

16.02%

0.78%

2.28% 3.13%

4.02

2010

14.81%

0.88%

1.96% 2.59%

2.93

2011

14.80%

1.16%

4.26% 3.28%

2.84

2012

14.53%

1.36%

4.25% 5.48%

4.01

2013

Source: Author’s computations based on data gleaned from Mauritius Bankers Association, various issues

Profitability assessment 3.10 Net interest income/ Net fee and commission income Profits/Total assets 2.69% Net interest income/ 3.03% Total assets 0.98% Net fee and commission income/ Total assets Equity assessment Capital and reserves/ 12.84% Total assets

Year

Table 2.8 (continued)

14.61%

0.73%

2.51% 3.45%

4.70

2014

46  Economics and Finance in Mauritius

2  Banking Sector Analysis 

47

investments, total assets, deposits, loans, net profits, total equity, and cash and cash equivalents. The data were gleaned from the Mauritius Bankers Association Profile of Banks, 2015 edition. To level the playing field, all data were converted into rupees in case of US dollar denominations. The different features of the faces capture the following data. In essence, Chernoff faces are able to cater for 15 components simultaneously. Height of face: Net interest income Width of face: Net fees and commission Structure of face: Salaries Height of mouth: Net profits Width of mouth: Cash and cash equivalents Smiling: Investments Height of eyes: Total assets Width of eyes: Loans Height of hair: Deposits Width of hair: Total equity Style of hair: Net interest income Height of nose: Net fees and commission Width of nose: Salaries Width of ear: Net profits Height of ear: Cash and cash equivalents A quick visual analysis (Fig. 2.2) clearly shows that the Mauritius Commercial Bank Limited substantially outperforms all of the other banks in the Mauritian banking sector. For instance, the bank makes use of robust deposits funding as shown by its height of hair height being conspicuously taller than that of any other bank in the Mauritian banking sector. Readers can go through the different faces and compare among them as an exercise to best understand the dynamics and characteristics of the various banks which make up the local banking sector.

48 

Economics and Finance in Mauritius

Fig. 2.2  Chernoff faces for the Mauritian banking sector (Source: Author’s Illustration)

5 A Network Analysis of the Sector-Wise Distribution of Credit in Mauritius In the period following the US subprime crisis, it became widely recognised among financial experts that strong connections could be the source of significant upside risk to financial stability. In order to unleash a full assessment of the connections among different sectors of the Mauritian economy, it is considered important to conduct a network assessment. Table 2.9 captures the various sectors involved, along with their respective code names. The data were gleaned from the monthly statistical bulletins issued by the Bank of Mauritius.

2  Banking Sector Analysis 

49

Table 2.9  Sectors considered in analysing the sector-wise distribution of credit in Mauritius Code name

Sectors

AGR MANUF TOUR TRANS CONS TRAD ICT FIN INFRA PUB FREE HEAL PERSO PROF EDUC MEDI

Agriculture & Fishing Manufacturing Tourism Transport Construction Traders Information Communication and Technology Financial and Business Services Infrastructure Public Nonfinancial Corporations Freeport Enterprise Certificate Holders Health Development Certificate Holders Personal Professional Education Media, Entertainment and Recreational Activities

To assess the impact of one sector on another, granger causality p-values are employed and adjusted so that lower p-values translate into higher values and thereby directly convey stronger impacts. The sample is split into two different periods: namely, June 2005 to December 2010 and from January 2011 to July 2016. The edge values are adjusted graphically to show larger sizes for stronger relationships. Large sectors are defined as those which represent at least 10% of the total credit granted for a given month and are represented as red nodes in the network. In a parallel manner, risky sectors are captured as green nodes in the credit network. The cyclical components are obtained for the various sectors using 1600 for lambda under Hodrick–Prescott analysis. Risky sectors are those which exhibited outside normal range fluctuations in their cyclical components. Sectors which were both large and risky are being captured by light blue nodes. Finally, all the rest of the sectors are represented as maroon coloured nodes. The results are depicted in Figs.  2.3 and 2.4, respectively, for the first and second periods of analysis.

50 

Economics and Finance in Mauritius

2005M06−2010M12

PUB ICT TRANS EDUC

FREE MANUF

ICT

INFRA

ICT

PROF

HEAL

FREE TOUR

CONS FIN TRAD

PERSO FIN

PROF

EDUC MEDI

MEDI

AGR

AGR

AGR HEAL

TRAD TOUR CONS

PUB

PERSO

MANUF TRANS

INFRA

HEAL

FREE

MANUF HEAL

Fig. 2.3  Network analysis for period June 2005 to December 2010

(Source: Author’s Illustration)

EDUC

51

2  Banking Sector Analysis 

2011M01−2016M07 ICT MANUF PERSO PROF

AGR

PROF INFRA

TRAD HEAL TRANS FIN

FREEEDUC

AGR

CONS ICT

PERSO

PERSO

TOUR PUB

INFRA

AGR

MANUF

MEDI

MANUF

Fig. 2.4  Network analysis for period January 2011 to July 2016 (Source: Author’s Illustration)

ICT

52 

Economics and Finance in Mauritius

Bibliography Bank of Mauritius. (2015a). Annual Report. Bank of Mauritius. (2015b, September). Financial Stability Report. Mauritius Bankers Association reports, various issues.

3 Developing a Credit Risk Model for Mauritian Bankers

The objective of this chapter is to develop a credit risk model which gauges the repayment capacity of household borrowers in Mauritius. The research is innovative as it differs from prior credit risk models by moving away from default modelling to repayment capacity modelling through the use of the Debt Service Coverage Ratio (DSCR). The model was developed by the author while he was working for an international bank. The chief benefit of our modelling approach is that it is not only practical, but also induces a sense of proactivity as the lender often has knowledge pertaining to the underlying forces which either boost or undermine DSCR. To trigger a holistic assessment, a whole set of models is employed such as continuous (Ordinary Least Squares), binary (Probit and Logit) and ‘discretisized’ (Ordered logit) models for DSCR in the study. To unleash richer results, our model provides due consideration to the interaction variables. Findings show that DSCR is positively influenced by margin cover and the loan tenor. Negative forces are conspicuously noted in the case of the cost of the loan and dummy variables such as loan for construction and borrowers who are employed by the public sector. Some evidence prevails as to the interaction between town dummy and margin cover, both of which have an © The Author(s) 2017 I. Ramlall, Economics and Finance in Mauritius, DOI 10.1007/978-3-319-39435-0_3

53

54 

Economics and Finance in Mauritius

effect on DSCR. Poor evidence is found in the case of age, arrears and town dummy. In general, findings advocate that a careful approach should be taken by banks whenever they compute the margin cover as it is found to constitute a particularly vital risk-mitigating mechanism when it comes to alleviating the moral hazard problem that is inherent in any lending activity.

1

Introduction

What are the potential drivers which impact on the repayment capacity of borrowers who contract mortgage loans? In this chapter, a focus is laid on the relationship between the repayment capacity1 of borrowers with their respective attributes together with other variables. As a matter of fact, most credit risk models have been inspired from theoretical foundations such as the structural and the reduced-form credit risk models, with emphasis being laid on default. This is where the current chapter contributes to the credit risk literature. First, the model employed builds on prior models by focusing on a credit risk model which directly befits the practical needs of the banker/lender by analysing the underlying forces which endorse the repayment capacity of a borrower. The main advantage of such a model over the prior well-established default forecasting models is that it breeds proactivity in the sphere of credit risk management as lenders usually know the underlying positive and negative forces which impact on the repayment capacity of their clients. Second, the proposed model is considered to be universal as it can be applied to any type of country as long as the ultimate objective is geared towards assessment of mortgage loans. The chapter is organized as follows. Section 2 briefly discusses on the credit risk literature. In Sect.  3, the dataset and the variables are thoroughly being discussed together with the econometric model. In Sect. 4, the empirical results are presents. Finally, Sect. 5 concludes.  Throughout the rest of the chapter, the terms repayment capacity and DSCR will be used interchangeably. 1

3  Developing a Credit Risk Model for Mauritian Bankers 

2

55

Related Literature

Risk is considered as an important concept which cuts across a wide range of different fields, including such diverse activities as health and safety, construction engineering and software engineering. In finance, risk2 is often considered as the identification and measurement of the probability of an event which generates loss. One of the most vital characteristics in finance pertains to continuous and evolving development achieved in the area of credit risk modeling. Considered as a classical structural credit risk model, the Merton (1973, 1974) model was developed under the flavor of the Black and Scholes (1973) option pricing theory. The Merton model is based on default risk measurement and constitutes an overwhelmingly used model among both practitioners and academicians in the world. Technically speaking, the Merton model argues that a firm is unlikely to default as long as the value of its liabilities is lower than the market value of the company’s assets.3 With time, the literature on credit risk modelling advanced to accommodate for the deficiencies in the structural model, namely the absence of information asymmetry between the lender and the borrower, default prevailing only at the maturity date, zero coupon bond and flat term structure of interest rates. The reduced form models are tilted towards a distinct information set relative to that of the structural form model. Black and Cox (1976) proposed a model whereby default could occur at any point in time. Geske (1977, 1979) extended the model to account for a coupon bond instead of the zero coupon bond. Similarly, Nielsen et  al. (1993) and Longstaff and Schwartz (1995) included stochastic interest rates, in particular a Vasicek (after Vasicek 1977) process for interest rate instead of the flat term structure of interest rates. Finally, the third-­generation credit risk models consist of KMV and CreditMetrics.  Risk refers to downside risk in this case. The counterpart of downside risk is known as upside risk.  The value of the firm is tantamount to the sum of its liabilities and equity value. Should liabilities exceed the firm’s value, then, equity value becomes negative so that it is better for the company to default, bearing in mind the limited liability feature of a company. In contrast, when market value exceeds liabilities, this increases the residual payoff of the shareholders so that default has lower propensity to occur. 2 3

56 

Economics and Finance in Mauritius

The Kealhofer, McQuown and Vasicek (KMV) model constitutes a version of the Merton model with the distinction that it is geared towards the probability of default in lieu of debt valuation. On the other hand, the CreditMetrics model pertains a portfolio model, developed by JP Morgan, to evaluate credit risk. The credit risk models discussed are chiefly used for corporate loans so that they are not much useful in the case of mortgage loans. The reason is based on the fact that a company is imbued with limited liability while a person has less incentives to default even if the value of his/her housing loan exceeds the market value of all his/her assets as it refers to a personal ownership which is at stake. Associated with that, there is an inner desire to own a property in one’s lifetime so that repayment becomes a committed responsibility. Bearing in mind these caveats, the present study adheres to a different but more pragmatic approach to credit risk modelling by employing a unique database from a Mauritian financial institution which is involved in the housing loan business.

3

Econometric Methodology and Hypothetical Relationship for Variables

This study employs a unique database on 15,600 borrowers who contracted housing loans from a financial institution4 in Mauritius. All data points gathered refer to the period ending October 2010. The data set consists of distinct features of borrowers along with information relating to the loan amount approved and disbursed. Subsequently, information pertaining to when the loan was disbursed, the interest rate applied thereon, the maturity period over which the loan repayment has been planned, the value of security furnished, the purpose of the loan, the borrower’s monthly income and the sector in which the borrower is being identified, and his/ her place of residence. To gauge the repayment capacity of the borrower,  The obtained data captured customers as numbered units because the data was already cleaned of any name of the borrower to shun off any confidentiality/data violation issues as the sole aim in using such data was for research purposes. 4

57

3  Developing a Credit Risk Model for Mauritian Bankers 

recourse is made towards the DSCR. The latter is calculated as the ratio of monthly gross income of the borrower divided by his/her Equal Monthly Instalment (EMI). Technically speaking, the DSCR should be at least equal to 1.0 to reflect a truly solvent borrower. However, in some circumstances, DSCR values can hover below one, reflecting purely co-borrowed loans. To overlook such a feasible bias in the sample, another estimation is effected by focusing exclusively on data points which reflect DSCR values beyond 1.0. The econometric model is specified below.5 DSCR i = β0Tenori + β1MarginCoveri + β2 Agei + β3Cost i + β4 Arrearsi + β5TownDumi + β6 PublicDumi + β7 ConstructionDumi + β8MarginCoveri * TownDumi + β9Cos ti * PublicDumi + β10Ten ori * ConstructionDumi + β11TownDumi * ConstructionDumi + ε i

(3.1)

Where DSCR Tenor MarginCover Age Cost Arrears TownDum PublicDum

Debt Service Coverage Ratio Loan Maturity Security Margin Cover Age of borrower Interest cost of the loan Loan arrears Locational dummy variable for the borrower Public employment dummy variable for the borrower ConstructionDum  Dummy variable where purpose of loan is for construction All the variables are defined in Table 3.A.1 in the appendix section. Distinct measures for DSCR are employed. As a matter of fact, under the continuous metric, DSCR is calculated as the monthly income over EMI with OLS estimation technique being used for the analysis. The ­advantage  Various versions of tenor, age, cost and margin cover, all squared, have also been attempted in the regressions but none is found to entail robust economic significance compared to when their values are employed without squaring them. Ironically, when they are employed in the analysis, the subsequent Variance Inflation Factors become exceedingly high, symptomatic with the presence of multicollinearity in the estimation process. 5

58 

Economics and Finance in Mauritius

of using the continuous measure is that it directly captures information pertaining to the real degree of the borrower’s repayment capacity.6 Three distinct model versions are being considered for the analysis. Model (1) factors in all the values for DSCR while model (2) focuses exclusively on DSCR higher than one. Finally, under the same continuous metric, natural logarithm is employed for the dependent variable, tenor, margin cover, age and cost of the loan to disentangle the elasticity coefficients in model (3). Results are depicted in Table 3.1. Afterwards, recourse is made towards a binary version of DSCR. In that instance, DSCR is given a value of zero for DSCR values which lie between 1 and 15 and one for values moving beyond 15. Such an analysis is highly warranted to allow bankers to demarcate bad borrowers from good borrowers and thus derive more insightful implications as to how the abovementioned factors differ for borrowers imbued with strong repayment capacity (15 ≤ DSCR < 31) relative to those found in the lower range (1 ≤ DSCR < 15). In brief, this exercise is akin to filtering down the forces which impact on the repayment capacity of borrowers. For these analyses, recourse is made towards logit and probit estimation models. Finally, the Ologit model is employed to unleash further enhanced analytical insights. In this case, the values of DSCR are “discreticized” as follows: 1 in case 1 ≤ DSCR< 5, 2 in case 6 ≤ DSCR< 10, 3 in case 11 ≤ DSCR< 15, 4 in case 16 ≤ DSCR< 20, 5 in case 21 ≤ DSCR< 25 and 6 in case 26 ≤ DSCR< 31. The aim for considering the Ologit model is to analyse the forces at distinct discrete values of DSCR and is considered to lie in-between the OLS and probit/logit models. Subsequently, the use of these three modelling approaches is anticipated to trigger a holistic assessment of repayment capacity and thus yield enriching insights for sound policies to be undertaken. The hypothetical relationship between each independent variable to the dependent variable is discussed below.

6  The best analogy that can be taken pertains to the hedging literature whereby binary or ­continuous metrics have been employed.

Table 3.1  Econometric results under OLS approach Model (1)

Tenor Margin Cover Age of borrower Interest cost Arrears on loan Borrower – Town Borrower – Public Loan-Construction MarginCoverTown TownDumConstructionDum Intercept Adj R-squared F(10, 13713) Prob > F Observations Root MSE

Model (2)

Model (3)

OLS: DSCR < 1

OLS: DSCR ≥ 1

OLS-Natural Logarithm

0.2212 (9.76)*** 0.6206 (70.55)*** 0.0094 (2.57)** −0.0661 (−3.85)*** −0.3566 (−2.04)* 0.0799 (0.88) −0.3278 (−3.96)*** −1.7155 (−21.42)*** −0.0158 (−1.24) −0.0406 (−0.32) 2.3839 (6.17)*** 0.4528 1136.45 0.0000 13724 3.3132

0.2263 (9.92)*** 0.6184 (70.22)*** 0.0121 (3.26)*** −0.0621 (−3.61)*** −0.3293 (−1.87)* 0.0879 (0.96) −0.3265 (−3.93)*** −1.7230 (−21.45)*** −0.0165 (−1.30) −0.0438 (−0.34) 2.2642 (5.19)*** 0.4530 570.70 0.0000 13636 3.3126

0.1845 (7.66)*** 0.2488 (60.68)*** 0.0218 (0.98) −0.1557 (−6.84)*** −0.0777 (−3.01)*** −0.1276 (−10.02)*** −0.0492 (3.71)*** −0.4405 (−38.28)*** 0.0278 (17.26)*** 0.0668 (3.56)*** 1.3809 (9.47)*** 0.4089 1152.98 0.0000 13636 0.4908

Note: The table presents the multivariate analysis of the impact of different factors on the DSCR for borrowers who contracted mortgage loans. The table reports point estimates of the coefficients followed by their t-values in parentheses. The definitions and constructions of the variables are in the Appendix. *, ** and *** denotes statistical significance at the 10%, 5% and 1% level, respectively. Model (1) considers values of DSCR even less than one while model (2) is estimated based on DSCR values exceeding one and subsequently all models are based on the same observations. (The rationale is that there were only few observations less than one under DSCR and also that, DSCR should technically be above one for a lender to enter into a transaction with the borrower). Finally, model (3) considers the natural logarithm (Natural logarithms could not be applied to variables having zero values and in these cases, the coefficients collapse into semi-elasticities coefficients for Loan construction dummy, town dummy, public dummy and arrears on loan) of DSCR, tenor, margin cover, age of borrower and interest cost, respectively to compute the elasticity coefficients in the case that DSCR is above one

60 

3.1

Economics and Finance in Mauritius

Tenor (+)

A priori, the longer the maturity period or the tenor of any credit, the lower the amount that the borrower will be required to pay. Thus, longer maturity systematically engenders an enhanced repayment capacity. A positive relationship is thereby anticipated to prevail between the loan maturity and DSCR.

3.2

Margin Cover (+)

The purpose of having recourse towards sufficient margin cover is to ensure that there is no such thing as moral hazard post-granting the loan so that there are very low incentives for the borrower to willingly default. In fact, for any banker who already disburses a loan, the main worry is embedded in the moral hazard risk component – some borrowers who originally pass through the screening process of the credit processing mechanism may eventually and intentionally default once they receive the loan amount. Thus, bankers wield margin cover as a vital tool to deter such defaults from the borrowers. More specifically, bankers take a higher security value which is being encumbered in their favour under security purview, well in excess of the loan amount taken so that the borrower knows that default would be synonymous with losing more than the loan amount taken. A positive relationship is thereby expected to manifest between margin cover and DSCR.

3.3

Age of the Borrower (+)

As the borrower’s age scales up, this boosts his repayment capacity in the form of yearly salary increments, additional qualifications obtained leading to more increments, and also following promotions as the borrower drifts up the rung of the ladder in his/her organization. All these factors induce a positive relationship between DSCR and the age of the borrower.

3  Developing a Credit Risk Model for Mauritian Bankers 

3.4

61

Cost of the Loan (−)

The cost for any loan is principally captured by the interest rate. The higher the interest rate, the larger will be the EMI and, therefore, the lower the DSCR. But it is anticipated that the effect of such a factor is susceptible to be of a lower magnitude than that of the margin cover by virtue of the fact that interest rate resetting at any financial institution is principally effected through steady changes in the main direction of monetary policy as credibility in monetary policy decisions is key ingredient to a strong monetary policy transmission mechanism.

3.5

Loan Arrears (−)

Loan arrears pertain to the inability of the borrower to make good for the periodical payments as planned by the lender. The higher the number of missed planned repayments, the larger will be the value for the arrears and the larger will be the readjusted EMI based on the same loan tenor. In that respect, a negative relationship is thereby expected to prevail between loan arrears and DSCR.  In a nutshell, loan arrears are like missed repayments which are carried forward but based on a preserved loan maturity period so that DSCR is automatically being squeezed unless the lender resorts to some form of loan restructuring.

3.6

 ummy Variables: TownDum(+); PublicDum(−) D & ConstructionDum(−)

A locational dummy is employed to categorize borrowers who reside in the district of Port-Louis and Plaine Wilhems, which represents the city and comprise all the towns present in Mauritius. It is hypothesized that

62 

Economics and Finance in Mauritius

people who live in the city or towns are likely to be better off not only in terms of wealth, but also in terms of greater repayment capacities through highly paid jobs than those who live in villages. It is anticipated that people who work in the public sector are likely to have lower repayment capacity relative to those who work in the private sector because the latter offers the best salary packages in an attempt to attract the most effective, efficient and productive human capital. Finally, a last dummy variable is included to accommodate for loan specificity. Indeed, housing loan has many variants such as loan for construction, loan for renovation and loan for extension as well. Since the focus of the study is geared towards construction loans only, a dummy variable is included and it is anticipated that such a loan type will generate bearish pressure on repayment capacity as the loan amount is often higher.

3.7

Interaction Dummies/Variables

Based on vital interactions, the model can potentially be mis-specified. For example, it can be conjectured that margin cover is higher for borrowers who have their houses located in the town so that the use of both variables can cater for such a state of affairs. In fact, the value of land in Mauritius tends to be particularly high in towns relative to non-town locations so that the use of MarginCover*TownDum is expected to enhance the DSCR of the borrower. In the same vein, since the loan value for construction loans tends to be higher relative to those loans used for non-construction purposes, the former loan type is expected to trigger higher loan maturity levels. Thus, Tenor*ConstructionDum is used as another independent variable in the credit risk model. Because borrowers working in the public sector are most likely to benefit from higher credibility which will eventually translate into lower borrowing costs, it becomes interesting to assess the effect of an interaction dummy variable that includes Cost*PublicDum as one variable. Finally, the last interaction variable represents an interaction dummy which is a composition of TownDum*Construction Dum. Unfortunately, while analysing the correlation coefficients, it becomes evident that Cost*PublicDum and Tenor*ConstructionDum demonstrate considerable correlations

3  Developing a Credit Risk Model for Mauritian Bankers 

63

with PublicDum and ConstructionDum, respectively, making them ineligible for the analysis. Subsequently, only MarginCover*TownDum and TownDum*ConstructionDum variables are being considered for the analysis.

4

Empirical Results

The summary statistics for the dependent and independent variables are shown in Table 3.A.2 in the appendix section. The minimum time period for the housing loan hovers around two years while the highest maturity period stands around 19. As per the correlation coefficients in Table  3.A.3, it transpires that multicollinearity does not constitute an issue of concern. Before running the regression, each variable is being assessed for feasible outlier effects. After cleaning the data for all outliers, the final sample consists of 13,724 borrowers. Afterwards, as focus is laid towards DSCR exceeding the value of one, this leads to a total of 13,636 observations. Within each model analysed, whether OLS or any of the variants of the limited dependent variable model, margin cover is found to systematically engender a positive impact and this holds at the 1% significance level and second-highest economic significance, independent of whether attention is drifted towards the continuous or the binary model. In the case of the elasticity coefficient under the log-linear7 model, margin cover posts strong impacts, i.e., DSCR changes by 0.28% following a 1% change in margin cover. In the case of the OLS model, a 1% change in margin cover triggers around 61% change in DSCR.  For the logit and probit models, the marginal effects are found to be also strong and hover around 0.60% and 0.48%, respectively. These findings buttress the notion that DSCR is positively affected by margin cover. In a parallel manner, under the Ologit model, robust marginal effect is found to prevail and which hovers around 0.23%. Thus, independent of the model under focus, there is widespread consensus among the various models as to margin cover unleashing strong and positive influences on DSCR. Such a finding  These models are also referred to as log-log or double-log models.

7

64 

Economics and Finance in Mauritius

concurs with the need for banks to cling to margin cover as an important weapon to mute down any possible moral hazard effects likely to impregnate any credit granted. Under the OLS approach for model (1) and model (2), a 1% change in the age of the borrower entails approximately a 1% change in DSCR. But, in the case of the log-linear model, age is found to be statistically insignificant. In the same vein, under the logit, probit and Ologit models, the marginal effects are found to register low economic significance. Thus, age variable is not necessarily associated with higher earnings. Under the OLS approach, a 1% change in tenor generates approximately a 22% change in DSCR. The elasticity of tenor with respect to DSCR is also positive and hovers around 0.18%, i.e., DSCR changes by 0.18% following a 1% change in tenor. Under the probit and logit models, a 1% change in tenor engenders a 0.46% and 0.15% change, respectively, in the probability that DSCR lies in the range of 15 to 31. Under the Ologit model, it occurs that a 1% change in tenor occasions a positive impact of 7.74% on DSCR. In general, the loan maturity period unleashes a positive effect on DSCR. This can be accounted for by the fact that the longer the loan maturity, the lower the repayments the borrower needs to make under his/her Equal Monthly Loan Instalment. The cost of loan as captured by the interest rate on the loan exerts downward pressures on the repayment capacity of the borrower. While the impact is around −6% and −16% under the OLS and elasticity models, under the limited dependent variable model, the effect appears to be statistically insignificant under the probit and logit models but not under the Ologit model where the impact hovers around −3%. The statistical insignificance under the probit and logit model implies that interest rate does not matter when DSCR is assessed in terms of binary level. Such a finding bolsters the need to resort towards a continuous DSCR to best disentangle the impact of interest rate on DSCR. Actually, any central bank in the world does not recklessly change its key interest rate so that interest rate changes happen more like a smooth process to safeguard the credibility of the monetary policy mechanism. Consequently, since changes in interest rates happen to be more of a smooth one, its impact is best captured when the dependent variable also happens to be a continuous measurement.

3  Developing a Credit Risk Model for Mauritian Bankers 

65

In spite of the fact that loan arrears variable registers the anticipated negative sign, it is found to be statistically insignificant, irrespective of the model under consideration in Table 3.2, except for the Ologit approach. Consequently, this signifies that arrears do not really negatively affect DSCR. This can reflect the fact that the institution is already sticking to highly sound credit policy schemes to keep at bay arrears problems. In the case of the independent variable which captures as to whether a borrower hails from the public sector, a negative impact is noted, implying that borrowers who work in the public sector are imbued with a rather suppressed repayment capacity. Under OLS, a 1% change for borrowers in the public sector causes around a 32% decline in DSCR with an effect of −0.04% in the case of the elasticity model. As anticipated, the impact is more pronounced under the Ologit model (−13.50%) than under the probit/logit model (−0.46%). This can signify that borrowers from the private sector are susceptible to availing themselves of higher salaries as the private employers vigorously attempt to lure the best workers to h ­ arness maximum profits. Such a finding adds lustre to the notion that the promotional exercise in the public sector is particularly tilted towards the level of seniority achieved relative to the private sector where it is chiefly based on the experience, productivity and intelligence of the staff. The locational factor of the borrower is not found to generate any impact on the repayment capacity of the borrower. Such a finding is highly interesting as it shows that there tends to be now lower difference in town and villages as the latter have been subject to substantial developments in terms of comparable amenities and standard of living to those living in towns. The purpose of the loan variable, in particular, loans which have been contracted mainly for construction in lieu of extensions/renovations of houses, is found to have rejoice over both strong economic and statistical significance. Under the probit and logit models, the marginal effects hover around 3.65%. Such a finding occurs on the back of larger funds being granted in the case of loans meant for construction purposes so that they occasion substantial pressures on the repayment capacity of the borrower. Finally, none of the interaction variable/dummy is found to be statistically significant in Table 3.1 for model (1) and model (2) so that they are being altogether ignored in Table 3.2.

66 

Economics and Finance in Mauritius

Table 3.2  Econometric results for limited dependent models

Tenor Margin Cover Age of borrower Interest cost Arrears on loan Borrower – Town Borrower – Public Loan-Construction Intercept Pseudo R2 Wald chi2(8) Prob > chi2 Observations Log likelihood

Model (4)

Model (5)

Model (6)

Probita

Logit

Ologit

0.0046 (5.58)*** 0.0060 (21.91)*** 0.0004 (3.51)*** −0.0002 (−0.44) −0.0064 (−1.01) 0.0021 (0.5) −0.0046 (−2.08)** −0.0365 (−8.66)*** −0.1635 (−11.38)*** 0.3147 743.94 0.0000 13636 −1899.50

0.0015 (6.54)*** 0.0048 (24.27) *** 0.0004 (3.56)*** −5.9e-05 (−0.11) −0.0046 (−0.79) 0.0046 (0.80) −0.0046 (−1.60) −0.0365 (−8.17)*** −0.1362 (−11.07)*** 0.3073 834.73 0.0000 13636 −1920.01

0.0774 (6.33)*** 0.2292 (40.64)*** 0.0067 (3.40)*** −0.0268 (−3.00)*** −0.1813 (−1.82)* 0.0101 (0.32) −0.1350 (−3.07) *** −1.2990 (−30.20)*** Na 0.2251 3058.63 0.0000 13636 −10767.455

Note: The table presents the multivariate analysis of the impact of different factors on the DSCR for borrowers who contracted housing loans. The table reports the derived marginal effects followed by their z-values in parentheses. The definitions and constructions of the variables are in the Appendix. *, ** and *** denotes statistical significance at the 10%, 5% and 1% level, respectively. Model (4) resorts a probit model whereby DSCR between 1 and 15 is censored to zero and DSCR between 15 and 31 is censored to 1. Model (5) used the same binary method as per model (4) but now under logit estimation approach. Finally, model (6) resorts to a “discretisized” limited dependent model à la Ologit philosophy as follows: 1 in case 1 in case 1 ≤ DSCR< 5, 2 in case 6 ≤ DSCR< 10, 3 in case 11 ≤ DSCR< 15, 4 in case 16 ≤ DSCR< 20, 5 in case 21 ≤ DSCR< 25 and 6 in case 26 ≤ DSCR< 31. At first glance, regression coefficients for Ologit model tends to lie in-between those of the OLS and the probit/logit model a Since the coefficients under the probit, logit and Ologit models are not directly interpretable as under OLS approach, their marginal effects have been computed (See Brooks 2008) and reported in the table. Under the logit and Ologit model, recourse is made towards the cumulative logistic distribution and in case of probit, the cumulative normal distribution is used

3  Developing a Credit Risk Model for Mauritian Bankers 

5

67

Conclusion

Conventional and recent developments in credit risk models have principally been focused towards modelling the probability of default. This chapter contributes to this literature by proposing a credit risk model which directly suits the needs of the lenders via a direct focus on all the forces susceptible to impound on the repayment capacity of a borrower and thereby stimulates a higher degree of proactiveness in credit risk management. The importance of margin cover as a propelling mechanism for higher repayment capacity implies that it effectively acts as a moral hazard mitigation mechanism so that any credit policy should include it as an important credit risk management tool. Loan maturity also drifts up repayment capacity. Based on the detrimental impacts of the US subprime crisis on the exporting arm of Mauritius, it appears considerate for the lenders/bankers in Mauritius to undertake loan restructurings via loan tenor readjustment to maintain or enshrine the repayment capacity of financially bruised borrowers. The fact that borrowers from the public sector are imbued with poor repayment profiles signifies that the credit risk manager should diversify his portfolio via inclusion of borrowers from the private sector. The negative relationship between the interest cost and DSCR justifies the bearish monetary policy stance clung by the Bank of Mauritius following the US subprime crisis in view of mitigating financial causalities. But, in the case that the upward interest rates adjustment manifests at a faster rate than the income growth of the borrowers, then, this will reduce the effectiveness of any policy envisaged to put the economy back on normal growth path. The low economic significance for age signifies that borrowers do not often get fresh increments on a regular frequency. Or it can best signify that any yearly increment based on experience or increments pertaining to the acquisition of new certificates is not substantial enough as to considerably leverage on their income. Because housing loans channelled towards construction exert the highest negative impact on DSCR, this signifies that bankers need to be very careful when they are dealing with such types of loans. This situation explains why, in many countries,

68 

Economics and Finance in Mauritius

­ ousing loans for construction are basically granted over longer time h periods, stretching beyond one decade in most cases. Finally, no major evidence is found as to arrears and town dummy variables impacting on the repayment capacity of the borrowers, suggesting that the institution already adhered to sound arrears management. As change is only constant feature in life, the current credit risk model can be further enhanced in the future.

6

Policy Recommendations

Based on the strong effect of margin cover on the repayment capacity of borrowers, it is recommended that a general system of asset valuation should be established in Mauritius. The assets can be rated along with regular valuations. Besides, compared with advanced economies where data are easily available for research, in Mauritius, there is still a lack of connection between the academic world and the industry. Authorities should work on setting up a generate credit bureau with an online database on credit matters. Researchers would then use then to be able to propose policies to the government, which would be all to the benefit of Mauritius in terms of higher output. Confidentiality issues do not manifest once the data are anonymized. Such a database would definitely leverage on not only the quality of research but also in terms of best policies to be envisaged. Acknowledgement  I would like to thank Mr Naim Maudabokus for the data and comments from participants at the First International Conference on Credit Analysis and Risk Management, held in Rochester, Michigan, USA in July 2011.

3  Developing a Credit Risk Model for Mauritian Bankers 

69

Appendix Table 3.A.1  List of variables Variable

Definition

DSCR

DSCR defined as monthly income of the borrower divided over his/her Equal Monthly Loan Instalment Interest Cost of the loan (Security Value – Loan Amount)/Loan Amount 1 in case borrower works in the public sector, 0 otherwise 1 in case borrower resides in a town, 0 otherwise 1 in case the purpose of the loan is for construction, 0 otherwise Outstanding loan amount/Original loan amount Age of the borrower Arrears/Oustanding loan amount

Cost Margin Cover PublicDum TownDum ConstructionDum Outstanding Loan Age of borrower Arrears

Table 3.A.2  Summary statistics Variable Tenor Margin Cover Age of borrower Interest cost of loan Arrears on loan Borrower-Town Borrower-Public Loan-Construction DSCR

Mean 7.8527 3.6980 41.1209 9.9577 0.0724 0.3919 0.1374 0.3214 5.5506

St. dev. 1.8370 4.5968 8.4288 2.2090 0.1655 0.4882 0.3443 0.4670 4.4774

Min

Max

1.76 0.02 20.28 3.0 0 0 0 0 1.0064

19.11 54.64 86.82 14.0 3.9585 1 1 1 30.9622

Note: The table presents summary statistics for the sample used in the multivariate analysis. Constructions and definitions of the variables are spelt out in the Appendix. The first column shows the mean values while columns (2), (3) and (4) depicts the standard deviation, the minimum value and the maximum value for each variable

DSCR Tenor Margin Cover Age of Borrower Cost of loan Arrears on loan Borrower-Town Borrower-Public Loan-Construction

1

DSCR 0.1058 1

Tenor 0.644 −0.2818 1

Margin Cover 0.1189 −0.3749 0.1933 1

Age of borrower −0.0188 −0.6609 0.1462 0.1462 1

Cost of loan

Table 3.A.3  Correlation matrix of independent variables used in the model

−0.0071 −0.1812 0.0446 0.1147 0.1369 1

Arrears on loan 0.0238 −0.0539 0.0361 0.0805 0.0419 0.0093 1

Borrower-­ Town

−0.0173 −0.0433 0.0165 0.0539 0.0631 −0.075 −0.025 1

BorrowerPublic

−0.3117 0.208 −0.2421 −0.0931 −0.1014 −0.0368 −0.0262 −0.0084 1

LoanConstruction

3  Developing a Credit Risk Model for Mauritian Bankers 

71

Bibliography Akerlof, G. (1970). The market for “Lemons”: Quality uncertainty and the market mechanism. Quarterly Journal of Economics, 89, 488–500. Black, F., & Cox, J.  C. (1976). Valuing corporate securities: Some effects of bond indenture provisions. Journal of Finance, 31(2), 351–367. Black, F., & Scholes, M. (1973). The pricing of options and corporate liabilities. The Journal of Political Economy, 81(3), 637–654. Brooks, C. (2008). Introductory econometrics for finance (2nd ed.). Cambridge: Cambridge University Press. Duffie, D., & Singleton, K. J. (1999). Modeling term structures of defaultable bonds. Review of Financial Studies, 12(4), 687–720. Geske, R. (1977). The valuation of corporate liabilities as compound options. Journal of Financial and Quantitative Analysis, 12, 541–552. Geske, R. (1979). The valuation of compound options. Journal of Financial Economics, 7, 63–81. Jarrow, R., & Turnbull, S. (1992). Credit risk: Drawing the analogy. Risk Magazine, 5(9). Leland, H., & Toft, K.  B. (1996). Optimal capital structure, endogenous bankruptcy, and the term structure of credit spreads. Journal of Finance, 51(3), 987–1019. Longstaff, F. A., & Schwartz, E. S. (1995). A simple approach to valuing risky fixed and floating rate debt. Journal of Finance, 50(3), 789–819. Merton, R.  C. (1973). Theory of rational option pricing. Bell Journal of Economics and Management, 4, 141–183. Merton, R. C. (1974). On the pricing of corporate debt: The risk structure of interest rates. Journal of Finance, 29(2), 449–470. Nielsen, L. T., Saá-Requejo, J., & Santa-Clara, P. (1993). Default risk and interest rate risk: The term structure of default spreads. Working Paper, INSEAD. Ramlall, I. (2011). Developing a Practical Credit Risk Model for Bankers in the Case of Mortgage Loans Portfolio in Mauritius. Chapter 15: First International Conference on Credit Analysis and Risk Management, Cambridge Scholars Publishing, Chapter 15. Vasicek, O. (1977). An equilibrium characterization of the term structure. Journal of Financial Economics, 5, 177–188.

Part III The Stock Market in Mauritius

4 Stock Market Analysis

1

Introduction

Regulated by the Financial Services Commission, the Stock Exchange of Mauritius (SEM) constitutes one of the key components of the various types of institutions which comprise the structure of the Mauritian financial system. The SEM started to operate in 1989 as a private limited company before being converted into a public company in 2008. SEM operates under two major arms – namely, the official market and the Development and Enterprise Market (DEM). The former is usually considered to be imbued with greater liquidity and stronger levels of transactions relative than the DEM, which was launched in 2006 and chiefly deals with small and medium sized enterprises. For instance, foreign investors have greater taste for stocks listed on the official market relative to the DEM market, with bank and hotel stocks being the most coveted by foreigners. Foreigners avail of certain important benefits when they invest in SEM, such as not having to pay withholding tax on dividends and no tax being imposed on capital gains. To buy or sell shares, investors need to contact brokers or investment dealers (see Table 4.A.1 in the Appendix section). The list of issuers on the official market is also provided in Table 4.A.2 in the appendix to this chapter. © The Author(s) 2017 I. Ramlall, Economics and Finance in Mauritius, DOI 10.1007/978-3-319-39435-0_4

75

76 

Economics and Finance in Mauritius

Like any other stock market, the SEM fulfils some important functions, including: (a) Acting as an economic barometer to reflect the general economic conditions of the Mauritian economy, although this does not seem to really work for SEM by virtue of the poor participation rate of Mauritians propelled by the lack of financial literacy, let alone weak coverage of firms in terms of a rather unchanged number of locally listed firms on the official market over the past decade. (b) Providing pricing and valuation of securities to thereby stimulate trading in stocks. (c) Ensuring soundness of the listed companies via stringent listing rules and financial reporting. (d) Enshrining diversification opportunities to both local and foreign investors (international portfolio diversification). (e) Assisting in personal and corporate wealth management in the case of individual and corporate investors, respectively. (f ) Raising of equity financing by companies to help them realise their investment projects thereby scaling down any feasibly underinvestment projects chiefly triggered by lack of bank financing. (g) Accommodating for a diverse set of traders in the market such as hedgers, speculators and arbitrageurs, all working together to ensure sound liquidity to prevail in the market. Below are some of the distinctive features to note about SEM: • SEM became the first exchange in Africa to trade equity products in US dollars in 2011. • SEM constitutes a full-fledged member of the World Federation of Exchanges. • SEM won, for the third time in five years, the “Most Innovative African Stock Exchange of the year Award” organized by Africa investor in 2015. • Since 2010, SEM has provided real-time data to global vendors, including Thompson Reuters, Financial Times and Bloomberg.

4  Stock Market Analysis 

77

• In 2013, SEM began to trade in exchange traded funds-securities which track indices. • To boost market activities and induce liquidity, SEM slashed its brokerage fees by a hefty amount of 88% from 1.25% to 0.15% applicable on turnaround trades – second level transaction on securities prior to the settlement of the first transaction. • In 2015, SEM launched the SEMSI, geared to track the price-­ performance of listed firms (both official and DEM markets) which show vehement commitments to sustainability practices in terms of environmental, social and governance criteria. In spite of the above developments, it is important to note that equity financing is not widely desired by listed companies. Ironically, the number of locally listed companies on the official market has not changed drastically over the years, highlighting the extent to which the democratization of equity financing still constitutes a hard nut to crack for a small economy. Most importantly, the composition of the official market does not reflect the general economic health of Mauritian firms. More listings of key firms would definitely leverage on the quality of SEM in acting as an economic barometer of the Mauritian economy. The next section discusses each of the distinct aspects of the official market in SEM.

2

 ectoral Market Capitalisation S on the Official Market

As depicted in Fig. 4.1, it is clear that the banks, insurance and other finance sector companies account for the lion’s share of the total market capitalization in SEM. Thus, if banks and insurance companies post remarkable performances, this will automatically generate a commendable performance of the overall index of SEM, namely, SEMDEX. The next largest sector emanates from investment companies. Based on the fact that investment companies derive their returns from the returns of other companies, such a sector will act as a drag on SEMDEX performance during a global recession, sparking substantial risk aversion and inducing investors to retreat into cash.

Industry 6.57%

Foreign 0.06%

Commerce 6.78%

Banks, Insurance & Other Finance 43.37%

Transport 0.56%

Fig. 4.1  Market capitalization on official market as at end December 2015 (Source: Author’s illustration based on data gleaned from the Stock Exchange of Mauritius Factbook 2016)

Investments 28.20%

Leisure & Hotels 11.35%

Property & Development 0.78%

Sugar 2.33%

78  Economics and Finance in Mauritius

4  Stock Market Analysis 

79

The third-largest sector is the leisure and hotels sector, whose performance is highly dependent on the performance of the tourism sector in Mauritius. Overall, these three sectors account for around 83% of the total market capitalization as at end December 2015. Such a state of affairs implies that the performance of SEMDEX is, to some extent, skewed towards certain key sectors of the Mauritian economy such as banks, insurance companies, hotels and investment firms. Such a state of affairs is highly interesting as it implies that even if SEMDEX posts astounding performances, this may not reflect the general trend of the economy as certain key sectors are being conspicuously overlooked. These include SMEs and manufacturing firms, which not only make up the largest chunk of total firms composition in Mauritius but also represent a large absorption component of employment. In that respect, it is anticipated that SEMDEX will never reflect the true economic barometer of the Mauritian economy because its firms’ composition does not reflect the general trend of what most firms experience in Mauritius. Nonetheless, based on the fact that Mauritius is imbued with a bank-based financial system, and dominated by two banks, both of which are listed on SEM, some degree of connection can be expected to manifest between the real and the artificial economy. However, the hitch to this is that these two banks avail themselves of substantial power such as exorbitant interest rate spreads as to dampen the connecting link between the real and the artificial economy. For instance, Ramlall (2015) found that the interest rate spread hovered around 7%1 with the banking system’s profitability being immune to the US subprime crisis.

3

Bank Financing Versus Equity Financing

Figure 4.2 creates a demarcation line between bank financing and equity financing for Mauritius. In this model equity financing is computed as the value of rights issues over GDP while bank financing is calculated as  Data found on the World Bank website showed an interest rate spread of 1.4% for the year 2014 for Mauritius. Yet such a figure does not reflect the true interest rate spread for Mauritius. Indeed, if the mid-values of savings and loan interest rates are used for various sectors from the recent monthly statistical bulletin issued by the Bank of Mauritius, the interest rate spreads for the various sectors are found to be substantially much higher than 1.4%. 1

80 

Economics and Finance in Mauritius

6.00%

180% 148.35%

5.00%

144.27%

154.06%

156.67% 151.22%

158.67% 146.11%

140%

124.85% 4.00% 3.00%

120%

130.34%

100%

109.36% 4.71%

2.00%

5.07%

4.68%

80% 60% 40%

1.00% 0.00%

160%

0.05%

0.00%

0.04%

0.00%

0.59%

0.00%

2006

2007

2008

2009

2010

2011

20%

1.33%

0% 2012

2013

2014

2015

YEAR Equity financing

Bank financing

Fig. 4.2  Bank financing versus equity financing (Source: Author’s illustration based on data gleaned from the Stock Exchange of Mauritius Factbook 2016 and Bank of Mauritius Statistical Bulletins, ­various issues)

the loans and advances2 of the banking sector over GDP. In 2015, bank financing accounted for around 146% of GDP while equity financing hovered around 4.7%. Such a finding clearly corroborates the fact that Mauritius is endowed with a bank-based financial system. In that respect, it is of paramount significance that the Bank of Mauritius adheres to robust regulatory measures along with proactive risk management policies to consolidate on the stability of the banking sector in Mauritius. For the years 2006–2011, equity financing constituted roughly less than 1% of GDP. This signifies that, though listed, many firms shun equity financing to avoiding distributing returns of lucrative investments to outsiders. Above all, even listed firms in Mauritius are endowed with high levels of inside ownership, that is, where managers also happen to be major shareholders of the company. Though increases in equity financing are noted for the years 2013 onwards, its funding capacity still remains very weak. Many reasons can be advanced for the low levels of equity funding in Mauritius. First, many firms, though listed, are still controlled by major  Data were gleaned from various issues of the monthly statistical bulletins issued by the Bank of Mauritius. 2

4  Stock Market Analysis 

81

shareholders, who do not want to distribute profits to outsiders. Second, it could imply that the listed firms already exploit their considerable market power so that they are less likely to engage in new products to gain further market share. Third, research and development is practically nonexistent in Mauritius so much so that firms are less likely to be in need of major funds for investments. Fourth, being large and listed, these firms can employ considerable negotiating power in bank loans so that the cost of bank loans becomes cheaper than equity financing. After all, under the pecking order theory, equity is used only as a last resort. One eye-catching point to note in Fig.  4.2 is that bank financing underwent a decline on the back of the crisis. The same trend seems to repeat itself for the period 2014–2015. It can thus be conjectured that the banking sector in Mauritius had been affected by the crisis and is still subject to some stressful conditions, which are likely to undermine the earnings potential of the Mauritian banking sector. Thus, looming risks still permeate over the economic and financial horizons. Another analysis is undertaken to compare the banking sector with the equity sector on the official market by comparing the total assets of the banking sector with the levels of market capitalization. As depicted in Fig. 4.3, it is evident that the assets of the banking sector are much greater than the total value of market capitalization. Interestingly, as at the end of 2015, the banking sector assets represent around 600 times the total market capitalization on SEM. Such a finding plainly shows the significance of the banking sector in Mauritius. In 2007, the market capitalization over GDP posted its highest value, hovering around 70%. Such an astounding performance of SEM ironically acted as a safety valve in limiting the detrimental impacts of the crisis which acted as a drag on its performance in 2008. It is important to bear in mind, even if market capitalization over GDP were to exceed 100%, such an event should be considered with caution. First, by virtue of the fact that only a few local firms are listed in the official market, this implies that the sound performances of these firms may not necessarily be generalised for the whole Mauritian economy. On the other hand, increases in the total assets of the banking sector entail far-reaching impacts on the general health of the Mauritian economy. Second, market capitalization is computed as the number of shares

82 

Economics and Finance in Mauritius

315%

8.00

280% 0%

7.00

245% 5%

6.00

210% 0%

5.00

175% 5%

4.00

0% 140%

3.00

105% 5%

2.00

70% 0%

Total assets of banking sector/GDP

Market Capitalisation (O)/GDP

2015

2014

2013

2012

2011

2010

2009

0.00

2008

0% %

2007

1.00 2006

35% 5%

Total assets of banking sector/MarketCapitalisation (O)

Fig. 4.3  Banking sector assets versus market capitalisation (Source: Author’s illustration based on data gleaned from the Stock Exchange of Mauritius Factbook 2016 and Bank of Mauritius Statistical Bulletins, ­various issues)

o­ utstanding multiplied by the share price. Consequently, it can happen that increases in the share prices are driven more by purchasing stamina in lieu of fundamentals so that the increases in market capitalization are more like fictitious events.

4

Price-Earnings Analysis in SEM

The price–earnings (PE) ratio is widely used by investors worldwide to gauge the level of confidence imbued in a given stock. Figure 4.4 captures the PE ratio for the official market for the period spanning from 1998 to 2015. In essence, it is possible to identify five distinct periods or trends: a negative trend for the period 1998–2002; a positive trend for the period 2002–2007; a downward trend for the period 2007–2008; a reversal in trend for the period 2008–2013; and, finally, a relapse into negative territory for the period 2013–2015, albeit with a hike noted in the last year. Technically speaking, the PE ratio does not reflect a pure metric of assessment by virtue of the fact that the denominator and the numerator happen to be measuring elements

83

4  Stock Market Analysis  16 14 11.95

12 11.56 10 8 6 4

14.16

14.05

13.26

11.83

9.93

10.74

8.98 7.43 6.4

5.91

11.29 11.3 9.92

7.98 6.17

5.33

2 0

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

YEAR

Fig. 4.4  Price-earnings ratio (Source: Author’s illustration based on data gleaned from the Stock Exchange of Mauritius Factbook 2016)

which occur at distinct time horizons. For instance, price reflects the level of confidence showed by investors so that it is more of a forward-looking assessment of the state of a company. On the other hand, earnings pertain to last quarterly reports so that while price acts as a lead indicator, earnings act as a lag indicator. Thus, the simultaneous use of both a lead and a lag indicator may not convey full information. Nonetheless, suffice it to say that PE can still be used as an indicative tool for making a confidence assessment in any stock market. As per Fig. 4.4, it is clear that the PE ratio underwent the most drastic decline in 2008, when it lost almost half of its value. As a matter of fact, in light of the crisis, two factors deter foreigners from investing in a foreign stock market – namely, the heightened level of risk aversion (usually captured by VIX – a volatility index) and undermined international portfolio diversification, all of which induced foreign investors to retreat to cash. Consequently, following a combination of maintained EPS based on past report and declined share prices, an abrupt fall in PE manifested.

84 

Economics and Finance in Mauritius

5

Foreign Investment Analysis in SEM

Foreign investors represent part of the market players who buy and sell shares on the official market of SEM. Figure 4.5 clearly demonstrates the steady preference for local stocks by foreigners up to 2007 when things veered off-course. Indeed, the up/down bars reflect the gradual taste for local stocks until 2007 when net foreign investments began to post sustained disinvestments with 2015 registering the largest amount of disinvestments. It is widely known that foreigners on the official market concentrate mostly on bank and hotel stocks. Such disinvestments reflect periods of uncertainty still manifesting at the global economy level chiefly when China, the world’s second-largest economy, experienced dwindling exports. If such a trend persists, this would be symptomatic with a state of trading activity being concentrated towards domestic investors so that foreign market penetration becomes low. In case that happens, more foreign disinvestments are expected to manifest and banks and hotels shares are likely to bear the brunt of such disinvestments since they are so favoured by foreign investors.

10,000 8,000 6,000 4,000

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

-2,000

2000

0

1999

2,000 1998

MILLIONS OF RUPEES

12,000

-4,000 -6,000

YEAR Foreign purchases (O) (in millions Rs)

Foreign sales (O) (in millions Rs)

Net foreign purchases (O) (in millions Rs)

Fig. 4.5  Foreign investments (official market) on the Stock Exchange of Mauritius (Source: Author’s illustration based on data gleaned from the Stock Exchange of Mauritius Factbook 2016)

85

4  Stock Market Analysis 

6

Turnover and Volume Assessment

Millions

As illustrated in Fig.  4.6, it is clear that the annual turnover on SEM began to experience real bullish forces from 2005 onwards. Years 2007, 2011 and 2015 represented important peaks in the level turnover of generated on SEM. In the case of volume, no major changes are noted for the period between 1998 and 2012. For the years 2013, 2014 and 2015, however, substantial hikes in volume are noted. Such a state of affairs clearly indicates higher volume exchanges in the last three years. Probing deeper, it transpires that the considerable increases in volume are due to SBM Holdings Limited (the bank investment holding company of SBM Group Limited), which accounted for more than 80% of the total volume exchanged on SEM for the years 2013, 2014 and 2015, based on the low denomination of share value which thereby requires larger volume levels.

Millions

20,000

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

18,000 16,000 14,000 12,000 10,000 8,000 6,000 4,000

Annual Turnover (Rs)

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

0

1998

2,000

Annual Traded Volume

Fig. 4.6  Turnover and volume analysis on official market (Source: Author’s illustration based on data gleaned from the Stock Exchange of Mauritius Factbook 2016)

86 

Economics and Finance in Mauritius

7

 erformance of Indices on the Official P Market

The performances of various known indices are depicted in Fig.  4.7. SEMDEX, an index based on the overall performance of the official market, registered a sustained period of positive returns till 2007 after which it experienced a marked decline in 2008. Thereafter, SEMDEX posted an upward momentum in its performance till 2014 to then relapse into territory of negative returns. Similar state of performance is noted for SEMTRI which consists of both capital gains/losses and gross dividends on the stocks. In general, since 1998 till 2015, the major indices of SEM registered maintained upward performances in spite of some bouts of bearish effects noted for some years such as 2008 and 2015.

