Financial Regulation and Bank Performance (Contributions to Finance and Accounting) 9811635080, 9789811635083

Citation preview

Contributions to Finance and Accounting

Shaofang Li

Financial Regulation and Bank Performance

Contributions to Finance and Accounting

The book series ‘Contributions to Finance and Accounting’ features the latest research from research areas like financial management, investment, capital markets, financial institutions, FinTech and financial innovation, accounting methods and standards, reporting, and corporate governance, among others. Books published in this series are primarily monographs and edited volumes that present new research results, both theoretical and empirical, on a clearly defined topic. All books are published in print and digital formats and disseminated globally.

More information about this series at http://www.springer.com/series/16616

Shaofang Li

Financial Regulation and Bank Performance

Shaofang Li Southeast University Nanjing, China

ISSN 2730-6038 ISSN 2730-6046 (electronic) Contributions to Finance and Accounting ISBN 978-981-16-3508-3 ISBN 978-981-16-3509-0 (eBook) https://doi.org/10.1007/978-981-16-3509-0 Jointly published with Shanghai Jiao Tong University Press The print edition is not for sale in China (Mainland). Customers from China (Mainland) please order the print book from: Shanghai Jiao Tong University Press. © Shanghai Jiao Tong University Press 2021 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 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 publishers, 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 publishers 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. The publishers remain neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Preface

The book evaluates the impact of financial regulatory on bank performance in normal times and during the global financial crisis. Following the 2008–2009 global financial crisis, the wide-ranging international financial regulatory reforms and the reshape of financial regulation in most countries have sought to strength financial stability through both improved and new standards. In 2010, the Basel Committee on Banking Supervision (BCBS) incorporated the revised capital regulatory framework, known as the Basel III accord, to encourage enhanced regulation of bank capital. On December 7, 2017, the Basel Committee on Banking Supervision (BCBS) finalized the reforms outlined in Basel III, informally known as Basel IV, mandating significantly higher minimum capital requirements for banks. The Financial Stability Board’s Key Attributes of Effective Resolution Regimes for Financial Institutions set out “the core elements to allow authorities to resolve institutions in an orderly manner without [resorting to the] taxpayer” (FSB, 2014). The regulatory changes in the last decade have been investigated and well documented by the World Bank database Bank Regulation and Supervision (see Barth, Caprio, and Levine (2013)). However, less is known about the implementation of the regulatory rule. Even though countries share the same regulations, the implementation across countries may differ. Our focus will therefore be on the implementation of financial regulation and how well the supervisors pursue and implement financial regulation. Regulators might allow for regulatory forbearance in which banks would inflate the levels of their capital (see Huizinga and Laeven (2012)) or simply the regulators would postpone the decision to close insolvent banks (see Brown and Dinç (2011)). By using the most recent data in banking industry, Chap. 2 empirically investigates how bank capital and competitive conditions affect bank risk-taking. Using financial data of 7620 banks on 118 countries from 2001 to 2016, we show that banks with a higher Tier 1 ratio and a lower Tier 2 ratio are lower risk-takers. A bank with greater market power in a banking system tends to reduce its risk-taking activities. Our findings also highlight that the negative relationship between Tier 1 ratio and bank risk is more pronounced in more competitive conditions. During the financial crisis, Tier 1 capital acted as a stable funding source and reduced bank risk, but the v