8,000

2,500

7,000 2,000

6,000 5,000

1,500

4,000 1,000

3,000 2,000

500

1,000 0

0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Year SEMDEX (End of Period)

SEM-10 (End of Period)

SEMTRI

Fig. 4.7  Performance of indices on the official market (Source: Author’s illustration based on data gleaned from the Stock Exchange of Mauritius Factbook 2016)

4  Stock Market Analysis 

8

87

 redicting the Performance of Stocks P Listed in the Stock Exchange of Mauritius

For many years some specific stocks on the Stock Exchange of Mauritius, widely coveted by both local and foreign investors, were considered to be blue chips. The stocks involved were Sun Limited (SUN), New Mauritius Hotels Ltd (NMH), Lux Island Resorts Ltd (NRL), MCB Group Limited (MCB), SBM Holdings Ltd (SBM), ROGERS and IBL. An analysis is thereby undertaken to gauge the value of one rupee investment in these stocks as at the first trading day in January 2006 to July 2016 – an almost decennial analysis. More specifically, all initial stock prices are normalized to be of one rupee at the start of the period of analysis. Thereafter, any changes in the stock prices are incorporated into the original normalized stock values. As depicted in Fig. 4.8, before the year 2010, strong co-movements or coupling effects were manifested among the abovementioned stocks. This could be explained by the fact that, prior to the onset of the real effects of the crisis onto the Mauritian economy, investors clung to these stocks because of their high returns, strong liquidity levels and robust market capitalization features. However, such state of affairs veered off-course from 2010 onwards. For instance, there was a strong underperformance in hotel stocks, which are SUN, NMH and NRL. This implies that one rupee invested in January 2016 generated a value well below one rupee, as glaringly captured by their values, which hover well below the vertical dotted line in Fig. 4.8. To sieve out the new perspective in the evolution of prices, an orange-coloured vertical rectangle line is drawn. It is clear that MCB, IBL and ROGERS stocks posted positive performances. Intriguingly, SBM, despite being a bank and holding a relatively large part of market share, registered poor performance on the back of its restructuring activities. Interestingly, it transpires that the evolution of the normalized values for the various stocks, in particular, MCB, tend to emulate the distinct wave patterns which prevail under Elliott Wave Theory, as shown in Fig.  4.9. Hence, it can be concluded that SEM’s evolution of prices can be forecasted by recourse to proper wave counting. Based on these findings, it is expected that a slight downturn can be anticipated, compatible with the last wave which manifests under the corrective phase of the Elliott Wave analysis. This analysis can be highly lucrative for long-term investors in the SEM.

Fig. 4.8  Analysing conventional blue chips stocks in SEM under normalised values (Source: Author’s illustration)

88  Economics and Finance in Mauritius

2

Impulsive phase

4

Fig. 4.9  Transposing Figure 4.8 into an Elliott Wave analysis (Source: Author’s illustration)

1

3

5

Corrective phase

a

b

c

4  Stock Market Analysis 

89

90 

9

Economics and Finance in Mauritius

Challenges Present to SEM

(a) As at date, although many changes had been realized by SEM, yet there are still only a very small number of locally listed firms on the official market. The changes made are more like horizontal improvements while vertical improvements akin to the real democratization of equity financing still constitute a hard nut to crack. Some local companies which are doing fairly well are not listed in SEM.  The government could come up with a policy of a lower tax rate such as 5% in lieu of the 15% being applied to fast-growing companies who get listed on SEM. (b) Vigorous campaigns should be initiated to induce individuals to invest. There should be an increased investment in financial education. The high interest rate spread charged by local banks is to the result of easy deposit funding on the back of Mauritians having a strong appetite for bank deposits. To break such eased access to deposits, people should be induced to hold equities. For instance, the government might introduce a policy of an income tax of 10% in lieu of 15% for all individuals who invest at least 10% of their monthly income in SEM. Apart from equities, individuals can also invest in other products such as exchange traded funds. Another feasible strategy would be to put in place an investment allowance for individuals who invest in local equities, for, say, a period spanning over more than three years. (c) Stock ownership is yet another problem in SEM. As a matter of fact, in many companies, major shareholders who own more than 50% to have a say in the management of the company. The move from block ownership to more individual ownership would be another development in promoting the democratization equity investments in Mauritius. This will be highly helpful in cushioning any feasible collusive behaviour in view of inciting prices to fictitiously enhance portfolio values. (d) SEM should open up more to start-up companies, which are imbued with innovative ideas but are all too often starved of the funds to turn them into viable projects. This could be done via quarterly calls for projects by fresh university graduates.

4  Stock Market Analysis 

91

Appendix  Simplified Equity Risk Model A of the Mauritian Economy The aim of this section is to furnish a simplified equity risk analysis of the Mauritian economy as depicted in Figure 4.A.1. In essence, such an analysis is vital to gauge the direct and indirect effects of a major negative global stock market shock such as that experienced during the US subprime crisis of 2007. In spite of the fact that the direct impacts of such a crisis took time to manifest on the real side of the Mauritius economy, yet, the transmission of any shock from global stock markets onto the local stock market should be quicker based on strong coupling effects among international stock markets. Household

Non-listed corporates with no equity investments

Banking

Stock market shock

Insurance Listed corporates

Non-listed corporates with strong equity investments

Global stock markets shock transmitting to SEM

Fig. 4.A.1  Diagrammatical view of the interactions among various sectors (Source: Author’s Illustration)

92 

Economics and Finance in Mauritius

Components Used for Analysis • Four sectors under scrutiny: Household, Corporate, Banking and Insurance • Non-financial sectors: Corporates and Households • Financial sectors: Banking and Insurance • Main objective of the analysis is to sieve out the transmission channels of a shock from the stock market onto the real economy. Such an analysis is vital to unleash insights in terms of how a possible stock market shock can transmit into real effects in the economy. • First round effect pertains to the direct impact of a shock. • Second round effect assesses the just after direct impact of a shock. • The extent of damage borne by the real sector from a shock in the stock market (financial market) will depend heavily on the network features of the financial system, i.e., the degree to which the four sectors are interlinked. The width of the connecting edges captures the extent of linkages among the various units under investigation. Based on the strong impact of global stock markets on the local equity market, a negative shock

Table 4.A.1  List of ATS (Automated Trading System) operators (January 2016)

Investment dealers Associated Brokers Ltd Capital Market Brokers Ltd AXYS Stockbroking Ltd Brammer Capital Brokers Ltd Swan Securities Ltd MCB Stockbrokers Ltd Ramet & Associes Ltée SBM Securities Ltd LCF Securities Ltd Prime Securities IPRO Stockbroking Ltd Source: Stock Exchange of Mauritius, http://www. stockexchangeofmauritius.com/find-a-broker

4  Stock Market Analysis 

93

Table 4.A.2  List of issuers on official market Symbol MCBG.N0000 SBMH.N0000 BBCL.N0000 MUA.N0000 SWAN.N0000 MEI.N0000 CIM.N0000 CMPL.N0000 GIDC.N0000 HML.N0000 HWF.N0000 IBL.N0000 SHEL.N0000 GCL.N0000 GOLI.N0000 MBL.N0000 MCFI.N0000 MOR.N0000 PIM.N0000 UBP.N0000 ALT.N0000 BMHL.N0000 CIEL.N0000 CAUD.N0000 FINC.N0000 SAVA.N0000 SAVA.P0001 MDIT.N0000 NITL.N0000 PAD.N0000 POL.N0000 ROGE.N0000 TERA.N0000 UTDL.N0000 ASL.N0000 NMHL.N0000 NMHL.P0000 LOTO.N0000

Company name Banks & insurance and other finance MCB Group Limited SBM Holdings Ltd Bramer Banking Corporation Ltd Mauritius Union Assurance Co. Ltd Swan General Ltd (formerly Swan Insurance Company Ltd) Mauritian Eagle Insurance Co. Ltd Cim Financial Services Ltd Commerce Compagnie des Magasins Populaires Ltée ENL Commercial Limited Harel Mallac Ltd Innodis Ltd Ireland Blyth Ltd Vivo Energy Mauritius Limited Industry Gamma Civic Ltd Go Life International Ltd (formerly Go Life International PCC) Phoenix Beverages Ltd Mauritius Chemical & Fertilizer Industry Ltd Mauritius Oil Refineries Ltd Plastic Industry (Mtius) Ltd The United Basalt Products Ltd Investments Alteo Limited Belle Mare Holding Ltd CIEL Limited Caudan Development Ltd Fincorp Investment Ltd ENL Land Ltd ENL Land Ltd (Preference Shares) The Mauritius Development Investment Trust Co. Ltd National Investment Trust Ltd Promotion and Development Ltd P. O. L. I. C. Y Ltd Rogers & Co Ltd Terra Mauricia Ltd United Docks Ltd Leisure & hotels Automatic Systems Ltd New Mauritius Hotels Ltd New Mauritius Hotels Ltd (Preference Shares) Lottotech Ltd (continued)

94 

Economics and Finance in Mauritius

Table 4.A.2  (continued) Symbol

Company name

NRL.N0000 SUN.N0000

Lux Island Resorts Ltd Sun Limited Property development BlueLife Limited Sugar Omnicane Ltd Transport Air Mauritius Ltd Debt Lux Island Resorts Ltd F/P Convertible Bonds MCB Group Limited (Floating Rate Subordinated Notes) The United Basalt Products Ltd (Unsecured Floating Rate Bonds) SBM Holdings Ltd (Class A 1 Series Floating Rate Senior Unsecured Bonds) SBM Holdings Ltd (Class B 1 Series Floating Rate Senior Unsecured Bonds) Foreign Dale Capital Group Limited Exchange traded funds NewGold Issuer Limited (RF) – Gold Bullion Debentures NewGold Issuer Limited (RF) – Platinum Debentures S&P GIVI South Africa Top 50 Index ETF Portfolio Structured products Absa Bank Limited (Series ASN.D0001 Credit Linked Notes) Absa Bank Limited (Series ASN.D0002 Credit Linked Notes) Global and specialised funds IPRO Growth Fund Ltd Global Investment Opportunities Fund (Class A, Class C, ClassI) Kotak India Equity Fund – Series 1 (A cell of Kotak Investment Opportunity Fund Ltd) Kotak India Equity Fund – Series 2 (A cell of Kotak Investment Opportunity Fund Ltd) ACM India Focus Fund Ltd Global Diversified Cell Number One (A cell of Global Diversified Fund PCC) Africa Sustainability Fund Novare Africa Property Fund One (A cell of Novare Africa Fund PCC) ACM Aussie Ltd ACM European Ltd Triangle Real Estate India Fund LLC (Class A)

BLL.N0000 MTMD.N0000 AIRM.N0000 NRL.D0701 MCBG.D2023 UBP.D0018 SBMH.D2024 SBMH.D2021

DCPL.N0000 NGLD.N0001 NPLT.N0001 NERA.N0001 ASN.D0001 ASN.D0002

(continued)

4  Stock Market Analysis 

95

Table 4.A.2 (continued) Symbol

SARE.N0101

MUA.D2124 NMHL.D0717 NMHL.D0718 NMHL.D0719

ROCK.N0000 ALTP.N0000 NFP.N0000 GFP.N0000 DEL.N0000 CMBI.N0000 ATIL.N0000 TREV.P0000 TAD.N0000

Company name Imara African Opportunities Fund Limited Imara Global Fund Limited Summit Fund PCC – Balanced Fund (A cell of Summit Funds PCC, formerly Fleming Financial Trust Global Funds PCC) RSJ Prop, PCC Sanlam Africa Core Real Estate Investments Limited AIGO UK Residential Property Fund (A cell of AIGO Holdings PCC) AIGO Commercial Property Fund (A cell of AIGO Holdings PCC) AIGO Natural Resources Fund (A cell of AIGO Holdings PCC) Global Windsor Capital Fund Limited IPRO Funds Ltd (IPRO African Market Leaders Fund – Class (I2) Institutional Class) Universal Golden Fund AfrAsia Special Opportunities Fund (Class A, Class B) Specialised Debt securities TC Mauritius Holdings Omnicane Limited Multicurrency Medium Term Note Programme The Mauritius Union Assurance Cy. Ltd – Floating Rate Subordinated Notes New Mauritius Hotels Ltd – EUR Fixed Rate Notes New Mauritius Hotels Ltd – MUR Floating Rate Tranche A Notes New Mauritius Hotels Ltd – MUR Floating Rate Tranche B Notes Global business companies Colina Holdings Limited Bayport Management Limited Rockcastle Global Real Estate Company Limited Atlantic Leaf Properties Limited New Frontier Properties Ltd Green Flash Properties Ltd Delta Africa Property Holdings Limited (formerly Delta International Property Holdings) CMB International Limited Astoria Investments Limited Trevo Capital Ltd (Preference Shares) Tadvest Limited

96 

Economics and Finance in Mauritius

abroad quickly transmits into a negative stock in SEM. Consequently, the drastic fall in equity prices generate stronger effects on listed firms, insurance companies and non-listed firms which have strong equity investments. The impacts on the banking and household sectors are found to be low due to the Banking Act 2004 (as a result of which the banks are not allowed to have more than 10% equity investments) and the current low level of financial literacy among Mauritian households when it comes to sophisticated investment products such as equities. Listed firms are instantly affected as based on plummeting equity share prices, they are subject to the low level of equity market value so that their debt-to-equity ratio undergoes a substantial increase. Consequently, banks may request these higher debt-to-equity companies to either undertake quicker repayment to maintain the stipulated level of debt-to-equity ratio.

Bibliography Stock Exchange of Mauritius Factbook 2016 Monthly Statistical Bulletins, Bank of Mauritius, various issues.

5 The Behaviour of Foreign Investments in the Stock Exchange of Mauritius

This chapter addresses the effects of the crisis on foreign investments in SEM, a highly coveted African market. The results show poor interaction between foreign investments and the bank rate, endorsing the predominance of local banks as main bidders. The long-run pre-crisis elasticity coefficients of purchases and sales, to the US dollar, hover around −0.72 and 0.10, respectively, showing a strong desire for local stocks by foreigners. The findings depict only weak evidence of feedback trading, signifying that foreigners intervened by purely reacting to fundamentals in lieu of past returns. No evidence is found in favour of base-broadening or price pressure effects, except in the case of MCB stocks. Nearly all of the hypotheses under investigation call forth the need for specificity of analysis not only in terms of net flows but also in terms of the blue-chip stocks.

1

Introduction

With the onset of globalization, foreign investments now constitute a major source of international flows in many countries across the world. However, as a result the impact of crises can unleash a major source of © The Author(s) 2017 I. Ramlall, Economics and Finance in Mauritius, DOI 10.1007/978-3-319-39435-0_5

97

98 

Economics and Finance in Mauritius

tensions to policy-makers when it comes to establishing sound ­policies to obviate any possible adverse effects of foreign investments. For instance, during times of crisis, many investors retreat into cash so that there is massive disinvestments in equity markets, trailing behind heightened market risk which may eventually translate into credit risk. In that respect, trading by foreigners is usually closely watched as their transactions can exert considerable effects on prices and on the balance of payments. The rationale is that foreigners can induce high volatilities so as to scale up the level of credit risk for loans pledged by shares. In the extreme case where a large proportion of total advances in a country is made up of loans pledged by shares, then, this can cause the whole financial system to stall. Hence, studies on foreign investments are usually undertaken to generate a sound assessment of feasible financial stability risks. As a matter of fact, foreign investments have gained momentum in the Stock Exchange of Mauritius (SEM) on the back of sustained growth in the level of crossborder equity investments. As at date, no study has been carried out with respect to foreign investments in Mauritius. In essence, this study presents a comprehensive assessment of certain key aspects related to foreign investments. Basically, the chapter addresses three key questions. First, focus is laid as to whether foreigners buy and sell simultaneously so that this would unleash cointegration between foreign sales and foreign purchases. The underlying rationale for such a state of affairs emanates from the fact that foreign investors undertake a restructuring of their portfolios so that they can sell to buy back in order to channel them to other portfolios. Such an investigation is based on the practical experience of the author. Second, focus is laid as to whether there subsist pronounced effects of foreign investments on the Treasury bill market and also on the rupee price of US dollar. Such an analysis is vital in view of gauging any feasible spillover effects of foreign investors’ transactions onto other markets. Third, focus will be laid on well-established analytical frameworks which characterize the behaviours of foreign investors based on theoretical foundations such as feedback traders, base-broadening and price pressure hypotheses. In a nutshell, this is the very first study that addresses all vital aspects of foreign investments in the Mauritian stock market. The main benefit is that

5  The Behaviour of Foreign Investments in the Stock Exchange... 

99

with better knowledge, policy-makers are better equipped to come up with better policies to enhance the level of trading in the stock market. Hence, while one set of analysis is purely practical, the other two sets of analyses are based on substitution effects and theoretical foundations. This chapter is organized as follows: Section  2 provides a literature review of studies about impact of foreign investments on stock markets. Section 3 focuses on the data and methodology parts. Section 4 discusses the empirical results. Finally, Section 5 concludes.

2

Literature Review

The literature on foreign investments covers a wide range of distinct scenarios involving many different types of countries. However, most studies usually begin by drawing a demarcation line between the blessings and drawbacks of foreigners in a given stock market. Among the benefits that are cited are the, transfer of knowledge and innovation which are widely considered to be key ingredients in fostering higher levels of foreign investment in a country. Indeed, Neto and Veiga (2013) found that foreign direct investments did indeed impact on both productivity and GDP growth through the diffusion of technology and innovation. Another important blessing attached to foreign investment relates to the rekindling of the level of liquidity in the market via a more broadened ownership level. Such a state of affairs has been the subject of study by many researchers. For example, Henry (2000), Bekaert and Harvey (2003), and Kim and Singal (2000) pointed out that capital flow liberalization is vital to ensuring the enhanced liquidity and efficiency of a stock market. In other words, by inducing ownership from outsiders, foreign investments rekindle the quality of trades and stock ownerships, thereby trailing behind an improved level of risk-sharing. Moreover, the very presence of foreign investors in a stock market signals the need for the regulatory authorities to keep abreast with international norms, meaning that robust and comprehensive rules and regulations are being established. For instance, Evans et al. (2002) stated that the very presence of f­ oreigners

100 

Economics and Finance in Mauritius

in a stock market required local regulatory bodies to impose better rules and regulations so that malpractices were utterly shunned off or reduced to a minimum level. While most studies have identified the many benefits attached to foreign investments, some drawbacks have also been identified. For example, Barbopoulos et al. (2013) probed into the shareholder wealth effects of foreign direct investment announcements by UK firms. They found evidence that the strongest gains accrued to acquirers who invested in countries which were endowed with high ratings for both political risk and corruption. In a parallel manner, Stiglitz (1998) stated that there was pressing need for capital flow controls in developing countries because they were highly sensitive and hence vulnerable to any switch in international flows. Above all, studies pointed out that foreign investments could destabilize a stock market through their effect on three channels: price pressure due to poor liquidity, positive feedback and herding behaviour. One important strand of the literature on foreign investments focuses on feedback traders, who are also known as trend chasers. In essence, feedback traders are technical analyst-type investors as they allocate their assets not based on fundamentals, but instead by relying on past returns as as documented by Choe et al. (1999), Borensztein and Gelos (2000), Kim and Wei (2002), Bonser-Neal et al. (2002), Karolyi (2002), Griffin et al. (2004), and Richards (2005). There are two versions of feedback traders, namely, positive feedback traders and negative feedback traders. The positive feedback hypothesis points out that investors channel funds into the market following rising returns so that foreign investment flows in the present are determined by past positive market returns. Alternatively stated, under the positive feedback hypothesis, investors buy past winners and sell past losers. Another version of feedback trading is based on the work of De Bondt and Thaler (1985) contrarian investment strategy, also known as negative feedback trading, whereby past losers are bought while past winners are sold. In the case of a negative feedback hypothesis, investors buy past losers and sell past winners. In that respect, foreign investment flows now are determined by past negative market returns. In a study that focused on US pension funds, Lakonishok et al. (1992) found evidence of positive feedback trading, but only in the case

5  The Behaviour of Foreign Investments in the Stock Exchange... 

101

of small capitalization stocks. By contrast, Shefrin and Statman (1995) and Odean (1999) found evidence of negative feedback trading in the case of individual investors. Hou and Li (2014) showed that the CSI 300 stock index futures market intensified positive feedback trading in the underlying spot market. Many studies incorporated feedback trading analysis. For example, Bohn and Tesar (1996) found a positive relationship between foreign equity flows and returns in emerging markets using monthly and quarterly data. Such a relationship was further confirmed via daily and weekly data by Froot et  al. (2001). Moreover, Froot et  al. (2001) also found evidence that flows into a market were positively correlated with lagged returns in that market. They stated that such positive feedback trading might be evidence of some foreign investors resorting towards returns to extract information about future returns. Jackson (2003) showed that individual investor flows demonstrated negative feedback trading with respect to recent returns. In the case of the USA, Griffin, Harris and Topaloglu (2003) pointed to the existence of positive feedback trading for institutions while individuals exhibited negative feedback trading. It is important to note that the activities of the positive feedback traders have the potential of destabilizing stock prices. For instance, DeLong et al. (1990) focused on the feasible destabilizing effects of positive feedback trading in a theoretical model. They ended up with the fact that, in the presence of feedback trading, even rational speculators might “jump on the bandwagon” so that prices might substantially deviate from their fundamentals. The base-broadening hypothesis constitutes another major strand in the foreign investment literature. Under the hypothesis, lagged flows are not taken into consideration so that good (bad) news regarding the equity market led to positive (negative) returns and to flows into (out of ) equity funds. As a matter of fact, Clark and Berko (1997) pointed out that the presence of foreign investors in a stock market conferred some interesting benefits, one of which had been labeled as the base-broadening hypothesis. The rationale is that, with foreigners merging with local investors, there is an automatic diversification effect and hence a decline in the risk premium which subsequently unleashes a corresponding

102 

Economics and Finance in Mauritius

increase in the equity share price via risk pooling, as discussed by Merton (1987). Warther (1995) found positive evidence for base broadening in the case of mutual fund and security returns in the USA. In a parallel manner, Clark and Berko (1997) ended up with evidence in favor of the hypothesis in the case of foreign equity purchases in Mexico and market returns. There is a broader version of the base-broadening hypothesis, labelled the price pressure hypothesis, which states that in a case where the demand for equity is not fully elastic, a large flow into (out of ) equity funds will push security prices up (down), and this will be reversed in subsequent periods. Consequently, lagged positive flows should predict negative returns, and vice versa. The reason is based on the fact that, if the price pressure hypothesis holds, then prices are expected to return to their fundamentals when the sentiment is no longer there. Warther (1995), Edwards and Zhang (1998) and Fant (1999) found evidence of a significant positive contemporaneous relationship between aggregate net monthly fund flows and equity market returns. Clark and Berko (1997) evaluated price pressure for foreign investors in the Mexican stock market. However, they did not find evidence of any price pressure in the Mexican market. Similarly, Dahlquist and Robertson (2004) found no evidence of price pressure for the Swedish market: foreigners’ net inflows were coupled with significant increases in prices, but there was no price reversion after these price increases. The empirical literature has witnessed an increasing amount of research from African markets. Indeed, with the effects of the crisis and the quest for fresh markets, African markets are now deemed to be an integral part in terms of international portfolio diversification. This development has been propelled by the robust resource endowments of African economies. The critical role played by African economies in luring foreign direct investment has also been the subject of many studies. For instance, Sakyia et al. (2015) analysed the effect of foreign direct investment and trade openness on economic growth in Ghana. Their findings showed that growth was propelled chiefly on the back of strong interactions between foreign direct investment and exports. Furthermore, Tuomi (2011) used

5  The Behaviour of Foreign Investments in the Stock Exchange... 

103

­ icrodata and interviews to analyse the forces which impounded on m foreign direct investment in South Africa in an attempt to find the critical roles played by political and regulatory issues, skills, currency volatility and labour market regulation. In addition, Drogendijk and Blomkvist (2013) undertook an interesting study which attempted to gauge on the extent of Chinese investments in Africa. Their findings showed that African countries were subject to the higher propensity of FDI outflows relative to the rest of the world. Most importantly, they found that Chinese firms invested in Africa mainly for the purpose of natural resources and strategic assets allocations. Mijiyawa (2014) examined the underlying rationale behind which Ghana attracted more foreign direct investment than its West African counterparts. Findings disclosed that enshrined openness to global markets, human capital, sound infrastructures, political stability and institutional quality played preponderant roles in Ghana’s ability to lure higher foreign direct investments.

2.1

Stock Exchange of Mauritius

Established in 1989, the Stock Exchange of Mauritius (SEM) underwent a series of reforms in its attempts to attract foreign investments. Major reforms in the financial sector, capital flows liberalization, a switch to a

Fig. 5.1  An overview of SEM features

104 

Economics and Finance in Mauritius

floating exchange rate regime along with the promotion of political and economic stability, constituted the main ingredients in endorsing foreign investments in SEM. Furthermore, additional developments such as the introduction of electronic trading, daily trading and the extending of trading hours all contributed towards further boosting the stock market’s performance. As depicted in Fig. 5.1, the number of listed companies has remained more or less constant from 1996 onwards. In the case of market capitalization over GDP, a steady increase is noted from 2001 onwards. As far as the level of confidence is concerned, this is captured by the price– earnings ratio which decreased from 1997 to mid-2002, after which an upward trend is noted. Figure 5.2 depicts that net purchases on the SEM had been rising consistently from 1994 to 1997 to then retreat in 1998 until 2002 where some momentum in purchases appeared. Thereafter, a sustained increase in net purchases is noted. The equity market of Mauritius makes an interesting case to study the impact of the US subprime crisis on foreign investments for a plethora

Fig. 5.2  Foreign transactions in SEM

5  The Behaviour of Foreign Investments in the Stock Exchange... 

105

of reasons. First and foremost, Mauritius is now considered to be one of the best African economies in terms of economic performance, political stability and the ease of doing business. For instance, in the year 2007, the Mauritian stock market achieved remarkable results. Consequently, it would be apt to gauge how the crisis impacted on foreign investments in Mauritius so that based on the findings, the suggested policies could, to a certain extent, also be applicable for other African economies. In a nutshell, Mauritius could be used as a benchmark. Second, foreign investments showed a sustained rising trend implying that, year in and year out, foreign investors have a desire for Mauritian stocks. Consequently, from the viewpoint of policy makers, it is essential to have a proper background assessment of foreign investments onto the local market. Equipped with such knowledge, better policies can then be envisaged. Third, the stocks bought by foreigners tend to be of blue chips type, i.e. those stocks which tend to be imbued with stronger liquidity and a more robust market capitalization base. As such, these stocks are subject to the interaction of both local and foreign buyers, making the analysis more interesting. Indeed, if foreigners were to purchase stocks which were illiquid, the analysis done would suffer from liquidity effects to such an extent that the ensuing results would be misleading.

3

Data and Methodology

The use of weekly time series data is considered pertinent, being neither too long nor too short (daily) as to lose vital information when gauging the effects of foreign investments in SEM. Data on foreign investments were gleaned from the web site of a local investment company, namely Cim Stockbrokers Ltd. The data was collected in excel on a weekly basis from April 2005 to January 2009 from the weekly review pdf reports issued by Cim Stockbrokers Research. Foreign stock market indices were taken from Yahoo Finance. The definition of the variables collected is shown in Table  5.1. January 2009 is considered to be the end period since from that period Cim Stockbrokers ceased to publish the same level of information. The post-crisis period is captured from January 2008

106 

Economics and Finance in Mauritius

Table 5.1  Definition of the variables Variable

Definition

Net purchases Purchases

Difference between purchases and sales over market capitalisation Value of purchases by foreign investors over market capitalisation Value of sales by foreign investors over market capitalization Stock Exchange of Mauritius Index Mauritius Commercial Bank Limited State Bank of Mauritius Limited Ireland Blyth Limited Rogers Limited New Mauritius Hotels Limited SUN Resorts Limited Weighted average yield of 91-days, 182-days and 364-days Treasury bills Rupee price of US dollar

Sales SEMDEX MCB SBM IBL ROGERS NMHL SUN Bank rate US dollar

onwards whereas the pre-crisis period spans from April 2005 to January 2008. Based on the fact that foreigners were only involved in the purchase of certain specific shares, the following blue chips comprising of MCB (Mauritius Commercial Bank Limited), SBM (State Bank of Mauritius Limited), NMHL (New Mauritius Hotels Limited), SUN (SUN Resorts Limited), IBL (Ireland Blyth Limited) and ROGERS (Rogers Limited) have been considered for the analysis.

3.1

Econometric Model Specification

Based on the interactions between two variables under scrutiny, a VAR (Vector Autoregressive Model) system is applied. The rationale behind the use of VAR arises from the fact that VAR is highly suitable in the case that a model needs endogenous variables. Another blessing of VAR is that it can unleash the feasibility of gauging on any feasible long-term relationship, which is technically labelled as cointegration. Should the latter prevail, then, a restricted VAR, known as the Vector Error Correction Model (VECM) is used. In that respect, recourse is made towards the Johansen and Juselius technique (1990). In the case that cointegration is absent, then, the VAR model is used, as specified below.

107

5  The Behaviour of Foreign Investments in the Stock Exchange... 

X 1,t = A0 + A1,1 X 1,t− 1 + A1, 2 X 1,t− 2 + …. + A1, p X 1,t− p + A2,1 X 2,t− 1 + A2, 2 X 2,t− 2 + .... + A2, p X 2,t− p + ε1,t

(5.1)



X 2,t = B0 + B2,1 X 2,t−1 + B2,2 X 2,t− 2 + …. + B2, p X 2,t− p + B1,1 X1,t−1 + B1,2 X 2,t− 2 + …. + B1, p X 2,t− p + ε 2,t



(5.2)

Where X1,t and X2,t constitute the two main variables under examination. The value p, known as the lag order, is determined via the use of various information criteria such as the Akaike and Schwartz-Bayes criteria. The following three hypotheses are investigated in this study:

3.1.1  Are Foreign Sales and Purchases Cointegrated? This first hypothesis is tested following the experience of the author as a fund manager. Indeed, in many of the foreign trades, higher purchases effected on a given single day were followed by higher sales on the same day. The identified reason relates to the fact that fund managers may be subject to portfolio readjustments within the same company so that sales and purchases have to be effected as identified separate as different portfolios constitute different entities based on their various styles of operations. In that respect, X1,t and X2,t are posited to capture purchases and sales, respectively.

3.1.2  D  o Foreign Investments in SEM Impound on the Treasury Bill Market and on the Rupee Price of USD? The second hypothesis is specified to draw out any substitution/interacting effects of foreign investments on the Mauritian Treasury Bill market and also on the value of the exchange rate. The ultimate objective is geared towards assessing any spillover effects from the actions of foreign investors in the Mauritian stock market.

108 

Economics and Finance in Mauritius

Indeed, if foreign investors believe that the local stock market will be underperforming and do not want to convert back their investments, already in rupees, into foreign currencies while simultaneously trying to avail of some decent returns, they can channel their monies into the local Treasury Bill market. Another plausible rationale for such a state of affairs could be to shift part of the realized gains from equities onto the fixed income market in order to generate a fixed source of return while simultaneously betting on strong equity price increases. Under such a perspective, X1,t and X2,t are posited to capture foreign investments and the bank rate, respectively. Foreign investors who buy and sell on the local stock market are directly involved in currency transactions so much so that their transactions can trigger feasible effects on the exchange rate market. In that respect, two versions of X1,t and X2,t are considered. First, X1,t and X2,t are posited to capture foreign sales and the rupee price of the US dollar, respectively. Second, X1,t and X2,t are posited to capture foreign purchases and the rupee price of the US dollar, respectively.

3.1.3  D  o Foreigners Behave Like Feedback Traders or as per the Base-Broadening Approach or as per Price Pressure Hypotheses? The third hypothesis is purely based on theoretical foundations as mentioned in the literature review. Three distinct strands of the literature are taken into consideration, namely the feedback trading, base-broadening effects and price pressure effects. The model to be applied is discussed below.

Feedback Trader Hypothesis

Net t = α 0 + α1 Rt + α 2 Rt −1 + α 3 Rt −2 + α 4 Rt −3 + α 5 Rt −4 (5.3) MktCap t −1

109

5  The Behaviour of Foreign Investments in the Stock Exchange... 

Where: Net: Net Purchases by Foreigners MktCap: Market Capitalization of SEM R: Returns of SEMDEX α 2 ,α 3 ,α 4 ,α 5 > 0 ; positive feedback trading α 2 ,α 3 ,α 4 ,α 5 < 0 ; negative feedback trading Feedback traders (also known as trend chasers) do not base their asset allocation decisions on market fundamentals, but rather on the history of past returns (see Froot et al. (2001). This is somewhat akin to the strategy deployed by technical analysts. The positive feedback hypothesis points out that investors channel funds into the market following rising returns so that flows must lag returns. Feedback trading is also known as pull factors which affect stock market returns. Under the positive feedback hypothesis, investors buy past winners and sell past losers. In the case of a negative feedback hypothesis, investors buy past losers and sell past winners.

Base-Broadening Hypothesis

Rt = α 0 + α1

(5.4)

Net t MktCap t −1

Where: Net: Net Purchases by Foreigners MktCap: Market Capitalisation of SEM R: Returns of SEMDEX α1 > 0 in case the hypothesis holds true.

Price Pressure Hypothesis (Encapsulates Base-Broadening Hypothesis) Rt = α 0 + α1

Net t Net t Net t Nett t + α2 + α3 + α4 MktCap t −1 MktCap t −2 MktCap t −3 MktCap t −4



(5.5)

110 

Economics and Finance in Mauritius

Where: Net: Net Purchases by Foreigners MktCap: Market Capitalisation of SEM R: Returns of SEMDEX Flows at time t is scaled by market capitalization at time t − 1. Clark and Berko (1997) pointed out that the presence of foreign investors in a stock market conferred some interesting benefits, one of which had been labeled as the base-broadening hypothesis. The rationale is that, with foreigners merging with local investors, this generated diversification effect and hence a fall in risk premium so that this entailed a permanent rise in the equity share price via risk pooling (Merton 1987). Warther (1995) found positive evidence for base-broadening in case of mutual fund flows in US and security returns. In a parallel manner, Clark and Berko (1997) ended up with evidence in favor of the hypothesis in the case of foreigner equity purchases in Mexico and market returns. However, Clark and Berko (1997) stated that even if the base-broadening hypothesis were to hold, nonetheless, α1 was susceptible to underestimate the effect of foreign inflows on the host country’s stock prices. Based on the fact that Eq. (5.5) encapsulates Eq. (5.4), only the former is used for estimation. Table 5.2  Summary statistics for stock market returns Mean Median Maximum Minimum Std. Dev. Skewness Kurtosis Jarque– Bera

CAC-40

DJIA

FTSE

JSE

NASDAQ NIKKKEI SEMDEX

−0.0014 0.0024 0.1243 −0.2505 0.0344 −2.0849 17.601 1873.5

−0.0012 0.0007 0.1069 −0.2002 0.0271 −1.773 18.208 1981.6

−0.0007 0.0025 −0.0013 0.0014 0.0068 0.0013 0.1258 0.1603 0.1036 −0.2363 −0.0963 −0.1660 0.0309 0.0337 0.0303 −2.0233 −0.1986 −1.1167 21.707 5.849 8.9672 2976.6 67.232 329.84

−0.0016 0.0023 0.0022 0.0034 0.1144 0.0871 −0.2788 −0.1572 0.0367 0.0276 −2.4374 −1.1270 19.188 9.6285 2322.3 398.27

The above table shows the summary statistics for the local and foreign stock market returns. Deviation of values from 3 under kurtosis and 0 under skewness, denote the existence of non-normality conditions

5  The Behaviour of Foreign Investments in the Stock Exchange... 

4

Results

4.1

Summary Statistics

111

Summary statistics are presented in Table 5.2. None of the stock market returns are normal; they exhibit negative skewness and excess kurtosis. The non-normality conditions are substantiated by the Jarque–Bera statistic. The first-order autocorrelation in weekly net purchases stands at 0.43. Substantial positive autocorrelation is expected in case of inflows as is found with the results of Froot et al. (2001) and Richards (2005). Such a positive autocorrelation is attributed to the fact that specific investors set up their positions slowly to mitigate market impact. In the case of correlation between net purchases and SEMDEX returns, the value hovers around 0.06, indicative of the fact that, a priori, pull forces are unlikely to account for the behavior of foreign investors in SEM. Table 5.3  ADF unit root tests Augmented Dickey-Fuller tests Net purchases Purchases Sales FTSE CAC-40 DJIA NASDAQ JSE NIKKEI SEMDEX MCB SBM IBL ROGERS NMHL SUN Bank rate US dollar

Level

First difference

Phillips Perron tests First difference Level

−5.6850* −6.2862* −2.1242 −0.5401 0.4442 −0.2628 −0.2009 −2.5340 0.0189 −1.5743 −1.8990 −1.4477 −1.5373 −1.5283 −1.3113 −0.7013 −1.6511 −1.3183

−11.944* −13.594* −11.605* −16.705* −16.925* −15.685* −14.296* −15.337* −8.428* −13.625* −12.903* −15.149* −15.400* −12.560* −14.147* −15.392* −10.442* −8.870*

−9.2795* −11.719* −8.0884* −0.8500 0.0782 −0.0228 −0.3152 −2.5544 0.0108 −1.5470 −1.8586 −1.4510 −1.4318 −1.4558 −1.3025 −0.7362 −1.6834 −1.2023

−57.188* −51.137* −62.067* −16.686* −16.900* −15.636* −14.295* −15.269* −15.183* −13.676* −12.923* −15.098* −15.542* −12.771* −14.148* −15.323* −10.877* −8.8717*

The above table shows the unit root tests with * denoting statistical significance at the 1% level.

112 

4.2

Economics and Finance in Mauritius

Unit Root Tests

Both Augmented Dickey–Fuller and Phillips–Perron unit root tests are undertaken with results shown in Table 5.3. These tests are conducted on the natural logarithmic of all the series. The net purchases variable is computed as a ratio of net purchases over market capitalization, while all other variables are taken in their logarithmic versions.1 As shown in Table 5.3, except for net purchases, purchases and sales, all the rest of the variables are not stationary in levels but stationary at first differences. In that respect, the series are said to be integrated to order one I (1), making them eligible for any eventual cointegration analysis.

4.3

Focusing on the Three Hypotheses

4.3.1  A  re Foreign Purchases and Foreign Sales Cointegrated? The Akaike and Schwartz-Bayes criteria are employed to investigate the appropriate lag length. Two optimal lag orders are used. The results show that purchases and sales by foreigners are both stationary in level but with prevalence of cointegration under two cointegrating equations. Further analysis is therefore not warranted since, for n variables, there can be at least n − 1 cointegrating equations. Relative to the pre-crisis period, for the post-crisis period, under the optimal lag of two, it surfaced that only one cointegrating equation prevailed. The results are shown in Table 5.4. A 1% change in sales lagged by one period trails behind a 0.57% change in purchases in the current period. Hence, evidence is found of sales impacting on purchases in the post-crisis period. No evidence is found of purchases affecting sales in the short run. The long-run elasticity coefficient is of a negative sign. Hence, over the long run in the post-crisis period, higher sales led to lower purchases; indicative of foreigners selling for good. Such a finding bodes well with the fact that the crisis has undermined the scope for international portfolio diversification.  Technically speaking, logarithmic cannot be applied to a rate. However, the bank rate is just the term used to denote the weighted average yields on 91-day, 183-day and 364-day Treasury Bills issued on a weekly basis by the Central Bank of Mauritius. 1

5  The Behaviour of Foreign Investments in the Stock Exchange... 

113

Table 5.4  Results for cointegration between purchases and sales for the post-­ crisis period Cointegration equation 1 Constant ΔPurchasest−1 ΔPurchasest−2 ΔSalest−1 ΔSalest−2 Adjusted R2 F-Statistic Cointegration equation 1 Purchasest−1 1.0000

ΔPurchasest

ΔSalest

−0.0473 (−2.8861)* −0.0749 (−0.4920) −0.8425 (−5.5963)* −0.3759 (−2.4850)** 0.5762 (2.4399)** 0.3242 (1.6322) 0.4445 9.9640

−0.0545 (−4.1198)* 0.01358 (0.1106) −0.1784 (−1.4702) −0.1986 (−1.6288) 0.1288 (0.6764) 0.1931 (1.2056) 0.3634 7.3955

Salest−1 13.964 (4.2436)*

Constant −57.238

The above table depicts the relationship between purchases and sales in the post crisis period. * and ** denotes statistical significance at the 1% and 2% level, respectively.

4.3.2  D  o Foreign Investments in SEM Impound on the Treasury Bill Market and on the Rupee Price of USD? Do Foreigners Investing in the Stock Exchange of Mauritius Consider the Treasury Bill Market to be a Substitute for Their Investments? It can be hypothesized that foreigners who invested in SEM can affect the local Treasury Bill market. For instance, in the instance that foreigners perceive that the local fixed income market is providing interesting yields, they can channel their proceeds from sales of shares to these fixed income instruments. In the case of net purchases and the bank rate during the pre-crisis period, one cointegrating equation is found. The VECM is run but no relationship is established as to whether the bank rate affects net p ­ urchases

114 

Economics and Finance in Mauritius

or vice versa. Though there is a negative effect of net purchases on the bank rate in the long run, yet the result is economically insignificant, as depicted in Table 5.5. Hence, this finding suggests that foreigners that deal in SEM are not concerned about the local fixed income instruments. Their objective is chiefly geared towards international equity diversification with a view to ironing out systematic country risks. Similarly, for the post-crisis period, under the optimal lag of one, no cointegration Table 5.5 Results for cointegration between bank rate and NetPurchases: pre crisis period Cointegration equation 1 ΔBank Ratet−1 ΔBank Ratet−2 ΔNetPurchasest−1 ΔNetPurchasest−2 Constant Adjusted R2 F-Statistic Cointegration equation 1 Bank Ratet−1 1.0000

ΔNetPurchasest

ΔBank Ratet

−0.8375 (−6.0323)* 0.0002 (0.3110) −0.0005 (−0.7060) −0.0404 (−0.3449) −0.0169 (−0.1943) 0.0000006 (0.0332) 0.4193 20.501

−0.9444 (16.624)* 0.2397 (2.7232)* 0.1631 (1.8546) −5.7290 (−0.4084) −13.190 (−1.2670) 0.0023 (0.9936) 0.0790 3.3171

NetPurchasest−1 Constant 0.0002 −0.0007 (2.4901)** Results for VAR between bank rate and NetPurchases: post crisis period ΔBank Ratet NetPurchasest ΔBank Ratet−1 0.2147 −0.0008 (1.6019) (−0.7294) NetPurchasest−1 −0.5.5957 0.5910 (−0.4415) (5.2491)* Constant −0.0044 0.0023 (−1.0875) (0.9936) Adjusted R2 0.0162 0.3261 F-Statistic 1.4633 14.553 The above table shows the cointegration results in the case of bank rate and net purchases in the pre and post crisis periods. * and ** denotes statistical significance at the 1% and 2% level, respectively

5  The Behaviour of Foreign Investments in the Stock Exchange... 

115

prevailed and VAR results corroborated the impotency of the relationship between the bank rate and net purchases. However, based on the fact that net purchases may be symptomatic to some noisy data. The analysis is refined by having recourse to purchases and sales, separately, with the bank rate. For purchases and the bank rate, one cointegrating equation manifests in the pre-crisis episode, as illustrated in Table 5.6. VECM results show no interaction between purchases and the bank rate. In case of purchases lagged by three weeks, though it generates a positive effect on the bank rate, yet the coefficient is not economically significant. Similar results are obtained in the post-crisis period, i.e., there is no interaction identified between these two variables. Hence, there is robust evidence that foreigners who invest in SEM are not really concerned about Treasury Bill investments; they are merely trying to harness benefits from international equity portfolio diversification. In the case of foreign sales and the bank rate, cointegration prevails under an optimal lag of one period irrespective of the pre- and post-crisis episodes. Interestingly, it transpires that a long-run positive relationship subsists between the bank rate and sales for both periods, hovering around 0.32 and 0.29, respectively, for the pre- and post-crisis periods, as demonstrated in Table 5.7. Such a finding is highly interesting as it shows that, in the long run, higher foreign sales are associated with a bank rate, unleashing some bouts of equity and fixed income market interactions. It is only in the pre-crisis period that short-run effects are found whereby the bank rate has a negative impact on sales. Such a negative relationship clearly shows that foreign investors in SEM did not consider the local Treasury Bill market as an alternative for their investments. Had foreigners who invested in SEM also been interested in the Mauritian Treasury Bill market, then a rise in the bank rate should have led to strong sales with proceeds being then channeled in the fixed income market. From these findings, it can also be deduced that the investor base in the local Treasury Bill market is susceptible to being dominated by local investors, in particular, banks, which usually constitute vigorous bidders.

116 

Economics and Finance in Mauritius

Table 5.6  Results for cointegration between bank rate and purchases: pre crisis period Cointegration equation 1 ΔPurchasest−1 ΔPurchasest−2 ΔPurchasest−3 ΔPurchasest−4 ΔBank Ratet−1 ΔBank Ratet−2 ΔBank Ratet−3 ΔBank Ratet−4 Constant Adjusted R2 F-Statistic Cointegration equation 1 Bank Ratet−1 1.0000

ΔPurchasest

ΔBank Ratet

−0.7435 (−4.5420)* −0.1194 (−0.7823) −0.0292 (−0.2122) 0.0115 (0.0968) 0.0253 (0.2724) 0.1118 (0.0348) −1.7156 (−0.5386) 1.9626 (0.6200) −6.0862 (−1.9806)*** 0.0308 (0.3664) 0.4096 11.255

−0.0062 (−1.3450) 0.0052 (1.3540) 0.0052 (1.3540) 0.0092 (2.7542)* 0.0016 (0.6274) 0.2426 (2.6850)* 0.1668 (1.8578) −0.0295 (−0.3308) −0.0418 (−0.4830) 0.0025 (1.0885) 0.1169 2.9567

NetPurchasest−1 Constant −0.4203 −2.2750 (−0.9472) Results for VAR between bank rate and purchases: post crisis period Purchasest ΔBank Ratet Purchasest−1 0.2579 0.0040 (1.8760) (1.2466) ΔBank Ratet−1 2.7860 0.1996 (0.4983) (1.5028) Constant 2.8465 −0.0205 (5.0466)* (−1.5339) Adjusted R2 0.0356 0.040 F-Statistic 2.0340 2.1772 The above table demonstrates the results obtained for the cointegration analysis between the bank rate and purchases by foreigners in the pre and post crisis periods. * and *** denotes statistical significance at the 1% and 5% level, ­ respectively

5  The Behaviour of Foreign Investments in the Stock Exchange... 

117

Table 5.7  Results for cointegration between bank rate and sales: pre crisis period Cointegration equation 1 ΔBank Ratet−1 ΔSalest−1 Constant Adjusted R2 F-Statistic Cointegration equation 1 Bank Ratet−1 1.0000

ΔBank Ratet

ΔSalest

−0.0018 (−0.2971) 0.2756 (3.1152)* 0.0002 (0.1955) 0.0030 (1.2532) 0.0507 3.3147

2.5032 (6.8663)* −11.475 (−2.1835)*** −0.0764 (−0.8669) 0.0251 (0.1712) 0.4290 33.568

Salest−1 Constant −0.3197 −1.7793 (−8.3322)* Results for cointegration between bank rate and sales: post crisis period ΔBank Ratet ΔSalest Cointegration equation 1 −0.0601 2.6525 (−3.1861)* (4.2105)* ΔBank Ratet−1 0.1098 2.5799 (0.8534) (0.6005) ΔSalest−1 −0.0072 −0.0086 (−1.6549) (−0.0594) Constant −0.0053 0.0325 (−1.4027) (0.2578) Adjusted R2 0.1612 0.3472 F-Statistic 4.5877 10.931 Cointegration equation 1 Bank Ratet−1 Salest−1 Constant 1.0000 −0.2890 −1.0057 (−6.2030)* The above table demonstrates the results obtained for the cointegration analysis between the bank rate and sales by foreigners in the pre and post crisis periods. * and *** denotes statistical significance at the 1% and 5% level, respectively

Do Foreigners Influence the Rupee Price of the US Dollar While Investing in the Stock Exchange of Mauritius? The buying/selling actions of foreigners can trail behind exchange rate fluctuations in a given stock market. An analysis is undertaken to isolate the impact of foreign investment on the rupee price of US dollar.

118 

Economics and Finance in Mauritius

In the case of the pre-crisis period, cointegration systematically ­manifests for net purchases, sales and purchases. In case of net purchases and the US dollar, there is no bidirectional causality in the short run, independent of the pre- and post-crisis periods. However, in the long run, though there is a negative effect of net purchases on the US dollar, yet the effect is not economically significant. In the case of purchases and the US dollar, it transpires that, in the long run, the elasticity of the US dollar to purchases is −0.72% in the pre-crisis era. Such a negative sign bodes well with the fact that higher purchases unleashes a rise in demand for rupee relative to the US dollar so that the US dollar should depreciate vis-à-­vis the rupee. As far as the short-term effects are concerned, purchases are found to positively impact on the US dollar; however, the result is not economically significant. VAR results in the post-crisis period furnish evidence of the impotence of foreigners’ actions on the value of the US dollar. In case of sales and the US dollar in the pre-crisis period, it occurred that in the long run, the elasticity of the US dollar to sales is 0.10%. Such a finding again befits the fact that higher sales should technically generate an appreciating US dollar vis-à-vis the rupee. No impact is noted post the onset of the crisis. It is interesting to note that, in relative terms, the elasticity is relatively higher in the case of purchases relative to sales (0.72 versus 0.10), plainly signaling that prior to the crisis, foreigners had mainly been purchasers rather than sellers as shown in Fig. 5.2 whereby net purchases had consistently risen from year 2005 to 2007 (Tables 5.8, 5.9, and 5.10).

4.3.3  D  o Foreigners Behave Like Feedback Traders or as per the Base-Broadening Approach or as per Price Pressure Hypotheses? Do Foreigners Behave Like Feedback Traders in the Stock Exchange of Mauritius? As far as SEMDEX is concerned, signs of positive feedback trading are noted in the post-crisis period. While the variable net purchase lagged by one week is positive for SEMDEX in both periods, it is positive only

5  The Behaviour of Foreign Investments in the Stock Exchange... 

119

Table 5.8  Results for NetPurchases and USDOLLAR: pre crisis Cointegration equation 1 ΔUSDOLLARt−1 ΔUSDOLLARt−2 ΔNETt−1 ΔNETt−2 Constant Adjusted R2 F-Statistic Cointegration equation 1 USDOLLARt−1 1.0000

ΔNET

ΔUSDOLLAR

−0.8650 (−6.1204)* 0.0066 (1.0887) 0.0021 (0.3142) −0.0204 (−0.1718) −0.0100 (−0.1125) −0.0000002 (−0.0129) 0.4277 21.181

−3.4219 (−1.6891) 0.4636 (5.2960)* 0.1975 (2.0467)*** 0.7315 (0.4289) 0.2229 (0.1745) −0.00005 (−0.2679) 0.3381 14.793

NETt−1 0.0019 (3.4610)* Results for NetPurchases and USDOLLAR: post crisis ΔUSDOLLAR ΔUSDOLLARt−1 0.3750 (2.7056)* ΔUSDOLLARt−2 0.00003 (0.0002) NETt−1 0.5832 (0.0867) NETt−2 −1.9927 (−0.2978) Constant 0.0014 (0.8138) Adjusted R2 0.0775 F-Statistic 2.1777

Constant −0.0068

NET −0.0003 (−0.1109) −0.0040 (−1.4716) 0.4257 (3.1552)* 0.2542 (1.8935) 0.000003 (0.0936) 0.37012 9.2267

The above table demonstrates the results obtained for the cointegration analysis between the net foreign purchases and US dollar by foreigners in the pre and post crisis periods. * and *** denotes statistical significance at the 1% and 2% level, respectively

in the post-crisis period and negative in the pre-crisis period. However, in the case of the blue chips, there is no strong indication for any feedback trading effects. Although positive feedback trading is noted in some cases, yet their signs are not economically significant. This plainly shows

120 

Economics and Finance in Mauritius

Table 5.9  Results for PURCHASES and USDOLLAR: pre crisis Cointegration equation 1 ΔUSDOLLARt−1 ΔUSDOLLARt−2 ΔPURCHASESt−1 ΔPURCHASESt−2 Constant Adjusted R2 F-Statistic Cointegration equation 1 ΔUSDOLLARt−1 1.0000

ΔUSDOLLAR

ΔPURCHASES

−0.0023 (−3.5991)* 0.3912 (4.4522)* 0.1324 (1.4346) 0.0012 (2.9527)* 0.0004 (1.5698) −0.00008 (−0.30575) 0.3772 17.357

−1.0551 (−5.3727)* 10.045 (0.3808) −53.717 (−1.9382) −0.0989 (−0.80387) −0.0189 (−0.2051) 0.0148 (0.1836) 0.4196 20.523

ΔPURCHASESt−1 0.7209 (6.8791)* Results for PURCHASES and USDOLLAR: post crisis ΔUSDOLLAR ΔUSDOLLARt−1 1.3808 (10.803)* ΔUSDOLLARt−2 −0.3907 (−2.9670)* PURCHASESt−1 0.00005 (0.0401) PURCHASESt−2 −0.0007 (−0.5417) Constant 0.0346 (0.3804) Adjusted R2 0.9663 F-Statistic 373.35

Constant −5.7336

PURCHASES 16.960 (1.4296) −18.864 (−1.5432) −0.7638 (−5.8167)* −0.2886 (−2.2250)*** 6.2974 (0.7460) 0.3910 9.9906

The above table demonstrates the results obtained for the cointegration analysis between the purchases and US dollar by foreigners in the pre and post crisis periods. * and *** denotes statistical significance at the 1% and 5% level, ­ respectively

5  The Behaviour of Foreign Investments in the Stock Exchange... 

121

Table 5.10  Results for SALES and USDOLLAR: pre crisis Cointegration equation 1 ΔUSDOLLARt−1 ΔUSDOLLARt−2 ΔSALESt−1 ΔSALESt−2 Constant Adjusted R2 F-Statistic Cointegration equation 1 ΔUSDOLLARt−1 1.0000

ΔUSDOLLAR

ΔSALES

0.0045 (1.8323) 0.4477 (4.8342)* 0.1084 (1.0591) 0.0004 (1.9866)*** 0.0001 (0.6418) −0.00009 (−0.3340) 0.3386 14.008

7.7368 (6.2191)* −138.15 (−2.9604)* −40.664 (−0.7880) −0.0681 (−0.5930) 0.0201 (0.2276) −0.0255 (−0.1714) 0.4274 19.962

ΔSALESt−1 −0.1080 (−7.2317)* Results for SALES and USDOLLAR: post crisis ΔUSDOLLAR ΔUSDOLLARt−1 0.3679 (2.5266)** ΔUSDOLLARt−2 −0.000009 (−0.00006) ΔSALESt−1 −0.0002 (−0.1245) ΔSALESt−2 −0.0006 (−0.3359) Constant 0.0049 (0.5327) Adjusted R2 0.0788 F-Statistic 2.1986

Constant −3.3213

SALES 17.313 (1.6877) 1.6105 (0.1539) 0.2782 (1.9185) 0.0816 (0.5897) 2.4271 (3.7417)* 0.0678 2.0190

The above table demonstrates the results obtained for the cointegration analysis between the sales and US dollar by foreigners in the pre and post crisis periods. * and *** denotes statistical significance at the 1% and 2% level, respectively

122 

Economics and Finance in Mauritius

Table 5.11  Dependent variable: net purchases under SEMDEX pre crisis and post crisis results ΔSEMDEXt ΔSEMDEXt−1 ΔSEMDEXt−2 ΔSEMDEXt−3 ΔSEMDEXt−4 NETt−1 Constant Adjusted R2 Durbin-Watson Sta F-Statistic

−0.0010 (−1.1234) 0.0014 (1.5320) −0.0019 (−1.9840)*** 0.0003 (0.3161) 0.0007 (0.8060) 0.2113 (2.4302)** 0.0001 (5.6078)* 0.0448 2.0101 2.0416

0.0004 (0.5507) 0.0026 (2.9721)* 0.0021 (2.1744)*** 0.0020 (2.0392)*** 0.0006 (0.6784) 0.3989 (3.0676)* 0.00006 (1.7015) 0.4227 2.1336 7.8339

The above table demonstrates the results obtained for the feedback trader hypothesis for the pre and post crisis episodes under SEMDEX, respectively. *, ** and *** denotes statistical significance at the 1%, 2% and 5% level, respectively

Table 5.12  Dependent variable: net purchases under MCB pre crisis and post crisis results ΔMCBt ΔMCBt−1 ΔMCBt−2 ΔMCBt−3 ΔMCBt−4 NETt−1 Constant Adjusted R2 Durbin-Watson Sta F-Statistic

0.0012 (1.5032) −0.0008 (−1.0568) −0.0001 (−0.1304) 0.0008 (0.9953) 0.00002 (0.0307) −0.4327 (−5.3960)* −0.00001 (−0.3761) 0.1865 2.2507 6.0820

0.00004 (0.0550) 0.0025 (3.1311)* 0.0012 (1.4491) 0.0013 (1.5251) 0.0009 (1.0311) 0.4509 (3.5757)* 0.00003 (0.9753) 0.4103 2.1583 7.4955

The above table demonstrates the results obtained for the feedback trader hypothesis for the pre and post crisis episodes under MCB, respectively. *denotes statistical significance at 1% level

5  The Behaviour of Foreign Investments in the Stock Exchange... 

123

that foreigners intervened in SEM by purely reacting to fundamentals of locally quoted companies and not resorting to historical trends (Tables 5.11, 5.12, 5.13, 5.14, 5.15, 5.16, and 5.17).

Do Foreigners Behave in a Manner Consistent with the Base-­ Broadening Hypothesis and the Price Pressure Hypothesis? In the case of contemporaneous flows, irrespective of the dependent variable considered (whether stock market returns or specific stock ­ returns), it transpires that none of the coefficients is statistically significant. This signifies no presence of any base-broadening effect on SEM. Moreover, none of the lagged flows are statistically significant in the case of ROGERS, IBL and SUN and this holds irrespective of the pre- and post-­crisis periods. Consequently, there is also no evidence in favor of the price pressure effects.

Table 5.13  Dependent variable: net purchases under SBM pre crisis and post crisis results ΔSBMt ΔSBMt−1 ΔSBMt−2 ΔSBMt−3 ΔSBMt−4 NETt−1 Constant Adjusted R2 Durbin-Watson Sta F-Statistic

0.0004 (0.5982) −0.0005 (−0.7665) 0.0008 (1.2147) −0.0009 (−1.2459) −0.0006 (−0.9339) −0.4364 (−5.4195)* 0.000007 (0.2735) 0.1935 2.2329 6.3204

0.0005 (0.8978) 0.0022 (3.4803)* 0.0006 (0.8734) 0.0007 (1.1012) 0.0006 (0.9903) 0.4296 (3.2608)* 0.00004 (1.1785) 0.4168 2.0957 7.6717

The above table demonstrates the results obtained for the feedback trader hypothesis for the pre and post crisis episodes under SBM, respectively.*denotes statistical significance at 1% level

124 

Economics and Finance in Mauritius

Table 5.14  Dependent variable: net purchases under SUN pre crisis and post crisis results ΔSUNt ΔSUNt−1 ΔSUNt−2 ΔSUNt−3 ΔSUNt−4 NETt−1 Constant Adjusted R2 Durbin-Watson Sta F-Statistic

0.0004 (0.6957) −0.0005 (−0.9134) 0.0003 (0.5049) 0.0001 (0.2184) −0.00001 (−0.0174) −0.4495 (−5.6929)* −0.000006 (−0.2398) 0.1887 2.1786 5.9254

−0.0001 (−0.3758) 0.0011 (2.1176)* 0.0006 (1.2858) 0.0005 (1.0860) 0.0007 (1.3325) 0.5195 (4.2884)* 0.00005 (1.2341) 0.3549 2.1794 6.1359

The above table demonstrates the results obtained for the feedback trader hypothesis for the pre and post crisis episodes under SUN, respectively.* and *** denotes statistical significance at the 1% and 5% level, respectively Table 5.15  Dependent variable: net purchases under NMH pre crisis and post crisis results ΔNMHt ΔNMHt−1 ΔNMHt−2 ΔNMHt−3 ΔNMHt−4 NETt−1 Constant Adjusted R2 Durbin-Watson Sta F-Statistic

0.0011 (1.7294) 0.0003 (0.5068) 0.0004 (0.6846) 0.0008 (1.1809) −0.0011 (−1.6213) −0.4378 (−5.5896)* −0.00001 (−0.5288) 0.2099 2.2625 6.8895

−0.0002 (−0.5145) 0.0007 (1.5561) 0.0008 (1.7058) 0.0007 (1.4588) 0.0002 (0.5796) 0.5253 (4.2750)* 0.00003 (0.9059) 0.3657 2.2461 6.3814

The above table demonstrates the results obtained for the feedback trader hypothesis for the pre and post crisis episodes under NMH, respectively.*denotes statistical significance at 1% level

5  The Behaviour of Foreign Investments in the Stock Exchange... 

125

Table 5.16  Dependent variable: net purchases under ROGERS pre crisis and post crisis results ΔROGERSt ΔROGERSt−1 ΔROGERSt−2 ΔROGERSt−3 ΔROGERSt−4 NETt−1 Constant Adjusted R2 Durbin-Watson Sta F-Statistic

0.0018 (2.1246)*** −0.0002 (−0.3292) 0.0009 (1.0556) −0.00004 (−0.0562) −0.0014 (−1.6249) −0.4443 (−5.6162)* −0.000008 (−0.5288) 0.2126 2.2573 6.9854

0.0016 (1.3143) 0.0013 (1.0472) 0.0016 (1.2565) 0.0003 (0.2787) 0.0009 (0.7326) 0.4564 (3.5471)* 0.00005 (1.2761) 0.3458 2.1382 5.9353

The above table demonstrates the results obtained for the feedback trader hypothesis for the pre and post crisis episodes under ROGERS, respectively. * and *** denotes statistical significance at the 1% and 5% level, respectively Table 5.17  Dependent variable: net purchases under IBL pre crisis and post crisis results ΔIBLt ΔIBLt−1 ΔIBLt−2 ΔIBLt−3 ΔIBLt−4 NETt−1 Constant Adjusted R2 Durbin-Watson Sta F-Statistic

0.0000009 (0.0015) −0.0006 (−0.9880) −0.0001 (−0.3199) 0.0003 (0.6040) 0.0005 (0.8195) −0.4483 (−5.6400)* −0.0000002 (−0.0119) 0.1755 2.2689 5.7207

0.0003 (0.4874) 0.00005 (0.0663) 0.0006 (0.8951) 0.0007 (1.0406) −0.0004 (−0.5999) 0.5801 (4.7289)* −0.00001 (0.2888) 0.2987 2.1912 4.9771

The above table demonstrates the results obtained for the feedback trader hypothesis for the pre and post crisis episodes under IBL, respectively.*denotes statistical significance at 1% level

126 

Economics and Finance in Mauritius

Nevertheless, in case of the banks, some interesting findings are noted. For MCB, in the pre-crisis period, net flows lagged by three weeks triggered a positive effect (24%) while for the post-crisis period, it becomes clear that flows lagged by four weeks generated a much higher effect of −64%. In the case of SBM, while no positive effect emanates from any lagged flows during the pre-crisis period, in the post-crisis period, more specifically lagged flows by fours week entail behind an effect of about −69%. This shows some sort of convergence in performance during the post-crisis period for the banks sector. Moreover, based on the fact that no lagged positive coefficients are found either for the hotels sector or for the commerce sector, it can be inferred that foreigners are principally interested in the banking sector, with MCB stocks being the most attractive ones in their portfolios (Tables 5.18, 5.19, 5.20, 5.21, 5.22, 5.23, and 5.24). Table 5.18  Dependent variable: D(SEM) under SEM pre crisis and post crisis results NET NETt−1 NETt−2 NETt−3 NETt−4 NETt−5 NETt−6 D(SEM(−1)) Constant Adjusted R2 Durbin-Watson Sta F-Statistic

−2.0750 (−0.2678) −2.5770 (−0.3298) 6.618 (0.8415) 16.745 (2.1167)*** −2.9229 (−0.3606) 12.514 (1.5746) −9.4400 (−1.2084) 0.0842 (0.9370) 0.0001 (0.0791) −0.0396 2.0020 328.49

−7.6735 (−0.9764) −8.3430 (−1.0510) 1.5333 (0.1936) 12.113 (1.5288) −7.3050 (−0.9031) 6.8848 (0.8552) −15.317 (−1.9272) 0.0469 (0.5274) 0.0110 (2.6451)* 0.0079 2.0205 1.1325

The above table demonstrates the results obtained for price pressure hypothesis for the pre and post crisis episodes under SEMDEX, respectively. * and *** denotes statistical significance at the 1% and 5% level, respectively

5  The Behaviour of Foreign Investments in the Stock Exchange... 

127

Table 5.19  Dependent cariable: D(MCB) under SEM pre crisis and post crisis results NET NETt−1 NETt−2 NETt−3 NETt−4 NETt−5 NETt−6 D(MCB(−1)) Constant Adjusted R2 Durbin-Watson Sta F-Statistic

−9.8046 (−0.8762) −1.5450 (−0.1367) −1.9643 (−0.1746) 24.226 (2.1431)*** −8.1843 (−0.7042) 2.9604 (0.2583) −16.140 (−1.4241) 0.1113 (1.2516) 0.0116 (1.9657)*** 0.0038 1.9994 1.0632

−17.640 (−0.7616) 26.845 (1.1108) −4.4744 (−0.1861) 44.149 (1.8737) −64.196 (−2.6553)* 31.454 (1.3023) −3.3479 (−0.1609) 0.1222 (0.7787) −0.0059 (−1.0625) 0.0515 1.9356 1.3801

The above table demonstrates the results obtained for price pressure hypothesis for the pre and post crisis episodes under MCB, respectively. * and *** denotes statistical significance at the 1% and 5% level, respectively

5

Conclusion

This study delves into the behaviour of foreigners in a highly coveted African stock market prior to and post the US subprime crisis. The research is comprehensive as it resorts to a diverse set of hypotheses under investigation. No evidence is found as to foreigners considering the local Treasury Bill market as an alternative investment channel. Such a finding adds lustre to the fact that the liquidity position of Mauritian banks constituted the main driver for the movements in the bank rate. Furthermore, foreign investments are not found to influence the rupee price of the US dollar in the post-crisis period. However, in the pre-crisis era and in the long run, the actions of foreigners triggered a depreciation

Table 5.20  Dependent variable: D(SBM) under SEM pre crisis and post crisis results NET NETt−1 NETt−2 NETt−3 NETt−4 NETt−5 NETt−6 D(SBM(−1)) Constant Adjusted R2 Durbin-Watson Sta F-Statistic

5.9496 (0.4695) −18.700 (−1.4625) 2.1384 (0.1660) 11.756 (0.9150) 8.9113 (0.6826) 22.797 (1.7496) −21.834 (−1.6837) −0.0851 (−0.9660) 0.0087 (2.9488)* 0.0110 2.0295 1.1851

3.7076 (0.1221) 15.643 (0.5079) 26.220 (0.8498) 41.573 (1.3585) −68.597 (−2.2681)*** 40.989 (1.3064) −10.568 (−0.3889) −0.1132 (−0.7086) −0.0034 (−0.4539) 0.0822 1.9953 1.6271

The above table demonstrates the results obtained for price pressure hypothesis for the pre and post crisis episodes under SBM, respectively. * and *** denotes statistical significance at the 1% and 5% level, respectively Table 5.21  Dependent variable: D(SUN) under SEM pre crisis and post crisis results NET NETt−1 NETt−2 NETt−3 NETt−4 NETt−5 NETt−6 D(SUN(−1)) Constant Adjusted R2 Durbin-Watson Sta F-Statistic

7.2321 (0.4319) −5.7131 (−0.3348) −18.887 (−1.1084) 6.7903 (0.3953) −10.532 (−0.6146) 4.8275 (0.2810) −20.841 (−1.2271) −0.0859 (−0.9461) 0.0062 (1.6069) −0.0278 2.0054 0.5637

−27.946 (−0.7190) 38.633 (0.9017) 4.7168 (0.1099) 53.970 (1.2874) −67.055 (−1.6063) 50.352 (1.2107) −48.500 (−1.3077) −0.1122 (−0.7771) −0.0144 (−1.4052) 0.0414 2.0093 1.3024

The above table demonstrates the results obtained for price pressure hypothesis for the pre and post crisis episodes under SUN, respectively

5  The Behaviour of Foreign Investments in the Stock Exchange... 

129

Table 5.22  Dependent variable: D(NMH) under SEM pre crisis and post crisis results NET NETt−1 NETt−2 NETt−3 NETt−4 NETt−5 NETt−6 D(NMH(−1)) Constant Adjusted R2 Durbin-Watson Sta F-Statistic

−19.351 (−1.5261) −25.889 (−1.9980)*** 8.8917 (0.6880) 22.410 (1.7362) −19.721 (−1.5281) −5.055 (−0.3828) −19.721 (−1.5281) −0.2113 (−2.4268)** 0.0243 (3.5533)* 0.0855 2.0517 2.5432

−42.096 (−1.0176) 35.227 (0.7635) 15.643 (0.3390) 29.056 (0.6415) −43.480 (−0.9842) 44.541 (1.0343) −42.262 (−1.0650) 0.0513 (0.3525) −0.0128 (−1.2905) −0.0695 1.9740 0.5450

The above table demonstrates the results obtained for price pressure hypothesis for the pre and post crisis episodes under NMH, respectively. *, ** and *** denotes statistical significance at the 1%, 2% and 5% level, respectively

of the US dollar vis-à-vis the rupee under purchases. Conversely, under sales, they generated an appreciating effect. But, in relative terms, the depreciating effect is considerably much higher than the appreciating effect, substantiating the fact that foreigners were mainly purchasers rather than sellers prior to the onset of the crisis. In addition, no evidence is found in favor of feedback trader hypothesis, showing that foreigners did not act as trend chasers but merely based their purchases on fundamentals. Likewise, the base-broadening effects are impotent.

130 

Economics and Finance in Mauritius

Table 5.23  Dependent variable: D(ROGERS) under SEM pre crisis and post crisis results NET NETt−1 NETt−2 NETt−3 NETt−4 NETt−5 NETt−6 D(ROGERS(−1)) Constant Adjusted R2 Durbin-Watson Sta F-Statistic

−10.814 (−1.0263) −11.993 (−1.1272) −6.8279 (−0.6438) 8.6075 (0.8098) −0.9183 (−0.0851) 8.7521 (0.8144) −10.201 (−0.9580) −0.0380 (−0.4215) 0.0143 (2.5661)** −0.0183 1.9966 0.7025

20.128 (1.3198) 4.2883 (0.2590) −3.9767 (−0.2354) 19.494 (1.1759) −23.799 (−1.4546) 15.841 (0.9924) −14.099 (−0.9702) 0.1304 (0.8988) −0.0073 (−1.7960) 0.0291 1.9626 1.2104

The above table demonstrates the results obtained for price pressure hypothesis for the pre and post crisis episodes under ROGERS, respectively. ** denotes statistical significance at the 2% level

This study contributes to the empirical literature on foreign investment analysis for an upper-income developing country. The analysis is rigorous as it caters for specificity when analyzing flows. The rationale is that the net purchases variable may muffle out important relationships. Such specificity of analysis has been underscored in almost all the hypotheses under investigation. The above findings show that it defeats the purpose of having recourse towards a tax on trade by foreigners in SEM.

5  The Behaviour of Foreign Investments in the Stock Exchange... 

131

Table 5.24  Dependent variable: D(IBL) under SEM pre crisis and post crisis results NET NETt−1 NETt−2 NETt−3 NETt−4 NETt−5 NETt−6 D(IBL(−1)) Constant Adjusted R2 Durbin-Watson Sta F-Statistic

−12.277 (−0.7979) 4.7241 (0.3032) −3.8313 (−0.2465) 19.090 (1.2222) −0.7672 (−0.0481) −0.7977 (−0.0504) −7.4675 (−0.4769) −0.0145 (−0.1616) 0.0044 (0.5409) −0.0444 2.0008 0.2980

11.147 (0.4082) 28.553 (0.9456) −24.520 (−0.7900) 8.4883 (0.2784) 12.456 (0.4231) 38.094 (1.3454) −37.281 (−1.3940) −0.2539 (−1.862) −0.0090 (−1.2527) 0.0211 2.0716 1.1512

The above table demonstrates the results obtained for price pressure hypothesis for the pre and post crisis episodes under IBL, respectively

6

Policy Recommendations

Research which assesses foreign investments on the stock exchange of Mauritius should be based on sales and purchases, separately, in order to generate better and refined results. Studies which simply cling to net purchases are susceptible to being imbued with muffling interactions as to trigger unreliable findings. Besides, since foreigners invest mainly based on fundamentals, any detrimental impacts on the economic conditions of listed Mauritian firms will drive down the levels of foreign investment. Thus, the authorities should put in place proper mechanisms to ensure that in periods of economic downturn, massive foreign disinvestments do not substantially drive down the prices of local equities.

132 

Economics and Finance in Mauritius

Acknowledgement  I would like to thank Mr Kishen Nadassen for providing the weekly reports (issued by Cim Stockbrokers Limited) in case of missing data to ensure full coverage of the time period under consideration.

Appendix Unit Root Tests

Bibliography Barbopoulos, L., Marshall, A., MacInnes, C., & McColgan, P. (2013). Foreign direct investment in emerging markets and acquirers’ value gains. International Business Review, 23, 604–619. Bekaert, G., & Harvey, C. (2003). Emerging market finance. Journal of Empirical Finance, 10, 3–56. Bohn, H., & Tesar, L. L. (1996). U.S. equity investment in foreign markets: Portfolio rebalancing or return chasing? American Economic Review, 86(2), 77–81. Bonser-Neal, C., Jones, S. L., Linnan, D., & Neal, R. (2002). Herding, feedback trading of foreign investors (Working paper). Bloomington: Indiana University. Borensztein, E. R., & Gelos, R. G. (2000). A panic-prone pack? The behavior of emerging market mutual funds (Working paper). Washington, DC: International Monetary Fund. Brennan, M. J., & Cao, H. H. (1997). International portfolio investment flows. Journal of Finance, 52(5), 1851–1880. Choe, H., Kho, B. C., & Stulz, R. M. (1999). Do foreign investors destabilize stock markets? The Korean experience in 1997. Journal of Financial Economics, 54, 227–264. Clark, J., & Berko, E. (1997). Foreign investment fluctuations and emerging market stock returns: The case of Mexico (Federal Reserve Bank of New York, Staff Report No. 24). New York: Federal Reserve Bank of New York. Dahlquist, M., & Robertson, G. (2004). A note on foreigners’ trading and price effects across firms. Journal of Banking and Finance, 28(3), 615–632. De Bondt, W. F. N., & Thaler, R. (1985). Does the stock market overreact? The Journal of Finance, 40(3), 793–805.

5  The Behaviour of Foreign Investments in the Stock Exchange... 

133

DeLong, B. J., Shleifer, A., Summers, L. H., & Waldmann, R. J. (1990). Positive feedback investment strategies and destabilizing rational speculation. Journal of Finance, 45, 379–395. Drogendijk, R., & Blomkvist, K. (2013). Drivers and motives for Chinese outward foreign direct investments in Africa. Journal of African Business, 14(2), 75–84. Edwards, F. R., & Zhang, X. (1998). Mutual funds and stock and bond market stability. Journal of Financial Services Research, 13, 257–282. Evans, G., Russell, G., & Sullivan, R. (2002). An international regulatory framework? In G. Evans, J. Goodman, & N. Lansbury (Eds.), Moving mountains: Communities confront mining and globalisation (pp.  207–222). New York: Zed Books. Fant, F. L. (1999). Investment behavior of mutual fund shareholders: The evidence from aggregate fund flows. Journal of Financial Markets, 2, 391–402. Frankel, J.  A., & Schmuckler, S.  L. (1996). Country fund discounts and the Mexican crisis of December 1994: Did local residents turn pessimistic before international investors? Open Economies Review, 7(Suppl. 1), 511–534. Froot, K. A., O’Connell, P. G. J., & Seasholes, M. S. (2001). The portfolio flows of international investors. Journal of Financial Economics, 59, 151–193. Griffin, J.  M., Nardari, F., & Stulz, R.  M. (2004). Daily cross-border equity flows: Pushed or pulled? Review of Economics and Statistics, 86, 641–657. Griffin, J.M., Harris, J.H. & Topaloglu, S., 2005. The Dynamics of Institutional and Individual Trading. Journal of Finance, 58(6), pp. 2285-2320.. Hamao, Y., & Mei, J. (2001). Living with the ‘enemy’: An analysis of foreign investment in the Japanese equity market. Journal of International Money and Finance, 20, 715–735. Henry, P. B. (2000). Stock market liberalization, economic reform, and emerging market equity prices. Journal of Finance, American Finance Association, 55(2), 529–564. 04. Hou, Y., & Li, S. (2014). The impact of the CSI 300 stock index futures: Positive feedback trading and autocorrelation of stock returns. International Review of Economics and Finance, 33, 319–337. Jackson, A. (2003). The aggregate behavior of individual investors (Working paper). London: London Business School. Karolyi, A. G. (2002). Did the Asian financial crisis scare foreign investors out of Japan? Pacific-Basin Finance Journal, 10, 411–442. Kim, E. H., & Singal, V. (2000). Stock market openings: Experience of emerging markets. Journal of Business, 73, 25–66.

134 

Economics and Finance in Mauritius

Kim, W., & Wei, S. J. (2002). Foreign portfolio investors before and during a crisis. Journal of International Economics, 56, 77–96. Lakonishok, J., Shleirfer, A., & Vishny, R. (1992). The impact of institutional trading on stock prices. Journal of Financial Economics, 32, 23–44. Merton, R.  C. (1987). A simple model of capital market equilibrium with incomplete information. Journal of Finance, 42(3), 483–510. Mijiyawa, A. G. (2014). Policy dynamics and foreign direct investment inflows in Ghana: What are the lessons for West African countries? Journal of African Business, 15(1), 1–12. Neto, D.  G., & Veiga, F.  J. (2013). Financial globalization, convergence and growth: The role of foreign direct investment. Journal of International Money and Finance, 37, 161–186. Odean, T. (1999). Do investors trade too much? American Economic Review, 89, 1279–1298. Richards, A. (2005). Big fish in small ponds: The trading behavior and price impact of foreign investors in Asian emerging equity markets. Journal of Financial and Quantitative Analysis, 40, 1–27. Sakyia, D., Commodoreb, R., & Opoku, E. E. O. (2015). Foreign direct investment, trade openness and economic growth in Ghana: An empirical investigation. Journal of African Business, 16(1–2), 1–15. Shefrin, H., & Statman, M. (1995). The disposition to sell winners too early and ride losers too long: Theory and evidence. Journal of Finance, 40, 777–792. Stiglitz, J. (1998, March 25). Boats, planes and capital flows. Financial Times. Stock Exchange of Mauritius Factbooks. Tuomi, K. (2011). The role of the investment climate and tax incentives in the foreign direct investment decision: Evidence from South Africa. Journal of African Business, 12, 133–147. Warther, V. A. (1995). Aggregate mutual fund flows and security returns. Journal of Financial Economics, 39, 209–235.

6 The Impact of US Subprime Crisis on SEMDEX

1

Introduction

It is widely accepted in the world that global stock markets act as the main driving force for the evolution of stock markets in both developing and emerging markets. For instance, Pagan and Soydemir (2000) found evidence that US stock markets generated robust impacts on Latin American markets with pronounced effects noted in the case of Mexico. SEMDEX is no exception to this state of affairs. An analysis is made to gauge the effects of foreign stocks markets on SEMDEX based on the use of various econometric techniques such as granger causality, vector error correction models, variance decomposition and impulse response analyses. To sieve out any distinctive impacts both prior to and post the crisis, two specific time periods were used  – namely, the pre-crisis period which spanned from January 2001 to January 2008. The post-crisis era occurred during the period February 2008 to January 2009. This latter is the final time period as the study was undertaken in the period January 2009. Nonetheless, the main objective is to bring to light how the crisis might have a significant impact on the performance of SEMDEX in just over one year. The following stock markets were considered for the a­ nalysis: © The Author(s) 2017 I. Ramlall, Economics and Finance in Mauritius, DOI 10.1007/978-3-319-39435-0_6

135

136 

Economics and Finance in Mauritius

DJIA, NASDAQ, FTSE, CAC-40, DAX, NIKKEI 225 and JSE.  The ultimate aim of the analysis is to assess whether international portfolio diversification was subject to bearish forces once the crisis broke out. In fact, during crisis times, stock markets tend to co-move with stronger momentum relative to non-crisis periods so that the ability of investors to harness international portfolio diversification benefits is significantly being muted down.

2

Brief Literature Review

Stock markets exist in a wide variety of countries across the world. As a result of the effects of globalization, financial integration across the world underwent a major push. In that respect, stock market interdependence also witnesses positive momentum through this process as not only local forces but also foreign forces drive the evolution of these markets. Ripley (1973) argued that those markets which were more open to capital investments from other countries experienced greater stock market interdependence. Many studies have chosen to focus on the analysis of a specific major event to gauge whether stock markets behave differently. Taylor and Tonks (1989), for example, assessed the impact of the UK abolition of exchange controls in October 1979 on UK stock market performance with respect to that of the rest of the world. They found cointegration effects with foreign stock markets after the abolition date, adding momentum to free market forces acting as a propelling mechanism in inducing stronger stock market interdependence. In a parallel manner, Roca (1999) assessed indices for a sample of countries – namely, Australia, Hong Kong, Singapore, the UK, the USA, Korea and Taiwan  – with stress being laid on the financial deregulation period in 1983 and 1987 in Australia. Roca (1999) found considerable cointegration with different markets. Allen and Macdonald (1995) focused on 16 countries over a 21-year analysis in an attempt to determine the diversification benefits for Australian investors. Their findings showed that the Australian stock market was highly cointegrated with those of Canada, France, Germany, Hong Kong, Switzerland and the UK. Eun and Shim (1989), King and Wadhwani (1990), Koch and Koch (1991), and Chowdhury (1994) have

6  The Impact of US Subprime Crisis on SEMDEX 

137

all ­investigated the i­nterrelationship among the stock index of distinct countries based on data from the 1980s. They found short-run interrelationships among national equity prices. As shown in Table 6.1, none of the stock markets returns are normally distributed as they exhibit positive/negative skewness and excess kurtosis.

3

Granger Causality (Lead-lag Effects)

Granger causality analysis is widely employed in stock markets analysis with the lag length being derived mainly from the optimal lag length from a VAR model. The results of the granger causality analysis are given in Table  6.2. A statistically significant p-value implies that the stated hypothesis of one variable not granger causing another variable does not hold true. Alternatively stated, a statistically p-value implies that the first variable does granger cause the second variable. Findings show that in the case of the European markets, namely FTSE, DAX and CAC-40, SEMDEX lagged behind these markets. SEMDEX is also found to lag behind BSE. However, in the cases of DJIA, NASDAQ, NIKKEI and JSE, it appears that there was bidirectional causality, which makes the granger causality analysis doubtful. This could be due to exceptionally stronger co-movements experienced during the crisis period. Nevertheless, in the case of DJIA, NASDAQ and JSE, the statistical significance was relatively higher in the case where SEMDEX lagged behind each of these markets. As argued by Brooks (2002), the main significance of granger causality is that it refers only to a correlation between the present value of a variable and the past values of others. To undertake a more robust analysis, cointegation analysis should be applied and before that, it is vital to undertake unit root tests.

4

Unit Root Tests

Recourse is made to Augment Dickey–Fuller and Phillips–Perron unit root tests, with the results being shown in Table  6.3. These tests were conducted on the natural logarithmic of all the series. The findings show

Mean Median Maximum Minimum Std. Dev. Skewness Kurtosis Jarque-­Bera Probability

−9.02E-05 0 0.098388 −0.088483 0.013337 0.07441 10.64752 5048.64 0

FTSE

−3.31E-05 2.18E-05 0.110802 −0.078733 0.01271 0.269378 13.35611 9279.739 0

DJIA −5.32E-05 0.00021 0.141732 −0.091424 0.01723 0.362663 9.299886 3470.191 0

NASDAQ −9.58E-05 0 0.141507 −0.114064 0.016123 −0.102784 10.95362 5462.462 0

NIKKEI

Table 6.1  Summary statistics for stock market returns −2.61E-05 0.000543 0.114022 −0.084923 0.016641 0.214923 8.541972 2666.259 0

DAX −0.000168 0 0.16 −0.114613 0.016268 0.531977 13.95841 10460.17 0

CAC-40

0.000563 0.000542 0.107711 −0.104854 0.014074 0.066975 10.7352 5164.662 0

JSE

0.000563 0.000978 0.082541 −0.111385 0.016655 −0.404367 7.768028 2018.201 0

BSE

SEMDEX 0.000579 0.000337 0.065043 −0.063011 0.007748 −0.101233 22.18122 31751.89 0

Table 6.2  Results for granger causality January 2008–January 2009 Null hypothesis FTSE does not GC SEMDEX SEMDEX does not GC FTSE DJIA does not GC SEMDEX SEMDEX does not GC DJIA NASDAQ does not GC SEMDEX SEMDEX does not GC NASDAQ NIKKEI does not GC SEMDEX SEMDEX does not GC NIKKEI DAX does not GC SEMDEX SEMDEX does not GC DAX CAC40 does not GC SEMDEX SEMDEX does not GC CAC40 JSE does not GC SEMDEX SEMDEX does not GC JSE BSE does not GC SEMDEX SEMDEX does not GC BSE

F-statistic 11.7163 1.1649 24.3822 3.0504 22.5733 4.2827 3.9226 3.9570 18.2657 1.6682 10.4308 1.2108 7.2523 3.0288 11.6637 1.3845

Probability 1.3E-05* 0.3135 1.9E-10* 0.0490* 8.8E-10* 0.0147* 0.0091* 0.0087* 3.7E-08* 0.1905 1.7E-06* 0.3062 0.0001* 0.0299* 3.3E-07* 0.2478

*Denotes statistical significance at the 5% level Table 6.3  Unit root tests ADF unit root tests Jan 2001 to Jan 2008 Level First difference FTSE −1.3149 −45.2251* DJIA −1.1367 −43.7599* NASDAQ 0.7746 −23.3460* NIKKEI −1.1267 −43.6904* DAX −0.5989 −43.5648* CAC-40 −1.3508 −45.9661* JSE 0.2672 −43.2220* BSE 1.3740 −30.8670* SEMDEX 2.6394 −32.4618* Phillips-Perron unit root tests Jan 2001 to Jan 2008 FTSE Level First difference DJIA −1.0934 −45.8959* NASDAQ −1.0810 −43.8553* NIKKEI −1.4087 −43.1264* DAX −1.0882 −43.7151* CAC-40 −0.5562 −43.6365* JSE −1.2942 −46.6402* BSE 0.3655 −43.3058* SEMDEX 1.3124 −39.8499* 2.7341 −32.5677* *Significant at 1% level

Jan 2008 to Jan 2009 Level First difference −1.1925 −7.9713* −0.2888 −15.2145* −0.7040 −18.8703* −0.8899 −13.6759* −0.9875 −17.2637* −0.9831 −18.4589* −0.8740 −16.9252* −0.8740 −15.6497* 0.3677 −14.5324* Jan 2008 to Jan 2009 Level First difference −0.9769 −17.9887* −0.3080 −19.9339* −0.4193 −18.9950* −0.8460 −16.2131* −0.8788 −17.2949* −0.7024 −18.6078* −0.6371 −17.0476* −0.8906 −15.5797* 0.2486 −14.4955*

140 

Economics and Finance in Mauritius

that the first differences of the logarithmic transformations of the series were stationary. Thus, the series are said to be integrated to order one I (1), making them eligible for the cointegration analysis.

5

Cointegration Analysis

The cointegration tests results are depicted in Table 6.4 in the appendix section under a bivariate analysis. Based on the pre- and post-crisis period analysis, the optimal lag (m,n) is used with m representing the lag derived for the pre- and post-crisis era, respectively. Technically speaking, the number of cointegrating vectors are determined from two sources, namely, the eigenvalue and the trace statistics. However, inconclusive results are obtained in the case that the results are not the same from these two analyses. For instance, inconclusive (0,1) means that trace statistic found one cointegrating equation while eigenvalue found no presence of a cointegrating equation. Besides, since a bivariate analysis is undertaken, results are not interpretable in the case two cointegrating vectors are identified (synonymous to full rank matrix). Recourse is made towards a group analysis based on the number of identified cointegrating vectors. FTSE, DAX and CAC-40 were classified in Group 1 because one cointegrating equation prevailed in both periods. Vector Error Correction Models were implemented for Group 1. Group 2 consisted of markets which had one cointegrating vector in the post-crisis era and comprised of DJIA and BSE. Finally, group 3 consisted of markets which had no cointegration in both periods such as NIKKEI 225 and JSE. Results for the various groups are depicted in Tables 6.5, 6.6 and 6.7 in the appendix section.

6

Results

For Group 1, which consisted of FTSE, CAC-40 and DAX, an increase is noted for the long-run coefficient for DAX while a fall is noted for FTSE and CAC-40. Nonetheless, for all stock markets, an increase in the speed of adjustment coefficient is noted. In fact, dur-

6  The Impact of US Subprime Crisis on SEMDEX 

141

ing the crisis period, it occurred that there was a shift in the speed of adjustments in the VECM tests for FTSE, CAC-40 and DAX, shifting from a positive low value figure to a negative and relatively higher figure. The existence of cointegration for DJIA and BSE (Group 2) post the crisis showed clearly that diversification benefits had been undermined. In the case of Group 3 (JSE and NIKKEI 225), VAR is used due to absence of cointegration. Neither NIKKEI 225 nor JSE impacted on SEMDEX prior to the onset of the crisis. Post the crisis, JSE influenced SEMDEX. Intriguingly, evidence is also found as to SEMDEX impacting on these markets. Such bi-causality effects during the crisis merely reflected the degree to which the crisis had induced greater synchronicity among markets.

7

Conclusion

The very presence of the cointegration of SEMDEX with foreign stock markets signifies undermined benefits from international portfolio diversification. The current analysis proves that financial market integration is not only time-varying, but also can be dependent on the occurrence of a crisis in which case stock markets tend to have robust coupling effects. Consequently, the gradual foreign disinvestments noted in SEM post the crisis could be attributed to evaporated benefits from international portfolio diversification.

8

Policy Recommendations

Since portfolio diversification benefits faded away on the local stock market due to stronger co-movements with foreign markets during crisis conditions, it is recommended that some sound measures be established to ensure that foreign disinvestments do not unleash substantial price declines. For instance, sectoral foreign investment limits could be established so that in the case of massive foreign disinvestments, no specific sector bears most of the full brunt of such detrimental impacts.

142 

Economics and Finance in Mauritius

Appendix Table 6.4  Cointegration results under monthly horizon analysis

H0 Jan 2001–2008 Jan 2008–2009 Jan 2001–2008

H1

Trace Eigenvalue statistic

Max Eigenvalue statistic

FTSE and SEMDEX (Optimal lags = 2,1) r = 0 r > 0 0.0537 15.1375 15.0847* r = 1 r > 1 0.0001 0.0528 0.0528 r = 0 r > 0 0.0691 19.7650** 19.5749* r = 1 r > 1 0.0006 0.1900 0.1900 DJIA and SEMDEX(Optimal lags = 2,3) r = 0 r > 0 0.0080 19.6558* 14.6139*

r = 1 r > 1 0.0027 5.0419* 5.0419* Jan r = 0 r > 0 0.0722 20.5061** 20.4818** 2008–2009 r = 1 r > 1 8.93–05 0.0243 0.0243 NASDAQ and SEMDEX(Optimal lags = 2,2) Jan r = 0 r > 0 0.0087 20.8654* 15.4947* 2001–2008 r = 1 r > 1 0.0028 5.0914* 3.8414* Jan r = 0 r > 0 0.0468 13.1584 13.0906 2008–2009 r = 1 r > 1 0.0002 0.0678 0.0678 NIKKEI and SEMDEX(Optimal lags = 2,3) Jan r = 0 r > 0 0.0075 16.4421* 13.5699 2001–2008 r = 1 r > 1 0.0015 2.8721 2.8721 Jan r = 0 r > 0 0.0522 11.6951 11.6939 2008–2009 r = 1 r > 1 4.43–06 0.0012 0.0012 DAX and SEMDEX(Optimal lags = 2,2) Jan r = 0 r > 0 0.0092 17.8214* 16.8008* 2001–2008 r = 1 r > 1 0.0005 1.0205 1.0205 Jan r = 0 r > 0 0.0776 22.0902* 22.0603* 2008–2009 r = 1 r > 1 0.0001 0.0298 0.0298 CAC-40 and SEMDEX (Optimal lags = 2,3) Jan r = 0 r > 0 0.0087 17.7472* 15.9036** 2001–2008 r = 1 r > 1 0.0010 1.8435 1.8435 Jan r = 0 r > 0 0.0697 19.7502** 19.7426** 2008–2009 r = 1 r > 1 2.79–05 0.0076 0.0076 JSE and SEMDEX(Optimal lags = 2,3) Jan r = 0 r > 0 0.0044 10.9552 8.0733 2001–2008 r = 1 r > 1 0.0016 2.8819 2.8819 Jan r = 0 r > 0 0.0258 7.1494 7.1386 2008–2009 r = 1 r > 1 3.96–05 0.0108 0.0108

Cointegrating equations 1 1

Inconclusive (2,0) 1

2 Zero

Inconclusive (1,0) Zero

1 1

1 1

Zero Zero (continued)

6  The Impact of US Subprime Crisis on SEMDEX 

143

Table 6.4 (continued)

H0

H1

Trace Eigenvalue statistic

Max Eigenvalue statistic

Cointegrating equations

BSE and SEMDEX(Optimal lags = 2,3) Jan r = 0 r > 0 0.0054 17.4019* 2001–2008 r = 1 r > 1 0.0041 7.5088** Jan r = 0 r > 0 0.0770 22.1158* 2008–2009 r = 1 r > 1 0.0008 0.2343

7.5088** 21.8814* 0.2343

Critical values for statistical significance Critical values Trace statistic H0 ; H1 5% 1% r=0;r>0 15.4947 19.9371 r=1;r>1 3.8414 6.6348

Max Eigenvalue statistic 5% 1% 14.2646 18.52 3.8414 6.6348

9.8930

Inconclusive (2,0) 1

* Significant at 1% level. ** Significant at 5% level.

Table 6.5  Results for Group 1 (FTSE, DAX and CAC-40) Results for FTSE-Ex ante Cointegration equation 1 ΔSEMDEX t–1 ΔSEMDEX t–2 ΔFTSE t–1 ΔFTSE t–2 Constant Adjusted R2 F-Statistic

∆SEMDEX 0.0009 (2.6194)* 0.2592 (10.9802)* −0.0162 (−0.6871) 0.0034 (0.2841) −0.0101 (−0.8468) 0.0006 (4.8185)* 0.0696 27.9461

ΔFTSE 0.002 (2.8813)* 0.0375 (0.8150) −0.0366 (−0.7943) −0.0626 (−2.6659)* −0.0286 (−1.2194) 2.26E-05 (0.0849) 0.0069 3.5152 (continued)

144 

Economics and Finance in Mauritius

Table 6.5 (continued) Results for FTSE-Ex ante Cointegration equation 1 ΔSEMDEX t–1 1.0000

ΔFTSEt–1 −1.9542 (−3.4296)*

Constant 10.2271

ΔSEMDEX −0.0419 (−3.9478)* 0.0554 (0.8845) 0.0730 (1.8244) −0.0014 (−1.6257) 0.0847 9.3316

ΔFTSE 0.0164 (0.9596) −0.1364 (−1.4085) −0.0530 (−0.8246) −0.0019 (−1.3498) 0.009 1.8193

ΔFTSEt–1 −1.4481 (−11.1299)*

Constant 5.2811

ΔSEMDEX 0.0010 (2.7474)* 0.2586 (10.9531)* −0.0164 (−0.6980) 0.0009 (0.1098) −0.0025 (−0.2874) 0.0006 (4.8243)* 0.0695 27.9419

ΔDAX 0.0029 (3.0118)* 0.0223 (0.3509) −0.0083 (−0.1314) −0.0280 (−1.1924) −0.0001 (−0.0050) 0.0001 (0.3099) 0.0031 2.1402

ΔDAX t–1 −0.9790 (−3.1341)*

Constant 1.8417

Results for FTSE-Ex post Cointegration equation 1 ΔSEMDEX t–1 ΔFTSE t–1 Constant Adjusted R2 F-Statistic Cointegration equation 1 ΔSEMDEX t–1 1.0000 Results for DAX-Ex ante Cointegration equation 1 ΔSEMDEX t–1 ΔSEMDEX t–2 ΔDAX t–1 ΔDAX t–2 Constant Adjusted R2 F-Statistic Cointegration equation 1 ΔSEMDEX t–1 1.0000

(continued)

6  The Impact of US Subprime Crisis on SEMDEX 

145

Table 6.5 (continued) Results for FTSE-Ex ante Results for DAX-Expost Cointegration equation 1 ΔSEMDEX t–1 ΔSEMDEX t–2 ΔDAX t–1 ΔDAX t–2 Constant Adjusted R2 F-Statistic Cointegration equation 1 ΔSEMDEX t–1 1.0000

ΔSEMDEX −0.0560 (−4.4700)* 0.0317 (0.5249) −0.1207 (−2.0521)** 0.1057 (2.6901)* 0.0541 (1.3858) −0.0014 (−1.6673) 0.1352 9.4473

ΔDAX 0.0153 (0.7329) −0.1155 (−1.1442) −0.1162 (−1.1849) −0.0390 (−0.5950) −0.0566 (−0.8698) −0.0027 (−1.9411) 0.0092 1.5067

ΔDAX t–1 −1.2436 (−14.2282)*

Constant 3.4254

ΔSEMDEX 0.0009 (2.7512)* 0.2588 (10.9615)* −0.0160 (−0.6788) −0.0004 (−0.0492) −0.0083 (−0.8992) 0.0006 (4.8183)* 0.0699 28.0865

ΔCAC-40 0.0024 (2.8948)* 0.0381 (0.6410) −0.0288 (−0.4838) −0.0823 (−3.5017)** −0.0332 (−1.4157) −3.61E-05 (−0.1048) 0.0095 4.4659

Δ CAC-40 t–1 −1.4897 (−3.2987)*

Constant 6.0007

Results for CAC-40-Exante Cointegration equation 1 ΔSEMDEX t–1 ΔSEMDEX t–2 Δ CAC-40 t–1 Δ CAC-40 t–2 Constant Adjusted R2 F-Statistic Cointegration equation 1 ΔSEMDEX t–1 1.0000

(continued)

146 

Economics and Finance in Mauritius

Table 6.5 (continued) Results for FTSE-Ex ante Results for CAC-40-Expost Cointegration equation 1 ΔSEMDEX t–1 ΔSEMDEX t–2 ΔSEMDEX t–3 Δ CAC-40 t–1 Δ CAC-40 t–2 Δ CAC-40 t–3 Constant Adjusted R2 F-Statistic Cointegration equation 1 ΔSEMDEX t–1 1.0000

ΔSEMDEX

ΔCAC-40

−0.0600 (−4.2215)* 0.0458 (0.7343) −0.1166 (−1.8844) −0.0117 (−0.1939) 0.0709 (1.8193) 0.0440 (1.1363) −0.0141 (−0.3751) −0.0015 (−1.7472) 0.1063 5.5885

0.0030 (0.1215) −0.0763 (−0.7022) −0.1194 (−1.1083) −0.1453 (−1.3743) −0.1466 (−2.1608)** −0.1118 (−1.6562) −0.0808 (−1.2324) −0.0034 (−2.2360)** 0.0324 2.2924

Δ CAC-40 t–1 −1.1635 (−14.1394)*

Constant 2.3097

*Denotes statistical significance at 1% level * and ** denote statistical significance at 1% and 5% level, respectively Table 6.6  Results for Group 2 (DJIA and BSE) Results for DJIA-Expost Cointegration equation 1 ΔSEMDEX t–1 ΔSEMDEX t–2 ΔSEMDEX t–3 ΔDJIA t–1

ΔSEMDEX −0.0617 (−4.2094)* 0.0961 (1.6115) −0.0891 (−1.4863) −0.0081 (−0.1413) 0.1321 (3.1664)*

ΔDJIA 0.0117 (0.4954) −0.1399 (−1.4463) −0.1865 (−1.9182) −0.0850 (−0.9170) −0.1684 (−2.4879)* (continued)

6  The Impact of US Subprime Crisis on SEMDEX  Table 6.6 (continued) Results for DJIA-Expost ΔDJIA t–2

−0.0083 (−0.2100)

−0.1432 (−2.1177)**

ΔDJIA t–3

0.0314 (0.7817) −0.0013 (−1.6295) 0.1537 8.0074

0.1407 (2.1550)** −0.0028 (−2.0233)** 0.0746 4.1094

ΔDJIAt–1 −1.2790 (−14.1312)*

Constant 4.4971

ΔSEMDEX −0.0622 (−4.5723)* 0.0870 (1.4412) −0.1537 (−2.6026)* −0.0430 (−0.7210) 0.0004 (0.0143) 0.0817 (2.5430)* −0.0163 (−0.5022) −0.0015 (−1.8153) 0.1196 6.2422

ΔBSE 0.0131 (0.4881) 0.1098 (0.9168) −0.2091 (−1.7839) −0.0546 (−0.4613) 0.0436 (0.6756) −0.0042 (−0.0665) −0.0730 (−1.1286) −0.0031 (−1.8289) 0.0015 1.0614

ΔDJIAt–1 −0.8286 (−14.7544)

Constant 0.4934

Constant Adjusted R2 F-Statistic Cointegration equation 1 ΔSEMDEX t–1 1.0000 Results for BSE-Expost Cointegration equation 1 ΔSEMDEX t–1 ΔSEMDEX t–2 ΔSEMDEX t–3 ΔBSE t–1 ΔBSE t–2 ΔBSE t–3 Constant Adjusted R2 F-Statistic Cointegration equation 1 ΔSEMDEX t–1 1.0000

* denotes statistical significance at 1% level. * and ** denote statistical significance at 1% and 5% level, respectively.

147

148 

Economics and Finance in Mauritius

Table 6.7  Results for Group 3 (NIKKEI 225 and JSE) Results for NIKKEI225-Exante ΔSEMDEX t–1 ΔSEMDEX t–2 ΔNIKKEI225 t–1 ΔNIKKEI225 t–2 Constant Adjusted R2 F-Statistic

ΔSEMDEX 0.2643 (11.2094)* −0.0116 (−0.4961) −0.0123 (−1.2361) 0.0026 (0.2705) 0.0006 (4.7554)* 0.0670 33.3339

ΔNASDAQ −0.0293 (−0.5263) −0.0296 (−0.5320) −0.0302 (−1.2811) −0.0263 (−1.1159) 0.0001 (0.3407) −0.0002 0.8826

ΔSEMDEX 0.1152 (1.7956) −0.1345 (−2.0908)** −0.0345 (−0.5317) 0.0204 (0.5879) 0.0592 (1.7385) 0.0329 (0.9535) −0.0014 (−1.5961) 0.0180 1.8250

ΔNASDAQ 0.2940 (2.5055)** −0.2549 (−2.1660)** −0.1930 (−1.6273) −0.0394 (−0.6200) −0.1237 (−1.9845)** −0.0651 (−1.0314) −0.0029 (−1.7380) 0.0652 4.1387

ΔSEMDEX 0.2636 (11.1723)* −0.0105 (−0.4477) 0.0033 (0.2945) −0.0055 (−0.5155)

ΔJSE −0.0062 (−0.1263) −0.0155 (−0.3155) −0.0150 (−0.6413) 0.0251 (1.0688)

Results for NIKKEI225-Expost ΔSEMDEX t–1 ΔSEMDEX t–2 ΔSEMDEX t–3 ΔNIKKEI225 t–1 ΔNIKKEI225 t–2 ΔNIKKEI225 t–3 Constant Adjusted R2 F-Statistic Results for JSE-Ex ante ΔSEMDEX t–1 ΔSEMDEX t–2 ΔJSE t–1 ΔJSE t–2

(continued)

6  The Impact of US Subprime Crisis on SEMDEX  Table 6.7 (continued) Results for NIKKEI225-Exante Constant Adjusted R2 F-Statistic

0.0006 (4.7455)* 0.0663 32.9949

0.0007 (2.4544)* −0.0012 0.4256

ΔSEMDEX 0.0922 (1.4688) −0.1289 (−2.0943)** 0.0277 (0.4486) 0.0712 (1.8400) 0.1364 (3.5487)* 0.0004 (0.0109) −0.0014 (−1.5611) 0.0577 3.7577

ΔJSE −0.0908 (−0.8904) −0.2553 (−2.5545)*** −0.1288 (−1.2820) −0.0521 (−0.8293) −0.0329 (−0.5275) −0.0230 (−0.3635) −0.0020 (−1.4392) 0.0246 2.1373

Results for JSE-Expost ΔSEMDEX t–1 ΔSEMDEX t–2 ΔSEMDEX t–3 ΔJSE t–1 ΔJSE t–2 ΔJSE t–3 Constant Adjusted R2 F-Statistic

* Denotes statistical significance at 1% level ** Denotes statistical significance at 1% level * and *** denote statistical significance at 1% and 2% level, respectively

149

150 

Economics and Finance in Mauritius

Bibliography Allen, D. C., & Macdonald, G. (1995). The long-run gains from international equity diversification: Australian evidence from cointegration tests. Applied Financial Economics, 5, 33–42. Brooks, C. (2002). Introductory econometrics for finance. Cambridge/New York: Cambridge University Press. Chowdhury, A. R. (1994). Stock market interdependencies: Evidence from the Asian NIEs. Journal of Macroeconomics, 16, 629–651. Eun, C. S., & Shim, S. (1989). International transmission of stock market movements. Journal of Financial and Quantitative Analysis, 24, 241–256. King, M. A., & Wadhwani, S. (1990). Transmission of volatility between stock markets. Review of Financial Studies, 3, 5–33. Koch, P. D., & Koch, T. (1991). Evolution in dynamic linkages across daily national stock indexes. Journal of International Money and Finance, 10, 231–251. Pagan, J. A., & Soydemir, G. (2000). On the linkages between equity markets in Latin America. Applied Economics Letters, 7, 207–210. Ripley, D. M. (1973). Systematic elements in the linkage of national stock market indices. The Review of Economics and Statistics, 55(3), 356–361. MIT Press. Roca, E. D. (1999). Short-term and long-term price linkages between the equity markets of Australia and its major trading partners. Applied Financial Economics, 9(5), 501–511. Taylor and Francis Journals. Taylor, M. P., & Tonks, I. (1989). The internationalisation of stock markets and the abolition of U.K. Exchange Control. The Review of Economics and Statistics, 71(2), 332–336. MIT Press.