vi

Preface

evidence on Tier 2 capital shows that a higher Tier 2 ratio results in a higher level of risk and increases bank instability. Chapter 3 aims to investigate the impact of both bank reform and competition on stability by analyzing the empirical relationships among banking sector reform, competition, and stability in transition economies. By using unbalanced data on 1121 banks from 22 transition countries between 1998 and 2016, our findings confirm the negative relationship between market power and stability, supporting the view that competition increases bank stability. The results also confirm that bank reform increases banking stability, supporting the view that better financial development leads to a more stable banking system. Both higher activity restrictions and more explicit guidelines for asset diversification increase bank stability, but this positive effect significantly weakens for banks with higher market power. More stringent capital requirements in combination with higher market power increase the risk of bank insolvency. Declaring insolvency power, private monitoring, financial statement transparency, and deposit insurance have only a direct impact on bank stability. Chapter 4 aims to examine the impact of bank regulation and supervision on competitive conditions in the banking sector in emerging economies and examines whether the unique characteristics in emerging markets shed light on the relationship between bank regulation and competition from different perspectives. Using a sample of 1,629 banks in 23 emerging economies between 1996 and 2016, this study employs the Panzar and Rosse (1987) methodology and constructs the competitive variable H-statistic as a measure of competition in the banking system. We also compute the Lerner index (see also Coccorese (2009); Koetter et al. (2012)) and the Boone indicator (2008) as alternative measures of competition. We investigate how competition evolved under different types of bank regulation and supervision. As our sample period covers several banking crises in some emerging economies as well as the recent global financial crisis, we also examine whether the relationship between bank regulation and competition changed during the banking crisis. Considering the different roles played by domestic banks and foreign banks, we also study whether the impact of bank regulation on foreign banks exhibits different patterns. Chapter 5 investigates the impact of bank competition and regulation on bank efficiency in the Asia-Pacific region during 2001–2016. The result reveals that market power is positively related to bank efficiency. We also find that stringent activity restrictions, strong official supervisory power, and low capital requirements are associated with high bank efficiency. Furthermore, market power has a stronger efficiency-increasing effect in a banking system characterized by the activity restrictions, supervisory power, and capital requirements described above. Foreign banks operating under increased activity restrictions in a host country with strong official supervisory power have relatively high efficiency. Nanjing, China

Shaofang Li

Preface

vii

References Barth, J. R., Caprio, G. J., & Levine, R. (2013). Bank regulation and supervision in 180 countries from 1999 to 2011. Journal of Financial Economic Policy, 5(2), 111–219. Brown, C. O. & Dinç, I. S. (2011). Too many to fail? Evidence of regulatory forbearance when the banking sector is weak. The Review of Financial Studies, 24(4), 1378–1405. Coccorese, P. (2009). Market power in local banking monopolies. Journal of Banking & Finance, 33(7), 1196–1210. Huizinga, H. & Laeven L. (2012). Bank valuation and accounting discretion during a financial crisis. Journal of Financial Economics, 106(3), 614–634. Koetter, M., Kolari, J. W., & Spierdijk, L. (2012). Enjoying the quiet life under deregulation? Evidence from adjusted Lerner indices for US banks. Review of Economics and Statistics, 94(2), 462–480. Panzar, J. C. and Rosse, J. N. (1987). Testing for “monopoly” equilibrium. The Journal of Industrial Economics, 35(4), 443–456.

Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Motivation of This Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Research Questions Addressed in This Study . . . . . . . . . . . . . . . . . . . 1.3 Structure and Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 1 2 4 5

2 Quality of Bank Capital, Competition, and Risk-Taking . . . . . . . . . . . . 2.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Data and Variable Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.1 Data Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.2 Variable Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.3 Descriptive Statistics and Correlation . . . . . . . . . . . . . . . . . . . 2.5 Research Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6 Empirical Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6.1 Basic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6.2 Impact of Competition on the Relationship Between Capital Structure and Risk-Taking . . . . . . . . . . . . . . . . . . . . . . 2.6.3 Impact of the Financial Crisis on the Relationship Between Capital Structure and Risk-Taking . . . . . . . . . . . . . . 2.6.4 Robustness Checks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

9 9 9 11 13 13 14 16 19 25 25

3 Banking Sector Reform, Competition, and Bank Stability . . . . . . . . . . 3.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Banking Sector Reform in Transition Countries and Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

51 51 51

29 36 40 45 46

53

ix

x

Contents

3.3.1 Banking Sector Reforms in Transition Countries . . . . . . . . . . 3.3.2 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 Data and Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.1 Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.2 Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.3 Summary Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5 Empirical Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6 Empirical Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6.1 Basic Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6.2 Impact of Bank Regulation on the Relationship Between Competition and Stability . . . . . . . . . . . . . . . . . . . . . 3.6.3 Robustness Checks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