Part IV Monetary Policy in Mauritius

7 Monetary Policy Analysis

1

Introduction

Established in September 1967, the Bank of Mauritius is now responsible for the conduct of monetary policy in Mauritius. In a rigorous rating exercise devised by Ramlall (2015a), the Bank of Mauritius was ranked 55th out of 114 central banks. Later, Ramlall (2016) proved that the quality of a central bank is an important determinant of both sovereign ratings and credit default swaps spreads across the world. It is therefore apparent that knowledge about a country’s central bank is important, in particular, its monetary policy mechanism. Interested readers can refer to the book entitled The Mauritian Economy: A Reader for more historical background information on monetary policy. This section will focus more on the present practice of monetary policy in Mauritius. The current monetary policy framework in Mauritius is based on the Key Repo Rate and is described below. The distinct tools of monetary policy are discussed. A schema of the general process of interactions among households, banks and public debt is also represented so that the root cause of the problems which trigger structural excess liquidity can be addressed via structural measures. Finally, the various transmission mechanisms and their effectiveness are covered in the last part. © The Author(s) 2017 I. Ramlall, Economics and Finance in Mauritius, DOI 10.1007/978-3-319-39435-0_7

153

154 

2

Economics and Finance in Mauritius

 urrent Monetary Policy Framework C in Mauritius

In December 2006, the Bank of Mauritius adopted a new framework for the conduct of monetary policy in Mauritius involving two central pillars. While the first pillar focused on economic analysis which captures short- and medium-term risks to price stability, the second pillar gauges monetary developments and the related long-term risks to price stability. Above all, the previously used Lombard Rate was replaced by the Key Repo Rate, effectively set at 8.50% in December 2006. To guide on its monetary policy stance, the central bank resorts to the Taylor Rule. For instance, in April 2014, the Bank began to cling to Taylor Rule estimates as a guide to monetary policy making. For the year 2014–15, the monetary policy stance in Mauritius adhered to a rather accommodative mode, unleashed by dwindling inflation driven by falling food and commodity prices. In that respect, the Key Repo Rate was unchanged at 4.65%. To enhance the effectiveness of the transmission mechanism of monetary policy, recourse is being made towards the development of a new framework of monetary policy in Mauritius. However, as at end of February 2016, nothing has yet been disclosed by the Bank of Mauritius with respect to the new operational framework of monetary policy. Section 4(1) of the Bank of Mauritius Act 2004 points out that the chief aim of the Bank is to maintain price stability and to promote an orderly and balanced economic development. To conduct monetary policy in Mauritius, a Monetary Policy Committee, which was established on 23 April 2007, is used to act as the core element in such an analysis and implementation. The Monetary Policy Committee assesses the upside risk of inflation to then adjust the Key Repo Rate, if such a need arises. More specifically, the Bank of Mauritius clings to the Key Repo Rate to signal its monetary policy stance which eventually transmits through to the savings and lending rates of banks via changes in the banks’ Prime Lending Rates. Technically speaking, the committee resorts to both local and foreign developments when gauging the trade-off between economic growth and inflation in order to determine the future direction of the Key Repo Rate. Once a decision is made about the Key Repo Rate, the latter is disseminated to the public on the same day via a communiqué.

7  Monetary Policy Analysis 

155

The central bank governor, who also acts the chairperson of the Monetary Policy Committee, then holds a press conference to furnish arguments pertaining to the change or unchanged monetary policy stance. Two weeks after the decision, more details, such as the minutes of the meeting and voting patterns, are made available to the public. In normal circumstances, the Monetary Policy Committee meets on a quarterly basis but it can also be convened as and when required. Liquidity management in the Mauritian monetary circuit is vital not only to manage the excess liquidity caused by natural transactions but also the excess liquidity triggered by the actions of the Bank of Mauritius itself. For instance, in 2015, following the purchase of foreign currencies, the central bank had to intervene to sterilize the effects of such interventions. The central bank issued securities to bring down excess liquidity in the monetary circuit. Such excess liquidity also acts as a hurdle in the effective transmission mechanism of monetary policy, chiefly when the central bank is adopting a contractionary monetary policy stance to curtain any feasible inflationary pressures. As a matter of fact, excess liquidity constitutes a rather significant problem which has always been plaguing the transmission mechanism of monetary policy in Mauritius. It is a known fact that the Key Repo Rate is disconnected from the local money market interest rates. Indeed, short-term interest rates, such as yields on Treasury Bills and interbank market rates deviate significantly from the Key Repo Rate on the back of excess liquidity in the banking system. To rekindle the transmission channel of monetary policy, it becomes imperative to mop up the excess liquidity in the banking system. In its 2015 annual report, the Bank of Mauritius stated “The MPC also commended the proposal for a new monetary policy framework, which would steer money market interest rates around the policy rate and enable an effective transmission of monetary policy.” (Bank of Mauritius Annual Report 2015) Such a statement signifies that the effective transmission of monetary policy occurs when changes in policy rates are fully transmitted to market interest rates. However, it is a fact that in Mauritius changes in policy rates, which eventually impound on prime lending rates and on deposit and borrowing rates, do not actually generate changes in the level of either deposits or loans. Such a state of affairs implies that the

156 

Economics and Finance in Mauritius

real effectiveness of the monetary policy transmission mechanism will manifest only when real repercussions are noted on deposits and loans. To see these impacts, there is a need to further develop the local market structures and induce financial literacy, without which monetary policy changes will continue to be ineffective, at least under the interest channel of the transmission mechanism. The high interest rate spreads which have been prevailing for many years also impede on an effective transmission mechanism of monetary policy because banks have already availed themselves of the considerable margin for manoeuvre in their interest rate policies. For instance, in its 2014 annual report, the Bank of Mauritius pointed out that “The Monetary Policy Transmission Mechanism, already weakened by the excess liquidity that had been plaguing the system for the past few years, broke down when some banks adjusted their savings rates downwards although the MPC had decided to leave the KRR on hold.” (Bank of Mauritius Annual Report 2014) Such high interest rate spreads will continue to manifest as long as Mauritians behave like conventional depositors, providing substantial funding power to banks which thereby neutralize the effective transmission mechanism of monetary policy. If Mauritians were financially more educated and sophisticated enough as to consider alternative assets, this would definitely gnaw at the funding power of banks which would then provide higher deposit rates, automatically squeezing out the already burgeoning interest rate spreads.

3

 ools for Monetary Policy T Implementation

There are presently two main types of monetary policy tools wielded by the Bank of Mauritius: direct and indirect tools. The direct tool is the minimum cash ratio reserve requirement which is applied on banks’ deposits. For instance, in May 2014, the Bank of Mauritius scaled up the cash reserve ratio from 8.0% to 9.0% to obviate the problem of excess liquidity in the system. The indirect tools consist of open market operations, standing facility and repo/reverse repo transactions. The open market operations are usually effected via Bank of Mauritius instruments in view of mopping up excess liquidity in the monetary circuit.

7  Monetary Policy Analysis 

157

A Repo (also known as repurchase agreement-indirect tool) represents the case when a bank gives out its securities as collaterals in return for money to a central bank to be held for a short-term period after which the securities are being bought back again by the bank by paying back the money plus applied interest thereon. The applied interest rate on such a facility is labelled as the repo rate. The repo rate is like a short-term overdraft rate applied to a bank which is in dire need of cash. A reverse repo (also known as reverse repurchase agreement) arises when a bank (which is usually the borrower) now becomes the lender to the central bank. Alternatively stated, the reverse repo rate reflects the interest rate that a bank obtains when it lends money to the central bank. Technically speaking, the repo rate is higher than the reverse repo rate. It is of paramount significance to note that only high-quality securities are being considered as collaterals for the repo and reverse repo transactions. It is most often the case that none of the tools of monetary policy are used in isolation. For example, in June 2010, the Bank of Mauritius mopped up excess liquidity in the monetary circuit through the issuance of Bank of Mauritius Bills, Special Deposits Facility, reverse repurchase transactions and scaling up the minimum Cash Reserve Ratio.

4

 chema of the Interactions Among S Households, Banks and Public Debt in Mauritius

This section focuses on the structural aspects of the Mauritian economy in which consumers or households are paying the costs while banks and the government are both gaining via higher profits and the lower costs of public debt (Fig. 7.1). Most importantly, such a state of affairs is actually impeding on a fluid interest rate channel of monetary policy transmission mechanism in Mauritius. As a matter of fact, Mauritian households are conventional depositors by virtue of risk aversion, a lack of knowledge about equities and lack of alternative assets that are easily understood by households. Consequently, banks avail themselves of easy funding structures based on stable deposits from households which they use under the ­financial intermediation process to generate high profits, propelled by very high interest rate spreads. Thus, banks tend to be awash with funds which

Key Repo Rate

Prime Lending Rates

Exorbitant borrowing costs to households

(RL)

LOANS

Conventional banking

PUBLIC DEBT

TREASURY BILLS

Low borrowing costs but skewed

Tilted towards shortterm maturity

Fig. 7.1  Schema of the interactions among households, banks and public debt in Mauritius (Source: Author’s Illustration)

(RD)

Households Government

DEPOSIT S

Financial intermediation

Strong interest rate spread inducing higher profits but weak innovation

Financial illiteracy, lack of alternative assets

BANKS

INVESTMENTS

158  Economics and Finance in Mauritius

7  Monetary Policy Analysis 

159

they need to invest to remunerate the depositors. Consequently, banks often participate aggressively in the local Treasury Bill markets in order to drive down the yields. Such bearish yields benefit the government via undermined cost of public debt and this is bolstered by the fact that the redemption profile of the Mauritian public debt is skewed towards short-­ term maturity. Based on exorbitant interest rate spreads which stimulate higher banks’ profits, this unleashes further funds to the banks which have added incentives to participate aggressively in the bills auctions. The above state of affairs has prevailed for many decades and has undermined the interest rate channel of monetary policy. On the deposits side, in spite of changes in the Key Repo Rate, deposits levels are noted to be insensitive to these changes. To bring about the change, it is vital to adopt vehement financial literacy campaigns, provide stronger fiscal incentives to households1 who invest in local equities and expand the level of alternative assets available to households. The new alternatives will have to have a comparable risk to that of bank deposits as most Mauritian households tend to be highly risk-averse. On the loans side, the same insensitivity is noted with respect to the Key Repo Rate. It is high time to curtail the really high interest rate spreads which benefit banks in Mauritius compared to even European and American banks.

5

 onetary Policy Channels M of Transmission Mechanism in Mauritius

Technically speaking, there are five main channels pertaining to the transmission mechanism of monetary policy in any country: the interest rate channel, the exchange rate channel, the credit channel, the assets price channel and the expectations channel. Each of these distinct ­channels is being discussed with relevance for the Mauritian economy. The strongest channels are discussed first before we move on to discuss the weakest channels of the transmission mechanism of monetary policy  Since dividends are already tax-free, the fiscal incentives should take more like finer income tax rates being applied to households who hold a certain stipulated level, say 10%, of equity holdings. 1

160 

Economics and Finance in Mauritius

in Mauritius. It is of paramount significance to note that the power of monetary policy lies in the short run only by virtue of the money neutrality conditions which prevail in the long run. In that respect, the definition of effectiveness of the transmission mechanism will hinge on two conditions, the degree of instantaneous impacts and on the extent of effects on the real variables (Fig. 7.2). To gauge the effectiveness of each channel, recourse is made to a 10-scale rating process. In the case of the exchange rate and interest rate channels, empirical studies are used such as Ramlall (2015b) and Tsangarides (2010). In the case of the expectation channel, a benchmark is used, namely whether the central bank explicitly and formally adheres to inflation targeting. Added to this, the extent to which interest rate changes occur in a gradualist manner is being considered together with inflation expectations surveys. Under the credit channel, there are two variants of the channels, namely the bank-lending channel and the balance sheet channel. The latter is expected to be particularly robust for both households and corporates in Mauritius because a hike in interest rates trails generates a rise in the repayment costs. In the case of the ­bank-­lending channel, an instant increase in the cash ratio reserve requirement will automatically constrain the level of credit availability. Under the asset Credit channel Exchange rate channel

Expectation channel

Interest rate channel

Asset price channel

Fig. 7.2  Effectiveness of the various channels of transmission mechanism in Mauritius (Source: Author’s illustration)

7  Monetary Policy Analysis 

161

price channel, a weak effect is posited to manifest itself based on lack of financial knowledge by Mauritians who do not heavily hold equities, let alone strong level of inside ownership among listed firms.

5.1

Strong Exchange Rate Transmission Channel

It is of paramount significance to note that monetary policy is inherently linked to the exchange rate system of a country. When it comes to Mauritius, being a small open economy, undeniably, the exchange rate channel of the transmission mechanism of monetary policy should be highly robust compared to any other channel of the transmission mechanism of monetary policy. The exchange rate channel works as follows. Suppose the Bank of Mauritius scales up the Key Repo Rate. Such a hike in interest rates will lure high capital flows so that there is a robust demand for the Mauritian rupee, which causes the currency to appreciate. The latter development subsequently impacts on the trade channel by deterring exports and inducing imports, both of which paves the way towards an undermined aggregate demand function. In that respect, it is obvious that the previously targeted policy of suppressing economic activities to subdue inflationary pressures is eventually being assisted via compatible impacts on the trade channel which eventually ricochet on the aggregate demand function.2 Ramlall (2015) pointed out that “The weights obtained from estimation are 0.83 and 0.17, respectively, for the exchange rate and interest rate components. The relatively high weight econometrically obtained for the exchange rates bodes well with the fact that the exchange rate channel constitutes a crucial force which impacts on aggregate demand in a small open economy like Mauritius. Similarly, the low weight obtained in the case of the interest rate channel is compatible with the fact that transmission mechanism via interest rate is poor in Mauritius. Such a finding is substantiated by impotency of interest rates on both deposits and loans in Mauritius.” (pp. 235–271)

 After all, AD = C + I + G + (X − M).

2

162 

5.2

Economics and Finance in Mauritius

Weak Interest Rate Transmission Channel

Following Ramlall (2015), ratios of 0.17, 2 marks are attributed to the interest rate channel following findings of a rather impotent interest rate transmission channel. In a parallel manner, Tsangarides (2010) found a weak interest rate channel for the transmission channel of monetary policy in Mauritius. A decline in the Key Repo Rate will first impound on the Prime Lending Rate of commercial banks. Thereafter, the commercial banks will eventually adjust their respective savings rate and lending rates downwards. Consequently, a lower savings rate will deter people from savings so that there should be an increase in consumption in order to boost the level of economic activities. Conversely, the fall in lending rate will induce people to contract an increasing number of loans and to induce higher investments and correspondingly higher levels of economic activities, all unleashing enhanced economic conditions. Unfortunately, the interest rate transmission channels simply starts from the change in the Key Repo Rate, before feeding into the Prime Lending Rate and ending with changes to the savings and lending rates, without entail real effects on the Mauritian economy in terms of expected falls in savings and increases in loans. The chief reason is that households in Mauritius are highly conventional and like to put all of their deposits in banks without being concerned about alternative investment modes due to the lack of knowledge in financial products. Besides, a lack of other innovative investment products may be part of the reason. Increasing the level of knowledge on financial products and enlarging the scope of financial products to them will definitely assist in “greasing on” the interest rate channel of monetary policy in Mauritius. It is vital to note that the real effectiveness of the interest rate channel for monetary policy transmission mechanism does not limit itself to changes in the Key Repo Rate being transmitted to other interest rates such as money market rates or saving and lending rates; real effectiveness hinges on the final impacts on savings and lending and hence real repercussions on the economy.

7  Monetary Policy Analysis 

5.3

163

Weak Asset Price Channel

The asset price works mainly via the effect of a change in interest rate on valuation of equities which then feed into the wealth effects of various economic units such as households. Besides, in the case of corporates which hold considerable equities, this leads to undermined values for equities as to reduce the value of Tobin’s q to hover below one (market value of firm/value of assets) and thus squeeze on i­ nvestments. Despite the presence of a stock market in Mauritius, most households still decline to invest in shares. The main reason for this is their rather complacent investment behaviour which is heavily geared towards bank deposits, along with the fact, mentioned above, that households lack the proper financial knowledge to invest in shares. It is for these reasons that a hike in interest rates does not trigger a switch away from equities to bank deposits, and vice versa. Therefore, the asset price channel of monetary policy is likely to be particularly weak in Mauritius. Even in terms of Tobin’s q, the impacts are likely to be low since only a small number of firms are listed in the local stock market. Thus, in Mauritius, even if there are concerns about feasible asset price bubbles in the local stock market, hikes in the interest rates are unlikely to prick these bubbles based on weak interaction between the equities and the savings markets. Another hurdle which impacts on the effectiveness of the asset price channel in Mauritius pertains to the robust level of inside ownership which prevails even among listed firms in the official market. Such a state of affairs plainly signifies that there is poor scope of equities holdings among other economic units. Consequently, limited equities ownership hinders on an effective asset price transmission mechanism channel. In the case that the asset is based on houses in lieu of equities, then, again, no effects are expected to occur because Mauritius has been experiencing steadily rising house prices over recent years. These structural issues automatically blunt the effectiveness of the asset price channel of the monetary policy transmission mechanism in Mauritius.

164 

Economics and Finance in Mauritius

Ramlall (2012) found that SEMDEX does not affect broad money demand in Mauritius, either in the long run or the short run, bolstering the notion of a weak asset price channel for monetary policy in Mauritius.

5.4

Credit Channel

The credit channel of the transmission mechanism of monetary policy hinges on two distinct subchannels – namely, the bank-lending channel and the balance sheet channel.

5.4.1  Bank-Lending Channel-Lenders Viewpoint The bank-lending channel probes into the lenders’ financial status in terms of their ability to furnish loans to potential borrowers. For instance, a drain in the reserves of banks via contractionary monetary policy leads to a decline in loanable funds, thereby reducing the level of credit being granted to borrowers. Consequently, an increase in the cash ratio reserve requirement will act as a direct constraint on credit availability.

5.4.2  Balance Sheet Channel-Borrowers Viewpoint Part of the balance sheet channel rests on equities valuation and thereby relies on the asset price channel. The process is as follows: a hike in the interest rate leads to a switch away from equities so that declines in equities values undermine the net worth of corporates. However, since the asset price channel is found to be weak on the back of the low equities participation explained by both high inside ownership and the lack of financial knowledge, this automatically implies a weak balance sheet channel effect for most Mauritian corporates. However, there is another avenue of the balance sheet channel, via interest expenses. When a hike in interest rate occurs, this increases the repayment costs of firms so that there is a drain in cash which eventually leads to lower net worth. Such an undermined equity level leads to a higher debt to equity ratio which thereby reduces the scope for more external bank borrowing,

7  Monetary Policy Analysis 

165

impacting on investment projects. Such a channel is expected to operate in Mauritius chiefly when most of the firms happen to be either medium-sized or small.3 Larger firms usually avail themselves of a better banking relationship in order to maintain an unchanged borrowing state. However, the balance sheets of both corporates and households are being affected via the balance sheet channel. In essence, a hike in interest rates increases the repayment costs of household loans so that this entails behind undermined wealth effects as to reduce feasible investments projects. Overall, the credit channel is expected to be strong in Mauritius, though somewhat buffeted by the asset price channel in terms of net worth impact. Thus, a rating of 9 over 10 is attributed to this channel.

5.5

Expectations Channel

The expectations channel of monetary policy is considered to represent the overriding or highest level of monetary policy advancement under a policy of inflation targeting. In fact, anchoring inflation expectations constitutes the main task of any central banker so that it assists in properly accomplishing such a task. Though the Bank of Mauritius has not explicitly moved to inflation targeting, some move towards inflation targeting is considered to be high on the agenda. However, until no explicit inflation targeting framework is being discussed at the public level, nothing concrete can be said. Nonetheless, since the Bank of Mauritius has always adhered to a gradualist approach in its monetary policy stance, the expectations of economic players are unlikely to significantly diverge from the future path of interest rate setting. This is also endorsed via inflation expectations survey performed by the Bank of Mauritius. In that respect, a mild state is considered to be achieved under the expectations channel. Once inflation targeting is explicitly achieved, the expectations channel is anticipated to considerably drift from a mild state to a robust zone in which case monetary policy credibility will attain its pinnacle.  Nonetheless, some SMEs in Mauritius avail themselves of finer interest rates from loans granted by the Development Bank of Mauritius Limited. 3

166 

6

Economics and Finance in Mauritius

Conclusion

Overall, the analysis shows that the credit channel and the exchange rate channel tend to be the main forces for the impact of monetary policy in Mauritius. Once there is an explicit and formal move to inflation targeting, this will transform the expectations channel to full mode of operation. However, structural reforms will be required to rekindle both the interest rate channel and the asset price channel pertaining to the monetary transmission mechanism in Mauritius. Vigorous financial literacy campaigns, the creation of alternative asset channels to deposits, and ensuring a gradual decline in the exorbitant interest rate spreads will create a better atmosphere for the implementation of monetary policy in Mauritius.

Bibliography Bank of Mauritius Annual Reports, Various issues. The Mauritian economy: A reader. (2001). Palgrave Macmillan. Ramlall, I. (2012). Broad money demand in Mauritius with implications for monetary policy. Journal of Economics and Behavioral Studies, 4(8), 436–448. Ramlall, I. (2015a). Global central bank ratings: A new methodology for global excellence. Palgrave Macmillan. Ramlall, I. (2015b). Mauritius financial system stress index: Estimating the costs of the subprime crisis. Journal of African Business, 16(3), 235–271. Ramlall, I. (2016). Does central bank quality determine sovereign ratings and credit default swap spreads: Evidence from the world? Journal of Central Banking Theory and Practice, 5(3), 5–29. Tsangarides, C. (2010). Monetary policy transmission in Mauritius using a VAR analysis. IMF working paper, WP/10/36.

8 Money Demand Analysis in Mauritius

1

Introduction

Central banks worldwide attribute a special role to monetary aggregates as they entail important effects on inflation. Alternatively stated, based on the notion that inflation tends to reflect a monetary phenomenon, it becomes warranted to gauge the money demand function. Mauritius initially embarked on monetary aggregates as the tool for monetary policy management before moving to a policy of interest rate targeting and, with time, the ultimate focus is to shift to inflation-targeting framework to secure its credibility in monetary policy. Due consideration is also given to credit aggregates. Monetary policy constitutes a major pillar in the economic performance of a country. In economic parlance, demand management policies are considered to be the foundation stone in terms of tools which can be wielded by the authorities of a country to secure the desired level of economic performance. While fiscal policy is the preferred tool of the government, monetary policy is widely employed by the central bank. Fiscal policy principally relates to the use of taxes and government expenditures while monetary policy is tilted towards the use of either the cost © The Author(s) 2017 I. Ramlall, Economics and Finance in Mauritius, DOI 10.1007/978-3-319-39435-0_8

167

168 

Economics and Finance in Mauritius

of credit or the quantity of money supply. The present monetary policy framework in Mauritius clings to a mixture of open market operations and reserve adjustments for liquidity management. Interest rates rather than monetary targeting are used for monetary policy implementation. The intermediate target for monetary policy rests on the Key Repo Rate, which is adjusted based on interest rate differentials with respect to key trading partners, exchange rate fluctuations and changes in inflation rates. However, the main difficulty in relying on interest rates is that the transmission mechanism is still ambiguous. In essence, deposits and loans at large tend to be immune to changes in interest rates. The lag effect of the interest rate transmission mechanism is still not known for certainty even among the policy makers. Consequently, it becomes interesting to gauge on as to whether monetary aggregates can still be wielded in terms of long-term policy analysis. Tsangarides (2010) found that the transmission mechanism of interest rate entailed weak impact on output, adding lustre to our analysis of the money demand function. Simmons (1992) analysed the money demand function. However, the study undertaken is deemed to be altogether obsolete, bearing in mind some important structural changes in the sphere of the Mauritian financial system, namely, the managed floating exchange rate regime introduced in 1994, let alone the gradual development of the stock exchange of Mauritius since its inception in 1989. In that respect, an update on the study becomes warranted. Based on a rather ineffective interest rate transmission mechanism, it becomes important to assess monetary targeting under money demand analysis as a tool for long-term inflation analysis. In an instance when the money demand behaviour is found to be stable, this signifies that money growth rates can be used to predict future inflation and output trends, i.e., inflation can be controlled via monetary targeting. The objective of this chapter is to probe into the long-run and short-­ run components of the broad money demand for the Mauritian economy for the period 2000–2009. Since 1992, no study has been performed to gauge on the broad money demand function. Findings show that Mauritius is imbued with a stable money demand function, mostly propelled by lack of financial innovation and a low level of institutional or regulatory changes over the period under scrutiny. More specifically, M2

8  Money Demand Analysis in Mauritius 

169

is found to generate a positive elasticity value of 2.80% with respect to GDP. Such a finding implies that Mauritius is imbued with a still developing financial system by virtue of the fact that monetization moves faster than output. Most importantly, the low value of the adjustment coefficient under the vector error correction model corroborates the notion of weak financial innovation so much so that there is a dearth of alternative assets to M2, ironically adding strength to the argument pertaining to a lack of financial innovation in the Mauritian financial system. As a small economy, Mauritius is automatically subject to the vagaries of the global economy. This is confirmed via robust exchange rate channel effects based on the positive evidence found in favour of foreign asset substitution. The findings show no real interaction between money holdings and stock market performance, mostly attributable to the low level of financial literacy coupled with the risk-averse nature of Mauritian depositors. Overall, the main policy implication of the study is that the monetary policy can cling to money demand analysis chiefly for long-run policy assessments.

2

Theoretical Underpinning

A stable money demand is considered to be of vital significance by policy makers when it comes to a sound assessment of monetary policy whereby central banks can alter their monetary aggregates to unleash predictable effects on output, interest rates and prices. The theory of money demand states clearly that the demand for money is meant for real balances with various variables being utilized used to explain for the level of economic activity, inflation rate and a set of opportunity cost variables used to represent both local and foreign asset substitution. As far as the econometric methodology is concerned, VECM constitutes the coveted analytical tool with respect to money demand studies due to its ability to simultaneously capture both the long-run and short-run determinants. Financial innovations are considered to be important hurdles when it comes to measuring the stability of money demand. For instance, in the 1980s, financial innovations in key financial centres represented the chief reason behind the instability of narrow money demand.

170 

Economics and Finance in Mauritius

In essence, money serves four basic functions, acting as a medium of exchange, a store of value, a unit of account and a source for deferred payment. The literature on money demand consists of distinct theories. Classical economists such as Schumpeter (1954) deemed money to be merely an economic bridge to enable economic activities to manifest because it is used as a numeraire to price products but without entailing any real impact on the economic variables. Money assists in solving the problem of barter system which was hindered by a lack of double coincidence of wants. Thereafter the classical interpretation of money, the quantity theory of money took birth under two distinct versions and separately developed by Fisher (1911) and Pigou (1917). Fisher (1911) version is described as the equation MV = PT, i.e. the product of quantity of money in circulation and the velocity of circulation is equal to the product of the price level and the volume of transactions. In the case that V and T are presumed to be stable, any hike in M will trigger a proportionate increase in P. Pigou (1917) later developed his own version of money demand which is also referred to as the Cambridge approach where MV = Py where V refers to income velocity of circulation. Since V is fixed and y is at full employment, again, any increase in M will occasion a proportionate increase in P. After the Cambridge approach to money demand, Keynes (1930, 1936) revolutionized the money demand concept from “an all spending state” to “a hoarding state” because he identified three motives for holding money: the transactionary motive, the precautionary motive and the speculative motive. The transactionary demand for money manifests on the back of nonsynchronization of receipts and payments and is positively associated with the income level. In the same vein, the precautionary demand for money is also positively associated to the income level but it is based on the need to hold money to make good for any unforeseen contingencies as and when required. The most distinctive element of Keynes’sss theory pertains to the speculative money demand which is based on the interest rate level, all encapsulated under an inverse relationship. After this emerged the post-Keynesian theories of money demand which consisted of, among others, Baumol (1952) and Tobin (1956) inventory-theoretic approach. The next section focuses on some of the major ramifications which were brought to the monetary policy framework in Mauritius.

8  Money Demand Analysis in Mauritius 

3

171

 auritian Economy and Monetary Policy M Conduct

Since its independence in 1968, Mauritius has experienced many economic changes in its landscape as to become one of the best-performing African countries in terms of sound economic growth accompanied by a remarkable level of social, political and economic stability. The Stock Exchange of Mauritius started its operation in 1989 and had since been subject to ongoing changes in terms of daily trading, the extension in trading time and the listing of foreign companies although the number of listed companies on the official market did not change significantly. Mauritius had a population of about less than 1.3 million with a real GDP growth rate of 2.8% in 2009. As at end March 2010, the financial system in Mauritius consisted of 18 banks, 12 non-bank deposit taking institutions, 14 money changers and 5 foreign exchange dealers. Mauritius is endowed with a bank-based financial system as banks contribute around 70% or more of the total assets in the financial system. For the year 2009, the following sectors contributed more than 10% of total GDP – namely, the manufacturing sector, the wholesale and retail trade sector, the transport sector, the real estate sector and the financial intermediation sector, contributing about 19.42%, 12.02%, 10.91%, 11.86% and 11.65%, respectively. Per capita GDP scaled up by about 16.35%, from Rs 164,172 to Rs 191,008 in 2009. In 2007, Mauritius came out 81st among 182 countries under the UN Human Development Index. The Mauritian financial sector deepened itself gradually with the ratio of M2/ GDP increasing by about 28% from March 2000 to September 2009. As at February 2010, the import cover ratio in terms of the number of weeks stood around 47.5, implying that Mauritius was well poised to bear any external shocks. In spite of the fact that there is a single license for both Segment A and Segment B banks, yet, no conspicuous penetration was noted in each case although, as the US subprime crisis unfolded. The Mauritian financial system is deemed to be sophisticated based on its constituents such as banks, a stock market and a derivatives market. Mauritius had not been spared from the US subprime crisis, as evidenced by a decline in output growth from 4.2% in 2008 to less than 2%

172 

Economics and Finance in Mauritius

in 2009, which is attributable to a contraction in the tourism, textile and construction sectors. Nevertheless, the country performed pretty well following the adoption of a fiscal stimulus package acting like automatic stabilizers, inducing a fall in the Key Repo Rate by a hefty 250 basis points and a strong and resilient banking sector (which contains no toxic assets) in the form of sound capital adequacy, profitability and liquidity ratios. To impose a discipline on the public debt level, a public debt management act was issued whereby the aim was to scale down public debt to 50% of GDP by the year 2013. However, being a small open economy, Mauritius future output growth will heavily hinge on the strength of the international economic recovery. The chief aim and legal responsibility of the Bank of Mauritius is to maintain price stability and to stimulate an orderly and balanced economic development. At the beginning stage of its economic development, Mauritius clung to a system of direct controls with the introduction of credit ceilings in 1973 and administered interest rates. Thereafter, Mauritius had recourse to indirect monetary policy by introducing vital reforms, which incorporate, among others, the liberalization of interest rates in 1988, the phasing out of credit controls in 1993, the removal of exchange controls in 1994 and connecting the bank rate to the latest average bill rate to move close to market-determined interest rates. A Reserve Money Programme and a Liquidity Forecasting Framework was established in 1996. The monetary policy framework experienced an important change in 1999 when the Lombard Facility was introduced to depict the main stance of monetary policy in Mauritius. Before 1999, the monetary policy framework was regarded as highly deficient because it was not feasible to isolate the movements in the bank rate due to monetary policy changes from movements resulting from the excess liquidity prevailing in the monetary circuit. In fact, the bank rate, which was determined by the market at the weekly auctions of Treasury Bills, was also employed to reflect the stance of monetary policy. Unfortunately, it transpired that the movements in the bank rate failed to accurately ­capture the main stance of monetary policy. Thus, market participants could clearly differentiate between policy-induced interest rate changes from liquidity-induced interest rate changes.

8  Money Demand Analysis in Mauritius 

173

To obviate the above shortcoming, in December 2006, the Bank of Mauritius began with a new framework for its monetary policy with the Repo Rate replacing the Lombard Rate as the signalling stance for monetary policy. Under the new framework, the Bank of Mauritius pursued a monetary policy strategy based on two approaches. The first approach incorporates economic analysis which factors in both short- and medium-­ term risks to price stability and economic growth. The second approach assesses monetary developments and the associated long-term risks to price stability. In 2007, a corridor of 125 basis points above and below (previously set to 50 basis points) had been established to derive the ceiling and floor for overnight interbank interest rates. In spite of the shift to the Key Repo Rate, some major problems impeded the effective application of monetary policy in Mauritius. First, the money market is still not well developed while the secondary market still operates below full swing in terms of the amount and frequency of transactions.

4

Data and Methodology

Money stock definitions differ for various economies, all hinging on their level of financial development. For example, industrial economies have a broader money definition, such as M3, M4 and M5. Nevertheless, M1 and M2 are widely used. M1, also known as narrow money, captures the currency level with the public plus demand deposits with the banking system. Quasi-Money consists of savings deposits, time deposits and foreign currency deposits. M2 is equal to M1 plus Quasi-Money. The most distinctive element which prevails between narrow money and broad money is that, while the former is readily available and transferable in daily transactions, broad money includes other forms of money for portfolio needs. In this study, M2 is used as the level of money stock to be consistent with most of the empirical evidence on money demand. In the long run, the money market is said to be in equilibrium when the real money supply is tantamount to the real money demand. Money

174 

Economics and Finance in Mauritius

demand functions are assessed by having recourse to variables which can be either in logarithmic form or in levels. In the former case, the variables are directly deemed as elasticities while, in the latter case, they are considered as semi-elasticities. It is of paramount significance to note that the long-run income elasticity is greater than one for economies which are characterized by an underdeveloped financial system in which case monetization happens to move faster than the output growth. As pointed out by Knell and Stix (2004), in the case of industrial economies, the long-­ run income elasticity of demand for broad money is near one. Money demand theories state a relationship between the quantity of money demanded and a set of few key economic variables which relate money to the real sector of the economy. More specifically, the money demand functions are usually estimated in log-linear form with the monetary aggregates and scale variables being considered in their logarithmic forms while other variables enter the model either in levels or in logarithmic form. Most empirical papers point out that the money demand function triggers a positive relationship with GDP (Anuar 1986; Habibullah 1987; Habibullah and Ghafar 1989), industrial production index (Marashdeh 1997; Sriram 1999) and a negative relationship with the interest rate variable (see Civcir 2003; Laidler 1993). For the present study, real gross domestic product is employed as the proxy for income. The standard theory of money demand points out that:

M / P = ƒ ( Y, R )

(8.1)

Where M Nominal money demand, M2 P Price level Y Scale variable for income and output R a vector of returns on various assets – real, local and foreign – which represent the opportunity cost of holding money. The latter is often identified as part of the portfolio framework and thereby consists of the following variables:

8  Money Demand Analysis in Mauritius 

175

(a) For real assets, the anticipated inflation rate is used. Because there is no such data on expected inflation in Mauritius, the actual inflation rate is employed for the analysis. Friedman (1969) argued that inflation represented a negative return for holding its own money because it scaled down the purchasing power of money. Thus, on the back of higher inflation, it is anticipated that real broad money demand will decline as economic players substitute real assets for money to preserve their wealth levels. In fact, while the real value of money falls under higher inflation, the real value of assets remains unchanged. Therefore, real assets have a major role to play in individual’s portfolio decision chiefly in the case of developing countries. (b) Two proxies are used to capture local assets substitution, namely the 3-month Treasury Bill yield and the all-share index (SEMDEX) of the local stock market. For local assets, the savings rate is employed to capture its own rate of return on money holdings. A higher 3-month Treasury Bill yield is likely to unleash deterring impacts on money holdings. Based on asset substitution (as per Wu (2009)) and to sieve out a full-fledged model for the money demand function in Mauritius, SEMDEX is incorporated as additional variable in the model. In fact, as at end September 2009, the stock market capitalization over GDP stood around 64%, plainly depicting that the local equity market may be a vital substitute for money holdings. Sriram (2001) argued that the failure to provide due consideration to the choice of the correct variables and model specification generated poor results. (c) In the case of foreign assets, both foreign interest rates and the anticipated rate of depreciation of the local currency are employed. Based on the fact that Mauritius shifted to a managed floating system in 1994, an analysis of how integrated financial markets had been in terms of the money demand function, becomes warranted. In fact, as a small open economy with no capital controls, exchange rates trigger considerable impacts on the Mauritian economy. Sriram (1999) described the portfolio shift between local and foreign money as the direct substitution effect based on exchange rate movements. Bahmani-Oskooee (2001) resorted to the Real Effective

176 

Economics and Finance in Mauritius

Exchange Rate (REER) while others such as Chowdhury (1995) and Ericsson and Sharma (1996) employed the Nominal Effective Exchange Rate (NEER). Unfortunately, due to the absence of correct data on NEER for Mauritius, recourse is made towards the rupee price of the US dollar. The reason is based on the fact that the exchange rate system in Mauritius is focused on the US dollar. The current study uses LIBOR as the foreign interest rate. A negative relationship is expected to prevail because an increase in foreign interest rate return will stimulate Mauritians to increase their holdings of foreign assets by correspondingly diminishing their holdings of money under portfolio reshuffling effects. From that perspective, the failure to incorporate foreign considerations in the money demand function is likely to trail behind model misspecification which eventually distorts on the long-run elasticity which exists between GDP and broad money. Monthly data, which span from January 2000 to September 2009, have been employed for the analysis.1 SEMDEX pertains to the general broad market index for the Mauritian Stock Exchange Market. The US dollar is employed to identify the exchange rate effects and is denominated against the Mauritian rupee. To factor in the local interest rate evolution, 91-day Treasury Bill yields are employed while LIBOR is used to include the foreign interest rate counterpart. All data are depicted in Table  8.1. M2 pertains to the broad demand for money. CPI refers to the Consumer Price Index, widely employed to capture the inflation level. Multiplicative Table 8.1  Definition of variables

M2 GDP CPI SEM USD TB3 LIBORUSD

Natural logarithm of (M2/CPI) Natural logarithm of (GDP/CPI) Natural logarithm of (CPI) Natural logarithm of (SEMDEX) Natural logarithm of (USD) Natural logarithm of (TB3) Natural logarithm of (LIBORUSD)

 The analysis was undertaken just after 2009 and hence the study used all available data at that time period. 1

8  Money Demand Analysis in Mauritius 

177

extrapolation technique is applied to convert the quarterly GDP data into its monthly version. While gleaning the broad money data for Mauritius, it transpires that there is a structural change in the data as from December 2007 onwards. To make sure of sound use of the available data, the change in values for M2 for January 2008 onwards is calculated to adjust the data. Such a strategy helps to preserve the short-­term dynamics to thereby untangle the true essence of the relationships among the variables in the model. Descriptive statistics for all the variables are shown in Table 8.2 while Table 8.3 depicts the correlation coefficients which exist among the variables. Technically speaking, a stochastic process is said to be stationary if its mean and variance are constant over different time periods and the value of the covariance between the two time periods hinges only on the distance or gap between the two time periods and not the actual time at which the covariance is calculated. Prior to implementing any cointegration analysis, it is considered vital to test for stationary of each series under examination. Table 8.4 points out that each variable is nonstationary in level but stationary in the case that it is first-differenced, making the variables suitable for cointegration analysis. As Rasche and Johannes (1987) argued, the series are not seasonally adjusted to avoid any possible introduction of spurious autocorrelation from the standard seasonal adjustment techniques. In other words, the objective for not seasonally adjusting the data is to maintain the short-term dynamics. In fact, seasonal dummies are used in the analysis. Cointegration technique under Johansen’s (1991) approach is employed to disentangle out the long-run relationships which exist among the variables. Tests of cointegration are applied for all the variables under consideration. The lag order of the VAR is not known and is determined based on the Johansen procedure. The error correction model is used as the main econometric methodology to estimate the money demand function for Mauritius. The greatest benefit of a technique is that it enables the researcher to simultaneously capture both the long-run and the short-run dynamics of the model. Apart from capturing the long-run and short-run dynamics of the model, ECM also allows to gauge on the disequilibrium as a process of gradually drifting closer to the long-run equilibrium. This underlies the reason as to why ECM has been employed for a number of studies by various researchers, let alone the fact that it is now deemed as the work horse model the case of money demand analysis.

Mean Median Maximum Minimum Std. Dev. Skewness Kurtosis Jarque-Bera Probability Observations

7.423465 7.439815 7.688118 7.173953 0.152373 −0.097758 1.751478 7.785541 0.020389 117

M2CPI

Table 8.2  Descriptive statistics 6.106016 6.105997 6.333858 5.903194 0.105007 −0.011433 2.083655 4.096029 0.128991 117

GDPCPI 4.471525 4.447727 4.768988 4.198508 0.178734 0.245083 1.836595 7.769651 0.020551 117

CPI 6.567645 6.540352 7.607684 5.831648 0.556006 0.323122 1.750995 9.641020 0.008063 117

SEMR 2.024203 2.050270 2.558002 1.302913 0.290562 −0.365327 2.202480 5.703233 0.057751 117

TB3

3.389750 3.388055 3.546195 3.247681 0.071986 −0.000305 2.391088 1.807526 0.405043 117

USD

0.989943 1.068153 1.926072 −1.248691 0.706204 −0.669321 2.959684 8.743737 0.012628 117

LIBORUSD

178  Economics and Finance in Mauritius

M2CPI GDPCPI CPI SEMR TB3 USD LIBORUSD

1.000000 0.910874 0.960772 0.901881 −0.411281 0.621165 −0.207622

M2CPI 0.910874 1.000000 0.899863 0.854265 −0.279805 0.579008 −0.185618

GDPCPI

Table 8.3  Correlation coefficients 0.960772 0.899863 1.000000 0.941156 −0.252228 0.591757 −0.201610

CPI 0.901881 0.854265 0.941156 1.000000 −0.168363 0.426142 0.021218

SEMR −0.411281 −0.279805 −0.252228 −0.168363 1.000000 0.004631 0.585602

TB3

0.621165 0.579008 0.591757 0.426142 0.004631 1.000000 −0.128837

USD

−0.207622 −0.185618 −0.201610 0.021218 0.585602 −0.128837 1.000000

LIBORUSD

8  Money Demand Analysis in Mauritius 

179

180 

Economics and Finance in Mauritius

Table 8.4  Unit root tests

M2 GDP USD LIBORUSD CPI TB-3 SEMDEX

ADF Test in level

ADF test in first difference

−0.4518 −0.5021 −2.4477 0.2854 0.2976 −1.8214 −0.0531

−12.6294** −3.9966** −6.4889** −6.0443** −8.6743** −8.0540** −7.9800**

denotes statistical significance at the 1% level

**

5

Empirical Results

When eight lags are used, it transpires that there are two optimal lags under the Schwarz information criterion and the Hannan–Quinn information criterion. When these two optimal lags are applied under the first model, which consists of all the nonstationary variables in levels forms, only one cointegrating equation at the 5% is obtained under both the trace test and the maximum-eigenvalue test. Subsequently, the VECM model is run with the exogeneity tests depicting that LIBORUSD and SEMR should be deemed as exogenous variables. Thus, a new system is built with exclusion of these two variables. Unfortunately, again, it seems that TB3 and USD are not statistically significant as long-run elements of the model. Thus, with the objective of sieving out a parsimonious and enshrined model, a final VECM is run which comprises only of M2, CPI and GDP.  Results depict that, under two optimal lags, there is one cointegrating equation at the 5% level with exogeneity tests demonstrating that GDP is merely weakly exogenous. Because the data is not seasonally adjusted by virtue of the fact that such prefiltering may impact on the dynamics in line with Wallis (1974), Ericsson et al. (1994) and Ericsson and Sharma (1994), seasonal dummies have been included. But, none is found to be statistically significant. The final model is shown in Table 8.5 with only the statistically significant variables being reported.

8  Money Demand Analysis in Mauritius 

181

Table 8.5  Empirical results for Mauritius Speed of adjustment coefficient Impact of long-run components on M2 GDPCPI(−1) CPI(−1) Constant term Impact of short-run components on M2 Constant term D(M2CPI(−1)) D(GDPCPI(−1)) D(GDPCPI(−2)) D(GDPCPI(−3)) D(USD(−1)) D(TB3) R-squared Adj. R-squared F-statistic Log likelihood *

−0.0309 (−2.4512)** 2.7991 (6.9188)** −0.6468 (−2.8699)** 6.7774 0.0064 (2.6589)** −0.3374 (−3.0722)** 0.1552 (3.9113)** −0.1081 (−2.2892)** 0.1317 (2.0687)** 0.1261 (2.3653)** −0.0457 (−2.6726)** 0.3412 0.2701 4.8027 360.5200

and ** denote statistical significance at the 5% and 1% level, respectively

Based on the quantity theory of money, the long-run elasticity between GDP and broad money should hover around 1.0. The results demonstrate, however, that, although the demand for broad money in Mauritius is positively elastic with respect to GDP, nonetheless, it is higher than two, with an elasticity of 2.80%, showing a high level of monetization in Mauritius. Previous empirical studies also found a similar state of affairs. For instance, Baharumshah (2004) finds an income elasticity of 3.69% in the case of Malaysia with Dekle and Pradhan (1997), Majid (2004), and Dahalan (2004) also reporting income elasticity of money demand higher than unity for developing countries. Such a state of affairs means that, for the period under examination, changes in real income, on ­average, generate a more than proportionate increase in real broad money demand. The coefficient of the error correction mechanism (ECM) stands at around −0.03. This implies a slow speed of adjustment of 3% to the remaining disequilibrium. The negative impact implies that lagged excess money holdings generate smaller holdings of current money. The negative sign on the ECM term is synonymous with corrective actions initiated by market players in the present period in order for the preceding period’s disequilibrium in money balances to steadily converge to the

182 

Economics and Finance in Mauritius

long-term equilibrium. Wu (2009) also identified a low adjustment coefficient of 1.1% in the case of China while Taylor (1986) and Ericsson and Sharma (1996) noted larger adjustment coefficients for ECM to lie at −0.26, −0.15, −0.20, −0.12 and −0.08 for Netherlands, Germany, France, Argentina and Greece, respectively. The higher adjustment coefficients identified for the developed economies can be attributed to the prevalence of more sophisticated financial systems in which case the speed of adjustment for money will automatically be higher, explained by larger choices of financial assets and more financially literate citizens. The three-month Treasury Bill yield impacts on broad money in the short run but not the short run. Indeed, a 1% increase in the three-­ month Treasury Bill yield engenders around a −4.6% decline in M2 in the short run. The negative impact is consistent with the theoretical considerations whereby the higher return on alternative local assets such as Treasury Bills reduces the level of money holdings. Based on the fact that banks represent the main bidders in the auctioning for Treasury Bills, in the short run, their actions are highly likely to impound on the level of broad money in the system. Findings further demonstrate that SEMDEX is impotent in affecting M2 and this manifests independent of the period under consideration – whether in the long run or the short run. Common sense would signify that as the equity market posts higher returns, this unleashes portfolio assets reshuffling via a switch away from broad money holdings to equities. But, in Mauritius, the cultural features predominate over the financial characteristics. Indeed, only high net worth educated people incorporate equities as part of the investment portfolios while Mauritians adhere to bank deposits. Inflation rate is found to occasion a bearish long-run impact on broad money holdings with the elasticity coefficient standing around 0.65%. Such a negative relationship is attributable to the fact that as the inflation rate increases, people have a taste for physical assets as they represent the best hedges against inflation relative to money balances. In fact, in the case of Cameroon, Nachega (2001) argued that “high inflation elasticities are generally expected in developing countries as the range of financial instruments outside money is very limited and real assets represent a substantial part of the public’s portfolio”. In addition, Wu (2009) finds that

8  Money Demand Analysis in Mauritius 

183

inflation unleashes a substantially negative effect on the money demand in the case of China. Under foreign asset substitution analysis, irrespective of whether the focus is laid on the short run or in the long run, LIBORUSD is not statistically significant. Thus, foreign interest rate does not really impound on the level of broad money demand in Mauritius. But, in the short run, the one-period lagged difference of the USD triggers a negative effect on broad money. More specifically, a 1% change in the one-period lagged difference of USD occasions around 13% change in broad money. Most importantly, the economic significance is found to be much higher than that of the three-month Treasury Bill yield, plainly depicting that foreign exchange rate movements in the US dollar engenders direct effects on local monetary aggregates in the short-run. Such a state of affairs may be accounted for by the fact that, as Mauritians expect a depreciating rupee because the rupee has been structurally been depreciating over time, they demand more money to buy assets denominated in the US dollar. Overall, findings clearly show that foreign asset substitution chiefly works through the exchange rate channel in lieu of the interest rate medium. Parameter constancy represents a key analytical component whenever gauging on the money demand function. In fact, for an effective implementation of monetary policy, it is imperative to have a stable money demand function. Besides, Wu (2009) states that “what is being sought in a stable demand function is a set of necessary conditions for money to exert a predictable influence on the economy so that the central bank’s control of money supply can be a useful instrument of economic policy”. Parameter constancy is analysed using post-estimation diagnostic tests. Results show that the estimated VECM is stable because nearly all the roots have modulus less than one and lie inside the unit circle as illustrated in Fig. 8.A.1 in the appendix section. Based on this result, both impulse response and variance decomposition results become valid. In addition, the diagnostic test results on the residual point out no problem of serial correlation under serial correlation LM tests as depicted in Table 8.A.1 in the appendix section. Furthermore, Table 8.A.2 in the appendix part points out that no autocorrelation prevails in the case of the VEC residual portmanteau tests for autocorrelations. Above all, VEC residual heteroskedasticity tests demonstrate that residuals are not heteroscedastic irre-

184 

Economics and Finance in Mauritius

spective of whether cross terms are absent (Table 8.A.3 in the appendix) or cross terms are present (Table 8.A.4 in the appendix). In general, these post-estimation diagnostic tests mean that the Mauritian money demand estimation is not only strong but also stable for the period under review. Only normality conditions have been violated. An alternative avenue present to assess the short-run dynamics and co-­ movements among the variables is to analyse the impulse response and variance decomposition functions from the VECM model. This approach enables us to gauge on the effect of distinct types of shocks on the variables in the model. Each innovation (shock) is derived via a standard Choleski decomposition. In the case of the variance decomposition as illustrated in Fig. 8.2, it transpires that a shock in GDP explains around 20% of the variance in M2 which is higher than that of CPI (below 0.1%). In the case of impulse response function as shown in Fig. 8.1, it occurs that, though the adjustment time is same both for GDP and CPI, yet, the effect is again higher for GDP than for CPI. Results are found in Figs. 8.1 and 8.2, respectively.

6

Conclusion

The present study analyses the money demand function in Mauritius. Findings point out a stable money demand function in Mauritius with an estimated income elasticity of 2.80%. In the case of developing countries, it is overwhelmingly agreed by economists (see Fry 1978; Friedman and Schwartz 1963) that the income elasticity of demand for money was greater than one. The computed money demand function for Mauritius is found to be stable based on the properties of the VECM model. Absence of serial correlation, no heteroscedasticity, roots of AR function lying inside the unit circle and violation of normality condition like in prior empirical studies, all endorse a rather stable money demand function for Mauritius. Most importantly, the slow speed of adjustment coefficient for VECM is compatible with those obtained in the case of developing countries. Moreover, the negative effect of inflation plainly depicts that, in Mauritius, high inflation encourages the holdings of more physical assets as they represent the cushioning mechanism against increasing

8  Money Demand Analysis in Mauritius  Response of M2CPI to M2CIP

.012 .008 .004 .000 1

2

3

4

5

6

7

8

9

10

Response of M2CPI to GDPCPI

.012 .008 .004 .000 1

2

3

4

5

6

7

8

9

10

9

10

Response of M2CPI to CPI

.002 .001 .000 -.001 -.002

1

2

3

4

5

6

7

8

Fig. 8.1  Response to Cholesky One S.D. innovations

185

186 

Economics and Finance in Mauritius Percent M2CPI variance due to M2CPI

100 80 60 40 20 0

100

1

2

3

4

5

6

7

8

9

10

Percent M2CPI variance due to GDPCPI

80 60 40 20 0

1

2

3

4

5

6

7

8

9

10

Percent M2CPI variance due to CPI

.5 .4 .3 .2 .1 .0

1

2

3

Fig. 8.2  Variance decomposition

4

5

6

7

8

9

10

8  Money Demand Analysis in Mauritius 

187

prices in the long run. Foreign asset substitution is found to work chiefly via the exchange rate rather than interest rate movements with a positive effect noted because economic agents often expect a depreciating rupee vis-à-­vis the US dollar. Many vital policy implications can be obtained from the above findings. First, based on the existence of a stable money demand function in Mauritius, is still possible to employ monetary targeting. This may provide justification for the monetary authority to benchmark broad money and use it as first leg of assessment for monetary policy decisions, but this will occur only in the long run, not in the short run. As a matter of fact, this is due to the fact that the low adjustment coefficient under ECM implies that monetary targeting can at best be applied over a long-­ term rather than a short-term horizon. Interestingly, the low adjustment coefficient under ECM buttresses the shift made by Bank of Mauritius from a policy that focused on monetary aggregates towards interest rates targeting. However, Tsangarides (2010) showed the presence of a rather weak interest rate transmission mechanism in Mauritius. Thus, in spite of the shift from the Lombard Rate to the Key Repo Rate, the transmission mechanism of monetary policy is still problematic in Mauritius. After all, the implementation of monetary policy constitutes a hard nut to crack for a small country where structural changes manifest as a normal feature of the Mauritian economy. Third, the study points out, like Wu (2009), that the data should not be seasonally adjusted to maintain their short-term dynamics to be able to disentangle the true nature of the relationships which exist among the variables in the system. Rather, seasonal dummies have been employed but none have been found to be statistically significant. Fourth, the low adjustment coefficient under ECM signify a real deficiency in terms of asset sophistication in Mauritius, i.e., the lack of other forms of assets other than M2. Hence, this may imply that Mauritius is imbued with a low level of financial deepening. In a parallel manner, the high income elasticity found for Mauritius results mainly from a restricted scope for financial asset substitution. Alternatively stated, this also means that Mauritius is still subject to an underdeveloped or shallow financial system whereby monetization

188 

Economics and Finance in Mauritius

moves faster than output. Fifth, an interesting aspect of the results is that, even though the Mauritian economy has been subject to a number of vital structural changes during the period under scrutiny, such as the adoption of different monetary policy frameworks, the long-run relationships between the examined macroeconomic and financial aggregates are found to be fairly stable. This means that these structural changes have not really worked on the financial side of the Mauritian economy in terms of unleashing a plethora of asset types. Such a state of affairs can also be explained by financial literacy problems, which occur even for educated people. Beyond that, the results further point out that monetary aggregates of the Mauritian economy remained healthy even post the US subprime crisis, in line with the fact that none of our banks were exposed to toxic assets as they practice conventional banking. Finally, should Mauritius further develop its financial system, this is anticipated to trail behind a fall in income elasticity.

7

Policy Recommendations

Since inflation exerts a bearish pressure on money holdings as to positively affect physical assets holdings, this implies that inflation could be one of the positive forces which induce higher real estate prices in Mauritius. Thus, proper inflation control could feasibly reduce the extent of real estate price increases in Mauritius. The low adjustment coefficient under the error correction model implies that the government has to develop new assets or encourage the use of existing alternative assets such as equities among Mauritian households. Otherwise, the lack of financial deepening still buffets the Mauritian economy so that households simply invest in bank deposits, thereby unleashing excess liquidity to then translate into aggressive bidding patterns of banks as to induce a low-cost, but short-term skewed redemption profile of public debt. Thus, important structural changes should manifest to scale up financial deepening.

189

8  Money Demand Analysis in Mauritius 

8

Appendix 1.5

1.0

0.5

0.0

–0.5

–1.0

–1.5 –1.5

–1.0

–0.5

0.0

0.5

1.0

1.5

Fig. 8.A.1  Inverse roots of AR characteristic polynomial Table 8.A.1  VEC Residual serial correlation LM tests Null Hypothesis: no serial correlation at lag order h Date: 05/05/10 Time: 11:44 Sample: 2000M01 2009M09 Included observations: 114 Lags

LM-Stat

1 2 3 4 5 6 7 8 9 10 11 12

13.26033 16.70197 17.45934 12.04866 13.79470 8.315142 16.38850 17.92420 15.33668 9.164431 5.509251 133.4576

Probs from chi-square with 9 df

Prob 0.1512 0.0536 0.0420 0.2106 0.1298 0.5027 0.0592 0.0361 0.0821 0.4222 0.7879 0.0000

190 

Economics and Finance in Mauritius

Table 8.A.2  VEC Residual portmanteau tests for autocorrelations Null Hypothesis: no residual autocorrelations up to lag h Date: 05/05/10 Time: 11:44 Sample: 2000M01 2009M09 Included observations: 114 Lags

Q-Stat

Prob.

Adj Q-Stat

Prob.

df

1 2 3 4 5 6 7 8 9 10 11 12

1.462988 5.631072 16.02256 26.88150 39.36273 47.51795 62.53012 79.04241 91.44980 99.62288 104.6995 182.5057

NA* NA* 0.3806 0.3100 0.2065 0.2580 0.1292 0.0503 0.0366 0.0500 0.0951 0.0000

1.475935 5.718449 16.39079 27.64460 40.69836 49.30665 65.30092 83.05942 96.53031 105.4893 111.1080 198.0679

NA* NA* 0.3566 0.2753 0.1677 0.2041 0.0859 0.0260 0.0160 0.0208 0.0417 0.0000

NA* NA* 15 24 33 42 51 60 69 78 87 96

*The test is valid only for lags larger than the VAR lag order df is degrees of freedom for (approximate) chi-square distribution *df and Prob. may not be valid for models with exogenous variables

Table 8.A.3  VEC residual heteroskedasticity tests: no cross terms (only levels and squares) Date: 05/05/10 Time: 11:45 Sample: 2000M01 2009M09 Included observations: 114 Joint test: Chi-sq 142.4229

df 132

Prob. 0.2527

Individual components: Dependent

R-squared

F(22,91)

Prob.

Chi-sq(22)

Prob.

res1*res1 res2*res2 res3*res3 res2*res1 res3*res1 res3*res2

0.145070 0.428279 0.167820 0.148219 0.108327 0.153801

0.701885 3.098574 0.834153 0.719770 0.502517 0.751808

0.8267 0.0001 0.6766 0.8084 0.9661 0.7737

16.53800 48.82384 19.13151 16.89695 12.34930 17.53336

0.7883 0.0008 0.6372 0.7690 0.9497 0.7333

8  Money Demand Analysis in Mauritius 

191

Table 8.A.4  VEC Residual heteroskedasticity tests: includes cross terms Date: 05/05/10 Time: 11:45 Sample: 2000M01 2009M09 Included observations: 114 Joint test: Chi-sq 455.6236

df 462

Prob. 0.5749

Individual components: Dependent

R-squared

F(77,36)

Prob.

Chi-sq(77)

Prob.

res1*res1 res2*res2 res3*res3 res2*res1 res3*res1 res3*res2

0.484385 0.822576 0.650810 0.595951 0.620413 0.569875

0.439215 2.167583 0.871372 0.689584 0.764154 0.619437

0.9987 0.0059 0.6979 0.9125 0.8380 0.9596

55.21991 93.77368 74.19232 67.93836 70.72707 64.96578

0.9713 0.0939 0.5695 0.7601 0.6793 0.8342

References Anuar, A. R. (1986). The demand for money in Malaysia, 1965(1)–1984(4). Analisis, 1(1), 27–40. Baharumshah, A. Z. (2004). Stock prices and long run demand for money evidence from Malaysia. International Economics Journal, 18(3), 389–407. Bahmani-Oskooee, M. (2001). How stable is M2 money demand function in Japan? Japan and the World Economy, 13, 455–461. Baumol, W. J. (1952). The transactions demand for cash: An inventory theoretic approach. The Quarterly Journal of Economics, 66, 545–56. Chowdhury, A. R. (1995). The demand for money in a small open economy: The case of Switzerland. Open Economies Review, 6, 131–144. Civcir, I. (2003). Broad money demand and currency substitution in Turkey. The Journal of Developing Countries, 36(1), 1–19. Dahalan, J. (2004). The uncertainty of the U.S. and Japanese interest rates and its effect on money demand in Malaysia. International Journal of Management Studies, 11(1), 71–89. Dekle, R., & Pradhan, M. (1997). Financial liberalization and money demand in the ASEAN countries. International Journal of Finance & Economics, 4(3), 205–15. John Wiley & Sons, Ltd.

192 

Economics and Finance in Mauritius

Ericsson, N. R., & Sharma, S. (1996). Broad money and financial liberalization in Greece (IMF working paper 96/62). Washington, DC: International Monetary Fund. Ericsson, N. R., Hendry, D. F., & Tran, H. (1994). Cointegration, seasonality, encompassing and the demand for money in the United Kingdom. In C.  Hargreaves (Ed.), Non-stationary time-series analysis and cointegration (pp. 179–224). Oxford: Oxford University Press. Fisher, I. (1911). The purchasing power of money. New York: Macmillan. Friedman, M., & Schwartz, A. (1963). A monetary history of the United States. Princeton: Princeton University Press. Friedman, M. (1969). The optimum quantity of money. In The optimum quantity of money and other essays. Chicago: Aldine Publishing Company. Fry, M. (1978). Money and capital or financial deepening in economic development. Journal of Money Credit and Banking, 10(4), 464–475. Habibullah, M. S. (1987). Income expectations and the demand for money in Malaysia: Some empirical evidence. Analisis, 2(1), 61–71. Habibullah, M.  S., & Ghaffar, R.  A. (1989). The demand for currency and demand deposits in Malaysia: An empirical evidence. Analisis, 3(1), 111–122. Johansen, S. (1991). Estimation and hypothesis testing of cointegration vectors in Gaussian vector autoregressive models. Econometrica, 59(6), 1551–1580. Keynes, J. M. (1930). A treatise on money, volume 1, The pure theory of money. London: Macmillan. Keynes, J.  M. (1936). The general theory of employment interest and money. London: Macmillan. Knell, M., & Stix, H. (2004). Three decades of money demand studies, some differences and remarkable similarities (OeNB working paper, No. 88). Wien: Österreichische Nationalbank. Laidler, D. E. W. (1993). The demand for money: Theories evidence and problems. New York: Harper Collins College Publishers. Majid, M.  Z. A. (2004). Reassessing the stability of broad money demand in Malaysia. Kuala Lumpur: Bank Negara Malaysia Discussion Papers. Marashdeh, O. (1997). The demand for money in an open economy: The case of Malaysia presented at the Southern Finance Association Annual Meeting, 19–22 November 1997, Baltimore. Nachega, J. (2001). A cointegration analysis of broad money demand in Cameroon (IMF working paper, WP/01/26). Pigou, A. C. (1917). The value of money. Quarterly Journal of Economics, 32 (1917–1918). Reprinted in Readings in Monetary Theory, F. A. Lutz, & L. W. Mints (Eds.). Philadelphia, 2, 162–183.

8  Money Demand Analysis in Mauritius 

193

Ramlall, I. (2012). Broad money demand in Mauritius with implications for monetary policy. Journal of Economics and Behavioral Studies, 4(8), 436–448. Rasche, R.  N., & Johannes, J.  M. (1987). Controlling the growth of monetary aggregates. Boston, MA: Kluwer Academic Publisher. Schumpeter, J. A. (1954). History of economic analysis. London: Allen & Unwin. Simmons, R. (1992). An error-correction approach to demand for money in five African developing countries. Journal of Economic Studies, 19, 29–47. Sriram, S. S. (1999). Survey of literature on demand for money: Theoretical and empirical work with special reference to error-correction models (IMF working paper, No. WP/99/64). Washington, DC: International Monetary Fund. Sriram, S. S. (2001). A survey of recent empirical money demand studies. IMF Staff Papers, 47(3), 334–65. Taylor, M. P. (1986). From the general to the specific: The demand for M2 in three European countries. Empirical Economics, 11, 243–261. Tobin, J. (1956). The interest elasticity of transactions demand for cash. Review of Economics and Statistics, 38, 241–47. Tsangarides, G. T. (2010). Monetary policy transmission in Mauritius using a VAR analysis (IMF working paper 10/36). Washington, DC: International Monetary Fund. Wallis, K.  F. (1974). Seasonal adjustment and relations between variables. Journal of the American Statistical Association, 69, 18–31. Wu, G. (2009). Broad money demand and asset substitution in China (IMF working papers 09/131). Washington, DC: International Monetary Fund.

9 Impact of Debit and Credit Cards on Currency in Circulation in Mauritius

This chapter analyses the impacts of credit and debit cards on currency in circulation for Mauritius for the period 1999–2008. The analysis is undertaken in three versions, an aggregate version which combines notes and coins together to then move towards specificity of currency, that is notes being assessed separately from coins. Findings show that the number of both credit and debit cards exerts no influence on the amount of currency in circulation. However, the use of a second model with a different proxy shows that the use of debit cards behaves as a complement to, rather than a substitute for, notes in circulation. Compatible with prior empirical evidence in other countries, GDP is found to unleash the strongest economic impact on currency in circulation, independent of the model employed. Overall, strong evidence is found as to Mauritians using their debit cards to withdraw notes instead of effecting direct payments so that debit cards act as a complement in lieu of a substitute to notes.

© The Author(s) 2017 I. Ramlall, Economics and Finance in Mauritius, DOI 10.1007/978-3-319-39435-0_9

195

196 

1

Economics and Finance in Mauritius

Introduction

Transactions done via cards constitute a widely used mode of payment nowadays which subsequently reduces the amount of notes and coins in circulation. This implies the presence of a substitution effect whereby both credit and debit cards are anticipated to scale down the level of currency in demand. Payers have a strong preference to use cards for payments as they are not only an efficient payment mechanism but also a safer way of settlement transactions. In that respect, there has been an increasing amount of literature that attempts to gauge the impact of debit and credit cards on currency in circulation. The empirical literature, however, is skewed towards developed countries with very few studies considering the impact in developing countries. Nonetheless, many these studies converged in terms of their findings, i.e., the use of cards is symptomatic with the undermined levels of currency in circulation. Yazgan and Yilmazkuday (2007) found that a hike in the use of credit and debit cards generated a decline in the demand for Turkish with a feasible decline in seigniorage income. In that respect, it becomes highly interesting to undertake such an analysis for Mauritius to draw out any implications in terms of possible decline in Bank of Mauritius seigniorage income. This chapter analyses the use of the cards payment mode in Mauritius, an upper-income developing country, and assesses their impacts on the country’s financial system. At this stage, it is important to bear in mind two aspects of such an analysis. First, it is anticipated that the increased use of cards, whether debit or credit, will be akin to the lower use of currency in circulation. Second, should this hold, then, this triggers critical consequences on central bank’s revenues via seigniorage. In essence, seigniorage pertains to a central bank’s ability to raise revenue via its monopoly right to create money. And, based on the fact that inflation tax revenue constitutes a product of the inflation rate and the real money base, an increase in cards usage hinders growth in the real money base and thereby entails lower inflation tax revenue. As a matter of fact, inflation tax applies to non-interest bearing assets only such as cash or the monetary base. Since cards reduce the monetary base (currency in circulation), this implies that it is difficult to issue new currencies with costs of production being lower than their face values, all leading to undermined

9  Impact of Debit and Credit Cards on Currency in Circulation... 

197

seigniorage income to central banks. For example, Stix (2004) argued that the loss of seigniorage income might be considerable following the negative effect of credit and debit cards on currency in demand. The most widely used initial framework for conducting such an analysis relates to the monetary exchange equation whereby the product of equilibrium amount of money and velocity should be equal to the product of the price level and the level of transactions. Based on the premise that both velocity and transactions remain constant, any change in money will lead to a change in the price level; a proportionate change in money translates to a proportionate change in the price level. Because currency in circulation represents money, it is anticipated that, in the long-run, the chief propelling force for a change in money will be inflation. The aim of this chapter is to answer the question of whether debit and credit cards generate a downward impact on the level of currency in circulation in Mauritius. Unlike prior studies, in which notes and coins have been considered in aggregate, the present study analyses them separately. Such a decomposition is warranted, particularly in the case of non-developed countries. Another contribution of this research is that two distinct proxies are employed. First, the number of cards is used in line with Rinaldi (2001). Second, the value of transaction per number of cards is employed. The results obtained are found to be sensitive to the proxies used. The rest of the chapter is organized as follows. Section 2 focuses on the literature review. Section 3 provides descriptive information with respect to debit and credit cards in Mauritius. Section 4 spells out the econometric model to disentangle the impacts of debit and credit cards on currency demand. Section 5 discusses the results obtained while Sect. 6 concludes.

2

Literature Review

Credit cards work simultaneously as a form of borrowing and a means for alternate transaction while debit cards works mainly as a medium of liquidity and an alternate transaction medium. Thus, credit and debit cards both share a common function in the form of an alternate

198 

Economics and Finance in Mauritius

transaction medium. However, both cards require sound infrastructural facilities such as the presence of ATMs along with firms which accept credit and debit cards as forms of payments. Consequently, any study that assesses the substitution effect of noncash payments should make sure that the country is endowed with a well-developed banking system with proper network coverage. The idea that the credit card represents an alternate transaction medium was first discussed by Akhand and Milbourne (1986). They stated that credit cards enabled agents to maintain lower money balances and more bonds. Similarly, Duca and Whitesell (1995) studied the impact of credit cards on money demand for US households and concluded that credit card ownership triggered a downward impact on checking and money balances. Similarly, Blanchflower et  al. (1998) found that credit cards unleashed decline in households’ transactions and precautionary demand for money. Yazgan and Yilmazkuday (2007) undertook an analysis on the Turkish economy, a small open economy. Their findings showed that both credit and debit cards generated a bearish impact on the demand for money. Rinaldi (2001) considered the effect of credit and debit cards, POS and ATMs on currency in circulation in Belgium to end up with a negative effect of POS terminals and ATMs while the effects of cards were found to be rather weak. In the of the Austrian cash demand, Stix (2004) discovered that cash demand was substantially impacted by debit card usage. Attanasio et al. (2002) assessed the demand for money in Italy between 1989 and 1995 using both firms and households data. They concluded that the demand for money of households who held an ATM card was much more elastic to the interest rate relative to households who did not possess an ATM card. Their findings entailed implications for inflation in that the welfare loss of inflation was greater for households in possession of a more sophisticated transaction technology as the latter enshrined the interest rate sensitivity of the demand for money. The negative effect of cards on currency in demand is often accounted for by the level of transaction value. De Grauve et al. (2000) noted that the average cost of card payments hovered around 1.3% of the transaction value while it stood around 9 percent of the transaction value when cash was used.

9  Impact of Debit and Credit Cards on Currency in Circulation... 

199

The empirical literature is also replete with studies conducted not only for specific countries but also for a specific group of countries. For example, Snellman et al. (2001) found strong evidence of a substitution effect among cash, cards, ATMs and POS terminals in the case of European countries. Drifting towards OECD countries, Drehmann et al. (2002) concluded that the number of POS terminals and ATMs had substantial effects on cash.

3

 ebit and Credit Cards Usage D in Mauritius

The use of cards in Mauritius is intrinsically related to the level of economic growth achieved. It is important to bear in mind, however, that credit cards often emanate from strategic company schemes and are thereby not as widely used as debit cards. In addition, the screening mechanism employed to become eligible as holders of credit cards by banks, high interest rates, strong security features and the preference by Mauritians for debit cards to withdraw notes rather than making direct payments, all point to the overwhelming use of debit cards in Mauritius. In fact, the ratio of the number of debit cards to credit cards stood at around 4:1 over the past decade with a maintained upward trend conspicuously noted for the years 2006 to 2008. Such a state of affairs shows clearly that debit cards considerably outnumber credit cards in Mauritius. Indeed, debit cards constitute the most famous cashless mode of payment by Mauritians for local transactions while credit cards tend to be favoured over more international payments such as online Amazon purchases. The present study is vital for Mauritius which possesses a vibrant banking sector and commendable infrastructural frameworks such as a satisfactory number of ATMs which are spread throughout the whole island (Fig. 9.1). In the period under analysis, there was a near doubling in the number of ATMs, from just 196 in June 1999 to 382 in June 2008. Therefore, it is anticipated that this will trigger a future decline in the level of currency in demand. In fact, Markose and Loke (2003) noted that the degree of credit and debit cards on cash demand could be magnified in instances

200 

Economics and Finance in Mauritius

Fig. 9.1  Ratio of debit to credit cards (Source: Bank of Mauritius monthly statistical bulletins, various issues)

where there was an enhancement in the network coverage of these cards. Thus, it is strongly believed that the number of ATMs reflected the best proxy to incorporate such network coverage. While the number of transactions rose by 186%, the value of transactions registered an increase of 243% for the same period under scrutiny. Relatively speaking, the increase in debit cards has been higher (153%) than that of credit cards (79%). Such a difference simply reflects the fact that Mauritians overwhelmingly make use of debit cards rather than credit cards. Moreover, in Mauritius, it can be easily conjectured that credit cards are principally used for online and international payments while debit cards are chiefly employed both for the withdrawal of notes and as a direct means of payment. It is vital to bear in mind that savings in time and costs and security reasons account for the use of debit cards, made feasible via the mushrooming of ATMs throughout Mauritius. Cheque payments are often highly labour-intensive and costly to local banks. As far as the security element is concerned, people tend to employ debit cards both for cash withdrawals and also for a direct mode of payment. The ratio of the value of notes to coins hovers around 27.07:1 over the period, demonstrating, as in other countries, that notes represent the lion’s share of currency in circulation. Such a finding is compatible with the notion that coins are chiefly reserved for low transaction value. In general, the Mauritian situation in terms of debit and credit card usage tilts more towards the European pattern, whereby debit cards are by far more diffused and used relative to credit cards, than to the experience in the USA, where credit cards predominate.

9  Impact of Debit and Credit Cards on Currency in Circulation... 

4

Econometric Models

4.1

First Model

201

To gauge the extent to which other modes of payments replaced the demand for currency in circulation, the following currency demand equation is used, spanning over a period of nearly ten years, from June 1999 to June 2008. All variables used are in their respective log form.



NCt = λ0 + λ1 Debit t + λ2 Credit t + λ3 ATMst + λ4 CPI t + (9.1) λ5 IR t + λ6 GDPt + λ7 NCt –1 + ε t N t = λ0 + λ1 Debit t + λ2 Credit t + λ3 ATMst + λ4 CPI t + λ5 IR t + λ6 GDPt + λ7 Nt - 1+ ε t (9.2)

Ct = λ0 + λ1 Debit t + λ2 Credit t + λ3 ATMst + λ4 CPI t + λ5 IR t + λ6 GDPt + λ7 Ct −1 + ε t

(9.3)

where NC Notes and Coins Debit Number of Debit cards Credit Number of Credit cards ATMs Number of ATMs CPI Consumer Price Index IR Interest rate GDP Gross Domestic Product A Coins N Notes The econometric model adopted for the current study is similar to that found in Rinaldi (2001). However, in this case the lagged values of the dependent variables have been given due consideration to account for any adjustment impacts occasioned by some “overshooting” or “undershooting” effects. In addition, the model is split into coins and notes, respectively. Alternatively stated, this study distinguishes itself from prior empirical methodology by factoring in notes and coins, both together and also in isolation. The rationale is that, for a developing country, the

202 

Economics and Finance in Mauritius

effect of debit and credit cards can be different depending on the type of currency in circulation. Apart for GDP and CPI, all data are being gleaned from Bank of Mauritius Monthly Statistical Bulletins which are freely accessible on its website. Gross Domestic Product (GDP) is anticipated to trail behind a positive effect on currency demand. The underlying rationale is based on the fact that a country which is subject to consistent growth will require monetary growth along the same lines to endorse its activities level. Based on the fact that GDP data are available only on a quarterly basis from the Mauritian Central Statistics Office, Mauritius, linear interpolation has been used to convert the quarterly data into monthly data. Interest rate is expected to post a downward effect on currency demand. A hike in interest rate causes a higher opportunity cost of holding money so that people will be more inclined to scale down their holdings of currency in favour of more profitable assets. The interest rate variable is constructed based on the evolution of both the Lombard Rate and the Key Repo Rate (the new interest rate benchmark which replaced the Lombard Rate). Inflation, measured by the Consumer Price Index (CPI), is anticipated to unleash a positive impact on currency demand due to inflationary adjustments along with transaction purposes. The CPI data are adjusted to a new base as at June 1999. ATMs capture the number of cash-dispensing ATMs. The impact of ATM on currency demand can be either positive or negative. A positive effect manifests itself in the case that there is a rise in the ease of making cash withdrawals. Conversely, a bearish impact occurs as ATMs scale down the transaction costs to consumers. Finally, as pointed out by Yazgan and Yilmazkuday (2007), the constant component isolates the level of technological enhancements. In that respect, it is anticipated to unleash a negative sign on the back that an enhanced technology level enables the use of other modes of payments other than currency balances. Lagged values of the dependent variables are included to account for any “overshooting” or “undershooting” effects. The correlation coefficients for all the variables are shown in the Appendix.

9  Impact of Debit and Credit Cards on Currency in Circulation... 

4.2

203

Second Model

NCt = λ0 + λ1 DebitUsaget + λ2 CPI t + λ3 IR t + λ4 GDPt + λ5 NCt –1 + ε t  (9.4) N t = λ0 + λ1 DebitUsage t + λ2 CPI t + λ3 IR t + λ4 GDPt + λ5 NCt –1 + ε t Ct = λ0 + λ1 DebitUsage t + λ2 CPI t + λ3 IR t + λ4 GDPt + λ5 NCt –1 + ε t

(9.5) (9.6)

It can be argued that the first model is imbued with a shortcoming in the proxies that are meant to measure credit and debit card usage because the value of the transaction has been ignored. To circumvent such a deficiency, a second model is run with a cleaner proxy being employed subject to the availability of data. Due to the unavailability of data pertaining specifically to the transaction value linked to credit cards and debit cards, respectively, the following proxy (denoted DebitUsage) is applied to best capture the component of debit card usage. Since the total value of transactions related to credit cards, debit cards, ATMs and Merchant Point of Sale (MPOS) is available, debit card usage is calculated using the latter figure divided by the total number of debit cards. By virtue of the fact that debit cards outnumber credit cards and ATMs in terms of both the number and the value of transactions, this proxy is anticipated to be effective. Beyond that, as per the Mauritian culture, many people avoid credit cards not only because banks are selective in their approach but also that they are reluctant to pay cumbersome penalty interests. Therefore, this means that a large chunk of the total value of transactions will be accounted for by debit cards.

5

Empirical Results

The correlation coefficient (Table 9.A.1) in the Appendix section shows that multicollinearity does not represent an issue of concern. Each of the residuals of the regressions are tested and all are found to be stationary. The results (Table 9.1) show that debit cards are linked with the expected

204 

Economics and Finance in Mauritius

negative impact while credit cards trigger a positive effect, though the coefficients are not statistically significant. Thus, the results show no real substitution effect between cash and cards. The impotency of credit cards, debit cards and ATMs demonstrate that notes and coins still represent the most coveted form of instrument for transactions. Alternatively stated, this signifies, however, that there had been a hike in the number of both debit cards, credit cards and ATMs over the period 1999–2008; nonetheless, this did not engender a decline in currency balances, showing the degree to which Mauritians still prefer to cling to currency balances for their transactions (Table 9.1). Table 9.1 Using number of credit, debit cards and ATMs as independent variables Dependent variables Notes & coins Constant D(DEBIT) D(CREDIT) D(ATMS) D(CPI) D(INT) D(GDP) D(NOTE & COINS(−1))

−4.26E-05 (−0.0112) −0.0409 (−0.1358) 0.0956 (0.4196) −0.0336 (−0.0623) −0.3432 (−0.3256) 0.0062 (0.4711) 1.5323 (6.6822)*** −0.3569 (−4.3693)***

D(NOTE(−1))

Notes −0.0001 (−0.0419) −0.0422 (−0.1356) 0.0981 (0.4173) −0.0391 (−0.0702) −0.3436 (−0.3160) 0.0063 (0.4624) 1.5827 (6.6900)***

0.0005 (1.1201) −0.0374 (−0.9224) 0.0023 (0.0754) 0.1163 (1.6068) −0.2150 (−1.5256) 0.0011 (0.6344) 0.1852 (6.1440)***

−0.3599 (−4.4105)***

D(COINS(−1)) Adjusted R-squared Durbin–Watson F-statistic Prob(F-statistic)

Coins

0.3353 2.3685 8.6403 0.0000

*** denotes statistical significance at the 1% level

0.3364 2.3716 8.6769 0.0000

0.6451 (9.9503)*** 0.5656 2.0721 20.7175 0.0000

9  Impact of Debit and Credit Cards on Currency in Circulation... 

205

As far as GDP is concerned, this is found to cause a positive effect on currency balances. More specifically, its effect is highest in the case of notes relative to coins. In fact, a 1% increase in GDP trails results in a 1.58% increase in notes but only a 0.18% increase in coins. Such a state of affairs is consistent with the fact that as the economy is subject to the prospects of higher economic growth, higher-value transactions impregnate more of the activities through higher denomination currencies, warranting a pronounced use of notes relative to coins. Such a finding was also noted by Snellman et al. (2001) who found that the greatest determinant of currency holdings in Europe was GDP. The distinct magnitude impacts on notes and coins separately, all emphasize on the need for specificity of analysis. The negative effect of ATMs means that they generate a substitute cash provision method in the case of notes and coins, jointly considered, and only notes when considered in isolation. However, the results are statistically insignificant. Interestingly, ATMs cause a positive impact on coins at the 11% significance level, reflecting that fact that since ATMs do not provide coins, ATMs and coins are deemed as complements by virtue of the fact that ATMs provide notes to settle large-value transactions while coins are used to settle low-value transactions. Inflation is anticipated to unleash a positive effect while interest rates have a negative impact. But the reverse manifests with none of the variables being statistically significant. It can thereby be conjectured that a hike in interest rate, synonymous with higher opportunity cost in terms of forgone earnings related to the holdings of alternative assets other than money, does not really have an influence on currency balances in Mauritius. Such a finding is congruous with the Mauritian situation as it has often been found that there is no significant impact of monetary policy decisions on listed stocks. The negative impact of inflation seems counterintuitive because a rise in the price level will encourage people to add on additional cash balances to maintain their purchasing power. Such a state of affairs merely reflects the Mauritian culture whereby, independent of the evolution of the interest rate and the inflation rate, people stick to the same level of currency for their transactions. As far as the correction variables are concerned, they point to some “overshooting” effects in the case of notes and some “undershooting” effects in case of coins.

206 

Economics and Finance in Mauritius

Table 9.2  Using a proxy for debit card usage as independent variable Dependent variables Notes & coins Constant D(DebitUsage) D(INT) D(GDP) D(CPI) D(NOTE & COINS(−1))

0.0022 (0.9616) 0.3056 (11.7217)*** 0.0007 (0.0904) 0.6088 (3.7188)*** −0.5005 (−0.7649) −0.0836 (−1.4525)

D(NOTE(−1))

Notes 0.0022 (0.9177) 0.3152 (11.7131)*** 0.0007 (0.0800) 0.6287 (3.7195)*** −0.5066 (−0.7502)

0.0006 (1.4003) 0.0168 (3.5554)*** 0.0005 (0.3062) 0.1567 (5.2156)*** −0.1469 (−1.1395)

−0.0865 (−1.5024)

D(COINS(−1)) Adjusted R-squared Durbin–Watson F-statistic Prob(F-statistic)

Coins

0.7234 2.5512 56.4600 0.0000

0.7236 2.5539 56.5202 0.0000

0.6613 (10.7412)*** 0.6076 2.0955 33.8364 0.0000

*** denotes statistical significance at the 1% level

The use of debit card usage (Table 9.2) is found to constitute a better metric because it posts a considerable increase in the explanatory power of the model. Moreover, the adjusted R2 is found to be relatively higher in the case of notes relative to coins, further imposing on the need to decompose currency in circulation. Compatible with prior empirical results, neither inflation nor interest rates exert any impact on notes or coins. Such impotency of both the interest rate and the CPI, even after incorporating the value of t­ransactions, shows the degree to which cultural factors permeate the Mauritian economy. The majority of Mauritians are highly conservative as far as finance is concerned, as evidenced by their actions of simply depositing their funds in banks and are thereby not watchful of the changes in interest rates as to switch to the local stock market. GDP is again found to trigger the highest level of economic significance on notes, with a 1% increase in GDP causing approximately a

9  Impact of Debit and Credit Cards on Currency in Circulation... 

207

0.63% rise in notes. Relatively speaking, the effect on coins is lower in magnitude, merely of 0.16%. Such a state of affairs is compatible with the fact that GDP will impact on notes more than coins by virtue of the fact that higher-value transactions require the use of higher-­denomination currencies such as notes. The low impact registered in the second model relative to the first model shows that results are highly sensitive to the choice of the proxy-whether number or value to be employed. The proxy employed to show debit card usage triggers a positive impact independent of the dependent variable under consideration, though the effect is of a greater magnitude in the case of notes. Based on the fact that debit cards outnumber credit cards, it can be deduced that increasing the use of debit cards has caused a positive effect on notes, with a 1% increase in transaction value engendering a 0.31% increase in notes. Such a finding is compatible with the fact that the increasing number of debit cards have occasioned improved access to notes so that this has been synonymous with a catalyst for notes in circulation. Thus, far from the results obtained by Snellman et al. (2001) and Yazgan and Yilmazkuday (2007), in the case of Mauritius, it can be concluded that the use of debit cards constitutes a key component of the electronic payments systems and that they have incited the use of notes in circulation. In brief, debit card usage represents a complement to rather than a substitute for the use of notes in Mauritius as people more often use debit cards to withdraw notes to meet their daily or weekly needs rather than to effect direct payments. Finally, only some “undershooting” effects are noted in case of coins.

6

Conclusion

This study assesses the utilization of alternative forms of payment relative to notes and coins in Mauritius, an upper-income developing country. Such an analysis is deemed important to find any substitution effect –whether the use of cards reduces the need for cash in circulation. Compared with prior models, our research provides due consideration for the specificity of the currency component. Such a decomposition is highly warranted, bearing in mind possible interactions between notes and coins, which can be seen as analogous to capital structure analysis, where short-term debt is usually separated from long-term debt. Another

208 

Economics and Finance in Mauritius

contribution of this study is that two distinct proxies are used, one to cater for the number of cards and one to capture the value of transactions. Using similar proxies as per Rinaldi (2001), the results in the first model show that the utilization of electronic forms of payments such as the number of debit cards, credit cards and the number of ATMs have not influenced the flow of currency in Mauritius. But, being sceptical about the use of number as a proxy, a richer proxy which is based on the value of transactions is used to try to attempt to capture the level of debit card usage, based on the available data from Bank of Mauritius monthly statistical bulletins. The reason is that the number of cards is not as important as the value of transactions carried out using these cards. Thus, this study advocates the use of a cleaner proxy for debit cards to disentangle its effect on currency in circulation. The newly used proxy for debit cards usage appears to be robust since both the adjusted R2 and the F-statistics are all higher. Interestingly, it transpired that the new proxy induces a positive impact on both notes and coins, with the effect being both economically and statistically higher in the case of notes relative to coins. Such a result bolsters the need for the specificity of analysis when gauging the effect of alternative modes of payments on currency balances. Most importantly, such a finding implies that debit cards have acted as complements rather than substitutes in inducing a higher level of notes in circulation in Mauritius. The underlying rationale for such a state of affairs relates to the easy access to ATMs at any point in time so that people make ­widespread use of debit cards more for note withdrawals to meet their daily or weekly transactions in lieu of making direct payments. Neither interest rate nor inflation are found to exert any impact on the level of currency in circulation. Independent of the model used, it transpires that GDP constitutes a core determinant pertaining to the level of currency in circulation. Such a finding is compatible with previous empirical evidence, as in the case of Snellman et al. (2001), who also found that the greatest determinant of currency holdings is GDP in Europe. In that respect, should there be any detrimental impacts on the Mauritian GDP, triggered either locally or externally, then a decline in currency in circulation is inevitably anticipated to occur.

9  Impact of Debit and Credit Cards on Currency in Circulation... 

209

The positive impact of debit card usage on notes is a good omen for the government in the case of any inflation tax revenue by virtue of the fact that the monetary base is actually being boosted via the use of debit cards. It can be thus be stated that debit cards indirectly enhance the seigniorage-­earning capacity of the government. Consequently, the present research finds evidence of a positive association between debit card usage and notes, in stark contrast to the findings of Yazgan and Yilmazkuday (2007), who noted that debit cards reduced seigniorage income in Turkey. Like any study, the present research is subject to certain shortcomings. For instance, due to data limitations, it has not been possible to single out the separate impacts of credit card usage and debit card usage because transaction values are present only in aggregate form. Despite this limitation, however, the proxy used to depict debit card usage is considered to be economically healthy because during the whole period under analysis, there have been four debit cards for every credit card held. Indeed, debit cards predominate over credit cards because Mauritians are usually reluctant to use credit cards due to high interest rates and for security reasons. On that basis, the debit card usage proxy is deemed to be both healthy and valid.

7

Policy Recommendations

The use of debit cards positively impacts on currency in circulation in Mauritius. Thus, Mauritians tend to use their cards mainly for cash withdrawals rather than for direct payments. Accordingly, the government is expected to gain more from seigniorage income. To reverse the trend, it could be that the authorities come up with policies which tend to reduce the fees charged by banks on their electronic payment devices which are being used by traders. Or the best strategy to induce direct payments is to significantly reduce all fees charged by banks on their electronic payment devices which are being used by traders. Finally, these electronic devices could be made accessible to all Mauritian ­traders as some still do not use them.

210 

8

Economics and Finance in Mauritius

Appendix

Table 9.A.1  Correlation coefficients for the variables (ATMS) (COINS) (CPI) (CREDIT) (DEBIT) (GDP) (INT) (NOT_COI) (NOTE)

(ATMS) 1.000000 0.195987 0.203618 0.129447 0.159487 0.209931 −0.037776 0.081250 0.080527

(COINS) 0.195987 1.000000 −0.041964 0.028151 −0.068222 0.410020 0.096964 0.505946 0.501810

(CPI) 0.203618 −0.041964 1.000000 −0.150331 0.135252 −0.114056 0.123113 −0.135146 −0.134639

(CREDIT) 0.129447 0.028151 −0.150331 1.000000 0.155911 0.183068 0.000749 0.152536 0.152719

(DEBIT) 0.159487 −0.068222 0.135252 0.155911 1.000000 0.061399 0.061169 −0.044696 −0.044216

(ATMS) (COINS) (CPI) (CREDIT) (DEBIT) (GDP) (INT) (NOT_COI) (NOTE) (DebitUsage)

(GDP) 0.209931 0.410020 −0.114056 0.183068 0.061399 1.000000 0.095105 0.496506 0.495452 0.391100

(INT) −0.037776 0.096964 0.123113 0.000749 0.061169 0.095105 1.000000 0.072949 0.072366 0.069202

(NOT_COI) 0.081250 0.505946 −0.135146 0.152536 −0.044696 0.496506 0.072949 1.000000 0.999982 0.812941

(NOTE) 0.080527 0.501810 −0.134639 0.152719 −0.044216 0.495452 0.072366 0.999982 1.000000 0.813081

(DebitUsage) 0.516796 0.550384 0.505919 0.494566 0.509769 0.588899 −0.021575 0.679028 0.681077 1.000000

Bibliography Akhand, H., & Milbourne, R. (1986). Credit cards and aggregate money demand. Journal of Macroeconomics, 8(4), 471–478. Attanasio, O., Jappelli, T., & Guiso, L. (2002). The demand for money, financial innovation, and the welfare cost of inflation: An analysis with household data. Journal of Political Economy, 110(2), 317–351. Bank of Mauritius Monthly Statistical Bulletins from June 1998 to June 2008. Blanchflower, D. G., Evans, D. S., & Oswald, A. J. (1998). Credit cards and consumers, NERA Working Papers, December. De Grauwe, P., Buyst, E., & Rinaldi, L. (2000). The costs of cash and cards compared: The cases of Iceland and Belgium, mimeo. Drehmann, M., Goodhart, C., & Krueger, M. (2002). The challenges facing currency usage: Will the traditional transaction medium be able to resist competition form the new technologies? Economic Policy, 34, 195–227.

9  Impact of Debit and Credit Cards on Currency in Circulation... 

211

Duca, J. V., & Whitesell, W. C. (1995). Credit cards and money demand: A cross-sectional study. Journal of Money, Credit, and Banking, 27(2), 604–623. Markose, S., & Loke, Y. J. (2003). The microstructure of recent trends in cashlessness: U.K. and U.S.A. compared, mimeo. Ramlall, I. (2009). Do credit and debit cards induce an evaporation of cash in Mauritius? International Research Journal of Finance and Economics, 36, 16–24. Rinaldi, L. (2001). Payment cards and money demand in Belgium, CES Discussion Paper DPS 01.16, KU Leuven. Snellman, J., Vesala, J., & Humphrey, D. (2001). Substitution of noncash payment instruments for cash in Europe. Journal of Financial Services Research, 19, 131–145. Stix, H. (2004). How do debit cards affect cash demand? Survey data evidence. Empirica, 31, 93–115. Yazgan, M. E., & Yilmazkuday, H. (2007). Monetary policy rules in practice: Evidence from Turkey and Israel. Applied Financial Economics, 17(1), 1–8.

10 A Posthumous Note on the Lombard Rate in Mauritius

The current study sheds light on the Lombard Rate which had been used as the main stance for monetary policy in Mauritius prior to the introduction of the Key Repo Rate. Results show that yields on Treasury Bills were cointegrated with movements in the Lombard Rate. In the long run, yields for different maturities adjusted more or less by the same amount. Most importantly, a 1% change in the Lombard Rate triggered approximately a 0.37% change in the respective Treasury Bills yields, showing that local market players did not fully adjust their bidding yields patterns. The most plausible explanation relates to an excessively strong level of aggressiveness in the Mauritian Treasury Bill market. Based on the ingrained Mauritian culture of depositing money mainly in banks, banks are thereby usually awash in cash which compel them to invest aggressively in the Treasury Bill market to ensure comfortable returns. Policy-­ wise, this implies that full monetary policy transmission mechanism onto the Treasury Bill market will always be incomplete on the back of conventional depositors, which ironically sounds like a good omen to the government in terms of undermined cost for public debt.

© The Author(s) 2017 I. Ramlall, Economics and Finance in Mauritius, DOI 10.1007/978-3-319-39435-0_10

213

214 

1

Economics and Finance in Mauritius

Introduction

On 13 December 2006, the Bank of Mauritius released a communiqué to inform the public that it will use the Repo Rate rather than the Lombard Rate as the main policy rate to signal changes in its monetary policy stance. However, to date no study has been undertaken in view of gauging the transmission mechanism of monetary policy onto the Treasury Bill (T-Bills) market. The objective of this chapter is to analyse the extent to which the Lombard Rate impacted on the short-term money market, namely on the T-Bills market. A priori, it is expected that causality will run from the Lombard Rate to 91-day, 182-day and 364-day Treasury Bills instruments, the only three different types of maturities present for Treasury Bills. It is important to note that the empirical literature on treasury yields are specific for each country and hence are usually scarce, let alone for upper-income developing countries. In that respect, this study directly contributes to such a scarce empirical literature. Furthermore, important policy implications are generated so that local authorities are mindful of their initiated actions. The rest of this chapter is organized as follows. Section 2 deals with the literature review. Section 3 focuses on the data and methodology parts while Sect. 4 discusses the results obtained. Finally, Sect. 5 concludes.

2

Brief Literature Review

The term structure of interest rates, which depicts the relationship between short- and long-term interest rates, is significant for market participants. For investors, the term structure of interest rates engender valuable information in terms of putting forth the future rate of returns. One of the key elements of the term structure pertains to the transmission mechanism of monetary policy. A key theory embedded under the term structure of interest rates reflects the expectations hypothesis, which means that the long-term interest rate constitutes a weighted average of current and expected future short-term interest rates and any liquidity premium. However, the empirical literature is still poor when it comes to

10  A Posthumous Note on the Lombard Rate in Mauritius 

215

analysing developing countries. Most studies are tilted towards advanced or developed economies. Many studies have been undertaken in view of testing the validity of the expectations hypothesis. Indeed, most studies clung to linear cointegration approach to sieve out the term structure of interest rates, such as those of Campbell and Shiller (1987), Hall et al. (1992), Engsted and Tanggaard (1994) and Gerlach and Smets (1997) among many others. Certain studies have also applied sophisticated techniques such as nonlinear cointegration such as those of Clements and Galvao (2003), Clarida et al. (2006), and Mili et al. (2012). Many studies have been undertaken for the Eurozone. For instance, Musti and D’Ecclesia (2008) assessed the validity of the expectations hypothesis for Italy and Germany to find that interest rates for distinct maturities are inherently connected over time. In a parallel manner, Koukouritakis and Michelis (2008) assessed 28 countries of the European Union to find positive evidence.

3

Data and Methodology

The analysis uses weekly data for the period from 2000 to December 2006. All data were collected from the Bank of Mauritius website. As depicted in Table 10.1, neither the Lombard Rate nor the Treasury Bill yields follow a normal distribution shown by the Jarque–Bera tests. While skewness is positive for the Lombard Rate, it is systematically negative for the Treasury Bills. As far as kurtosis is concerned, none is close to 3 (Table 10.1).

3.1

Unit Root Tests

Prior to undertaking any cointegration analysis, it is important to undertake unit root tests. All series were taken in their logarithmic form. The results are shown in Table 10.2. The time series are not stationary in levels. However, the first differences of the logarithmic transformations of the series are stationary. In

216 

Economics and Finance in Mauritius

Table 10.1 Summary statistics for Lombard Rate and yields on different maturities Lombard Rate Mean Median Maximum Minimum Std. Dev. Skewness Kurtosis Jarque-Bera Probability Observations

Bank rate

182 Days

364 Days

91 Days

10.99919 8.211802 7.934235 8.768827 11.50000 8.263940 7.830000 8.730000 13.00000 12.04901 11.80000 12.33000 9.500000 2.563659 2.560000 2.540000 1.048166 2.098508 1.919887 2.273226 0.111015 −0.199146 −0.224527 −0.160169 1.841508 2.001803 2.027889 2.064204 17.79835 14.77482 14.66756 12.51448 0.000137 0.000619 0.000653 0.001917 307 307 307 307

Table 10.2  Unit root tests Lombard Rate Bank Rate 91 days 184 days 364 days *

7.459332 7.220000 11.56000 2.580000 1.871813 −0.076089 1.878024 16.39877 0.000275 307

Level

First difference

−0.3274 −0.9960 −0.5296 −0.4829 −0.9641

−17.3713* −17.6637* −12.7926* −13.6952* −8.5259*

Denotes statistical significance at the 1% level

that respect, the series are said to be integrated of order one I (1), making them eligible for the cointegration analysis.

3.2

 ointegration Tests Between Lombard Rate C and the Yields of Treasury Bills

The cointegration results generate one cointegrating equation for each individual pair of analysis done. A pair of analysis refers to analysing the relationship between the Lombard Rate and a given Treasury Bill yield maturity. The fact that the optimal lag was one week clearly shows that T-Bills tend to co-move together on a weekly basis (Table 10.3).

10  A Posthumous Note on the Lombard Rate in Mauritius 

217

Table 10.3  Results of the cointegration test under weekly horizon

H0

H1

Eigenvalue

Trace statistic

Max eigenvalue statistic

91 days and Lombard rate (Optimal lag = 1) r=0 r>0 0.0582 20.4148* r=1 r>1 0.0068 2.1111 182 days and Lombard rate (Optimal lag = 1) r=0 r>0 0.0552 19.0155** r=1 r>1 0.0056 1.7054 364 days and Lombard rate (Optimal lag = 1) r=0 r>0 0.0518 17.6470** r=1 r>1 0.0046 1.4172 Bank rate and Lombard rate (Optimal lag = 1) r=0 r>0 0.0586 20.1915* r=1 r>1 0.0058 1.7667

Cointegrating equations

18.3036** 2.1111

1

17.3100** 1.7054

1

16.2299** 1.4172

1

18.4247** 1.7667

1

Critical values for statistical significance Critical values Trace statistic H0; H1 r = 0; r > 0 r = 1; r > 1 *

5% 15.4947 3.8414

1% 19.9371 6.6348

Max eigenvalue statistic 5% 14.2646 3.8414

1% 18.52 6.6348

Significant at 1% level Significant at 5% level

**

4

Results

Based on the fact that cointegration prevailed, the error correction model or equilibrium correction models are presented for each maturity of the Treasury Bills. All the long-run coefficients were statistically significant except for the short-run relationships. Hence, comments are made only on the speed of adjustment coefficients and the coefficients describing the long-run relationships. All of the results are shown in the appendix section (see Tables 10.A.1, 10.A.2, 10.A.3, 10.A.4). The results obtained are compatible with the intuitive explanation that the speed of adjustment coefficient back to equilibrium is always much higher in case of the impact of Lombard Rate on the distinct T-Bills

218 

Economics and Finance in Mauritius

maturities as yields of T-Bills should follow the stance of monetary policy undertaken by the Bank of Mauritius. On a maturity-wise basis, the speed of adjustment for 91 T-bills is higher than that of 182 T-Bills which is higher, in turn, compared to the 364 T-Bills. Such a result suggests that market players tend to focus on the lower end maturity rather than on upper end maturity when bidding. Gauging on the long-run relationship, it transpires that the elasticity is not so different for the three maturities, whereby a 1% change in the Lombard Rate elicits approximately a 0.37% change in the distinct yields. The results derived for the bank rate are practically the same as for the Treasury Bill yields. Such a result bodes well given the fact that whenever there have been changes in the Lombard no full instantaneous adjustments manifested in the yields of Treasury Bills. For instance, it is a well-known fact that bidders from the banking sector are usually slow in adjusting their yields.

4.1

Predictive Power

Prior to analysing the predictive power of one specific maturity for another maturity, it was considered important to remove a break in the data that occurred between 26 March 2004 and 23 July 2004. The removal of such a break is not considered vital in the previous analysis since the previous econometric model is based between the Lombard Rate and each of the three maturities for Treasury Bills. However, in the current context, the analysis is based within the yields themselves under distinct treasury bills maturities. The cointegration tests shown in Table 10.4 confirm the presence of one cointegrating equation for all three pairs under investigation.

4.1.1  91 Days Treasury Bills and 182 Days Treasury Bills Results in Table 10.A.5 show that in the short run, only 91-day Treasury Bills yields affect those of 182-day yields and not vice versa, adding support to the previous finding that bidders basically adjust yields by

10  A Posthumous Note on the Lombard Rate in Mauritius 

219

Table 10.4  Results of the cointegration test under weekly horizon Trace statistic

Max eigenvalue statistic

H0 H1 Eigenvalue 91 days and 182 days (Optimal lag = 1) r=0 r>0 0.056 17.1649* 16.4845** r=1 r>1 0.0024 0.6803 0.6803 91 days and 364 days (Optimal lags = 4) r=0 r>0 0.0531 16.1644** 15.4827** r=1 r>1 0.0024 0.6817 0.6817 182 days and 364 days (Optimal lags = 4) r=0 r>0 0.0525 15.6258** 15.3126** r=1 r>1 0.0011 0.3132 0.3132 Critical Values for Statistical Significance

Cointegrating equations 1

1

1

Critical values Trace statistic H0; H1 r = 0; r > 0 r = 1; r > 1 *

5% 15.4947 3.8414

1% 19.9371 6.6348

Max eigenvalue statistic 5% 14.2646 3.8414

1% 18.52 6.6348

Significant at 1% level Significant at 5% level

**

moving from shorter maturities to higher maturities, and not vice versa. Above all, in the long run the elasticity of 91-day yields to 182-day yields is 1.05%, which is intuitively correct as yields are expected to undergo the same level of adjustment over time.

4.1.2  91 Days Treasury Bills and 364 Days Treasury Bills Results in Table  10.A.6 show that in the short run, 364-day Treasury Bills yields positively affect those of 91 day yields while 91-day Treasury Bills negatively impact on 364-day Treasury Bills yields. Above all, in the long-run the elasticity of 91-day yields to 182-day yields is 1.00%, which is again intuitively correct.

220 

Economics and Finance in Mauritius

4.1.3  182 Days Treasury Bills and 364 Days Treasury Bills In case of 182 and 364, it surfaced that, in the short run, 182 affects 364 up to three lags and 364 affects 182 up to three lags. However, while 182 negatively affects 364, 364 positively impacts on 182. The elasticity of 364 with respect to a 1% change in 182 is 1.04%, more or less similar to previous findings (Table 10.A.7).

5

Conclusion

The current study undertakes a posthumous analysis of the Lombard Rate which has been used as the main stance for monetary policy up to the year 2006. Results, in general, tend to suggest that market players tend to focus on the lower-end maturity rather than on upper-end maturity when bidding. Another major finding is that the speed of adjustment coefficient back to equilibrium is always much higher in case of the impact of Lombard Rate on the distinct Treasury Bills maturities, clearly showing the transmission of the monetary channel. The most important finding is that a 1% change in the Lombard Rate triggers approximately a 0.37% change in the respective Treasury Bills yields, showing that local market players do not fully adjust their bidding patterns once there is a change in the Lombard Rate. The most plausible explanation is that this may show the extent of aggressiveness in the Mauritian Treasury Bill market since it is a well-known fact that bidders from the banking sector tend to be awash with excess liquidity. Mauritians tend to be highly risk-averse and thereby tend to invest mainly as depositors in banks. A lack of knowledge of sophisticated tools constitutes another hindrance in view of tapping other markets such as the equity market, as evidenced by Ramlall (2014). Consequently, banks tend to be highly awash in cash and this has usually been evidenced by the excessive liquidity in the banking sector. Based on the need for banks to generate ample returns, it is of paramount significance that banks seek sound investments to trigger sustainable returns. Consequently, banks invest aggressively in the Treasury

10  A Posthumous Note on the Lombard Rate in Mauritius 

221

Bill market to ensure comfortable returns. Such a state of affairs signifies that full monetary policy transmission mechanism onto the Treasury Bill market will never take place. Ironically, this implies that the effect of monetary policy changes will never be fully felt by the government in terms of burgeoning costs of public debt. In a nutshell, Mauritians’ preference for bank deposits triggers excess cash to banks which compel them to invest massively in Treasury Bills which, in turn, benefits the government in terms of lower cost of public debt.