53 54 55 55 56 59 62 67 67

4 The Impact of Bank Regulation and Supervision on Competition . . . 4.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.1 Previous Studies on Competition . . . . . . . . . . . . . . . . . . . . . . . 4.3.2 Bank Regulation and Effects on Bank Competition . . . . . . . 4.4 Estimation Methodology and Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.1 Estimation of Competition Variables . . . . . . . . . . . . . . . . . . . . 4.4.2 Data Sources, Sample Selection, and Variable Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.3 Statistical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.4 Determinants of Bank Competition . . . . . . . . . . . . . . . . . . . . . 4.5 Empirical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5.1 Bank Regulation, Supervision, and the H-Statistic . . . . . . . . 4.5.2 Alternative Measures of Competition . . . . . . . . . . . . . . . . . . . 4.5.3 Impact of the Financial Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5.4 Further Tests: Foreign Banks Versus Domestic Banks . . . . . 4.5.5 Robustness Checks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

91 91 91 93 93 95 96 96

5 Bank Competition, Regulation, and Efficiency . . . . . . . . . . . . . . . . . . . . 5.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Related Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.1 The Impact of Competition on Bank Efficiency . . . . . . . . . . . 5.3.2 Literature Review on Bank Regulation and Bank Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

69 74 85 86

99 102 105 105 105 110 114 115 122 122 130 133 133 133 135 135 135

Contents

5.4 Variables Definition and Research Design . . . . . . . . . . . . . . . . . . . . . . 5.4.1 Bank Efficiency Estimation and Main Variables Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.2 Model Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.3 Data Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.4 Summary Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5 Empirical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.1 Basic Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.2 Impact of Bank Regulation on the Relationship Between Competition and Efficiency . . . . . . . . . . . . . . . . . . . . 5.5.3 Further Test: Foreign Banks Versus Domestic Banks . . . . . . 5.5.4 Robustness Checks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xi

137 137 141 141 145 145 145 152 156 163 170 176

6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 6.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 6.2 Summary of the Main Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179

List of Tables

Table 2.1 Table 2.2 Table 2.3 Table 2.4 Table 2.5 Table 2.6 Table 2.7 Table 2.8 Table 2.9 Table 3.1 Table 3.2 Table 3.3 Table 3.4 Table 3.5 Table 3.6 Table 3.7 Table 4.1 Table 4.2 Table 4.3 Table 4.4

Variable definition and data sources . . . . . . . . . . . . . . . . . . . . . . . Summary statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pearson correlations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Impact of capital structure and competition on bank risk-taking (non-performing loan ratio) . . . . . . . . . . . . . . . . . . . . Impact of capital structure and competition on bank risk-taking (log Zscore) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Impact of competition on the relationship between capital structure and bank risk-taking . . . . . . . . . . . . . . . . . . . . . . . . . . . . Impact of capital structure on bank risk-taking during the financial crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Robustness checks: alternative measures of competition and bank risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Robustness checks: subsamples . . . . . . . . . . . . . . . . . . . . . . . . . . . Variable definition and data sources . . . . . . . . . . . . . . . . . . . . . . . Summary statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bank reform, competition and stability: basic analysis . . . . . . . . Bank reform, competition and stability: alternative measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Relationship between bank competition, reform, regulation and stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The impact of bank reform and regulation on stability: through the channel of competition . . . . . . . . . . . . . . . . . . . . . . . . Impact of bank competition and reform on bank stability: robustness checks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Variable definition and data sources . . . . . . . . . . . . . . . . . . . . . . . Summary statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Impact of bank regulation and supervision on H-statistic . . . . . . Impact of bank regulation and supervision on competition. Alternative measure of competition: Lerner index and Boone indicator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

17 20 21 26 30 33 37 41 43 60 63 68 70 72 75 80 101 103 107

111 xiii

xiv

Table 4.5 Table 4.6 Table 4.7 Table 5.1 Table 5.2 Table 5.3 Table 5.4 Table 5.5 Table 5.6 Table 5.7 Table 5.8 Table 5.9