6

Policy Recommendations

The imperfect transmission channel of the previously used Lombard Rate onto the Treasury Bill market yields highlights the degree of aggressive bidding by banks. This indirectly substantiates one of the core structural problems in the Mauritian financial system – namely, that of excess liquidity. To tackle the root cause of the problem, the authorities should enlarge the asset investment types of households so that the latter do not keep their investments simply in bank deposits. More and more financial literacy campaigns with respect to investment products should be organized on a regular basis. Above all, fiscal policy could be used. For example, the application of lower income tax rates such as 10% in lieu of 15% to households who keep at least 10% of their monthly income in equities, could be a truly robust fiscal incentive tool that would help to stimulate non-deposit asset holdings in Mauritius. To cater for more and more households, listed firms would be required to have some stipulated thresholds with respect to the percentage of total shares being held by major shareholders, along with a reasonable number of floating shares. Ironically, the Mauritian stock exchange will benefit from a larger participation base which renders the market more liquid. In a parallel manner, the central bank will also benefit as a greater involvement in equities by households will generate switching effects between equities and bank deposits, hinging on the degree of interest rate changes, all working towards rekindling the interest rate transmission mechanism of monetary policy in Mauritius. However, a curtailing of the high interest rate spreads will also be required to further boost the interest rate channel of monetary policy in Mauritius.

222 

7

Economics and Finance in Mauritius

Appendix

Table 10.A.1  Results for 91 days ΔLombard Rate 0.0352 (3.0784)* −0.0584 (−0.9327) 0.0100 (0.6167) 0.0001 (0.2527) 0.0214 3.2174

Δ 91 days 0.1722 (3.9950)* 0.1783 (0.7541) 0.0448 (0.7290) 0.0004 (0.2034) 0.0559 6.9942

91 dayst − 1 −0.3757 (−8.9434)*

Constant −1.6512

ΔLombard Rate 0.0294 (2.8081)* −0.0451 (−0.7336) 0.0055 (0.3431) 0.0001 (0.2511) 0.0159 2.6386

Δ 182 days 0.1543 (3.9109)* 0.0370 (0.1599) 0.0968 (1.6038) 0.0003 (0.1803) 0.0538 6.7669

182 dayst − 1 −0.3851 (−8.0566)*

Constant −1.6082

Cointegration equation 1 ΔLombard ratet − 1 Δ91 dayst − 1 Constant Adjusted R2 F-Statistic Cointegration equation 1 Lombard ratet − 1 1.0000 Denotes statistical significance at 1% level

*

Table 10.A.2  Results for 182 days Cointegration equation 1 ΔLombard ratet − 1 Δ182 dayst − 1 Constant Adjusted R2 F-Statistic Cointegration equation 1 Lombard ratet − 1 1.0000

Denotes statistical significance at 1% level

*

10  A Posthumous Note on the Lombard Rate in Mauritius 

223

Table 10.A.3  Results for 364 days ΔLombard Rate 0.0246 (2.6600)* −0.0451 (−0.7498) 0.0088 (0.5794) 0.0001 (0.2539) 0.0139 2.4321

Δ 364 days 0.1341 (3.7337)* −0.0435 (−0.1870) 0.1389 (2.3539)** 9.67e-05 (0.0454) 0.0556 6.9758

364 dayst − 1 −0.3711 (−7.1744)*

Constant −1.6016

ΔLombard Rate 0.0287 (3.0480)* −0.0551 (−0.9322) 0.0102 (0.8246) 0.0001 (0.2562) 0.0214 3.2118

Δ Bank rate 0.1657 (3.7418)* 0.3944 (1.4219) −0.0312 (−0.5371) 0.0003 (0.0026) 0.0532 6.6911

Bank ratet − 1 −0.3776 (−7.8041)*

Constant −1.6120

Cointegration equation 1 ΔLombard ratet − 1 Δ364 dayst − 1 Constant Adjusted R2 F-Statistic Cointegration equation 1 Lombard ratet − 1 1.0000 *

Denotes statistical significance at 1% level

Table 10.A.4  Results for bank rate Cointegration equation 1 ΔLombard ratet − 1 Bank ratet − 1 Constant Adjusted R2 F-Statistic Cointegration equation 1 Lombard ratet − 1 1.0000 *

Denotes statistical significance at 1% level

Table 10.A.5  Results for 91 days and 182 days Cointegration equation 1 Δ91 dayst − 1 Δ182 dayst − 1 Constant Adjusted R2 F-Statistic Cointegration equation 1 91 dayst − 1 1.0000 *

Δ91 days 0.0873 (1.1786) 0.1928 (1.4826) 0.0747 (0.5590) 0.0004 (0.3097) 0.0716 8.3025

Δ 182 days 0.2285 (3.3366)* 0.2890 (2.4003)** −0.0885 (−0.7163) 0.0003 (0.2931) 0.1128 13.1190

182 dayst − 1 −1.0565 (−72.4255)*

Constant 0.1836

and **Denotes statistical significance at 1% and 5% levels, respectively

Table 10.A.6  Results for 91 days and 364 days Cointegration equation 1 Δ91 dayst − 1 Δ91 dayst − 2 Δ91 dayst − 3 Δ91dayst −4 Δ364 dayst −1 Δ364 dayst −2 Δ364 dayst −3 Δ364 dayst − 4 Constant Adjusted R2 F-Statistic Cointegration equation 1 364 dayst − 1 1.0000 *

Δ91 days 0.0512 (1.5055) −0.1786 (−1.6516) −0.4399 (−4.1071)* −0.2093 (−1.9234) −0.0446 (−0.4272) 0.3027 (3.4538)* 0.6539 (5.6580)* 0.3926 (3.1672)* 0.0532 (0.4303) 0.0009 (0.8328) 0.1986 8.7941

Δ 364 days 0.1029 (3.2852)* −0.2055 (−2.0602)** −0.3729 (−3.7756)* −0.1031 (−1.0269) −0.0686 (−0.7120) 0.3061 (2.9196)* 0.5341 (5.0119)* 0.2226 (1.9477) 0.0325 (0.2849) 0.0004 (0.3822) 0.1366 5.9763

182 dayst − 1 −1.0084 (−27.6438)*

Constant 0.1834

and **Denotes statistical significance at 1% and 5% levels, respectively

10  A Posthumous Note on the Lombard Rate in Mauritius 

225

Table 10.A.7  Results for 182 days and 364 days Cointegration equation 1 Δ364 dayst − 1 Δ364 dayst − 2 Δ364 dayst − 3 Δ364 dayst −4 Δ182 dayst −1 Δ182 dayst −2 Δ182 dayst −3 Δ182 dayst − 4 Constant Adjusted R2 F-Statistic Cointegration equation 1 364 dayst − 1 1.0000 *

Δ182 days −0.1250 (−3.1240)* 0.3613 (3.2295)* 0.6019 (5.3927)* 0.2430 (1.9807)** 0.0168 (0.1311) −0.2862 (−2.5709)* −0.4567 (−4.2323)* −0.1329 (−1.1809) −0.0297 (−0.2565) 0.0004 (0.4220) 0.1596 6.5908

Δ364 days −0.0596 (−1.4395) 0.3320 (2.8697)* 0.6901 (5.9787)* 0.4980 (3.9251)* 0.1859 (1.4035) −0.2503 (−2.1752)** −0.4995 (−4.4761)* −0.3509 (−3.0143)* −0.0883 (−0.7363) 0.0010 (0.9347) 0.2055 9.1324

182 dayst − 1 −1.0491 (−34.4774)*

Constant 0.0035

and **Denotes statistical significance at 1% and 5% levels, respectively

Bibliography Bank of Mauritius. (2006). A new framework for the conduct of monetary policy by the Bank of Mauritius. Communiqué, 13 December. https://www. bom.mu/monetary-policy/monetary-policy-framework/communiquenew-framework-conduct-monetary-policy-bank Campbell, J. Y., & Shiller, R. J. (1987). Cointegration and tests of present value models. The Journal of Political Economy, 95, 1062–1088. Clarida, R. H., Sarno, L., Taylor, M. P., & Valente, G. (2006). The role of asymmetries and regime shifts in the term structure of interest rates. Journal of Business, 79, 1193–1224.

226 

Economics and Finance in Mauritius

Clements, M. P., & Galvao, A. B. (2003). Testing the expectations theory of the term structure of interest rates in threshold models. Macroeconomic Dynamics, 7, 567–585. Engsted, T., & Tanggaard, C. (1994). A cointegration analysis of Danish zero-­ coupon bond yields. Applied Financial Economics, 4, 265–278. Gerlach, S., & Smets, F. (1997). The term structure of Euro-rates: Some evidence in support of the expectations hypothesis. Journal of International Money and Finance, 16, 305–321. Hall, S.  G., Anderson, H.  M., & Granger, C.  W. J. (1992). A cointegration analysis of treasury bill yields. The Review of Economics and Statistics, 74, 116–126. Johansen, S., & Juselius, K. (1990). Maximum likelihood estimation and inference on cointegration-with applications to the demand for money. Oxford Bulletin of Economics and Statistics, 52, 169–210. Koukouritakis, M., & Michelis, L. (2008). The term structure of interest rates in the 12 newest EU countries. Applied Economics,Taylor & Francis Journals. 40 (4), 479–490. Mili, M., Sahut, J. M., & Teulon, F. (2012). New evidence of the expectation hypothesis of interest rates: A flexible nonlinear approach. Applied Financial Economics, 22, 165–176. Musti, S., & D’Ecclesia, R. L. (2008). Term structure of interest rates and the expectations hypothesis: The eurozone. European Journal of Operational Research, 185, 1596–1606. Ramlall, I. (2014). Is there a pecking order in the demand for financial services/ product in Mauritius? Journal of African Business, 15(1), 49–63.

Part V The Business Sector and Its Financing Structure in Mauritius

11 What Drives the Capital Structure of Non-­Listed Mauritian Firms?

This chapter examines the capital structure of non-listed ­non-financial firms in Mauritius. Distinct metrics of leverage are given due consideration, including the use of liabilities, leases, loans and debt. Most importantly, all these distinct metrics of leverage are being assessed in their dual forms – short-term and long-term components, respectively – in order to gain enhanced insight of the capital structure. The findings show that the liquidity and size unleash bearish effects on leverage. While assets tangibility positively impacts on leverage, profitability, non-debt tax shield and growth variables are found to trail behind no impacts. The findings reveal it is feasible that there may be crowding-out effects of loans by leases based on investment which induces the use of leases but reduces the use of loans. The potency of the age variable is also noted. Overall, findings advocate a modified pecking order for Mauritian firms in which case short-term leverage precedes long-term leverage.

© The Author(s) 2017 I. Ramlall, Economics and Finance in Mauritius, DOI 10.1007/978-3-319-39435-0_11

229

230 

1

Economics and Finance in Mauritius

Introduction

Modigliani and Miller (1958) argued that, in the case of perfect market conditions, capital structure is entirely irrelevant. In the real world, however, various forms of market imperfections manifest to trigger three theories of capital structure: the static trade-off theory (Myers 1984a), the pecking order theory (Myers and Majluf 1984), and the agency cost theory. The static trade-off theory argues that the optimal capital structure is obtained at the point where the net tax benefit of debt financing is aligned to the leverage-related costs, such as financial distress and bankruptcy costs. The pecking order theory states that there is a well-specified order in the financing structure, beginning from retained earnings, before shifting to debt and only resorting to equity as a last resort. The agency cost theory pertains to agency problems which manifest between managers and shareholders. The agency cost theory tends to have over poor application in Mauritius because nearly all of the firms have their shareholders as managers. Such a state of affairs ironically applies even to listed firms which are characterized by high levels of insider ownership. In the same vein, the static trade-off theory is deemed to be of limited relevance to Mauritius because there prevail no robust signs as to firms resorting to a balance between tax saving occurring from debt use against any financial distress costs. Consequently, it becomes evident that the pecking order theory tends to offer the best reflection of the capital structure of nonfinancial Mauritian firms. But the pecking order theory has often been analysed solely in relation to listed non-financial firms which may somewhat bias the analysis in the case of developing countries. The may be because very few firms are usually listed on the stock market of a given developing country. Ironically, even if listed, the share price of the firm may not generate an authentic picture of the performance of the firm on the back of liquidity issues. In addition, in cases where there are only a few listed non-financial companies, this tends to scale up the extent of bias in the results obtained due to the presence of outliers. The latter are caused by few firms found in different sectors so that the distribution of the computed ratios are likely to

11  What Drives the Capital Structure of Non-Listed 

231

be highly widespread. Above all, the use of listed firms, of which there are only a small number, does not reflect the general trend of capital structure for the country since any analysis should include mostly non-listed companies, which are larger in number. In that respect, it is considered best to have recourse to non-listed non-financial firms so as to reduce any feasible bias in capital structure analysis. Subsequently, to analyse the capital structure of firms situated in a developing country, it is important to factor in non-listed firms to obtain a real and complete picture of the capital structure stance prevailing in that country. Another key contribution of the present study is that the focus is laid on various versions of leverage, such as liabilities, leases, loans and debt. Such a decomposition is deemed to be particularly vital because different attributes of the capital structure may entail distinct impacts, all of which are dependent on the type of leverage used. Finally, this study adds to the scarce literature on capital structure for developing countries based on the fact that most capital structure studies often focus on developed countries. In general, results advocate a modified pecking order theory but one that is removed of its equity component and where short-term leverage predominates over long-term leverage. The analysis probes into the determinants of capital structure for various non-listed Mauritian firms for the period 2005–2006. The next section focuses on the literature review pertaining to capital structure. The section after that specifies the econometric model along with the relationships likely to hold between leverage and the various independent variables. Therefore, the section afterwards discusses on the results obtained. Finally, section 5 concludes.

2

Pecking Order Theory

Deemed as one of the most influential theories of capital structure, the pecking order theory was introduced by Donaldson (1961), but it obtained its rigorous theoretical foundation by Myers and Majluf (1984). In brief, the pecking order theory pinpoints an order in the selection of finance on the back of distinct degrees of information asymmetry and related agency costs embedded in different sources of finance.

232 

Economics and Finance in Mauritius

Consequently, retained earnings are used first because they represent the cheapest means of funding, hardly being influenced by any information asymmetry. Second, debt is employed as there is weak information asymmetry by virtue of fixed obligations acting as an effective ­monitoring device. Finally, external equity is used only as a last resort because it transmits an adverse signaling impact as evidenced through event studies. Overall, the pecking order theory is also compatible with shareholder’s wealth maximization as it is geared towards reducing the cost of raising finance. Watson and Wilson (2002) assessed the empirical validity of the pecking order theory in the case of UK small and medium-sized enterprises (SMEs) which were given a threefold classification into high information asymmetry firms, low information asymmetry firms and closely-held firms. In addition, they decomposed debt into hire purchase liabilities, longterm debt, short-term debt and intra-group debt balances, respectively, in order to generate better insight of the various drivers which impacted on leverage. Their findings showed that the pecking order theory rejoiced over considerable support in the case of closely-held firms which were characterised by low information asymmetry. In addition, Watson and Wilson (2002) found a pecking order within debt because the explanatory power of their estimated models rose substantially when debt was split into its distinct components. Thus, their main findings emphasized on the need to analyse the pecking order theory under specificity of leverage. Bevan and Danbolt (2000) investigated the pecking order theory in the UK credit market from 1991 to 1997. They did this by splitting debt into its individual components in order to spawn a holistic analysis of capital structure decisions and also to avoid any feasible bias in the selection of the gearing ratio. They concluded that over time UK firms had moved gradually from debt finance to equity finance. They attributed such growth in equity to areas of high technology and the internet. Above all, they noted that bigger firms, which were historically more reliant on long-term debt, were now resorting to increased levels of equity finance. In lieu of contradicting the pecking order theory, they stressed the significance of supply forces which might cause the preference of equity to long-term debt. In the case of manufacturing SMEs in Australia, Zoppa and McMahon (2002) found robust empirical evidence in favour of the pecking order

11  What Drives the Capital Structure of Non-Listed 

233

theory. Nonetheless, they pointed out the need for a modified pecking order theory tailored to the prevalent characteristics of SMEs financing. The modified pecking order theory proposed by them was to utilize ­internal profits first, and then to make use of short-term credit such as trade credit and personal credit cards. Afterwards, recourse is made towards long-term debt and then equity capital from current owners; it is only as a last resort that fresh outside equity was used. Grinder and Gordon (1995) focused on the conflict between the pecking order theory and the managerial entrenchment hypothesis to draw out any relationship between capital expenditure and insider ownership. They based their analysis on the fact that the managerial entrenchment hypothesis is related to an inverse relationship between insider ownership and capital expenditures while the pecking order theory posits no such link. Hence, they noted positive evidence in favour of the pecking order theory because their research findings depicted no such association between capital expenditure and insider ownership. They thereby deduced that the reliance on internal cash flow for capital expenditure financing was chiefly the result of information asymmetry between managers and new shareholders in lieu of information asymmetry among insiders, which was compatible with the pecking order theory. They further concluded that shareholders did not need to strive for high insider ownership to increase their wealth because the managerial entrenchment hypothesis was empirically limited.

3

Model Specification

Data are collected from the Registrar of Companies in Mauritius for about 450 firms for the period 2005–2006. Because of the presence of outliers, some firms are being overlooked, meaning that a total of 395 firms are used for the analysis. The model used for the analysis is obtained from previous studies on the pecking order theory such as those of Ozkan (2001), Bevan and Danbolt (2000) and Titman and Wessels (1988). The selected model is believed to capture the essence of the subject under scrutiny.

234 

Economics and Finance in Mauritius

LEVi = β 0 + β1GRO i + β 2 SIZE i + β3 TANG i + β 4 PRO i + β 5 LIQ i + β6 NDTSi + β 7 AGE i + β8 INVi + U i

i captures a given non-financial firm. LEV represents Leverage, GRO represents growth, SIZE represents size, TANG represents the tangibility of assets, PRO represents profitability, LIQ represents liquidity, NDTS represents non-debt tax shield, AGE represents age of the company and INV represents investment. These variables are known as the theoretical features of capital structure. Bevan and Danbolt (2000) argued that there was a danger when analysing capital structure determinants based on overly aggregate metrics of gearing. Similarly, Hall et al. (2000) pointed out that the examination of both short-term and long-term components of leverage allowed differentiation among factors influencing each of them, respectively. Thus, the above equation was estimated for distinct dependent variables – namely, total liabilities, long-term liabilities, short-term liabilities, long-term leases, short-term leases, long-term loans, short-term loans, long-term debt and short-term debt. All of the variables are defined in Table 11.1 with the summary statistics depicted in Table 11.2. Capital structure is influenced by firm-specific features which are reflected by the various theoretical attributes of capital structure. These theoretical attributes are discussed for each of the explanatory variables.

3.1

Non-Debt Tax Shields (NDTS)

Firms which are subject to other tax shields such tax deductions for depreciation and investment tax credits have lesser incentives to exploit on the debt tax shield. Thus, under the pecking order theory a negative relationship is expected to prevail between NDTS and financial leverage. Nonetheless, Scott (1977) and Moore (1986a) discussed that considerable NDTS could behave as attractive collateral and so it could stimulate the intake of higher debt levels. Subsequently, in this case, a positive relationship is anticipated to manifest.

235

11  What Drives the Capital Structure of Non-Listed  Table 11.1  Definition of variables Variable

Definition

Non-debt tax shield

Depreciation over earnings before interest and tax Net profits over total assets Fixed assets over total assets Natural logarithm of total assets Percentage change in total assets Current assets over current liabilities Logarithm of number of years since date of incorporation Purchase of equipment over total assets Total liabilities over total asset Long term liabilities over total asset Short term liabilities over total asset Long term leases over total asset Short term leases over total asset Long term loans over total asset Short term loans over total asset Long term loans and leases over total asset Short term loans, leases and overdraft over total asset

Profitability Tangibility Size Growth Liquidity Age Investment Total liabilities Long-term liabilities Short term liabilities Long term leases Short term leases Long term loans Short term loans Long term debt Short term debt

Table 11.2  Summary statistics Variable

Mean

Non-debt tax shield Profitability Tangibility Size Growth Liquidity Age Investment Short term leverage Long term leverage Total leverage Short term leases Long term leases Short term loans Long term loans Short term debt Long term debt

0.4992037 0.0515638 0.2764015 17.73889 0.1554434 1.626323 2.672048 0.0565327 0.5503352 0.1613985 0.7117337 0.0104019 0.0248124 0.029667 0.0991975 0.1597168 0.1240099

Std. Dev. 1.197958 0.1633851 0.2279746 1.675925 0.2999875 1.388988 0.8087456 0.0797295 0.2696857 0.1806385 0.312365 0.020369 0.0476112 0.0569948 0.1677942 0.1617667 0.1696332

Min −2.369584 −1.488931 0 12.07862 −0.6196301 0.1155688 −1.227069 0 0.0352703 0 0.0428754 0 0 0 0 0 0

Max 7.444458 0.9951875 0.9289668 23.16809 1.848505 10.03107 4.665246 0.502 1.454132 1.097464 2.156737 0.1271789 0.3149749 0.3593782 1.097464 0.7828515 1.097464

236 

3.2

Economics and Finance in Mauritius

Size

Rajan and Zingales (1995) found that larger firms had lower leverage because the information asymmetry level was higher by virtue of their complex structures. Fama and French (2002) pointed out, however, that bigger firms could leverage more because their size was synonymous with the possession of robust financial resources. Nonetheless, Barclay et al. (1995) noted that a split of financial leverage into short-term debt and long-term debt triggered a positive association between long-term debt and size, but a negative relationship between short-term debt and size. The reason for this was the fact that small firms were inclined to borrow more short-term debt because of the lower fixed costs related to short-term borrowing. A large number of studies uncovered a significant positive association between size and debt ratio (Lasfer 1995; Rajan and Zingales 1995; Barclay and Smith 1996; Berger et.al. 1997). Kester (1986) and Remmers et al. (1974) noted no considerable impact of size on capital structure.

3.3

Growth (GRO)

Myers (1984a) identified an inverse relationship between growth and financial leverage on the back of high interest rates or restrictive covenants which deterred debt taking. Titman and Wessels (1988) attributed such a negative association to the bondholders’ reluctance to lend to equity-controlled firms as the latter were inclined to invest suboptimally to expropriate wealth from bondholders. Baskin (1989) noted a significant positive association between growth and leverage. In addition, according to the increased operating efficiency of Higgins (1977), such a negative association emerged from better-managed firms which depended less on outside financing.

3.4

Profitability (PRO)

Under the pecking order theory, profitability means ample cash being generated to reduce the need for financial leverage. Thus, the pecking order theory posits an inverse relationship between profitability and

11  What Drives the Capital Structure of Non-Listed 

237

financial leverage (Vasiliou et al. 2003). But Jensen (1986) favoured such a negative relationship in the case of an ineffective market for corporate control. The reason for this was the fact that, under an ineffective market for corporate control, even if a firm were to have high profits, lenders might be unwilling to lend as debt no longer acted as an effective monitoring device. Conversely, in the case of an effective market for corporate control, then a positive relationship was anticipated to manifest.

3.5

Tangibility of Assets (TANG)

Consistent with the pecking order theory, Rajan and Zingales (1995) and Frank and Goyal (2002) stated that TANG represented a form of secured collateral, thereby causing a positive impact on leverage. However, Grossman and Hart (1982) pointed out that, with high monitoring costs for shareholders of capital outlays for low tangibility of assets firms, there should be a correspondingly higher level of debt working as a cost-­effective monitoring mechanism. In that case, a negative relationship is expected to manifest between TANG and leverage. In addition, Titman and Wessels (1988) differentiated between tangibility (tangible assets/total assets) and intangibility (intangible assets/total assets) in order to forecast a positive association between tangibility and leverage and a negative association between intangibility and leverage. Most importantly, Van der Wijst and Thurik (1993), Chittenden and Hutchinson (1996), and Stohs and Mauer (1996) found a positive association between tangibility and long-term debt, but an inverse relationship between tangibility and short-term debt.

3.6

Liquidity (LIQ)

Liquidity is regarded to be negative debt as it scales down the need to take on debt. As per Ozkan (2001), such a negative association emanates from the feasible conflicts between shareholders and bondholders. The reason for this is based on the fact that the higher the liquidity level, the greater the ease with which shareholders can wield the liquid assets of the firms at the expense of bondholders. Nonetheless, liquidity can trigger an inducing impact in the case of high liquidity enshrining the availability of debt.

238 

3.7

Economics and Finance in Mauritius

Investment (INV)

Investment variable is included in the model to cater for the fact that higher investment may trail behind the greater need for financial leverage. Subsequently, investment is expected to engender a positive impact on leverage. But it can be pointed out that investment is related to the higher use of leases rather than loans. The underlying reason is based on the fact that it pays for firms to contract lease for their equipment that they require for a number of years, rather than incurring highly expensive purchases in the case of new equipment. By virtue of the fact that focus is being laid on non-listed firms in lieu of having recourse to the marketto-­book value as the proxy for investment, the purchase of equipment emanating from the cash flow statement of companies, is employed as the best alternative to capture for the purchase of equipment.

3.8

Age (AGE)

The Age element has also been incorporated to capture for the fact that an older company, after having established a solid market base, is susceptible to having recourse to lesser financial leverage. The rationale is based on the fact that mature companies are in a better position to best manage their cash flows. It can also be stated, however, that the age element unleashes a positive impact on leverage in those instances when older firms require an increase of funds to keep abreast with the latest ­technology. Sound relationship banking may also contribute towards the positive association between age and leverage.

4

Results

The correlation coefficients point out that multicollinearity does not constitute an issue of concern as depicted in Tables 11.3 and 11.4. The regression results are illustrated in Table 11.5. Profitability is found to be statistically significant, but only in the case of short-term debt. Such a state of affairs signifies that more profitable

NDTS

1 −0.0044 0.1861 −0.0732 −0.0114 −0.0689 −0.06 0.21 −0.0212 −0.0086 −0.0234 0.0625 0.0506 0.0134 0.0744 0.0066 0.0878

Variable

NDTS PRO TANG Size Growth Liquidity Age Investment STLev LTLev TLev STLeases LTLeases STLoans LTLoans STDebt LTDebt

1 −0.165 0.0726 0.1298 0.218 0.0265 −0.0299 −0.1999 −0.0876 −0.2233 −0.2063 −0.2006 −0.0206 −0.1085 −0.2501 −0.1637

PRO

Table 11.3  Correlation coefficients

1 0.1723 −0.1469 −0.1486 0.0774 0.3558 −0.2653 0.3793 −0.0097 0.2217 0.2271 0.0864 0.2925 0.0118 0.3531

TANG

1 0.0105 −0.0391 0.3033 −0.166 −0.2381 −0.0303 −0.2232 −0.1319 −0.1391 0.0989 −0.0442 −0.1495 −0.0827

Size

1 −0.0332 −0.1411 0.1374 0.0203 −0.0831 −0.0305 −0.0567 −0.0045 −0.0901 −0.073 −0.0474 −0.0735

Growth

1 −0.0389 −0.1766 −0.5622 −0.1054 −0.5467 −0.1856 −0.1747 −0.1428 −0.1052 −0.3114 −0.153

Liquidity

1 −0.0936 −0.1406 −0.0315 −0.1397 −0.0498 −0.0561 0.048 −0.1166 0.0532 −0.1311

Age

1 −0.0055 0.0787 0.0408 0.2637 0.3662 −0.0618 0.0419 0.0596 0.1443

Investment

11  What Drives the Capital Structure of Non-Listed 

239

STLev LTLev TLev STLeases LTLeases STLoans LTLoans STDebt LTDebt

Variable

1 −0.0812 0.8169 0.1092 0.0821 0.0842 −0.0759 0.4087 −0.052

STLev

1 0.5085 0.1424 0.1684 0.1137 0.7842 0.0921 0.8228

LTLev

Table 11.4  Correlation coefficients

1 0.1767 0.1684 0.1386 0.3882 0.4064 0.4312

TLev

1 0.7806 0.0778 −0.0904 0.2541 0.1298

STLeases

1 −0.0249 −0.1023 0.1703 0.1796

LTLeases

1 0.0756 0.4066 0.0678

STLoans

1 0.0663 0.9602

LTLoans

1 0.1134

STDebt

1

LTDebt

240  Economics and Finance in Mauritius

**

−0.0167 (−1.72) −0.1578 (−0.96) −0.0386 (−0.54) −0.0407 (−4.45) *** −0.0476 (−1.10) −0.127 (−9.26) *** −0.042 (−2.21) ** −0.3061 (−1.76) 1.8049 (11.50)*** 15.91 0 0.3915 395

−0.0132 (−1.97) ** 0.0105 −0.12 0.3501 (6.68) *** −0.0125 (−2.04) ** −0.0094 (−0.35) −0.0093 (−1.60) −0.0111 (−0.78) −0.2136 (−1.76) 0.3519 (3.35) *** 8.32 0 0.174 395

Long term liabilities −0.0034 (−0.54) −0.1684 (−1.68) −0.3888 (−6.65) *** −0.0281 (−4.19) *** −0.0382 (−1.01) −0.1177 (−10.09) *** −0.0309 (−2.06) ** −0.0924 (−0.67) 1.453 (12.60) *** 28.04 0 0.5028 395

Short term liabilities −0.0018 (−0.78) −0.0428 (−1.37) 0.0224 −1.8 −0.0028 (−2.11) ** −0.0023 (−0.34) −0.0027 (−2.51) ** −0.0005 (−0.23) 0.1815 (2.97) *** 0.0684 (2.96) *** 4.37 0 0.1914 395

Long term leases −0.0003 (−0.33) −0.0173 (−1.54) 0.012 (2.36) ** −0.0013 (−2.43) ** −0.0033 (−1.08) −0.0016 (−3.57) *** −0.0004 (−0.47) 0.0465 (2.22) ** 0.0346 (3.41) *** 4.87 0 0.1448 395

Short term leases

and ***denote statistical significance at the 5% and 1% level, respectively

F(8,386) Prob > F R-squared Obs

Constant

Investment

Age

Liquidity

Growth

Size

Tangibility

Non-Debt Tax Shield Profitability

Total liabilities

Independent variables

Table 11.5  Results of regression

0.0023 −0.25 −0.0275 (−0.33) 0.2474 (4.87)*** −0.008 (−1.36) −0.0135 (−0.54) −0.0094 (−1.63) −0.0278 (−2.02)** −0.2493 (−2.33)** 0.2798 (2.67)*** 4.59 0 0.1268 395

Long term loan 0.0005 −0.3 0.0098 −0.63 0.0196 −1.43 0.0021 −1.04 −0.0141 (−1.84) −0.0062 (−4.09)*** −0.0002 −0.967 −0.07 (−2.07)** 0.002 −0.06 4.1 0.0001 0.0473 395

Short term loan 0.0004 −0.05 −0.0703 (−1.03) 0.2698 (5.46) *** −0.0109 (−1.88) −0.0159 (−0.64) −0.0122 (−2.08) ** −0.0284 (−2.13) ** −0.0678 (−0.62) 0.3483 (3.39) *** 8.96 0 0.1789 395

Long term debt

−0.0018 (−0.32) −0.1807 (−2.72) *** −0.0297 (−0.86) −0.0163 (−3.62) *** −0.0133 (−0.45) −0.0327 (−7.24) *** 0.0196 −1.81 0.0138 −0.09 0.4699 (5.84) *** 11 0 0.1648 395

Short term debt

242 

Economics and Finance in Mauritius

firms have recourse to cheaper internal funds in lieu of contracting costly bank credit. The impotency of profitability on long-term debt means that Mauritian firms avoid long-term leverage. Non-debt tax shield is statistically significant, but only in the case of long-term liabilities. For the rest of the different metrics of leverage, growth, profitability and non-debt tax shield variables, however, are all found not to be statistically significant. These findings plainly point out that the use of specific leverage variables may simply bias results so that wrong policies are being taken. Thus, a decomposition is highly warranted to trigger richer insight. Nonetheless, the impotency of the growth variable entails vital implications. As a matter of fact, since all firms are non-listed firms, they are mostly closely-held so that higher growth deters the use of leverage in order to ensure a larger pie to be shared among the shareholders of the company. As far as the asset structure is concerned, a positive impact is noted, implying that higher assets tangibility induces higher collaterals value so that firms which are endowed with higher assets tangibility become eligible for a higher amount of leverage, compatible with the findings of Rajan and Zingales (1995). But the positive impact occurs only in the case of long-term liabilities, short-term leases, long-term loans and long term-debt. Alternatively stated, a negative relationship manifests in the case of total liabilities and short-term liabilities. Since liabilities constitute a broader measurement of leverage, such a finding again implies the need to undertake leverage analysis in specific forms. When comparing the impact of asset tangibility on leases and loans, it transpires that the effect is higher for long-term loans. Such a finding signifies that the greater the value of fixed assets, such as the land and buildings owned by a company, the greater the amount of long-term loan that the company can contract from a bank on the back of these assets, thereby increasing the safety net of the bank once they are being encumbered in its favour. In spite of the fact that size is found to be statistically significant for nearly all of the variants of leverage, its sign violates the theoretically consistent sign under the pecking order theory. Findings point out that larger firms tend to use less leverage. However, such a negative sign is consistent with the results of Rajan and Zingales (1995), who argued that larger firms select internal sources of funding by virtue of their complex structures which spurt out regular asymmetrical knowledge problems. In

11  What Drives the Capital Structure of Non-Listed 

243

the case of Mauritius, it can be argued that larger firms do not want to distribute profits to outsiders such as the banks. Consequently, Mauritian firms try their most to use internally generated sources of funds to harness maximum benefits from investments to the internal shareholders only who happen to be also the managers in the case of closely-held firms. Liquidity is found to engender bearish effects on the different versions of leverage. Such a finding is highly consistent with the pecking order theory in that firms which are imbued with robust liquid resources are less likely to contract external funding such loans or leases. The underlying rationale is based on the notion of costly external borrowings so that easily available liquid funds help to alleviate these high costs of external funding. Comparing the impact of liquidity on short-term loans and short-term leases, it occurs that a 1% change in liquidity generates around −0.62% change in short-term loans and a −0.16% change in short-term leases. The relatively higher effect on loans to leases is due to the fact that lease repayments tend to be lower than loan repayments so that enhanced liquidity conditions trigger more pronounced effects in the case of loans. As far as the age variable is concerned, its effect is statistically significant in the case of total liabilities, short-term liabilities, long-term loans and long-term debt. A 1% change in age causes around −2.78% and −2.84% change in long-term loan and long-term debt, respectively. Such a finding is of paramount significance as it clearly shows that mature and well-established companies avail themselves of strategic cash flow or funds management in order to reduce the demand for external borrowings. In that respect, the number of years established in a business is particularly important in decreasing the demand for external funding. Such a finding could plausibly imply that, with time in the same business, a firm can avail of some power such as availing of loyal customers and strong bonds with suppliers so that the working capital needs of the firm becomes better relative to an incumbent firm. Findings show that investment exerts a heavy impact on loans and leases only, again stating the need to undertake a decomposed analysis of leverage on the back of impotent effects noted for total liabilities. Deeper analysis reveals that investment positively impacts on leases, but negatively affects loans. Such a result means that Mauritian firms purchase equipment chiefly through leasing facilities so that there is a

244 

Economics and Finance in Mauritius

crowding-­out effect on loans. Above all, the effects are found to be particularly higher for long-term loans and long-term leases relative to their counterparts. Such a finding signifies that equipment purchases mainly impact on long-term borrowings as they are particularly expensive, all requiring years of repayments to induce an affordable repayment capacity to the firms.

5

Conclusion

This chapter provides a critical assessment of the capital structure features for non-listed non-financial firms in Mauritius. The analysis is holistic as it focuses on different versions of leverage such as loans, leases, debt and total liabilities. In essence, the analysis makes use of both broad and specific metrics of leverage. Findings depict that profitability, non-debt tax shield and growth do not account for leverage needs. Nonetheless, compatible with the pecking order theory, asset tangibility trails behind a positive effect on leverage while liquidity exerts downward pressures on leverage. Size is found to unleash a negative impact on leverage, meaning that larger firms tend to curtail the use of leverage. Findings disclose crowding-out effects by virtue of investment being positively linked to leases but negatively related to loans. Consequently, firms resort to leases as the financing mode for the purchase of equipment. In addition, age is found to generate bearish effects on loans and debt, in particular on their long-term components, thereby magnifying the importance of experience acquired in doing business. Another vital contribution of this study refers to the need to employ a modified pecking order theory, something that is highly appropriate in the Mauritian context. In fact, looking at the specificity of leverage, it transpires that firms should first use short-term leverage and then long-­ term leverage as the former carries lower costs than the latter. Now, with respect to loans and leases, the latter are expected to be used first and then loans. The underlying rationale is based on the fact that leases are endowed with low information asymmetry. Short-term component of leases is used first prior to having recourse to long-term leases because the agency problem may be lower for short-term debt than long-term

11  What Drives the Capital Structure of Non-Listed 

245

debt (Myers 1977; Barclay and Smith 1996, 1999; Michaeles et  al. 1999). Therefore, the proposed modified pecking order theory signifies that firms will first resort to short-term leases and then long-term leases because in the former case four variables (namely asset structure, size, liquidity, and investment) are found to be statistically significant relative to only three variables (size, liquidity, and investment) in the latter case. A priori, in the case of loans, it transpires that long-term loans are used first with asset structure, age and investment being statistically significant while short-term loans are used afterwards based on liquidity and investment being statistically significant. However, since information asymmetry is susceptible to being lower for short-term loans relative to long-term loans, the former should precede the latter. Thus, the final composition of the modified pecking order of capital structure for Mauritian firms consists of internal funds first, then short-term leases, then long-term leases, then short-term loans, then long-term loans and only as a last resort to bring in equity capital from existing owners. External equity financing is not considered because the firms are not quoted on the Stock Exchange of Mauritius. Overall, the findings are highly interesting as they do trigger a comprehensive assessment of the determinants of capital structure in the case of an upper-income developing country which happens to have relatively few listed firms (below 45  in number) on its official market. In that respect, any results drawn from assessing the capital structure of listed firms for developing countries should be taken with extreme care as the results cannot be generalized by virtue of the fact that relatively few firms are often listed on the stock markets for developing countries. Finally, the analysis clearly shows the need to undertake a specificity of leverage in order to entail fruitful results with sound implications. Supply-side forces may also influence the finding of short-term leverage predominating over long-term leverage. As a matter of fact, Mauritius is imbued with a bank-based financial system. Thus, as part of their asset– liability management, banks contract short-term deposits to advance long-term loans. In view of mitigating such inherent mismatch between assets and liabilities, banks are motivated to go for more short-term loans in lieu of very long-term loans.

246 

6

Economics and Finance in Mauritius

Policy Recommendations

Researchers who undertake studies which probe into the determinants of capital structure in Mauritius should best have recourse to non-listed non-­ financial firms in order to generate a truly holistic assessment. Moreover, the government should see to it that leases are provided under best conditions for equipment so that ample competitive forces prevail as to benefit Mauritian firms. The reason is that since most investments undertaken by firms are based on the purchase of equipment which is being sourced via leases, better leases conditions will automatically unleash stronger investments and thereby induce higher economic growth. Most importantly, based on relatively few listed firms in Mauritius, it is high time that the government undertakes vigorous measures to initiate the listings of more high-quality and promising firms which can then move beyond leverage to fund their investment activities. Finally, since asset tangibility exerts a strong effect on long-term leverage, this signifies that bankers need to undertake sound asset valuation assessments on a regular basis to ensure that loans do not turn into negative equity. Such a task can be fulfilled via a general system of asset valuation in Mauritius in which case ratings are being assigned so that high-rated assets can avail of higher loans amounts.

Bibliography Barclay, M. J., Smith, C. W., & Watts, R. L. (1995). The determinants of corporate leverage and dividend policies. Journal of Applied Corporate Finance, 7(4), 4–19. Barclay, M. J., & Smith, C. W. (1996). On financial architecture: Leverage, maturity, and priority. Journal of Applied Corporate Finance, 8(4), 4–17. Barclay, M. J., & Smith, C. W. (1999). The capital structure puzzle: Another look at the evidence. Journal of Applied Corporate Finance, 12(1), 8–20. Baskin, J. (1989). An empirical investigation of the pecking order hypothesis. Financial Management, 1(1), 26–35. Berger, P. G., Ofek, E., & Yermack, D. L. (1997). Managerial entrenchment and capital structure decisions. Journal of Finance, 52(4), 1411–1438. Bevan, A., & Danbolt, J. (2000). Dynamics of the determinants of capital structure in the UK (Working paper). University of Glasgow.

11  What Drives the Capital Structure of Non-Listed 

247

Chittenden, F. H., & Hutchinson, G. P. (1996). Small firm growth, access to capital markets and financial structure: Review of issues and an empirical investigation. Small Business Economics, 8, 59–67. Donaldson, G. (1961). Corporate debt capacity: A study of corporate debt policy and the determination of corporate debt capacity. Boston: Division of Research, Harvard Graduate School of Business Administration. Fama, E. F., & French, K. R. (2002). Testing tradeoff and pecking order predictions about dividends and debt. Review of Financial Studies, 15(1), 1–43. Frank, M. Z., & Goyal, K. V. (2003). Testing the pecking order theory of capital structure. Journal of Financial Economics, 67(2), 217–248. Griner, E. H., & Gordon, L. A. (1995). Internal cash flow insider ownership capital. Journal of Business Finance and Accounting, 22, 179–199. Grossman, S., & Hart O. (1982) Corporate financial structure and managerial incentives. In J. McCall (Ed.), The economics of information and uncertainty. Chicago: University of Chicago Press. Hall, G., Hutchinson, P., & Michaelas, N. (2000). Industry effects on the determinants of unquoted SMEs’ capital structure. International Journal of the Economics of Business, 7(3), 297–312. Jensen, M. C. (1986). Agency costs of free cash flow corporate finance and takeovers. American Economic Review, 76, 323–329. Kester, W. C. (1986). Capital and ownership structure: A comparison of United States and Japanese manufacturing corporations. Financial Management, 15, 5–16. Lasfer, M. A. (1995). Agency costs, taxes and debt: The UK evidence. European Financial Management, 1(3), 265–285. Michaelas, N., Chittenden, F., & Poutziouris, P. (1999). Financial policy and capital structure choice in UK SMEs: Empirical evidence from company panel data. Small Business Economics, 12, 113–130. Modigliani, F., & Miller, M. H. (1958). The cost of capital, corporation finance and the theory of investment. American Economic Review, XLVIII(3), 261–297. Moore, W. (1986a). Asset composition bankruptcy costs and the firm’s choice of capital structure. Quarterly Review of Economics and Business, 26, 51–61. Myers, C. S., & Majluf, S. N. (1984). Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics, 13(2), 187–221. Myers, S. C. (1977). Determinants of corporate borrowing. Journal of Financial Economics, 5(2), 147–175.

248 

Economics and Finance in Mauritius

Myers, S. C. (1984a). The capital structure puzzle. Journal of Finance, 39(3), 574–592. Ozkan, A. (2001). Determinants of capital structure and adjustments to long run target: Evidence from UK company panel data. Journal of Business Finance and Accounting, 28, 175–195. Rajan, R. G., & Zingales, L. (1995). What do we know about capital structure? Some evidence from international data. The Journal of Finance, 50(5), 1421–1460 Ramlall, I. (2009). Determinants of capital structure among non-quoted Mauritian firms under specificity of leverage: Looking for a modified pecking order theory. International Research Journal of Finance and Economics, 31, 84–92. Remmers, L., Stonehill, A., Wright, R., & Beekhuisen, T. (1974). Industry and size as debt ratio determinants in manufacturing internationally. Financial Management (Summer), 3, 24–32. Stohs, M.  H., & Mauer, D.  C. (1996). The determinants of corporate debt maturity structure. Journal of Business, 69(3), 279–312. Scott, J. H. (1977). Bankruptcy, secured debt, and optimal capital structure. The Journal of Finance, 32, 1–19. Titman, S., & Wessels, R. (1988). The determinants of capital structure choice. Journal of Finance, 43, 1–18. Van der Wijst, N., & Thurik, R. (1993). Determinants of small firm debt ratios: An analysis of retail panel data. Small Business Economics, 5, 55–65. Vasiliou, D., Eriotis, N., & Daskalakis, N. (2003). The determinants of capital structure: Evidence from the Greek Market. Paper presented at the 2nd annual meeting of Hellenic Finance and Accounting Association, Athens, pp. 1–16. Watson, R., & Wilson, N. (2002). Small and medium size enterprise financing: A note on some of the empirical implications of pecking order. Journal of Business Finance and Accounting, 29, 556–578. Zoppa, A., & McMahon, R. G. P. (2002). Pecking order theory and the financial structure of manufacturing SMEs from Australia’s business longitudinal survey. School of commerce, Research paper series: 02–2 ISSN-1441-3906.

12 Capital Structure Analysis of Exporting and Non-exporting Mauritian Firms: A Pre- and Post-crisis Investigation

This chapter probes into the impact of the crisis on the capital structure of Mauritian exporting and non-exporting companies by augmenting the determinants of capital structure through the use of strategic variables like tax, sales and equity. Two metrics of leverage (short-term leverage and ­long-term leverage) are used and two major periods of time (pre- and ­post-crisis). The weak impact of the age variable is noted while sales induce the use of short-term leverage but deter the use of long-term leverage. Evidence is found of Mauritian firms contracting longer-term debt during the crisis period to fund their purchases of fixed assets. Size effects reveal that the larger firms in Mauritius tend to make more use of short-­term debt rather than long-term debt. Liquidity acts as a major determinant in contracting short-term debt while asset tangibility acts as the major driver in taking on long-term debt. Profitability variable is consistent with static trade-off theory based on a positive relationship between profitability and long-term debt. Policy-wise, is the analysis identifies a need to democratize equity financing to smaller firms to ensure that bankruptcies are well contained during stressful conditions. Impotency of tax under short-term debt signifies no need for fiscal motives to instil a relieving operational manoeuvre to Mauritian companies. © The Author(s) 2017 I. Ramlall, Economics and Finance in Mauritius, DOI 10.1007/978-3-319-39435-0_12

249

250 

1

Economics and Finance in Mauritius

Introduction

The world financial systems experienced jerky periods of financial stress conditions. The difficulties spread out from financial systems into the real economy. Consequently, non-financial firms began to face detrimental states of affairs on the back of falling demand for their exports and sales. Mauritius, an upper-income developing country, is highly dependent on international trade for its economic activities. Many firms are so involved in exporting that it becomes useful to undertake an analysis of how the crisis reshaped the capital structure of Mauritian firms. This is the essence of this chapter, which addresses an important question with respect to how the capital structure of Mauritian firms evolved following the US subprime crisis. Indeed, with the onset of the crisis, it is expected that firms, with considerable international exposure, should be experiencing more stressful conditions at diverse levels such as sales, profits, ability to undertake investments, let alone the rising threshold of permissible debt-to-equity ratio stipulated by their bankers. An insight of the capital structure of firms before and after the crisis becomes extremely important for policy makers to ensure that firms do not generate rising layoffs and thereby unleash rising levels of unemployment, which will eventually choke off potential growth. To the author’s best knowledge, this is the first study in Mauritius to make a direct links between the crisis and the capital structure of Mauritius firms, with comparisons being made between exporting and non-exporting firms both before and after the crisis. Indeed, the financial crisis provides an opportunity to obtain more information about the vulnerabilities of Mauritian firms. To assess the effects of the crisis on the capital structure of exporting and non-exporting Mauritian firms, it is vital to consider the non-listed firms so that we can secure a true picture of the capital structure stance prevailing in Mauritius. Another important contribution of this study is that it factors in leverage in its short-term and long-term components. Finally, this study adds to the currently relatively scarce literature on capital structure for a developing country based on the fact that most capital structure studies are tilted towards developed countries.

12  Capital Structure Analysis of Exporting and Non-exporting... 

251

Capital structure constitutes a core component in finance. Interest in the subject dates back to the work of Modigliani and Miller (1958), who showed that in the case of perfect capital markets, capital structure is utterly irrelevant. However, in practice, different levels of market imperfections manifest, thereby giving birth to three theories of capital structure – namely, the Static Trade-Off, the Pecking Order (Myers and Majluf 1984), and the Agency Cost Theories. The Static Trade-Off Theory states that an optimal capital structure is obtained when the net tax advantage of debt financing balances leverage-related costs such as financial distress and bankruptcy. The Pecking Order Theory points out that there is a certain order in financing starting from retained earnings, then moving to debt and eventually drifting to equity financing. The Agency Cost Theory relates to agency problems that arise between managers and shareholders. The Agency Cost Theory is susceptible to being belittled as a theory in Mauritius since nearly all of the firms have their shareholders as managers. Above all, such a theory tends to also permeate listed firms with high insider ownership levels. In a parallel manner, the Static Trade-Off Theory is deemed to be of little use in explaining the situation in Mauritius since there are no signs that firms set up a balance between tax savings manifesting from the use of debt against any financial distress costs. However, it is common practice among local firms to initially exhaust their own funds before taking on loans since the latter are imbued with a higher cost of borrowing (See Ramlall (2009). It can be deduced, therefore, that the Pecking Order Theory is particularly well ingrained into the capital structure of non-listed non-financial Mauritian firms. This study analyses the determinants of capital structure for various firms in Mauritius with accommodation given to the type of firms; exporting versus non-exporting. Section 2 of the chapter focuses on the literature related to capital structure. In Sect. 3, the econometric model is specified and the relationships subsisting between leverage and distinct financial variables explained. Subsequently, Sect. 4 discusses the results obtained while Sect. 5 concludes.

252 

Economics and Finance in Mauritius

2

Literature Review

2.1

Underpinning Theories of Capital Structure

The influential paper of Modigliani and Miller on capital structure irrelevancy is deemed to be the key element in the development of theoretical and empirical research related to the optimal capital structure worldwide. In essence, Modigliani and Miller (1958) showed that the value of a firm is independent of the capital structure based under stipulated assumptions of perfect and frictionless capital markets. However, in reality market frictions predominate over both the economic and financial states of human activities. Distinct theories of capital structure saw the day on the back of different financial aspects under consideration such as agency costs, information asymmetry, transaction costs and bankruptcy costs. In essence, there are four main theories of capital structure – namely, the pecking order theory, the static trade-off theory, the agency cost theory and finally, the signalling theory, as depicted in Fig. 12.1. (i) Pecking Order Theory According to the Pecking Order Theory developed by Myers and Majluf (1984), the cost of financing is linked to the level of information asymmetry involved in the type of financing undertaken by a firm. Consequently, there is a hierarchy in the structure in which a firm will fund its investment activities. First, a firm will deplete all of Static Trade-Off Theory Pecking Order Theory

Agency Theory

Theories of Capital Structure

Fig. 12.1  Theories of capital structure

Signaling Theory

12  Capital Structure Analysis of Exporting and Non-exporting... 

253

its own cash resources since they are nearly costless, and they require no payment of interest. Second, firms will have recourse to bank borrowings because banks exercise significant monitoring mechanisms such as covenants to ensure that funds are properly used. Finally, firms, particularly listed firms, will resort to equity financing only as a last resort. The rationale is again based on the significantly higher level of information asymmetry imbued in the use of equity funding relative to internal of bank borrowings. (ii) The Static Trade-Off Theory The Static Trade-Off Theory attempts to strike an optimal balance with respect to the benefits and drawbacks of debt or leverage. The use of debt benefits firms through interest payments which reduce the taxable level of the firm and hence the level of taxes that have to be paid. Consequently, there are strong incentives for firms to have recourse towards increased leverage in order to avail themselves of higher and higher interest tax shields. However, assessing the other side of the coin, the use of overly high debt or leverage is very risky to the firm as it scales up its probability of going bankrupt. In order to maximize its value, a firm will borrow up to the point where the marginal tax benefit of debt is offset by the marginal costs of bankruptcy, as stated by Myers (1984). (iii) The Agency Theory The Agency Theory developed from the agency problem pointed out by Jensen (1986). Agency costs emanate from the separation of ownership and management in a company so that there is an inherent conflict of objectives. For instance, shareholders will always try to maximize the value of the firm while management will want to maximize sales. Jensen (1986) pointed out the classical case of the agency problem in the form of the free cash flow problem whereby managers of firms who are imbued with excess free cash flows will be tempted to over-invest at the expense of reducing the value of the company. To mitigate such a problem, the use of debt is highly recommended as it disciplines on the behaviour of the managers. Jensen and Meckling (1976) argued for another source of agency costs due to the fact that bondholders are paid prior to equityholders so that the latter have strong incentives to invest in risky projects.

254 

(iv)

2.2

Economics and Finance in Mauritius

The Signaling Theory Ross (1977) developed a model to show that firm value can be signalled to outside investors via different financing alternatives. Consequently, by using more debt, a firm is indirectly showing that it is able to meet additional debt obligations in the future on the back of higher profits and cash flows.

Empirical Evidence

The Pecking Order Theory of capital structure is widely considered to be the most practical theory in terms of real-life situations whereby the greater the level of information asymmetry, the higher the costs of borrowings. The Pecking Order Theory was first suggested by Donaldson (1961), but it effectively obtained its robust theoretical foundation following the works of Myers and Majluf (1984). Technically speaking, the Pecking Order Theory attributes the selection order in financing to the distinct degrees of information asymmetry imbued in those sources of finance. Retained earnings are found on the lower rung of the financing ladder because they constitute the cheapest source of funding. Debt financing is found on the second rung of the financing ladder as there is low information asymmetry due to fixed obligations acting as an effective monitoring device. On the last and final rung of the financing ladder, there is external equity, which is used only as a last resort financing component since it conveys adverse signaling effects. In general, the Pecking Order Theory is compatible with shareholder’s wealth maximization since it is inherently geared to minimize the costs of raising funds. Many studies have been undertaken with respect to the Pecking Order Theory. Watson and Wilson (2002) analysed the empirical validity of the Pecking Order Theory in the case of UK small and medium-sized enterprises (SMEs) split into high information asymmetry firms, low information asymmetry firms and closely-held firms. To generate a full-fledged analysis, they decomposed debt into hire purchase liabilities, long-term debt, short-term debt and intra-group debt balances, respectively. Their findings demonstrated that the Pecking Order Theory manifests strongly in the case of such firms. Moreover, they found evidence in favour of a

12  Capital Structure Analysis of Exporting and Non-exporting... 

255

pecking order within debt since the explanatory power of their estimated models rose considerably when debt was disentangled out into its several components. Consequently, their main conclusion emphasized the need to assess the Pecking Order Theory by looking at leverage in various dimensions. Bevan and Danbolt (2000) analysed the Pecking Order Theory in the UK credit market from 1991 to 1997 by decomposing debt into its individual components to generate a richer picture of capital structure decisions and also to avoid any potential bias in the choice of the gearing ratio. They concluded that UK firms steadily moved away from debt finance to equity finance. They attributed such growth in equity to areas of high technology and the internet. Above all, they found that large firms, which were traditionally more reliant on long-term debt, were now having recourse towards more equity finance. Consequently, rather than contradicting Pecking Order Theory, they addressed such an anomaly to supply factors that may trigger a preference to equity over long-term debt. Zoppa and McMahon (2002) found robust evidence in favour of the Pecking Order theory in the case of manufacturing SMEs in Australia. Furthermore, they emphasized the need to have an adjusted Pecking Order theory which is well adapted and customized to the features of the SMEs financing. The modified Pecking Order Theory, as pointed out by them, was to use internal profits first, then short-term credit alike trade credit and personal credit card. Thereafter, recourse was made towards long-term debt financing, then, equity financing from existing owners and, finally only as a last resort, that new equity financing from uninvolved parties, was used. Griner and Gordon (1995) focused on the disagreement between the Pecking Order Theory and the Managerial Entrenchment Hypothesis over whether or not there is any association between capital expenditures and insider ownership. They related their analysis over the fact that Managerial Entrenchment Hypothesis was based on an inverse association between insider ownership and capital expenditures while Pecking Order Theory was based on no association between insider ownership and capital expenditures. Consequently, their findings endorsed the Pecking Order Theory since their research finding demonstrated no such association between capital expenditures and insider ownership. They

256 

Economics and Finance in Mauritius

deduced that the reliance on internal cash flow for capital expenditure financing was mainly due to information asymmetry between managers and new shareholders rather than information asymmetry between insiders which was consistent with Pecking Order theory. Furthermore, they concluded that shareholders needed not strive for high insider ownership to maximize their wealth because Managerial Entrenchment Hypothesis was empirically limited. New studies on capital structure have emerged in light of the US subprime crisis of 2007/08. Proença et  al. (2014) investigated the determinants of capital structure in the case of Portuguese SMEs. Their findings showed that liquidity, asset structure and profitability constituted the most vital variables that influence capital structure. Yang et al. (2010) tackled the capital structure problem by simultaneously solving for capital structure and stock returns. Their findings showed that capital structure is mainly accounted for by stock returns, expected growth, uniqueness, asset structure, profitability and industry classification. In the case of stock returns, their determinants consist of leverage, expected growth, profitability, value and liquidity. The main caveat of their study pertains to the use of listed non-financial firms which may not reflect the general dynamics of firms in Taiwan. Chakraboti (2010) analysed the determinants of capital structure for 1169 Indian firms over the period 1995–2008. Findings show that capital structure of Indian firms is best explained by the Pecking order and Static trade-off theories in lieu of the agency cost theory. Kayo and Kimura (2011) focused on the influence of time, firm, industry country determinants of capital structure. Their findings showed that time and firm levels constitute the most vital determinants of forces that explain variances of leverage. Chang et  al. (2014) analysed the agency theory and disciplinary effects of debt on optimal capital structure. They found that overleveraged and underleveraged firms having weak governance adjust slowly towards their target debt levels. Mateev et al. (2013) undertook an analysis of micro, small and medium-­sized SME firms in Central and Eastern Europe. Their findings show that there is no strong evidence in favour of the Pecking Order Theory. Chang et al. (2014) reanalysed the true determinants of capital structure in China. Their findings showed that the main forces influ-

12  Capital Structure Analysis of Exporting and Non-exporting... 

257

encing capital structure comprise of profitability, industry leverage, asset growth, tangibility, size, state control and largest shareholding.

2.3

Rationale for Study and Associated Benefits

Acclaimed as one of the successful countries in the Indian Ocean, Mauritius is undeniably subject to the vagaries of international trade by virtue of being a small open economy. The US subprime crisis unleashed major adjustments worldwide with respect to the ways in which firms conducted their businesses. The most painful adjustment pertained to firms which were already highly leveraged so that with the crisis, the drastic decline in sales considerably undermined their debt repayment capacity, putting them at the forefront of failures. In that respect, it is considerate to implement a detailed analysis of capital structure determinants of Mauritian firms split into exporting and non-exporting counterparts. Such a study is of paramount significance, bearing in mind that Mauritius relies significantly on its tourism and textile sectors as the main drivers of foreign exchange accumulation. This study is expected to be fruitful to policy makers in a number of different ways, as stated below: 1. First and foremost, there is a clear-cut demarcation line between exporting and non-exporting firms so that there is a rich set of results which bankers can use of when attempting to ensure the lowest level of firms’ failures. This entails implications at the financial stability level in terms of ensuring that proper policies are taken at broader levels to mitigate the level of bankruptcies. 2. Second, this study paves the way towards identifying the various firms’ attributes susceptible to impacting on their leverage levels. 3. Third, leverage is decomposed into short-term and long-term components so that there is again a richer set of results that can be used by policy makers to scale down the extent of bankruptcies. 4. Fourth, the time periods are split into two parts – a pre-crisis and a post-crisis analysis – to better gauge on the extent of crisis-induced changes in firms’ attributes susceptible to impact on their different types of leverage.

258 

3

Economics and Finance in Mauritius

Model Specification and Data

All data were gleaned from the Registrar of Companies in Mauritius, with more than 250 companies being considered for the analysis. To capture the impact of the crisis, 2008 was used chosen as the cut-off year. As a matter of fact, despite the fact that the crisis began in 2007, due to lagged effects, 2008 is considered as the best period to capture the beginning of the crisis effects in the Mauritian economy. In that respect, all years before 2008 were pulled into the pre-crisis period. All of the years after 2008 (inclusive) were placed into the post-crisis period. Proper data pre-processing was undertaken to shun off outliers. The model was derived on the basis of previous studies on Pecking Order Theory such as Ozkan (2001), Bevan and Danbolt (2000) and Titman and Wessels (1988). However, to secure more refined results, it was considered vital to include certain strategic variables like sales, tax and the equity level of the companies. For instance, sales were incorporated to see how they fared up well before and after the crisis on the financial leverage of Mauritian companies. In the same vein, a tax variable was also included to control for any feasible stress exerted chiefly during the crisis episode. The econometric model, taken from Ramlall (2009), previously applied to Mauritian firms, is specified in Eq. (12.1) which, in essence, constitutes a pooled cross-sectional analysis. The chosen model is strongly believed to capture the essence of the subject under study. Hall et al. (2000) pointed out that the spitting of leverage into short-­ term leverage and long-term leverage assisted in distilling down forces susceptible to affect them. Therefore, the econometric model is run twice; short-term liabilities and long-term liabilities as components of leverage, respectively. Moreover, to gauge the effects of the crisis, two periods are considered: the pre- and post-crisis periods. Above all, the regressions are further refined by running separate regressions for exporting Mauritian firms and non-exporting Mauritian firms. In all, eight regressions are considered in this study to generate full-fledged results with respect to the crisis onto Mauritian firms.

12  Capital Structure Analysis of Exporting and Non-exporting... 

259

LEVi = β0 + β1AGE i + β2SALESi + β3 PROi + β4 NDTSi + β5 TAX i + β5 TANG i + β6 CAPi + β7 INVi + β8SIZE i + β9 LIQ i + ε i (12.1)

Where i LEV

denotes a specific non-financial firm Leverage, Short-Term Leverage (STL) and Long-Term Leverage (LTL) AGE Age of the company SALES Sales PRO Profitability NDTS Non-Debt tax shield TAX Tax TANG Tangibility of assets CAP Equity capital INV Investments in assets SIZE Size LIQ Liquidity All variables are defined in Table  12.1, with the summary statistics depicted in Table 12.2 and Table 12.3

Table 12.1  Definitions of variables Variables

Definitions

Non-Debt Tax Shield Profitability Tangibility Size Liquidity Age

Depreciation over earnings before interest and tax Net profits over total assets Fixed assets over total assets Natural logarithm of total assets Current assets over current liabilities Logarithm of number of years since date of incorporation Purchase of equipment over total assets Tax over total assets Sales over total assets Equity over total assets Current liabilities over total assets Long term liabilities over total assets

Investment Tax Sales Capital Short-term liabilities Long-term liabilities

260 

Economics and Finance in Mauritius

Table 12.2  Summary statistics for exporting and non-exporting firms in the pre crisis period Variable Age Sales Profitability Non-Debt Tax Shield Tax Tangibility Short-term liabilities Long-Term liabilities Capital Investment Size Liquidity

Mean 25.9062 1.6287 0.1173 0.0438 0.0207 0.3530 0.4907 0.1816 0.3364 0.0799 16.4901 0.5647

Standard deviation Minimum Maximum 11.4004 1.2635 0.1569 0.0452 0.0326 0.2724 0.2653 0.2096 0.2492 0.1388 2.9810 0.2858

5 0.0002 0 0 0 0 0 0 0 0 6.6148 0

62.8630 5.000 1.000 0.2000 0.2000 0.9000 0.9000 0.9000 0.9000 0.9000 23.3704 0.9000

Table 12.3  Summary statistics for exporting and non-exporting firms in the pre crisis period Variable Age Sales Profitability Non-Debt Tax Shield Tax Tangibility Short-term liabilities Long-Term liabilities Capital Investment Size Liquidity

Mean 25.2784 1.7016 0.1343 0.0382 0.0224 0.3208 0.4608 0.1337 0.4173 0.0606 17.0855 0.5924

Standard deviation 11.9457 1.3121 0.1643 0.0433 0.0291 0.2688 0.2606 0.1644 0.2678 0.1059 2.9700 0.2868

Minimum Maximum 5 0 0 0 0 0 0 0 0 0 8.6711 0

62.8630 5.000 1.000 0.2000 0.2000 0.9000 0.9000 0.9000 0.9000 0.9000 23.4327 0.9000

While working with the data, significant outliers were found in the variables so that lower and upper values were set. In the case of natural logarithm of sales, an upper cut-off value of 5 was applied to make the data smoother since outliers were found to have a significant distortion on the estimated regression coefficients.

12  Capital Structure Analysis of Exporting and Non-exporting... 

261

Capital structure is influenced by firm-specific features which are represented by the various theoretical attributes. These theoretical attributes are discussed for each independent variable in the specified regression model.

3.1

Non-debt Tax Shields (NDTS)

All studies undertaken in capital structure need to cater for benefits attached to the use of financial leverage. Indeed, firms which are subject to tax shields, such as tax deductions for depreciation and investment tax credits, have less need to exploit debt tax shields. Hence, in light of the Pecking Order Theory, a negative relationship is anticipated to occur between NDTS and financial leverage. Nevertheless, a positive relationship can also occur. For instance, Scott (1977) and Moore (1986) argued that considerable NDTS acted as attractive collaterals so that it induced high debt levels.

3.2

Size

Size is widely used to infer the level of strength endowed in a company in alleviating the extent of information asymmetry. In fact, Fama and French (2002) stated that larger firms were heavily leveraged because their sizes were synonymous with the possession of substantial financial resources. However, Rajan and Zingales (1995) found that bigger firms had lower leverage since the information asymmetry level is higher by virtue of their complex structures. Barclay et al. (1995) argued that the splitting of financial leverage into short-term debt and long-term debt led to a positive relationship between long-term debt and size but a negative relationship between short-term debt and size. The rationale was that small firms’ borrowings were skewed to short-term debt on account of the lower fixed costs associated with short-term borrowing. A plethora of studies found a robust positive relation between size and debt ratio (Lasfer 1995; Rajan and Zingales 1995; Barclay and Smith 1996; Berger et al. 1997). Kester (1996) and Remmers et al. (1974), however, found no significant effect of size on capital structure.

262 

3.3

Economics and Finance in Mauritius

Profitability

Under the Pecking Order Theory, profitability constitutes an important element as it signifies a strong possibility of generating ample cash that eventually scales down the need for financial leverage. Viewed in this perspective, a negative relationship should manifest between profitability and financial leverage like in the study done by Vasiliou et  al. (2003) under the Pecking Order Theory. However, Jensen (1986) attributed such a negative relationship to an ineffective market for corporate control. The reason was based on the fact that, under an ineffective market for corporate control, even if a firm had high profits, lenders might be unwilling to lend as debt no longer served as an effective monitoring device. Conversely, if an effective market for corporate control prevails, then, a positive relationship is expected to manifest. Proença et al. (2014) found a negative relationship between debt ratios and profitability, showing clearly that SMEs in Portugal clung to internal funding in lieu of external funding by virtue of costly information asymmetry levels.

3.4

Tangibility of Assets (TANG)

Consistent with the Pecking Order Theory, Rajan and Zingales (1995) and Frank and Goyal (2002) pointed out that TANG represented secured collaterals so that a positive relationship would prevail. However, a negative relationship can also manifest. For instance, Grossman and Hart (1982) argued that there should be a correspondingly higher level of debt acting as a cost-effective monitoring mechanism in the case of exorbitant monitoring costs for shareholders of low tangibility of assets. Similarly, Titman and Wessels (1988) differentiated between tangibility and intangibility, predicting a positive relationship between tangibility and leverage and a negative one between intangibility and leverage. Above all, Van der Wijst and Thurik (1993), Chittenden et  al. (1996) and Stohs and Mauer (1996) uncovered a positive relationship between tangibility and long-term debt but a negative relationship between tangibility and short-­ term debt.

12  Capital Structure Analysis of Exporting and Non-exporting... 

3.5

263

Liquidity

Capital structure models factor in the level of liquidity because it is considered as negative debt by relieving off the need to take on debt. According to Ozkan (2001), the posited negative relationship is based on conflicts between shareholders and bondholders – higher liquidity signifies greater ease in which shareholders can manipulate the liquid assets of the firms at the expense of bondholders. Nonetheless, liquidity can trail behind a positive impact in the case that high liquidity eases the availability of debt.

3.6

Investment

Investment variable is taken into account because a higher investment may generate high financial leverage. Therefore, investment is expected to cause a positive effect on debt. Nevertheless, investment can be associated with the higher use of leases relative to loans. The reason is that it is cost-­ effective for local companies to lease equipment that they may need for a number of years rather than incurring expensive purchase costs. Based on the fact that the focus is on non-listed firms, instead of using the marketto-­book value as the proxy for investment, the purchase of equipment emanating from the cash flow statement of companies is considered to be the best alternative.

3.7

Age

The Age element has been included to reflect the fact that an older company, after having built a stronger market base, has the propensity to use less debt. In fact, mature companies are able to better manage their cash flows. However, it can also be posited that age triggers a positive effect on leverage since older firms may need more funds to keep abreast with the latest technology. Most importantly, the age variable can also capture the extent of sound relationship established with banks in terms of securing finer rates for banking facilities.

264 

Economics and Finance in Mauritius

4

Results

4.1

Strategic Variables: Sales, Tax and Equity

To gain deeper insights as to the effects of the crisis that could potentially impact upon the financing structures of Mauritian firms, sales, tax and equity were introduced as additional explanatory variables. Sales are considered to filter out any crisis-induced decline in sales that could eventually impact on leverage. In the same vein, tax is taken aboard to check as to whether during stressful conditions, companies need to leverage to be tax solvent. Should this be the case, then, there would be need for fiscal adjustments to needy companies during hard times. Finally, equity is considered to analyse the critical conditions of firms during the crisis whereby banks usually require companies to respect their respective debt-to-equity ratios. In that respect, equity financing is considered an important driver to maintain ample access to leverage.

4.2

Results for non-strategic variables

Variance Inflation Factors did not disclose any multicollinearity problems for any of the regressions run. Similar findings were obtained in the case of correlation coefficients analysis.1 To cater for feasible heteroscedasticity problems, robust estimators were employed. Results are depicted in Tables 12.4 and 12.5, respectively. The explanatory powers of the models are found to be slightly higher in the crisis period. NDTS is found to play a major role chiefly in the case of long-term leverage for non-exporting firms in the pre and post crisis episodes. Under short-term leverage assessment, it transpires that size trails behind a positive effect in the case of non-exporting firms in the pre-­crisis period and exporting firms in the post crisis period. The positive sign is compatible with the notion that larger firms are likely to benefit lower information asymmetry problems so that they can avail themselves of larger debts in the short term. Such a finding bodes well with the s­ tudies  Results are not shown due to space but can be made available to any interested reader.

1

AGE SALES PRO NDTS

−0.36 TAX −2.42** TANG −30.04*** CAP −0.64 INV 3.02*** SIZE 6.92*** LIQ 12.54*** Cons F(10,731) Prob > F R- squared Obs

−0.01 3.77*** 1.00 1.71* 0.4223 −0.1057 −0.7543 −0.0631 0.0005 0.1099 0.7134 378.69 0.0000 0.7890 742

−0.0004 0.0210 −0.0389 −0.1837 1.41 −3.90*** −33.67*** −0.79 0.39 4.62*** 20.26***

−1.12 4.13*** −0.73 −1.02

t-value

TAX TANG CAP INV SIZE LIQ Cons F(10,407) Prob > F R-squared Obs

AGE SALES PRO NDTS −0.2846 −0. 0953 −0.6223 −0.0484 −0.0025 0.3276 0.5680 215.10 0.0000 0.7787 418

−0.0011 0.0323 −0.0792 0.3236

Coefficient

Pre-crisis export-CL

*, **, *** denotes statistical significance at the 10%, 5% and 1% level, respectively

TAX −0.0884 TANG −0.0798 CAP −0.7055 INV −0.0412 SIZE 0.0047 LIQ 0.2009 Cons 0.5219 F(10,1239) 299.20 Prob > F 0.0000 R-squared 0.6464 Obs 1250

−0.000004 0.0212 0.0592 0.2328

Coefficient

AGE SALES PRO NDTS

Post-crisis non-export-CL

Coefficient

t-value

Pre-crisis non-export-CL

Table 12.4  Results under short-term leverage

−1.16 −1.76* −16.35*** −0.75 −1.12 5.37*** 9.04***

−2.25*** 4.70*** −0.91 1.43

t-value

TAX TANG CAP INV SIZE LIQ Cons F(10,223) Prob > F R-squared Obs

AGE SALES PRO NDTS

−0.7087 −0.1857 −0.7149 −0.2204 0.0071 0.0903 0.5912 166.10 0.0000 0.8250 234

0.0009 0.0267 0.0345 −0.2861

Coefficient

Post-crisis export-CL

−1.44 −4.37*** −23.18*** −1.63 2.53** 2.11** 8.96***

1.69* 4.01*** 0.59 −1.16

t-value

AGE 0.0006 SALES −0.0228 PRO 0.1944 NDTS −0.1190 TAX 0.0343 TANG 0.1206 CAP −0.3944 INV 0.0498 SIZE 0.0004 LIQ −0.3491 Cons 0.4594 F(10,407) 38.99 Prob > F 0.0000 R-squared 0.5486 Obs 418

1.10 −2.56*** 1.87* −0.45 0.11 1.94* −8.74*** 0.66 0.16 −4.99*** 6.41***

Coefficient t-value

Pre-crisis export-LTL

*, **, *** denotes statistical significance at the 10%, 5% and 1% level, respectively.

3.03*** −3.05*** 1.77* 2.04*** −1.97** 4.18*** −10.32*** 2.41** −1.30 −4.84*** 6.83***

Coefficient t-value

Post-crisis non-export-LTL

AGE 0.0006 1.77* AGE 0.0012 SALES −0.0116 −2.05** SALES −0.0133 PRO 0.0684 1.31 PRO 0.1629 NDTS 0.4206 2.34** NDTS 0.3955 TAX −0.5431 −2.27** TAX −0.8329 TANG 0.1046 2.75*** TANG 0.1232 CAP −0.3398 −12.46*** CAP −0.2709 INV 0.0251 0.36 INV 0.1893 SIZE −0.0057 −3.39*** SIZE −0.0019 LIQ −0.2022 −6.04*** LIQ −0.1337 Cons 0.4506 9.73*** Cons 0.2754 F(10,1239) 37.85 F(10,731) 25.58 Prob > F 0.0000 Prob > F 0.0000 R-squared 0.3196 R-squared 0.3974 Obs 1250 Obs 742

Coefficient t-value

Pre-crisis non-export-LTL

Table 12.5  Results under long-term leverage

AGE −0.0010 SALES −0.0263 PRO −0.0536 NDTS 0.3184 TAX 0.9375 TANG 0.1505 CAP −0.3012 INV 0.2274 SIZE −0.0067 LIQ −0.1392 Cons 0.4503 F(10,223) 25.62 Prob > F 0.0000 R-squared 0.5686 Obs 234

−1.76* −3.57*** −0.84 1.24 1.66* 2.63*** −7.42*** 1.67* −2.33** −2.47*** 5.71***

Coefficient t-value

Post-crisis export-LTL

12  Capital Structure Analysis of Exporting and Non-exporting... 

267

of Lasfer (1995) and Rajan and Zingales (1995). Under long-term leverage, a negative sign is noted in the case of statistically significant results. Overall, therefore, a positive sign is obtained between short-term debt and size and a negative sign between long-term debt and size. It can therefore be conjectured that larger firms in Mauritius tend to harness a higher degree of short-term debt rather than long-term debt. In contrast to the theoretical foundations of the Pecking Order Theory, the current study does not find a statistically significant and negative relationship between profitability and leverage, i.e., the larger the internal funding base, the lesser the need to have recourse to external financing. The most plausible explanation for such a state of affairs is that most of the firms in the sample simply do not make enough profits to deter the use of leverage. Indeed, in the case of short-term leverage, independent of the pre- or post-crisis episodes and the type of firm under focus, all impacts of profitability are statistically insignificant. Ironically, in the case of long-term leverage, some bouts of positive effects are noted in the case of non-exporting firms in the post-crisis era and exporting firms in the pre-crisis period. The non-negative sign of profitability can be attributed to the fact that profits are not synonymous to instant cash. These findings contrast starkly with the findings in Vasiliou et al. (2003) and Proença et al. (2014). Nonetheless, the statistically significant positive signs could be attributed to the trade-off theory of capital structure whereby more lucrative firms are endowed with lower risks of bankruptcy so that they can avail themselves of higher leverage levels in view of exploiting the associated interest tax shields. Asset tangibility is found to generate negative pressures on short-term leverage but positive pressures on long-term leverage, consistent with the a priori foundations of the Pecking Order Theory. The positive relationship between asset tangibility and long-term leverage is compatible with the studies of Titman and Wessels (1988), Scott (1977) and Frank and Goyal (2002) whereby tangible assets take after strong collaterals which feed into the ability to take on more leverage. Interestingly, the effects are more pronounced in the post-crisis episode, irrespective of exporting or non-exporting firms. Such a result endorses the need to maintain solid collaterals to ensure availability of funds during difficult times. The negative relationship between short-term leverage and asset tangibility can be

268 

Economics and Finance in Mauritius

explained by the fact that the liquidity variable acts as the main driver for Mauritian firms to contract short-term leverage. Age variable is found to be statistically significant in some cases, albeit with only a small impact. Under short-term leverage, age exerts impacts only for exporting firms, but with the impact changing the sign – a negative effect prior to the crisis and a positive impact post the onset of the crisis. Such a finding could signify that relationship banking does help exporting firms to have better facilities under crisis conditions. This is also noted in the case of non-exporting firms, whereby there is a marginal increase in the size of the impact between the pre-crisis and the post-crisis periods. Overall, the impact of age is found to be low. Equity is found to unleash systematically robust downward pressures on the use of leverage; whether short-term leverage or long-term leverage. In the case of short-term leverage, the impacts hover in the range of −0.62 to −0.75 while the range of effects under long-term leverage is −0.27 to −0.39. There is strong evidence that stronger equity-based firms are less exposed to solvency risks based on undermined used of leverage. Interestingly, in the case of short-term leverage, there is a rise in the magnitude impact of equity on leverage from the pre-crisis era to the post-crisis period and this holds for both exporting and non-exporting firms. In the case of long-term leverage, the reverse manifests, i.e., there is a decline in the magnitude effect of equity on leverage from the pre-crisis episode to the post-crisis episode. Overall, findings do show that stronger equity-based firms tend to be subject to low leverage risk problems. Sales are found to trigger enshrining effects on short-term leverage to the order of 2–3% and this occurs irrespective of the type of firm or the period under investigation. Interestingly, in the case of long-term leverage, sales are found to trigger downward pressures on the use of leverage to the order of 1–2%. Therefore, sales tend to scale up the use of shortterm leverage but scale down the use of long-term leverage. Tax variable does not trigger any impact in the case of short-term leverage. Such a finding is interesting as it shows that taxes in Mauritius do not really hinder on financing structures of companies, mostly due to a particularly low and flat tax rate structure. In the case of long-term leverage, some negative impacts are mostly noted.

12  Capital Structure Analysis of Exporting and Non-exporting... 

269

Investments do not affect the financing structure of Mauritian firms under conditions of short-term leverage. The most plausible explanation is that these investments in fixed assets purchases are not high enough to increase the leverage level. Things become different under long-term leverage, however. In fact, for both exporting and non-exporting firms, a marked increase is noted with respect to the effects of investments on leverage with the results being statistically significant. Consequently, it can be conjectured that during the crisis periods, firms in Mauritius contracted longer term debt in view of financing their purchases of fixed assets. The liquidity variable is expected to capture the ability of a company to meet its short-term liabilities as they fall due and is thereby deemed to be a metric of short-term solvency. Results show that liquidity exerts positive impacts on short-term leverage while negative effects on long-term leverage in contrast to the findings of Laureano et al. (2012) and Proença et al. (2014). Such a finding emphasizes the need to consider leverage in dual forms – short-term and long-term components, respectively. Most importantly, this signifies that different countries tend to have different types of relationships when it comes to leverage assessment.

5

Conclusions

This chapter contributes to the emerging literature pertaining to the impact of the crisis on the capital structure of firms. Such research is particularly vital for policy makers. To gain increased insight, recourse is made towards both exporting and non-exporting firms. Above all, a dual dimensional analysis is made with respect to debt – short-term and long-term components, respectively. In light of the crisis, three strategic variables are used, namely, equity, tax and sales. Profitability variable is not found to unleash downward pressures on the use of leverage by Mauritian firms, the most plausible explanation for this being the fact that the sample consists of firms which do not make enough profits to scale down the use of leverage. The statistically significant positive signs, noted in the case of long-term leverage, could

270 

Economics and Finance in Mauritius

be attributed to the trade-off theory of capital structure whereby more lucrative firms are imbued with lower risks of bankruptcy so that they can avail themselves of higher leverage levels in view of exploiting the associated interest tax shields. Age variable is not found to unleash significant impacts on leverage of Mauritian firms, suggesting that relationship banking does not help significantly in securing higher levels of leverage. In addition, strong evidence is found for Mauritian firms having recourse towards longerterm debt in the post-crisis period in order to fund the purchases of fixed assets. In addition, liquidity exerts positive impacts on short-term leverage, but negative effects on long-term leverage. Moreover, size effects show that larger firms in Mauritius tend to harness more short-term debt than long-­term debt. Findings further show that liquidity acts as a major determinant of short-term Mauritian debt while asset tangibility acts as the chief driver in influencing the level of long-term debt. Such a finding is of particular interest as it shows the level of caution exercised by Mauritian banks in always adhering to strong collaterals whenever granting longer-term loans. Such policy is expected to render robust the level of financial stability in the Mauritian banking system. Three strategic variables are employed in the current study – namely, equity, sales and tax. Results show that strong equity-based firms tend to have lower levels of leverage to such an extent that the authorities should attempt to consolidate the equity positions of bruised firms in view of reducing potential bankruptcy problems. Such a finding adds momentum to the further democratization of equity-based financing in Mauritius by further enlarging the number of listed companies. The findings show that sales tend to induce the use of short-term leverage but scale down the use of long-term leverage. Tax tends to be impotent only in the case of short-term leverage. The impotency of the tax variable adds lustre to the fact that fiscal adjustments are not warranted to unleash some margin of manoeuvre to Mauritian firms.

12  Capital Structure Analysis of Exporting and Non-exporting... 

6

271

Policy Recommendations

Strong evidence prevails as to asset tangibility acting as a bullish force in inducing long-term leverage. This implies the need to put in place a general system of asset valuation in Mauritius. The use of tax as a tool to stimulate higher activities among Mauritian firms is unlikely to pay off. This could be due to the fact that tax is already at a low level in Mauritius. Thus, authorities should have recourse to alternative incentives to boost the activities of local firms such as lower interest rates. In fact, based on an already high interest rate spread, it could be high that that authorities impose a certain level of interest rate spread bands to ensure that there is no phasing out of low-return projects in the short run, but which trigger high-return impacts in the long run. The potential output of Mauritius would have been much higher today if these projects had been funded by banks. Above all, the registrar of companies should establish a rigorous online data system which incorporates all financials of Mauritian companies so that researchers can undertake not only cross-sectional but also time-series/panel data analyses to generate policy-oriented research. Although, such a data system exists, the data are not yet as comprehensive as those found in the annual reports of Mauritian firms. To ease the process, companies should be urged to submit both hard and soft copies (under an established template).

Bibliography Barclay, M. J., & Smith, C. W. (1996). On financial architecture: Leverage, maturity, and priority. Journal of Applied Corporate Finance, 8(4), 4–17. Barclay, M. J., Smith, C. W., & Watts, R. L. (1995). The determinants of corporate leverage and dividend policies. Journal of Applied Corporate Finance, 7(4), 4–19. Berger, P. G., Ofek, E., & Yermack, D. L. (1997). Managerial entrenchment and capital structure decisions. Journal of Finance, 52(4), 1411–1438. Bevan, A., & Danbolt, J. (2000). Dynamics of the determinants of capital structure in UK. Working paper 2000/9.

272 

Economics and Finance in Mauritius

Chakraboti, I. (2010). Capital structure in an emerging stock market: The case of India. Research in International Business and Finance, 24, 295–314. Chang, C., Chen, X., & Liao, G. (2014a). What are the reliably important determinants of capital structure in China? Pacific-Basin Finance Journal, 30, 87–113. Chang, Y.-K., Chou, R. K., & Huang, T.-H. (2014b). Corporate governance and the dynamics of capital structure: New evidence. Journal of Banking and Finance, 48, 374–385. Chittenden, F., Hall, G., & Hutchinson, P. (1996). Small firm growth, access to capital markets and financial structure: Review of issues and an empirical investigation. Small Business Economics, 8, 59–67. Donaldson, G. (1961). Corporate debt capacity: A study of corporate debt policy and the determination of corporate debt capacity. Boston: Division of Research, Harvard Graduate School of Business Administration. Fama, E. F., & French, K. R. (2002). Testing tradeoff and pecking order predictions about dividends and debt. Review of Financial Studies, 15(1), 1–43. Frank, M. Z., & Goyal, K. V. (2002). Testing the pecking order theory of capital structure. Journal of Financial Economics, 67, 1–30. Griner, E. H., & Gordon, L. A. (1995). Internal cash flow insider ownership capital. Journal of Business Finance and Accounting, 22, 179–199. Grossman, S., & Hart O. (1982). Corporate financial structure and managerial incentives: The economics of information and uncertainty (J.  McCall, Ed.). Chicago: University of Chicago Press. Hall, G., Hutchinson, P., & Michaelas, N. (2000). Industry effects on the determinants of unquoted SMEs’ capital structure. International Journal of the Economics of Business, 7(3), 297–312. Jensen, M., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3(4), 305–360. Jensen, M.  C. (1986). Agency costs of free cash flow, corporate finance and takeovers. American Economic Review, 76, 323–329. Kayo, E. K., & Kimura, H. (2011). Hierarchical determinants of capital structure. Journal of Banking and Finance, 35, 358–371. Kester, W. C. (1986). Capital and ownership structure: A comparison of United States and Japanese manufacturing corporations. Financial Management, 15(1), 5–16. Lasfer, M. A. (1995). Agency costs, taxes and debt: The U. K. evidence. European Financial Management, 1(3), 265–285.

12  Capital Structure Analysis of Exporting and Non-exporting... 

273

Laureano, L., Urbano, H., & Laureano, R. M. S. (2012). Debt maturity structure: Evidence from Greece, Ireland, Italy, Portugal and Spain. In Proceedings of the XXII Jornadas Luso-Espanholas de Gestão Científica (pp. 1–15). Mateev, M., Poutziouris, P., & Ivanov, K. (2013). On the determinants of SME capital structure in Central and Eastern Europe: A dynamic panel analysis. Research in International Business and Finance, 27, 28–51. Modigliani, F., & Miller, M. H. (1958). The cost of capital, corporation finance and the theory of investment. American Economic Review, XLVIII(3), 261–297. Moore, W. (1986b). Asset composition, bankruptcy costs and the firm’s choice of capital structure’. Quarterly Review of Economics and Business, 26, 51–61. Myers, S. C. (1984b). The capital structure puzzle. Journal of Finance, 39(3), 574–592. Myers, C.  S., & Majluf, S.  N. (1984). Corporate financing and investment ­decisions when firms have information that investors do not have. Journal of Financial Economics, 13(2), 187–221. Ozkan, A. (2001). Determinants of capital structure and adjustments to long run target: Evidence from UK company panel data. Journal of Business Finance and Accounting, 28, 175–195. Proença, P., Laureano, R., & Laureano, M. (2014). Determinants of capital structure and the 2008 financial crisis: Evidence from Portuguese SMEs. Procedia Social and Behavioral Sciences, 150, 182–191. Rajan, R. G., & Zingales, L. (1995). What do we know about capital structure? Some evidence from international data. The Journal of Finance, 50, 1–58. Ramlall, I. (2009). Determinants of capital structure among non-quoted Mauritian firms under specificity of leverage: Looking for a modified pecking order theory. International Research Journal of Finance and Economics, 31, 83–92. Remmers, L., Stonehill, A., Wright, R., & Beekhuisen, T. (1974). Industry and size as debt ratio determinants in manufacturing internationally. Financial Management, 3, 24–32. Scott, J. H. (1977). Bankruptcy, secured debt, and optimal capital structure. The Journal of Finance, 32, 1–19. Stohs, M.  H., & Mauer, D.  C. (1996). The determinants of corporate debt maturity structure. Journal of Business, 69(3), 279–312. Titman, S., & Wessels, R. (1988). The determinants of capital structure choice. Journal of Finance, 43, 1–18. Van der Wijst, N., & Thurik, R. (1993). Determinants of small firm debt ratios: An analysis of retail panel data. Small Business Economics, 5, 55–65.

274 

Economics and Finance in Mauritius

Vasiliou, D., Eriotis, N., & Daskalakis, N. (2003). The determinants of capital structure: Evidence from the Greek market. Paper presented at the 2nd annual meeting of Hellenic Finance and Accounting Association, Athens, Greece, pp. 1–16. Watson, R., & Wilson, N. (2002). Small and medium size enterprise financing: A note on some of the empirical implications of pecking order. Journal of Business Finance and Accounting, 29, 556–578. Yang, C.-C., Lee, C.-f., Gu, Y.-X., & Lee, Y.-W. (2010). Co-determination of capital structure and stock returns-A LISREL approach. An empirical test of Taiwan stock markets. The Quarterly Review of Economics and Finance, 50, 222–233. Zoppa, A., & McMahon, R. G. P. (2002). Pecking order theory and the financial structure of manufacturing SMEs from Australia’s Business Longitudinal survey. School of commerce, Research paper series: 02–2 ISSN-1441-3906.

Part VI Hedging in Mauritius

13 What Factors Drive Hedging Among Mauritian Firms?

This chapter fills a vital gap when it comes to assessing the drivers behind hedging in Mauritius. It is widely acknowledged by both practitioners and academicians that knowledge about derivatives’ use constitutes a major hurdle in their use. Prior empirical evidence on hedging is chiefly focused on developed countries with little being understood about the situation in developing countries. Using data for the period 2005–06, findings clearly show that managers’ incentives to hedge, the tax convexity motive to hedge along with financial and operational explanations underlying hedging, are not applicable to the Mauritian context. Important drivers for hedging are found to be a composition of the size and age of firms, buttressing the notion that high fixed costs and knowledge about derivatives act as the main propelling forces for a derivatives framework. These findings suggest that larger firms are the ones most susceptible to use derivatives. Consequently, smaller firms, though subject to currency risk, are less susceptible to endorse sound risk management programmes. The government is therefore advised to establish a derivatives house whereby smaller firms can have the possibility to hedge as per their size so that currency risk impacts are well contained in the economy.

© The Author(s) 2017 I. Ramlall, Economics and Finance in Mauritius, DOI 10.1007/978-3-319-39435-0_13

277

278 

1

Economics and Finance in Mauritius

Introduction

The empirical literature on determinants of hedging is mostly tilted towards the experience of developed economies. Different reasons are cited for the reasons for firms to seek to hedge, including the financial distress costs, the underinvestment hypothesis, and the convexity of tax structure, amongst others. Unfortunately, very little is presently known about hedging in the Mauritian context. It becomes interesting to gain an understanding as to how Mauritian firms hedge their currency risk exposures. However, prior to embarking on a full-fledged analysis, it is important to understand the features of the Mauritian aspect with respect to derivatives. First, local firms use forward contracts which are issued by commercial banks. The advantage attached to these forward contracts is that they impose no initial or maintenance margin account akin to futures contracts. In addition, the fees or commissions applied are relatively low because the bank provides a whole set of facilities such as overdraft, documentary credit, bank guarantees and loans. Second, practitioners are fully aware of the fact that the rupee has maintained a depreciating trend against the US dollar, euro and the pound sterling for many years – in particular, years prior to the onset of the US subprime crisis. Consequently, at that point in time, it became unjustifiable for exports to hedge their exposure as a depreciating rupee automatically leveraged on the value of foreign currency deposits. Such a state of affairs explains why many exporters preferred to practice currency hoarding during the period of rupee depreciation. This chapter analyses the determinants of hedging in Mauritius. It is an extensive analysis, covering listed companies, unlisted firms and the top 100 companies in terms of growth and turnover, respectively. Beyond that, the research is interesting because it probes into the decision to hedge for an upper-income developing country, the empirical evidence for which is practically nonexistent. The chapter proceeds as follows. Section 2 focuses on the rationale which induces firms to hedge. Section 3 provides an insight on the features of Mauritian firms along with sound reasons advanced to account for the degree to which ­hedging theories are applicable in Mauritius. Section  4 is geared towards the data analysis along with the econometric model. Section 5 discusses the results. Finally, Sect. 6 concludes.

13  What Factors Drive Hedging Among Mauritian Firms? 

2

279

Reasons Motivating Firms to Hedge

Traders such as arbitrageurs, hedgers and speculators account for all of the derivatives activities carried out around the world. In fact, derivatives markets have grown substantially over recent years and are still continuing on an upward trend. Under conditions of perfect capital markets, it does not make sense for firms to hedge by virtue of the fact that investors can themselves replicate corporate hedging activities. In practice, many frictions permeate the financial system. Consequently, nearly all the empirical literature on hedging starts with frictions being introduced in the ideal model of Modigliani and Miller (1958). The next section focuses on the different theories which prevail to account for the use of derivatives for hedging purposes.

2.1

Financial Distress Costs Hypothesis

If financial distress is expensive, firms will automatically increase their values through hedging activities because hedging scales down the present value of financial distress costs. Technically speaking, firms which are endowed with high gearing in their capital structure are more likely to be subject to financial distress costs. Indeed, Dolde (1995), Berkman et  al. (1997), Haushalter (2000), Gay and Nam (1998) and Graham and Rogers (2002), all pointed out a positive relationship between hedging and gearing. However, some studies found results that contradicted these findings. In essence, Nance et al. (1993), Geczy et al. (1997), and Allayannis and Ofek (2001) found no positive endorsement for this hypothesis. In the same vein, Joseph and Hewins (1997) found that the stress exerted on financial distress costs hypothesis was relatively weak for an analysis which was done for UK multinationals. Dionne and Garand (2003) proposed a mixture of leverage and liquidity to capture for the degree of the financial distress costs. Their variable was set to one in case a firm was subject to a debt and a quick ratio level which were above and below the industry’s mean, respectively.

280 

Economics and Finance in Mauritius

2.2

Underinvestment Hypothesis

Under the underinvestment hypothesis, it is widely acknowledged that corporates which are subject to important investment opportunities and are simultaneously buffeted by lack of funds, are most likely to suffer from underinvestment. Froot et  al. (1993) found that hedging could dampen down cash flow volatility so that firms were able to fund their projects during downturns in the economy through internally generated funds. But, Lin and Smith (2003) pointed out that different firms were subject to distinctive asset structures so that a high market-to-book ratio did not always send the signal of more valuable investment opportunities. To capture for investment opportunities, a second proxy was proposed in the literature, namely, research and development expenditure, scaled by total assets. Moreover, the empirical literature also resorted to the use of the number of financial analysts following the firm – the higher the level of public analysis on a firm, the lower the level of information asymmetry to the outsiders. Allayannis and Ofek (2001) employed dividend yield by virtue of the fact that firms could unleash more liquidity by cutting down their dividend payments. Borokhovich et al. (2004) adhered to the quick ratio while Nguyen and Faff (2003) resorted to both the current ratio and the ratio of cash and cash equivalents to the firm’s size. Gay and Nam (1998) had recourse to a dummy variable, which was equal to one in the case that a firm was simultaneously suffering from low cash but imbued with high growth opportunities. Similarly, Haushalter (2000) argued that firms which had rated debt were subject to more scrutiny so that they suffered less from information asymmetry.

2.3

Tax Convexity

The tax benefit hypothesis is based on the premise that firms which are subject to larger tax loss carry forwards have a greater propensity to hedge. In essence, the tax convexity motive to hedge is based on the notion that, if a firm is subject to an effective tax schedule which is convex, then a decline in the volatility of before tax income would diminish the amount of expected taxes, making it lucrative for the firm to hedge. The principal tax motive for

13  What Factors Drive Hedging Among Mauritian Firms? 

281

hedging is to lessen the variability of taxable income but not the variability of cash flow. Thus, under a convex tax function, hedging tends to leverage on the after-tax firm value. The most famous metric of the tax function convexity had been the amount of the tax loss carryforwards. A second metric, widely used for the tax function convexity, consists of a dummy variable, which indicates whether a firm’s pre-tax income is expected to be in the progressive region of the tax code. As a matter of fact, Smith and Stulz (1985) pointed out that hedging ironed out pre-tax income functions so that firms were able to increase the probability of using tax preference items.

2.4

Managers’ Motives

Managerial theories behind hedging are based on the premise that firms get involved in hedging to avoid capital market discipline. Many papers used the number of exercisable options held by managers as a proxy for the risk-taking incentives of managers. Furthermore, other theories try to link corporate hedging to managerial reputation such as DeMarzon and Duffie (1995) and Breeden and Viswanathan (1998). In that perspective, hedging can undermine the degree of informational asymmetry among shareholders, managers and the labour market. The reason is that managers who possess superior skills may have recourse to hedging to better communicate their skills to the market. Joseph and Hewins (1997) found evidence in favour of the aim of hedging being associated to risk reduction at a level with which the management felt comfortable.

2.5

Country-Specific Features

De Jong et al. (2002) pointed out that the use of derivatives was higher in the case of Dutch firms than in that of US firms. Such a distinction was attributed by them to the fact that the USA tends to be endowed with a more protective legal structure for shareholders’ rights compared than that in The Netherlands. Besides, Lel (2004) noted that firms which operated in a country with a developed financial market and robust corporate governance were more likely to hedge. Similarly, Bartram et al. (2004) found that the size of derivatives market and level of financial risk influenced the hedging decision.

282 

2.6

Economics and Finance in Mauritius

Size

The empirical literature is full of studies which show that lower size is related to the more extensive use of derivatives for hedging purposes. In fact, if hedging costs were inversely proportional to firm size, then the entire picture would change as smaller firms would be more inclined to hedge since they were the ones most concerned about financial distress costs. In addition, as per Froot et al. (1993), firms which were subject to expensive external sources of funding were more likely to hedge. Based on the notion that smaller firms are more likely to be affected by information asymmetry problems, they are expected to hedge under costly external sources of funding. Unfortunately, smaller firms tend to be deprived of ample resources to establish proper derivatives programmes. In that respect, it is argued that larger size entails greater hedging as this makes it possible to harness economies of scale in transactions. Besides, larger firms are more susceptible to hedge because they tend to be imbued with complex and geographically dispersed operations, all of which are likely to lead to a higher requirement for hedging purposes.

2.7

Corporate Governance

Corporate governance features are anticipated to influence the risk management policy through their role as a risk mitigation tool for the principal–agent problem. There are two main tools which are being used and they consist of ownership concentration and board characteristics. Under ownership concentration, firms which experience higher levels of ownership structures are less likely to be affected by agency conflicts as shareholders ensure that the best practices are maintained in the firm so as to maximize its value and thus hedging is highly appreciated. Basically, large shareholders have the proper resources to ensure that sound monitoring is being carried out on the managers’ activities to undermine any feasible management self-interest hedging motive. The variable percentage of shares held by blockholders is widely used to control for the firm’s ownership structure. Under board characteristics, it is presumed that board independence should play a pivotal role in the firm’s risk management

13  What Factors Drive Hedging Among Mauritian Firms? 

283

horizon. As a matter of fact, Whidbee and Wohar (1999) stated, in instances where there were low levels of insiders’ shareholdings, managers’ decision to hedge with derivatives was influenced by outside directors’ membership in the board.

3

Characteristics of Mauritian Firms

Before embarking on the methodology and research model to be deployed for the analysis, it is important to understand the characteristics of Mauritian firms. In fact, each country has its own specific characteristics which tend to impact on the hedging behaviour. For example, De Jong et al. (2002) pointed out that the use of derivatives was more widespread among Dutch firms relative to US firms by virtue of the stronger legal frameworks which characterize the U.S. economy.

3.1

Characteristics of Local Firms

The following components are given due consideration in order to draw a sound picture of the Mauritian companies, from the hedging perspective.

3.1.1  Concentrated Ownership Structure Mauritian firms are inclined to be endowed with concentrated ownership level so that agency costs are practically nonexistent. Such a state of affairs, being so blatant and predominant, tends to prevail even among the listed firms. Consequently, this signifies that a manager’s motives to hedge are unlikely to be applicable for Mauritian firms.

3.1.2  Derivative Products (Forwards) Collected from Firms The Mauritian financial system is inherently a bank-based financial system in which case bank financing predominates over any other source of financing. Ironically, bank financing is so widespread in Mauritius that

284 

Economics and Finance in Mauritius

even listed firms cling to bank financing relative to equity financing.1 Thus, local companies principally adhere to forward contracts which are issued by local banks to hedge their foreign exchange risk exposure. The two local banks, namely, SBM Ltd. and MCB Ltd., are found to be the main players in the local forward markets although there is no such thing as transparent prices being disseminated on the bank counters. The reason which induces firms to use forward contracts from banks to hedge is due to the fact that Mauritian banks provide this service as a whole package of banking facilities. For instance, a company tends to profit from lower fees being charged for a forward contract facility in the case that the same company is already availing itself of various facilities such as loans, overdraft or bank guarantees from the same bank, synonymous with economies of scale in transactions. Furthermore, the bid–ask spreads (the margin between the rate at which banks buy and the rate at which they sell) are not materially distinct from those for spot ­transactions. Therefore, recourse to forward contracts by local firms does not entail substantial costs on the local firms relative to a situation where foreign instruments would have been employed or only a forward contract facility is to be contracted form a bank. Moreover, since the local forward contracts are being issued by Mauritian banks, they are automatically denominated in rupees against a foreign currency. If foreign derivatives were to be used by Mauritian firms, that would bring additional risk.2 In a nutshell, it can be stated that the degree of financial market development entails a direct bearing effect on the choice of derivative instruments along with the costs of hedging. If local banks were not involved in the provision of forward contracts, then many companies would be unable to hedge their currency risk exposure or they would forgo hedging if it were too costly. Notwithstanding, Mauritian banks are highly selective in providing forward contracts because only healthy companies are eligible to access such a facility.

 Rights issues do not constitute a common practice on the Stock Exchange of Mauritius.  This could also lead to lower risk, it all depends on the movements in the two currencies vis-à-vis the rupee. 1 2

13  What Factors Drive Hedging Among Mauritian Firms? 

285

3.1.3  Chronic Depreciation on the Back of Ongoing Deficit3 The Mauritian rupee maintained a depreciating trend vis-à-vis major currencies for the period prior to the onset of the US subprime crisis. It was, therefore, prudent for exporters to maintain their revenues intact in a foreign currency deposit account to avail themselves of higher monetary values. Interestingly, not only exporters but also importers who were induced to hold foreign currency deposit accounts in banks with a view of suppressing their import costs. More specifically, the rupee had been subject to a depreciating trend against the euro, the pound sterling and the US dollar in the period from 1999 to 2006. Such a depreciation is inherently related to the balance of payments deficit, meaning that a weak currency can help to restore back equilibrium, in theoretical terms. In practice, the balance of payments conditions always showed current account deficits being funded by capital inflows.

3.1.4  Price-Takers Being a small open economy, Mauritian importers and exporters have no choice but to simply accept currency risk as something natural to and an intrinsic part of international trade.

3.1.5  C  omplacent Hedging Behaviour-Risk Borne by Consumers In case of the local car industry, many car dealers maintained bank accounts in foreign currencies so that any increase in costs based on currency risk are automatically being shifted to consumers. In that respect, car dealers are unlikely to hedge as they have the power to transfer any currency risk to the purchaser. However, a considerate car dealer can undertake proper risk management to ensure lower operating costs and to maximize profits. Unfortunately, many car dealers are unaware of derivatives, displaying a  Such a  depreciating rupee is highly appreciated among exporters as  it their revenue receipts denominated in rupees. 3

286 

Economics and Finance in Mauritius

lack of knowledge about such instruments. Such a need becomes warranted when one bears in mind the strong level of competition which characterizes the car dealer industry.

3.1.6  Underlying Risk Exposure Local firms do not often provide figures pertaining to the level of their foreign sales. In that respect, on the ground of data limitations, it is not feasible to gain knowledge about their underlying risk exposures. Nonetheless, this does not constitute an issue of concern because firms give risk management information as notes to their financial statements which set the stage in having the sample of forward hedgers. Moreover, the use of foreign sales variable in the empirical literature is considered as an important variable chiefly for multinationals. But, in Mauritius, many of the local firms do not have subsidiaries abroad, i.e., multinationals (locally established firm having subsidiaries abroad) are nearly absent in Mauritius. Otherwise, even the largest Mauritian firms are comparatively smaller than foreign peers.

3.2

Relevance of Hedging Theories in Mauritius

As the saying goes, “Cut your coat according to your cloth”. Similarly, at first sight, it appears improper to think that all the motives for hedging would be applicable in the Mauritian context. The underlying rationale emanates from country-specific drivers which need to be given due consideration. The following section analyses each of the motives to hedge and assesses its relevance in the Mauritian context.

3.2.1  Tax Hypothesis At the outset, it is important to note that, in Mauritius, the tax hypothesis is not applicable to the country by virtue of the fact that the local corporate tax system had been established at a flat rate of 15%. Therefore, the tax convexity motive which induces hedging in some countries will

13  What Factors Drive Hedging Among Mauritian Firms? 

287

not be applicable in the case of Mauritius. Such a motive for hedging can therefore be safely disregarded in our modelling framework.

3.2.2  Financial Distress Costs Hypothesis It can be argued that Mauritian firms which experience a high level of currency risk exposure, then, the financial distress costs motive can manifest as an important channel which drives firms to hedge. For example, with hedging, savings derived from undesired fluctuations in exchange rates can be employed to make good for interest payments, thereby mitigating the feasibility of the firm going bust. Thus, the hedging motive to weaken financial distress costs tends to be particularly relevant for many Mauritian firms, chiefly those, which are heavily strained by cash shortfalls during rising interest rates.