List of Tables

Bank regulation, supervision, bank crisis, and H-statistic . . . . . . Bank regulation, supervision, foreign ownership, and H-statistic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Impact of bank regulation and supervision on competition: robustness check . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Variables definition and data sources . . . . . . . . . . . . . . . . . . . . . . Summary of profit inefficiency and cost inefficiency according to countries and years . . . . . . . . . . . . . . . . . . . . . . . . . . Summary statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Impact of bank competition and regulation and bank efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Impact of bank regulation on the relationship between competition and efficiency . . . . . . . . . . . . . . . . . . . . . . . Impact of ownership on the relationship between bank competition, regulation and bank efficiency . . . . . . . . . . . . . . . . . Alternative measure of bank competition . . . . . . . . . . . . . . . . . . . Robustness checks based on alternative estimation methods . . . . Robustness checks based on different subsamples . . . . . . . . . . . .

116 119 123 142 146 148 149 153 157 164 167 171

Chapter 1

Introduction

1.1 Motivation of This Study Following the 2008–2009 global financial crisis, the wide-ranging international financial regulatory reforms and the reshape of financial regulation in most countries have sought to strength financial stability through both improved and new standards. In 2010, the Basel Committee on Banking Supervision (BCBS) incorporated the revised capital regulatory framework, known as the Basel III accord, to encourage enhanced regulation of bank capital (Basel Committee on Banking Supervision, 2010). On December 7, 2017, the Basel Committee on Banking Supervision (BCBS) finalized the reforms outlined in Basel III, informally known as Basel IV, mandating significantly higher minimum capital requirements for banks. The Financial Stability Board’s Key Attributes of Effective Resolution Regimes for Financial Institutions set out “the core elements to allow authorities to resolve institutions in an orderly manner without [resorting to the] taxpayer” (Financial Stability Board, 2012). The regulatory changes in the last decade have been investigated and well documented by the World Bank database Bank Regulation and Supervision (see Barth et al., 2013). However, less is known about the implementation of the regulatory rule. Even though countries share the same regulations, the implementation across countries may differ. Our focus will therefore be on the implementation of financial regulation and how well the supervisors pursue and implement financial regulation. Regulators might allow for regulatory forbearance in which banks would inflate the levels of their capital (see Huizinga & Laeven, 2012) or simply the regulators would postpone the decision to close insolvent banks (see Brown & Dinç, 2011). Given these concerns, this study investigates in the impact of the financial regulatory on bank behaviour. In particular, the primary objective of this book is to evaluate the impact of financial regulatory on bank performance in normal times and during the global financial crisis. To that end, this book proceeds from three perspectives: (1)

investigates how bank capital and competitive conditions affect bank risktaking. Particularly, based on the Basel Accords, this study classifies bank capital based on capital quality: Tier 1 capital with higher loss-absorbing

© Shanghai Jiao Tong University Press 2021 S. Li, Financial Regulation and Bank Performance, Contributions to Finance and Accounting, https://doi.org/10.1007/978-981-16-3509-0_1

1

2

(2)

(3)

(4)

1 Introduction

capacity and Tier 2 capital with lower loss-absorbing capacity, and examine the effects of bank capital structure on bank risk-taking; explores the impact of both bank reform and competition on stability by analyzing the empirical relationships among banking sector reform, competition, and stability in transition economies; examines the impact of bank regulation and supervision on competitive conditions in the banking sector in emerging economies and examine whether the unique characteristics in emerging markets shed light on the relationship between bank regulation and competition from different perspectives; tests the impact of competition and regulation on bank efficiency based on international evidence in the Asia–Pacific region.