3.2.3  Underinvestment A priori, underinvestment is likely to manifest on the back that external financing such as bank loans is always costly relative to internal ­financing such as the use of internal source of funds. Consequently, through unleashing additional funds through the process of hedging, this can reduce the underinvestment level of local firms. In fact, many Mauritian firms, although they do not engage in research and development, are always involved in the acquisition and disposals of new equipment and machinery to boost sales. Viewed from this perspective, hedging helps to suppress the burden of the acquisition costs related to the purchase of new equipment.

3.2.4  Managers’ Motives Like the tax hypothesis, where weak evidence is found as a motive for Mauritius firms to hedge, the managers’ motives to hedge are also found to be weak and this holds whether or not the firm is quoted on the Stock Exchange in Mauritius. The underlying reason is based on the notion

288 

Economics and Finance in Mauritius

that, due to the concentrated ownership level, it defeats the purpose for managers to hedge to signal about their management quality to the outside world. In a nutshell, among the diverse motives for hedging found in the empirical literature, it is believed that underinvestment and the expected financial distress costs hypotheses are the most likely to hold in the case of Mauritius.

3.3

Substitutes for Hedging in Mauritius

This section analyses the degree to which the different substitutes for hedging are likely to apply in the case of Mauritius. First, corporate governance, foreign currency debt and dividends are given due consideration in line with the empirical literature. Second, attention is laid on the practical side of the picture as to how Mauritian firms can utilize non-derivatives means to hedge their currency risk exposure.

3.3.1  Corporate Governance Corporate governance, as stressed by Lel (2006), constitutes a key substitute ingredient to hedge in the presence of strong agency costs. In Mauritius, however, even the quoted firms (both on the official market and on the development enterprise market) are imbued with a very concentrated ownership structure with most companies being managed by their principal owners. As far as unlisted firms are concerned, the owners are mostly the managers, synonymous with closely-held firms. Thus, the positive impact of corporate governance on hedging is unlikely to hold in the case of Mauritius. In general, since most firms are closelyheld firms, there is no need to have recourse towards corporate governance as a substitute metric for hedging in order to generate higher market values.

13  What Factors Drive Hedging Among Mauritian Firms? 

289

3.3.2  Foreign Currency Debt Foreign currency debt can also be utilized as an alternative mode of hedging device. This can happen either via borrowing from a local bank in a foreign currency or through direct borrowing in a foreign currency from a foreign financial institution. As argued by Clark and Judge (2007), it is preponderant to create a demarcation line between those hedging instruments suited to short-term transaction exposures such as forwards, futures and options and those which are more appropriate for long-term multiple period foreign currency exposure, such as currency swaps and foreign currency debt. The issue of whether short-term or long-term hedging instruments are important for Mauritian firms can best be analysed by focusing on their capital structure. As found by Ramlall (2009), Mauritian firms lean towards short-term financing. In that respect, it is anticipated that local firms are more prone towards the use of more short-­ term hedging tools such as forwards rather than foreign currency debt.

3.3.3  Dividends Based on the fact that local shareholdings structure is concentrated and managers tend to be owners of the firms, the use of dividends as a signalling device is not susceptible to act as an alternative mechanism to the use of derivatives for hedging. In fact, with most of the local firms being owned by their managers, irrespective of whether they are listed or non-listed, there is no need to use either direct directives or substitutes to derivatives such as dividends to trigger positive signal to the market.

3.3.4  Operational Hedging The way that exchange rate fluctuations impact on the cash flow of multinational corporations has been the subject of much empirical research in the sphere of economics and finance. As a matter of fact, exchange rate variations have a direct effect on firms’ cash flows, and the values of their assets and liabilities. Thus, most of the estimates of foreign exchange risk have been based mainly on the exposure of multinational companies. This had paved the way to an analysis as to whether operational

290 

Economics and Finance in Mauritius

hedging can manifest as either a substitute for or a complement to financial hedging. But such a testing does not suit the Mauritian context because local firms do not have subsidiaries abroad – there might be only a marginal number of multinationals. Thus, in these circumstances operational hedging is not of any major significance for local firms. In addition, as stated by Kim et al. (2006), “For most exporting firms, the need for financial hedging should be greater since they are not operationally hedged or less hedged.” Unlike previous studies which have focused exclusively on multinational corporations, this chapter focuses on foreign exchange exposure based on Mauritian firms.

3.3.5  Practical Side of the Picture One interesting finding noted while skimming through the financial statements is that the following ways are being mentioned by Mauritian firms as to how they hedge their foreign exchange risk exposure: 1. Cash buffers/profits since exchange rate is managed on an ongoing basis. 2. Maintenance of bank accounts in foreign currency.4 3. Use of forward contracts. 4. Use of debt denominated in a foreign currency.5 The above classification is entirely compatible with the fact that hedging is basically geared towards short-term risk rather than long-term risk management. The above classification is effected in the decreasing order of usefulness, i.e., the first one is widely employed by firms which are subject to currency risk. The last one is found to be being used by just a few firms. Based on the above list, the most widely coveted hedging device comprises of the extent to which the company is awash with cash or profits (eventually to be converted into cash) so that it can bear any currency risk. In fact, Nance et  al. 1993) pointed out that companies  The solution is to analyse whether they report their asset and liabilities in foreign currencies to make them eligible for the analysis. 5  But this holds only for a small number of companies. 4

13  What Factors Drive Hedging Among Mauritian Firms? 

291

might resort to their profits as substitutes for the use of currency derivatives. Thus, more lucrative firms would be less susceptible to the use of currency derivatives. In that respect, while gauging the importance of forward contracts among local firms, it becomes important to factor in its main substitutes, which come in the form of cash/profits level and bank accounts denominated in foreign currencies.

4

Data and Econometric Model

Prior to undertaking any analysis on the drivers for hedging, it is of paramount significance to level the playing field at the start of the analysis by including in the sample only those firms which are, ex-ante, exposure to currency risk. Alternatively stated, only those firms in which their activity renders them a likely eligible candidate for currency risk are considered. Accordingly, the study considers about 400 firms. Overall, this research sheds light on the motives for hedging by focusing on a sample of non-financial Mauritian companies for the year 2005–06. The analysis undertaken is extensive as it caters for a large set of firms, in particular, importers, exporters,6 listed and unlisted firms, respectively. All the data were collected in person from the Registrar of Companies by the author. Technically speaking, the sample under consideration should exclude financial firms (banks, investment and insurance companies). The reason for this is that hedging mostly characterizes the non-­ financial sector whereas speculation predominantly prevails among the financial counterpart.7 The main utilities,, the Central Water Authority, the Central Electricity Board, the State Trading Corporation and the Mauritius Sugar Syndicate, have all been excluded from the analysis since the government is involved in their operations. In addition to the fact it is a utility company, the Mauritius Sugar Syndicate is ignored since it represents the most active and dynamic hedger in Mauritius, clearly constituting an outlier in the data.  Technically speaking, the sample should be biased towards importers based on the depreciating rupee; because many exporters are also importers, however, they were incorporated into the analysis. 7  Both hedgers and speculators account for the main actors in the derivatives market. 6

292 

Economics and Finance in Mauritius

A large sample of importers and exporters are considered, with their financial statements being properly examined to know whether or not they hedge. However, it is considered improper to dichotomize the whole sample into importers and exporters, respectively due to the fact that many exporters were also importers in one way or another, with the most common instances pertaining to the imports of raw materials. While collecting the data, it is important to ensure that at no point has it been mentioned whether or not derivatives are being employed for speculative purposes, consistent with the fact that hedging represents the chief motive for hedging among non-financial firms. In fact, under the financial instruments section, the firm should disseminate information with respect to all risks which impact on the bottom line of the firm. The most common risk identified among local firms pertains to the credit risk related to trade payables.

4.1

Contribution to the Empirical Literature

The foreign exchange risk exposure and hedging features of Mauritian companies is being assessed in this research. Indeed, most of the prior studies analysed the determinants of hedging for firms found in developed countries, with greater stress being laid on developed countries, principally reflecting the availability of data from these countries. Moreover, in spite of the fact that some cross-country analyses have been carried out, yet, they have been criticized on the basis of there being too much heterogeneity in the way in which firms behave. Thus, it is considered better to focus on the experience of a specific country with a view of uncovering more reliable results with respect to the determinants of hedging.

4.2

Hedging Model

As per Gezcy et al. (1997), the hedging decision model is specified as follows with some modifications introduced in order to have cleaner proxies and also to be compatible with the Mauritian context.

13  What Factors Drive Hedging Among Mauritian Firms? 



293

 Research & Development, Liquidity, Profit, Size,  Di = ƒ   (13.1)  Tang, Age, Groowth, Distress 

Where: Size Natural logarithm of total assets Research & Development Capital expenditures over total assets Liquidity: Quick ratio and cash over total assets. Leverage Current plus long term debt over total assets Profit: Net profit over total assets Age Log of number of years of existence as from the year incorporated Tang Tangibility of assets denoted by fixed assets over total assets Growth Percentage change in total assets Distress Leverage over total assets, interest coverage ratio and a dummy variable that is equal to one in the case debt and a quick ratio level that is, respectively, above and below the sample’s mean.

Both probit and logit models are employed in the estimation process to spawn a holistic assessment. The dependent variable, Di in Eq. (13.1), reflects a binary variable, i.e., it takes the value of one in the case that the firm hedges and zero otherwise. The hedging decision is deemed to be a linear function of the independent variables. The advantage of using the binary variable is that it obviates the use of notional amounts of derivatives which tends to be buffeted by noisy disclosures resulting from aggregation and netting. Above all, not all local firms report on the notional amount of derivatives in use so that the binary variable happens to be a more encompassing approach to capture the hedging practices among Mauritian firms. The model ignores the accuracy variable because in Mauritius, there is no such variable in terms of analyst forecast accuracy to proxy for asymmetric information. Nonetheless, no major difference is expected to manifest, bearing in mind the high level of insider ownership which permeates Mauritian firms. Similarly, the tax variable has been overlooked by virtue of the fact that the tax function convexity is inapplicable in Mauritius as firms face a flat tax system in lieu of a progressive tax system.

294 

Economics and Finance in Mauritius

4.2.1  Financial Distress Cost Hypothesis To isolate financial distress cost, recourse is made towards leverage over total assets. A positive relationship is anticipated to prevail between hedging and financial distress cost. Leverage is defined as all loans and leases.

4.2.2  Expertise in Hedging To ascertain the expertise level, recourse is made to the age of the company based on the fact that expertise is considered to be important in the practice of hedging. As a matter of fact, age is measured as the natural logarithm of the number of years lapsed since the date of incorporation. A positive relationship is thereby posited to manifest between age and hedging.

4.2.3  Profitability It is anticipated that not only liquidity but also profitability may impact on either the hedging decision or the extent of hedging. Thus, it becomes important to include profitability in order to gauge on the degree to which it acts as an effective substitute to hedging with forward contracts. Therefore, profitability should trail behind a bearish effect on hedging.

4.2.4  Tangibility of Assets The more tangible the firms’ assets, the greater its ability to issue secured debt. Because foreign debtors are more susceptible to demand collateral, firms endowed with larger tangible assets are more likely to borrow from abroad. These firms would be more likely to keep unhedged positions so that a negative relationship is anticipated to prevail between tangibility and the use of currency derivatives.

13  What Factors Drive Hedging Among Mauritian Firms? 

295

4.2.5  Size Size is used to capture the fact that larger firms are more inclined to engage in risk management on the back of economies of scale in derivatives programme. Indeed, the economies of scale involved in setting up a hedging program constitutes an important channel for the relationship between size and hedging. Most importantly, there are fixed costs embedded in establishing a derivatives program. Prior empirical studies have found a strong positive relationship between the size of the firm and the likelihood of hedging activity (Geczy et al. 1997; Mian 1996; Haushalter 2000; Allayannis and Ofek 2001; Graham and Rogers 2002). Size is incorporated as the logarithm of total assets to control for the size effect. A positive relationship between size and hedging is anticipated to manifest. Besides, it is most likely that larger firms are imbued with a larger risk exposure relative to smaller firms.

4.2.6  Liquidity The quick ratio, defined as current assets less inventories divided by current liabilities, is included in an attempt to capture the availability of internal funds (Geczy et al. 1997). The quick ratio is employed to proxy for the state of liquidity of the firm. A negative relationship is anticipated to prevail between the quick ratio and hedging activities. Cash over total assets is also employed to proxy for another feasible substitute for hedging in Mauritius. In fact, in the instance that currency risk does not impact substantially on the assets or liabilities side of the company, the ample maintenance of cash constitutes the best avenue to bear the foreign exchange rate risk.

4.2.7  Growth Opportunities Froot et al. (1993) stated that given capital market imperfections, firms would hedge in order to reduce their underinvestment problem. Thus, firms with higher growth opportunities are more likely to use currency

296 

Economics and Finance in Mauritius

derivatives. Companies’ growth opportunities are defined as the ratio of capital expenditure to total assets. The empirical literature employed research and development costs over total assets to proxy directly for research and development. In Mauritius, however, research and development is practically nonexistent among the local firms. This state of affairs prevails even for larger firms. The best possible avenue present to capture such a variable refers to the level of capital expenditure, computed by the purchase of fixed assets. In general, Mauritian firms’ investment is mainly focused towards the acquisition of vehicles to harness a larger activity base level or to replace old and obsolete machinery. A positive relationship is expected to occur between hedging and capital expenditure in line with the underinvestment hypothesis. It is assumed that hedging, financing and investment decisions are exogenously determined so that there is no need to employ a simultaneous equations framework, as pointed out by Lin and Smith (2003).

5

Results

No multicollinearity problem is noted among the independent variables. The highest correlation noted hovers around 0.41 (between cash and quick ratio). Therefore, the estimation model is freed from any multicollinearity issue which can perturb the regression results. The positive correlation between size and the decision to hedge is attributed to considerable economies of scale and hedging transaction costs. Thus, indirectly, prior to undertaking any regression analysis, it can be conjectured that size and age will have significant effects on the decision to hedge in Mauritius. While glancing across the hedge dummy variable, it transpires that, among the distinct independent variables considered, size and age are both statistically significant. Under alternative forms of hedging, the use of foreign debt (Aabo 2006), Geczy et al. (1997) is not testable on ground that only four firms reported the use of such a type of debt in their financial statements. Alternatively stated, financial and operational hedges are intentionally overlooked as they do not play important roles in the Mauritian context. In addition, it does not make sense to employ dividend payout because

13  What Factors Drive Hedging Among Mauritian Firms? 

297

most of the firms in the sample are unlisted firms. The alternative forms of hedging which are tested relates to cash and liquidity.

5.1

Actual Sample Size and Heteroscedasticity

Of the 384 firms which are considered to be affected by exchange rate risk as per their activity level, further analysis on their financial statements reveal that under the financial instrument section, they do not mention exposure to currency risk although credit risk is found to be an overwhelming risk component. Thus, to have a proper sample, it is considered important to remove all those firms which either do not mention or report such exposure on the assumption that foreign exchange exposure is somewhat immaterial to them, having little impact on their cash flows. From this screening process, there are 286 firms, of which 60 firms are on the hedging side (constituting around 20.9% of the whole sample). Due to the outliers found in some of the specific variables, this adversely impacted on the use of all the identified hedgers. Thus, an elimination process is undertaken so that 61 firms (comprising both hedgers and non-hedgers) are being ignored, trailing behind a total of 225 firms, out of which 47 are hedgers, still constituting around 20.9% of the remaining sample. To cater for the problem of heteroscedasticity, robust options are employed for all estimation techniques.

5.1.1  Univariate Analysis The Pearson correlation coefficients are analysed between the dependent and the independent variables (Tables 13.1).

5.1.2  Multivariate Analysis Decision to Hedge Results of the regression analysis are shown in Table 13.2. Of the five statistically significant variables, size is found to unleash a positive effect on

1

0.1007 0.132 1

Net Profit

Source: Own computation

Cash

Growth

Age

Leverage

Size

Tang

Quick

Capital expenditure

Hedge Dum Net profit

Hedge Dum −0.0593 0.3763 0.0061 0.9277 1

Cap

Table 13.1  Pearson correlation coefficients

1

0 −0.1714 0.01 1

0.2325 1

Tang −0.0233 0.7286 −0.0835 0.212 0.3173

Quick 0.0194 0.7722 0.3097 0 −0.0799

Size

0.2299 0.118 0.0773 0.2385 0.0003 1

0.1588 0.0172 0.0478 0.4752 −0.0804

Leverage

0.0014 −0.2073 0.0018 0.369 0 −0.0148 0.8248 1

0.0888 0.1843 −0.2191 0.0009 0.2112

Age

0.0828 0.0371 0.58 0.1799 0.0068 0.2927 0 −0.0266 0.6917

0.1889 0.0045 0.0036 0.957 −0.1159

Growth

0.0007 0.0927 0.1658 −0.0675 0.3134 0.0437 0.5147 0.0405 0.5456 −0.184 0.0056 1

−0.0753 0.2607 0.0729 0.2761 0.2235

Cash

0.7107 0.4089 0 −0.1538 0.021 0.0082 0.9023 −0.1941 0.0035 −0.1836 0.0057 0.0486 0.4682 1

−0.0249

0.076 0.2653 0.3108

298  Economics and Finance in Mauritius

13  What Factors Drive Hedging Among Mauritian Firms? 

299

Table 13.2  Determinants of the decision to hedge under probit and logit Dependent Net profit Capital expenditure Quick ratio Tangibility Size Leverage Age Cash Growth Constant Wald chi2(9) Prob > chi2 Pseudo R2 Log likelihood Number of obs

Binary variable; 1 if hedges, 0 otherwise Probit

Logit

1.5539 1.99* −0.3373 −0.21 −0.1699 −0.92 −1.0652 −2.02* 0.1359 2.09* 2.4647 2.72** 0.4677 3.04** 1.416 1.71 −0.4838 −1.12 −4.7207 −4.08** 27.2 0.0013 0.1126 −102.327 225

2.6904 2.02* −0.4375 −0.15 −0.2844 −0.87 −1.7997 −1.96* 0.2349 2.01* 4.2905 2.57** 0.8041 2.93** 2.29 1.49 −0.8873 −1.11 −8.1199 −3.84** 24.68 0.0033 0.1106 −102.5583 225

Denotes statistical significance at 5% level and 1% level, respectively

*, **

the decision to hedge. Such a result is consistent with previous empirical evidence whereby size is found to account for the chief underpinning reason for firms to hedge. The justification is related to economies of scale and transaction costs in hedging. Such a positive relationship had been uncovered by Mian (1996). In a parallel manner, Tufano (1996) pointed out that managers would have an inclination to overinvest to maximise their utility so that hedgers would tend to be larger firms. The notion is that with hedging sending good signals for managers, the latter are able to consider projects without being subject to scrutiny from the capital markets. But, it is vital to note that, in the Mauritian context, many of these firms are unlisted ones. Thus, a ­feasible explanation to account for the size effect on

300 

Economics and Finance in Mauritius

the decision to hedge relates to the fixed costs embedded in establishing a hedging programme. However, such a reason may seem questionable due to the fact that, in Mauritius, forward contracts are being offered as a set of banking package facilities by banks to their customers. In that respect, the fees and commissions may not be expensive as banks avail of economies of scale in transactions. Factoring in these caveats into mind, the most reasonable explanation pertaining to the size effect is that, due to their higher activity base, larger firms are imbued with greater currency risk and thereby find it profitable to hedge with forward contracts-with the ultimate aim of saving funds from unexpected losses. Concerning age, a positive and statistically significant effect is noted. The positive impact bodes well with the fact that expertise represents a key driver for hedging so that, years after the date of incorporation, the firm is able to cope with any new derivative programme. This implies that there is a tendency for well-established companies to be part constituent of the sample which consists of the hedgers. Among the statistically significant variables, age manifests as the one embodying the highest statistical significance. This can signify that a new firm, although affected by currency risk, may find it somewhat risky to have recourse towards forward contracts. Alternatively stated, this can also mean that only after building up sufficient experience over time a firm can identify areas where losses can be avoided via the use of derivatives such as forward contracts. Previous research has stressed the importance of underinvestment in the sense that those firms which are subject to investment opportunities and constrained by funds are more likely to hedge (see Fazzari et al. (1988), Froot et al. (1993) and Kaplan and Zingales (1997)). In Mauritius, however, there is no scope for research and development as is the case for foreign firms. The best channel that R&D expenditure can be captured is through the purchase of fixed assets. Hence, information is gleaned on capital expenditure under the cash flow section of the financial statements of the firms. Results show that such a variable has no impact on the decision to hedge, independent of the logit or probit model used. Linking such a finding with the net profit result, it can be posited that the hedgers in Mauritius are mostly large firms which have established a firm and reputable business operational status so that they tend to have abundant funds, thereby generating little scope for the underinvestment

13  What Factors Drive Hedging Among Mauritian Firms? 

301

problem to manifest and eventually lesser incentives for them to hedge. Another feasible explanation relates to the investment opportunity in Mauritius whereby due to high upfront investments in certain industries, many firms have already ensured their market share so that there is weak chance for aggressive investment to consistently keep an edge over their competitors. Among the statistically significant variables, net profit emerges as that which has the second-highest economic significance immediately after leverage. This implies that profitable firms are more susceptible to hedge relative to less profitable ones. Such a finding, although not revealed by the prior empirical evidence, nonetheless is found to be particularly strong for Mauritius. Thus, this may show why firms, although affected by foreign exchange risk under a somewhat high exposure, may not have recourse to hedging because they are loss-making. Indeed, this research presents important finding about the determinants of hedging for a developing country. Moreover, it can also be conjectured that larger companies tend to well-established ones and therefore able to reap high net profits. Among the independent variables which are statistically significant, only tangibility posts a negative sign, something which is compatible with the previous empirical finding that firms with higher tangible assets are more likely to borrow from abroad so that they cling to a non-derivatives means of hedging their risks. However, while examining the financial statements of Mauritian firms, only some firms (five to six) acknowledged that they borrowed from abroad. Thus, the negative impact of tangibility is not directly explicable. The most plausible explanation can be that firms which have higher levels of fixed asset can easily absorb losses from currency risk so that they prefer not to hedge with forward contracts. Alternatively, a still better explanation can be that these large firms avail themselves of their monopoly power so that any exchange rate loss is simply shifted to the customers, blatantly defeating the purpose for hedging via derivatives. In fact, it is a common fact in Mauritius to note that car dealers prefer to pass on any exchange rate loss onto their customers via higher selling prices. The quick ratio trails behind a downward effect on the decision to hedge, which is theoretically compatible with the notion that liquidity acts as a non-derivative form of hedging (see Adam (2002)). The negative

302 

Economics and Finance in Mauritius

sign is consistent with the fact that firms which have greater liquidity can be considered as strong currency risk shock absorbers so that there is a lesser incentive for them to hedge. An interesting element of this result is that, firms, even those subject to foreign exchange rate risk, have no recourse to hedging with forwards because their exposure is not high enough as to warrant hedging. Thus, currency risk can easily be managed on an ongoing basis via the maintenance of sufficient liquidity. The quick ratio is not statistically significant, however, implying that liquidity does not really work as an effective alternative to hedging in Mauritius. The growth variable engenders a negative and statistically insignificant effect on the dependent variable. It can thus be argued that the underinvestment problem does not matter so much as to stimulate firms to hedge. A plausible explanation is based on the characteristic of these firms. Being unlisted, these firms try to optimize on resource allocation so that hedging does not act as a signalling tool to outsiders. Higher growth does not require hedging in the case the firm has low currency risk. Another explanation is related to the lack of knowledge and expertise related to the use of derivative instruments. It can be that these firms simply operate on traditional basis without taking account of any new financial tools that they could employ to enshrine their profits. Another plausible explanation relates to the fact that most firms simply transfer the exchange rate loss to their customers. As pointed out by Tufano (1996), the cash variable is also assessed to see whether cash buffers reduce the need for hedging with forwards. But, like the quick ratio, such a variable is found to be statistically insignificant. Thus, there is strong evidence that non-alternative forms of hedging do not work in Mauritius. Another feasible explanation is based on the size of the foreign exchange risk exposure. It may be the case that the hedgers are mainly firms which are heavily influenced by currency risk and forwards provide them with the possibility to scale down losses from the depreciating rupee. Finally, focusing on the leverage variable, it is found that such a variable generates not only the highest statistical but also the highest economic significance of all the independent variables. Higher leverage firms tend to hedge with derivatives to weaken the probability of financial ­distress costs. Thus, Mauritian firms are also subject to the financial distress cost hypothesis rationale for hedging. Nance et al. (1993) clearly stated that

13  What Factors Drive Hedging Among Mauritian Firms? 

303

hedging is propelled by two main factors – the probability that the firm is subject to financial distress costs if it does not hedge and also the corresponding costs of bankruptcy.

6

Conclusion

This study assesses the determinants of hedging in Mauritius. A priori, when analysing hedging, it is important to take into account the features of the local firms  – concentrated ownership structure and the flat rate tax system. Consequently, the estimation model does not provide due consideration to convexity in the tax structure and mangers’ incentives to hedge since these two motives are irrelevant for local firms. In a parallel manner, financial and operational hedges have intentionally been omitted by virtue of their irrelevance in the local context. Therefore, the other hypotheses such as underinvestment, financial distress costs hypothesis and alternative forms of hedging are given full analysis in this research. Results point out that under the binary metric, net profit, tangibility, size, leverage and age constitute important drivers which account for the decision to hedge. Compared with the statement advanced by Nance et al. (1993), evidence is found that profits do not exert a bearish impact on the use of currency derivatives, mainly due to a positive relationship between net profits and the decision to hedge. Moreover, the analysis also demonstrates, that, compatible with the previous empirical evidence, size does matter in the decision of whether or not to hedge. In addition, the fact that older firms are found to adopt higher incentive to hedge, this may be attributable either to a hike in their activity level or the higher level of acquired knowledge over time in the use of derivatives. The research makes the following important contributions. First, it provides evidence pertaining to the decision to hedge in case of an upper-­ income developing country where country-specific factors have to be given due consideration. Second, the age proxy has not been considered in previous empirical evidence. The most plausible explanation is that foreign firms are very dynamic so that age does not constitute a driving motive for hedging. In the case of a developing country, however, years

304 

Economics and Finance in Mauritius

of establishment are important in bolstering the learning curve process so that companies begin to have a sense of maturity in their operations. In addition, business cycles will often take after electoral cycle to the effect that successful establishment generates lesser risk to these problems. Second, this study contributes to the literature on hedging by proposing the use of capital expenditure as the best proxy to reflect R&D since the latter is practically nonexistent among Mauritian firms. But, focusing on the other side of the coin, a major problem of this study refers to the hedging metric applied. Due to data limitations, it is not possible to have a continuous hedging metric which will enable a full-fledged analysis or a general picture of whether hedgers are tilted more towards the long or the short position.: first, each position being assessed separately (the long and the short position, respectively), then the net long position (the long position minus the short position), and, finally, the total position (the long plus the short position). By implementing these diverse sets of analyses, it is expected that more insight can be obtained with respect to the determinants of hedging foreign exchange risk using forward contracts.

7

Policy Recommendations

Indirectly, this study has demonstrated that the degree of financial market development does matter when it comes to the use of derivatives. For example, the absence of operational hedging and convertible debt implies that other forms of hedging may be employed. In that respect, local hedgers make widespread use of forward contracts for a variety of reasons. First, these forwards are related to the rupee value and this may imply that they may be unwilling to use foreign currency futures which are linked to another currency. Second, forward contracts are provided through a whole bunch of banking facilities (letter of credit, overdrafts, loans and others) so that the bank can scale down its commissions. Third, companies often have some negotiating power whenever they contract forward contracts with banks, mainly when substantial amounts of foreign currencies are involved.

13  What Factors Drive Hedging Among Mauritian Firms? 

305

Since size is found to act as the main driver for hedging in Mauritius, this implies that smaller firms, which are subject to currency risk, do not hedge. To promote the use of derivatives among smaller firms, it may be high time that ongoing workshops are being organized to disseminate information about the benefits of derivatives’ use. Above all, to induce greater use of derivatives among these small firms, a derivative contract could be split into lower denominations. Moreover, a standard and well-organized reporting framework should be adopted for all companies which use derivatives. Indeed, to generate better results, it is best to employ the extent of derivatives use. Thus, notional and market values of derivatives amount should be made mandatory in financial statements reporting for the benefit of rigorous research which will, in turn, benefit the firms in terms of enhanced risk management positions, the government in terms of better policies and the country at large, in terms of better risk management and higher economic growth. As an end note, it is vital to bear in mind that hedging is not a panacea and can be devastating when it is not performed in the proper way, as had been the case of Air Mauritius Ltd and State Trading Corporation.

Bibliography Aabo, T. (2006). The importance of foreign debt as an alternative to currency derivatives in actual management of exchange rate exposure. European Financial Management, 12(4), 663–649. Adam, T. R. (2002). Risk management and the credit risk premium. Journal of Banking and Finance, 26, 243–269. Allayannis, G., & Ofek, E. (2001). Exchange rate exposure, hedging, and the use of foreign currency derivatives. Journal of International Money and Finance, 20, 273–296. Allayannis, G., & Weston, J. P. (2001). The use of foreign currency derivatives and firm market value. The Review of Financial Studies, 14, 243–276. Allayannis, G., Lel, U., & Miller, D. (2004). Corporate governance and the hedging premium around the world. Working paper, University of Virginia, Charlottesville.

306 

Economics and Finance in Mauritius

Amihud, Y. (1993). Evidence on exchange rates and equity shares, exchange rates and corporate performance. Irwin: Business One. Bartram, S., Brown, G., & Fehle, F. (2004). International evidence on financial derivatives usage. Working paper, Lancaster University, University of North Carolina at Chapel Hill and University of South Carolina. Berkman, H., Bradhury, M., & Magan, S. (1997). An international comparison of derivatives use. Financial Management, 26(4), 69–73. Borokhovich, K., Brunarski, K., Crutchley, C., & Simkins, B. (2004). Board composition and corporate use of interest rate derivatives. The Journal of Financial Research, 27(2), 199–216. Breeden, D., & Viswanathan, S. (1996). Why do firms hedge? An asymmetric information model. Working paper, Duke University, Durham. Breeden, D. T., & Viswanathan, S. (1998). Why do firms hedge? An asymmetric information model. The Journal of Fixed Income, 25(3), 7–25. Brown, G., Crabb, P., & Haushalter, D. (2003). Are firms successful at selective hedging? Working paper, The University of North Carolina at Chapel Hill and Northwest Nazarene University. Clark, E., & Judge, A. P. (2007). The determinants of foreign currency: Does foreign currency debt induce a bias. European Financial Management, 14(3), 445–469. De Jong, A., Ligterink, J., & Macrae, V. (2002). A firm-specific analysis of the exchange rate exposure of Dutch firms. ERIM Report Series Research Management. DeMarzo, P., & Duffie, D. (1995). Corporate incentives for hedging and hedge accounting. Review of Financial Studies, 95, 743–771. Dionne, G., & Garand, M. (2003). Risk management determinants affecting firm’s values in the gold mining industry: New empirical results. Econometric Letters, 79, 43–52. Dolde, W. (1995). Hedging, leverage, and primitive risk. Journal of Financial Engineering, 4, 187–216. Fazzari, S. M., Hubbard, R. G., & Peterson, B. C. (1988). Financing constraints and corporate investment. Brookings Papers on Economic Activity, 2, 141–206. Flood, E., Jr., & Lessard, D. R. (1986). On the measurement of operating exposure to exchange rates: A conceptual approach. Financial Management, 15, 25–37. Froot, K., Scharfstein, D., & Stein, J. (1993). Risk management: Coordinating investment and financing policies. Journal of Finance, 48, 425–458. Gay, D., & Nam, J. (1998). The underinvestment problem and corporate derivatives use. Financial Management, 27(4), 53–69.

13  What Factors Drive Hedging Among Mauritian Firms? 

307

Geczy, C., Minton, B. A., & Schrand, C. (1997). Why firms use currency derivatives. Journal of Finance, 52, 1323–1354. Graham, J.  R., & Rogers, D.  A. (2002). Do firms hedge in response to tax incentives? The Journal of Finance, LVII(2), 815–839. Hankins, K. W. (2007). How do firms manage risk? The interaction of financial and operational hedging. SSRN working paper. Haushalter, G. D. (2000). Financing policy, basis risk, and corporate hedging: Evidence from oil and gas producers. The Journal of Finance, 55(1), 107–152. Jorion, P. (1990). The exchange rate exposure of U.S. multinationals. Journal of Business, 63, 331–345. Joseph, N.  L., & Hewins, R.  D. (1997). The motives for corporate hedging among UK multinationals. International Journal of Finance and Economics, 2, 151–171. Kaplan, S. N., & Zingales, L. (1997). Do Investment-cash flow sensitivities provide useful measures of financing constraints? The Quarterly Journal of Economics, 112(1), 169–215. Kim, Y., Mathur, I., & Nam, J. (2006). Is operational hedging a substitute for or a complement to financial hedging? Journal of Corporate Finance, 12, 834–853. Lel, U. (2004). Currency risk management, corporate governance, and financial market development. Working paper, University of Indiana. Lel, U. (2006). Currency hedging and corporate governance: A cross-country analysis (International Finance discussion paper 858). Washington, DC: Board of Governors of the Federal Reserve System. Levi, M. D. (1994). Exchange rates and the valuation of firms. In Y. Amihud & R. Levich (Eds.), Exchange rates and the valuation of equity shares. New York: Irwin Publishing. Lim, S., & Wang, H.  C. (2001). Firm risk management policies: Financial hedging and corporate diversification. Academy of Management Proceeding, 20(1), 1–6. Lin, Chen-Miao., & Smith, S. (2003). Hedging, financing and investment decision: A theory and empirical test. Woking paper, Georgia State University and Federal Reserve Bank of Atlanta. Marston, R. (2001). The effects of industry structure on economic exposure. Journal of International Money and Finance, 20, 149–164. Mian, S. L. (1996). Evidence on corporate hedging policy. Journal of Financial and Quantitative Analysis, 31(3), 419–439. Modigliani, F., & Miller, M. H. (1958). The cost of capital, corporate finance and the theory of investment. American Economic Review, 32, 261–297.

308 

Economics and Finance in Mauritius

Nance, D. R., Smith, C. W., & Smithson, C. (1993). On the determinants of corporate hedging. The Journal of Finance, XLVIII(1), 267–284. Nguyen, H., & Faff, R. (2003). Can the use of foreign currency derivatives explain variations in foreign exchange exposure?: Evidence from Australian companies. Journal of Multinational Financial Management, 13(3), 193–215. Petersen, M. A., & Thiagarajan, S. R. (1997). Risk measurement and hedging, Working paper, North-Western University, Evanston. Ramlall, I. (2009). Determinants of hedging: An empirical investigation for Mauritius. The ICFAI University Journal of Financial Risk Management, 6(3/4), 99–120. ISSN 0972-916X, ZDB-ID 24871539. Reeve, T. (1994). Gold Hedge Survey. Burns Fry Limited. Smith, C. W., & Stulz, R. M. (1985). The determinants of firms’ hedging policies. Journal of Financial and Quantitative Analysis, 20(4), 391–405. Tufano, P. (1996). Who manages risk? An empirical examination of risk management practices in the Gold Mining Industry. The Journal of Finance, LI(4), 1097–1137. Whidbee, D.  A., & Wohar, M. (1999). Derivative activities and managerial incentives in the banking industry. Journal of Corporate Finance, 5, 251–276. Zingales, L. (1995). What determines the value of corporate votes? Quarterly Journal of Economics, 110, 1047–1073.

Part VII Financial Stability Risk Assessment in Mauritius

14 Developing a Financial Stability Model for Mauritius

The objective of this chapter is to develop a financial stability model for Mauritius. Indeed, in light of the crisis, there had been an increasing need to generate financial stability models to gauge the different risks likely to impact on both the economic and financial fronts of an economy. This chapter develops a broad financial system strength index model for Mauritius. The results show that Mauritius had been affected the crisis with the costs of undermined lost output standing in the range of 3.4–5.4%. Latent risks are noted under different channels, namely, public debt sustainability, tourist arrivals and earnings, central bank equity, quality of balance of payments sustainability, trade finance, net foreign investments on the Stock Exchange of Mauritius and future GDP growth paths in Europe and the USA. Findings further show the existence of an ineffective interest rate channel, a robust credit transmission channel and a vibrant exchange rate channel under the Monetary Condition Index. Interestingly, Mauritian banks’ profitability structure is found to have been unaffected by the crisis on the back of a maintained interest rate spread at 7% in spite of a fall in the TED spread. Policy-wise, the authorities should focus on the tourism and credit channels while scaling down the exorbitant interest rate spread to boost the effectiveness of monetary © The Author(s) 2017 I. Ramlall, Economics and Finance in Mauritius, DOI 10.1007/978-3-319-39435-0_14

311

312 

Economics and Finance in Mauritius

policy. In fact, the high interest rate spread may be acting like a hurdle in insulating the interest rate channel of monetary policy transmission mechanism in Mauritius.

1

Introduction

The aim of this chapter is to assess the effect of the US crisis on the Mauritian economy by having recourse to a financial stability tool known as the Financial System Stress Index (FSSI). The latter has been employed worldwide to identify individual risks. To the author’s best knowledge, the current chapter offers the first financial stability model for Mauritius. Alternatively stated, a financial stability model for Mauritius is developed based under 11 core components, including the central bank balance sheet, the banking sector, exchange rates, the tourism sector, the payments system, the balance of payments, government debt, the money market, monetary policy conditions, external shocks and the stock market, which reflect all the intricate features of the Mauritian economy. As a matter of fact, the philosophy under FSSI is to assess the various characteristics of an economy into a single component, usually labelled as the financial stress indicator. Financial crises occur when the financial stress indicator reaches extreme values so that an inverse relationship prevails between the financial stress indicator and the level of economic activity. Many institutions worldwide cling to their specific stress indicators. For instance, the Swiss National Bank publishes a composite stress indicator in its financial stability reports. Other institutions which adhere to financial stress indicators include the European Central Bank, the Organization for Economic Cooperation and Development and the Bank for International Settlement. The current analysis contributes to the empirical literature on FSSI on several fronts. First, it provides evidence for a developing country’s FSSI chiefly where data availability tends to be a major hitch such as the absence of a sound yield curve which is endorsed by a vibrant bond market, credit default swaps and corporate bond spreads. In the same vein, certain techniques which are applicable to developed countries may not be helpful for developing economies. For instance, principal component analysis is impotent in the case of developing countries since many

14  Developing a Financial Stability Model for Mauritius 

313

­ nancial variables are subject to lethargic states, not able to move strongly fi in the short-run. The analysis proposes a better technique, namely, VECM to properly gauge on the relationships among the variables under examination. Indeed, this is the second contribution of the analysis as it shows the way out to develop a FSSI by using long-run VECM coefficients as weights. Third, a dynamic public debt sustainability metric is being developed, based on the cyclical components of seasonally adjusted GDP, domestic and external debts. The public debt sustainability metric is captured as the cyclical component of GDP minus the cyclical component of domestic debt minus the cyclical component of external debt. Fourth, recourse is made towards a large set of financial variables to disentangle the best financial stress indicators, best reflective of the features of the Mauritian economy. Finally and most importantly perhaps, the study extends the usefulness of FSSI not only as a risk identification tool but also as a risk measurement tool in terms of furnishing an estimate of the costs of the crisis to the Mauritian economy.

2

Literature Review

The pioneering work in the area of FSSI emanated from the work of Illing and Liu (2006), who applied principal component analysis to disentangle a general financial stress indicator for Canada. Davig and Hakkio (2009) and Kliesen and Smith (2010) employed the same technique for computing the Kansas City Financial Stress Index (KCFSI) and the St. Louis Fed Financial Stress Index (STLFSI) Indices for the USA.  FSSI posits that the knowledge of certain variables can be useful in forecasting any negative events likely to impact upon the economic performance of an economy. Such knowledge is now ingrained in a famous financial stability tool known as Financial System Stress Index. The empirical literature points out no single definition for a financial stress, meaning that many definitions exist. Illing and Liu (2003) characterized financial stress as a component which triggers uncertainty and higher expected financial loss. Hakkio and Keeton (2009) defined financial stress as a distortionary force which impede on the sound functioning of financial markets. Balakrishnan et al. (2009) defined financial stress as a state when banks failed to fulfil their financial intermediation function.

314 

Economics and Finance in Mauritius

Grimaldi (2010), by contrast, defined financial stress as a situation when negative shocks transmit quicker than positive shocks in the financial system. Sandahl et al. (2011) defined financial stress as a disruptive episode that adversely impacts on the financial markets’ ability to act as an efficient intermediary between the borrowers and lenders. For different countries in the world, distinct FSSI were developed. As a matter of fact, different countries will have distinct features which require an entirely different modelling structure for the FSSI.  FSSI is often constructed from a set of main components. Kota and Saqe (2013), for example, built a FSSI for the Albanian economy by incorporating four key components, namely, the banking sector, the money market, the foreign exchange market and the housing sector. In the case of Hungary, Hollo (2012) constructed a financial stress indicator based on six components – namely, the spot foreign exchange market, the foreign exchange swap market, the government bond market, the interbank market, the equity market and the banking sector. Hollo (2012) noted that the banking sector and the foreign exchange swap market were strongly related to the financial stress index. Hollo (2012) went on to argue that the degree of financial damage borne by a financial system was dependent on four chief forces: the size of the shock, the extent of tensions built up, the reaction of policy makers to the shock and the expectations of economic agents. Sandahl et al. (2011) built a financial stress index for the Swedish economy using equally weighted stress indicators to generate an overall view of the level of financial stress in the markets. The relationship between a financial stress indicator and the level of economic activity has been widely explored. Davig and Hakkio (2010) found that the USA underwent a normal regime in the case of low financial stress and high economic activity. A distressed state manifested by virtue of high levels of financial stress and subdued economic activity. In the case of the euro area, Hollo et  al. (2012) found that the effect of a financial stress on economic activity was weak in low-stress periods while in the case of high-stress periods, there was strong disruption in economic activities. They employed a composite indicator of systemic stress, labelled as CISS. Bloom (2009) employed the VAR technique to gauge the link between the volatility index (VIX) of the S&P 500 and industrial production in the USA to uncover the considerable impact of stock market volatility on industrial production.

14  Developing a Financial Stability Model for Mauritius 

315

With time, increasing sophistication was brought to the FSSI to cater for various challenges, mostly linked to the dire need of providing due consideration to the specific features of the country under scrutiny. The most basic or standard FSSI models resorted to the variance–equal weight approach by averaging the various standardized variables as undertaken in studies by Bordo et al. (2001) and Cardarelli et al. (2009). In addition, Nelson and Perli (2007) and Grimaldi (2010) employed logit models to build a probability-based stress index. Recently, Hollo et al. (2012) included portfolio theory in the construction of FSSI. A state space representation of the FSSI problem was undertaken by Brave and Butters (2010).

3

Data and Methodology

To reflect the core features of the Mauritian economy, recourse was made towards a plethora of important variables as depicted in Table  14.1. As a matter of fact, the development of any FSSI is more of an art than a science as it required a refined incorporation of key variables. Since Mauritius is still classified as a developing country (an upper-income developing country), the link between financial and economic variables may not be as vibrant as that in advanced economies. For instance, the local stock market does not really act as an economic barometer in terms of the health of the economy on the back of lack of reactive forces on the SEMDEX when the key interest rates alter. Hence, each variable is initially analysed with respect to GDP to assess on its eligibility to enter the FSSI based on cointegration analysis as studied by Bloom (2009). Central bank balance sheets, the banking sector, exchange rates, the tourism sector, the payments system, the balance of payments, government debt, money markets, monetary policy conditions, external shocks, and the stock market, are all given due consideration, all geared to contribute towards a fully-fledged FSSI for Mauritius. Compared with Kota and Saqe (2013), the analysis does not include the housing sector because the latter has consistently been unleashing above-normal yearly returns, let alone no data availability. Similarly, contrary to the study of Oet et al. (2011), no recourse is made towards time-varying weights because the structure of the Mauritian economy has altered little over the years. This is bolstered by many facts such as the

316 

Economics and Finance in Mauritius

Table 14.1  Components of the Mauritian financial system stress index Components

Metrics

Central bank balance sheet

BoM Z-score BoM capital ratio MCB bad debts MCB Z-score MCB deposit gap MCB loan gap MCB stock beta MCB loans to deposits ratio Domestic credit of the banking sector GARCH (1,1) on USD, EUR and GBP Exchange market pressure index Tourist arrivals Tourist earnings Value of cheques cleared MACSS Current account deficits Capital account inflows Total government debt/GDP Domestic debt gap External debt gap Public debt sustainability metric Interbank interest rate Transactions amount on interbank market Bid-cover ratio TED spread Monetary condition index Currency in circulation Broad money multiplier Excess cash holdings Net international reserves Import cover Oil price Trade finance SEMDEX stress Net foreign investment stress index

Banking sector

Exchange rate Tourism sector Payment system Balance of payments Government debt

Money market

Monetary policy conditions

External shocks

Stock market

number of listed firms being or more less the same in size and number over the last ten years, the banking sector has always been the main pillar of the financial system and banks availing itself of the cheap funding base from conventional households who shun sophisticated investment products resulting from the weak level of financial literacy.

14  Developing a Financial Stability Model for Mauritius 

317

The portfolio-theoretic approach used by Hollo et  al. (2012) is not applicable in Mauritius because of several deficiencies in the country. First, the stock exchange of Mauritius fails to act as the economic barometer for Mauritius, at least in the short run. Second, the Mauritian economy is featured by an oligopolistic banking system with exorbitantly high interest rate spreads. Third, wealth holding is mainly achieved in the form of cash or deposits because only high net worth individuals resort to equity ownership. Fourth, the interest rate channel of monetary transmission mechanism is impotent as there is no real link between the Key Repo Rate, and deposit and credit levels. We undertake a more comprehensive analysis in terms of the FSSI components relative to those employed in prior studies. Illing and Liu (2006) developed a Canadian stress index by focusing on the following components – namely, the equity market, the bond market, the currency market and banking sector components. Cardarelli et al. (2009) built a financial stress index by using the following components – stock market returns, the volatility of stock returns and foreign exchange, liquidity, sovereign debt spreads, international reserves and the risk and earning capacity of the banking system. In the case of The Netherlands, Van den End (2006) developed a financial stability condition index for the country by using data on interest rates, exchange rates, real estate, the stock market, solvency and volatility state of the stock of the financial institutions. Overall, it becomes apparent that equity and bond markets (representative of the capital market) are employed in many studies. Dynamic weights based on VECM computations were used to derive the FSSI compared with the equal-weights system employed by Cardarelli et al. (2009) and Bordo et al. (2001). The VECM-based weights are deemed to be more realistic as it captures the true relationship of a given variable with respect to the GDP. When calculating the FSSI, the weight attached to a given variable is computed as its VECM weight over the sum of weights for all the variables in the system. The chief benefit attached to dynamic weights is that they avoid bias induced by an equal-weight system whereby important variables are underweighted while less important variables become overweighted. Statistically insignificant variables are deemed impotent and thereby overlooked for FSSI computation. Proper stationarity tests were undertaken prior to the running of any VECM model.

318 

4

Economics and Finance in Mauritius

 nalytical Parts: Discussion of FSSI A Components

This section focuses on each component of the FSSI with the metrics to be used.

4.1

Bank of Mauritius Z-Score and Capital Ratio

Defined as the ratio of the sum of equity and profits over total assets, divided by the standard deviation of net profits over total assets, the central bank z-score is used to capture stressful conditions in the balance sheet of the Bank of Mauritius. Graphical analysis shows that the central bank z-score underwent stressful conditions from October 2006, implying that the financial strength of the Central Bank of Mauritius had undermined. Consequently, this means that the central bank independence of BoM might be affected in terms of the less transparent and credible manner to conduct monetary policy, let alone the public debt of the government since profits made by BoM are eventually transferred to the government. It can be argued that this does not constitute an issue of concern bearing in mind that central banks in Chile and Czechoslovakia are doing well in spite of being subject to negative equity for many years. But, being a small economy with many structural problems, a low equity base can herald poor-quality functional activities. To complement the z-score, recourse is made towards the capital ratio, which is defined as the ratio of equity over currency in circulation (the largest chunk of liabilities for most central banks in the world). Figures 14.1 and 14.2 (Appendix) depict BoM z-score and capital ratio, respectively, in the Appendix section. It is vital to note that an increase in the stress index conveys accentuating stress conditions.

4.2

 anking Sector: Mauritian Commercial Bank Ltd B (MCB)

Banks which benefit from oligopolistic power in Mauritius consist of MCB Ltd. and SBM Ltd. The latter is part-owned by the government

14  Developing a Financial Stability Model for Mauritius 

319

so that any problem is likely to trigger government assistance. MCB Ltd., on the other hand, is a fully private bank. From that perspective, it becomes interesting to focus more on MCB Ltd. Nevertheless, it also becomes important to assess the economic health of SBM Ltd. However, the analysis for SBM Ltd. was not feasible due to lack of data on its website. Based on the availability of data from the website of MCB Ltd., it becomes possible to perform a refined analysis of the bank with respect to the various stress indices. Bad debts, z-score, deposit and loan gaps, stock market beta and loans to deposit ratio stress indices, are shown in Figs. 14.3, 14.4, 14.5, 14.6, and 14.7 (Appendix), respectively. Although the levels of bad debts at MCB are low, nonetheless, it is informative to assess its stress index in light of the crisis. The stress index for MCB bad debts generates a bearish trend for the period spanning from 2002 to March 2010, after which a rather upward trend can be detected, predominantly resulting from loan delinquencies propelled by the contagion effects of the crisis onto the Mauritian economy. Despite the fact bad debts represented a very low level of total assets of MCB, yet some minor risks did manifest. The MCB Z-score stress index lost momentum over the years indicative of robust equity capacity. But, as from September 2012, MCB Z-score stress level underwent an increase. MCB loan gap and deposit gap were derived as the difference between the HP-filtered values and actual values, as implemented by Kota and Saqe (2013) in the case of the Albanian economy. The derived values were standardized thereafter and indexed to generate their respective stress indices. Loan and deposit gap stress indices underwent stressful conditions between June 2002 and September 2007. A conspicuous finding pertains to the widening gap between the loan gap stress index and the deposit gap stress index, which is largely explained by the subdued dynamics in loans propelled by the crisis. Many studies probed into the sensitivity of banks stock to that of the market. An instance includes that of the Thailand financial stress index which was reported in the Central Bank of Thailand 2010 inflation report. Analysis of the MCB stock market beta did not disclose any stressful condition in spite of the prevalence of the crisis though a stagnating state is observable. In addition, from June 2008 onwards, the loan-to-deposit stress level showed a sustained rising trend occasioned by a dwindling

320 

Economics and Finance in Mauritius

funding gap-­lesser deposits funding base to support the loans. In the case of the total credit in the banking, some stress conditions are identified in the period from February 2009 to April 2010, as demonstrated in Fig. 14.8. Subsequently, the credit channel appeared to restore on its positive momentum. We also assess whether MCB loans and deposits are cointegrated.1 As shown in Table 14.A.1, both trace test and max-eigenvalue tests confirm one cointegrating equation at the 12% significance level with the long-­run VECM coefficient hovering at around 1.12%. Probing deeper, the variance decomposition analysis demonstrated that up to 50% of variance in loans was triggered by deposits, clearly accentuating the significance of the local funding base for local banks when it comes to the provision of loans. Under interest rates analysis, no long-run relationship under cointegration was found whether for deposits or for loans. Similarly, no short-run effects were noted. Such a result substantiates the notion of an impotent interest rate transmission mechanism for monetary policy in Mauritius.

4.3

Exchange Market Pressure Index

The Mauritian economy is highly sensitive to changes in the exchange rates. In that respect, we formulate the exchange rate market pressure under Eichengreen et al. (1995) as follows:



EMPIt =

1 1 1 ∆Et − 2 ∆Rt + 2 it 2 σE σR σI

(14.1)

EMPIt Exchange rate market pressure index at time t ΔEt 12-month percentage change in nominal exchange rate at time t, all under indirect quotes 12-month percentage change in international reserves ΔRt Annualized interest rates on bank rate it 2 Variance of the 3 (n) different series σn  Results are not shown due to space constraints but can be made available to the reader upon request. 1

14  Developing a Financial Stability Model for Mauritius 

321

The objective behind the Exchange Market Pressure Index (EMPI) is to depict how pressures in exchange rates are being absorbed into three distinct channels  – namely, the depreciation of the local currency, the dwindling of international reserves and a hike in interest rates to increase the demand for local assets to thereby reduce these exchange rate pressures. Using 80 as the limit value to identify exceptional periods of stress, it transpires that from December 2007 to February 2009, the Mauritian FOREX market experienced the most considerable level of stress conditions. Such a finding consolidates the notion that there is a robust exchange rate operating channel in the Mauritian economy, aligned to its inherent feature as a small open economy. It is widely known in finance to have recourse to the Generalized Autoregressive Conditional Heteroscedasticity (GARCH) models to capture currency risk volatility. Technically speaking, the GARCH model incorporates both the reaction and persistence parameters.

σ 2t = ω + βσ 2t –1 + αε 2t – 1



(14.2)

Where: σ2t

Conditional variance at time t Conditional variance at time t − 1 Shock at time t – 1 ε2t–1 ω > 0, β > 0, α > 0 , β and α