1.2 Research Questions Addressed in This Study This book is built on four main parts: Chap. 2 focuses on a direct link between quality of bank capital, competition, and banks’ risk-taking. Bank capital plays an important role in the financial system, because it acts as a financial cushion to absorb a firm’s losses and also reduces the incentive for banks to take excessive risks (Jokipii & Milne, 2011), but higher capital requirements would lead to a decrease in lending and would undermine banks’ profitability (Ben Naceur et al., 2018; Košak et al., 2015). Previous analyses also indicate that more competitive conditions among banks result in decreasing monopoly rents, which in turn leads to lower profits and capital adequacy. The latter renders the banks less able to absorb the losses and incentivizes excessive risk-taking (Keeley, 1990). However, les competition among banks may also lead to more risk-taking activities if large banks are deemed “too systemically important to fail” (Kelly et al., 2016). As the first research question, this part investigates how bank capital and competitive conditions affect bank risk-taking. Based on the Basel Accords, we classify bank capital based on capital quality: Tier 1 capital with higher loss-absorbing capacity and Tier 2 capital with lower loss-absorbing capacity. We then examine the effects of bank capital structure on bank risk-taking. We test whether bank risk-taking behaviour varies according to the quality of bank capital, and whether this relationship changes significantly in different competitive conditions and during the financial crisis. Chapter 3 focuses on the relation between bank sector reform, competition and bank stability. The recent global financial crisis between 2008 and 2010 influenced the global economy through banking systems across countries and led to unprecedented consequences. Some studies highlight that deregulation and excessive competition in financial sectors were important factors that drove the financial crisis (Brunnermeier, 2009; Carletti, 2008; Fernández et al., 2013; Llewellyn, 2008). The impact of bank regulation and competition on financial stability has thus attracted more attention from academics and policymakers since then (Acharya & Richardson, 2009; Beck et al., 2010, 2013; Schaeck & Cihák, 2014).

1.2 Research Questions Addressed in This Study

3

The banking systems in transition countries have undergone extensive reforms in the past two decades and this provides us with a unique setting in which to test the relationships among banking sector reform, competitive conditions, and financial stability. In this part, we investigate the impact of both bank reform and competition on stability by analyzing the empirical relationships among banking sector reform, competition, and stability in transition economies. Chapter 4 focuses on the impact of bank regulation and supervision on competition from emerging economies. Over the past two decades, because of the development of information technology, globalization, and deregulation, the banking system has undergone dramatic changes. Emerging economies have experienced considerable economic development and financial reforms, including financial liberalization, bank privatization, widening access to foreign banks, and restructuring national banking systems, which aimed to reshape competitive conditions in the banking sector. These changes decreased profitability in traditional bank activities and led to massive mergers and acquisitions (M&A) among financial institutions in emerging economies. While changes in the financial market are the main driving forces for bank consolidation in developed economies, financial supervisory authorities play an important role in the bank consolidation process in emerging economies (Gelos & Roldós, 2004). The overall differences in economic development, bank consolidation, and regulation between developed economies and emerging economies created distinctive features in competitive conditions in the banking industry in emerging economies. The goal of this study is to examine the impact of bank regulation and supervision on competitive conditions in the banking sector in emerging economies and examine whether the unique characteristics in emerging markets shed light on the relationship between bank regulation and competition from different perspectives. Chapter 5 focuses on the relationship between bank competition, regulation, and bank efficiency in the Asia–Pacific region. Since the 1980s, the global financial market has become increasingly integrated, and both the developed and developing countries have embarked on financial liberalization. In the process of financial liberalization, deregulation, and increased integration, banks have expanded their services and engaged in greater risk-taking activities to improve their productive efficiency (Luo et al., 2016). With increased competition following financial liberalization and deregulation, banks have to improve their efficiency (Andrie¸s & C˘apraru, 2014; Schaeck & Cihák, 2008). Since the 1990s, governments across the region have implemented financial reforms, including financial liberalization, bank privatization, widening access to foreign banks, and restructuring of the national banking systems, to enhance the competitive strength of the banking sector. Prudential norms on non-performing loans and capital requirements are also implemented to strengthen financial stability (Zhao et al., 2010). The banking sector in Asia–Pacific region underwent a restructuring process after the 1997–1998 Asian financial crisis and the recent global financial crisis (Noman et al., 2018; Williams & Nguyen, 2005). In this part, we examine how bank competition and regulation affect bank efficiency, using a sample of 1261 banks across 28 Asia Pacific countries for the period 2001–2016.

4

1 Introduction

1.3 Structure and Contents Apart from the Introduction and Conclusion, this book is organized in four chapters. Chap. 2 empirically investigates how bank capital and competitive conditions affect bank risk-taking. Particularly, based on the Basel Accords, we classify bank capital based on capital quality: Tier 1 capital with higher loss-absorbing capacity and Tier 2 capital with lower loss-absorbing capacity. We then examine the effects of bank capital structure on bank risk-taking. We test whether bank risk-taking behaviour varies according to the quality of bank capital, and whether this relationship changes significantly in different competitive conditions and during the financial crisis. Using financial data of 7620 banks on 118 countries from 2001 to 2016, the results show that banks with a higher Tier 1 ratio and a lower Tier 2 ratio are lower risk-takers. A bank with greater market power in a banking system tends to reduce its risk-taking activities. The main findings also highlight that the negative relationship between Tier 1 ratio and bank risk are more pronounced in more competitive conditions. During the financial crisis, Tier 1 capital acted as a stable funding source and reduced bank risk, but the evidence on Tier 2 capital shows that a higher Tier 2 ratio results in a higher level of risk and increases bank instability. Chapter 3 provides empirical analysis of the relation between banking sector reform, competition on bank stability in transition countries. The banking systems in transition countries have undergone extensive reforms in the past two decades and this provides us with a unique setting in which to test the relationships among banking sector reform, competitive conditions, and financial stability. This study investigates the impact of both bank reform and competition on stability by analyzing the empirical relationships among banking sector reform, competition, and stability in transition economies. By using unbalanced data on 1121 banks from 22 transition countries between 1998 and 2016, our findings confirm the negative relationship between market power and stability, supporting the view that competition increases bank stability. The results also confirm that bank reform increases banking stability, supporting the view that better financial development leads to a more stable banking system. Both higher activity restrictions and more explicit guidelines for asset diversification increase bank stability, but this positive effect significantly weakens for banks with higher market power. More stringent capital requirements in combination with higher market power increase the risk of bank insolvency. Declaring insolvency power, private monitoring, financial statement transparency, and deposit insurance have only a direct impact on bank stability. Chapter 4 focuses on the impact of bank regulation and supervision on competition in emerging economies. Using a sample of 1629 banks in 23 emerging economies between 1996 and 2016, this study employs the Panzar and Rosse (1987) methodology and constructs the competitive variable H-statistic as a measure of competition in the banking system. We also compute the Lerner index (see also Coccorese (2009), Koetter et al. (2012)) and the Boone indicator (2008) as alternative measures of competition. We investigate how competition evolved under different types of

1.3 Structure and Contents

5

bank regulation and supervision. As our sample period covers several banking crises in some emerging economies as well as the recent global financial crisis, we also examine whether the relationship between bank regulation and competition changed during the banking crisis. Considering the different roles played by domestic banks and foreign banks, we also study whether the impact of bank regulation on foreign banks exhibits different patterns. Our analysis shows that banking systems with higher concentration and fewer activity restrictions and entry barriers are more competitive. We also find that reducing foreign bank limitations and increasing capital strictness and official supervisory power also enhances competition in the banking sector. The results also provide evidence that competition in banking systems with fewer government-owned banks and fewer diversification guidelines tends to be more intensive. The results also confirm that higher private monitoring of banks and deposit insurance coverage significantly contribute to an increase in bank competition. Chapter 5 explores the relationship between bank competition, regulation, and efficiency in Asia–Pacific region. This part examines how bank competition and regulation affect bank efficiency, using a sample of 1261 banks across 28 Asia Pacific countries for the period 2001–2016. We employ a stochastic frontier analysis (SFA) model to estimate bank efficiency. Our findings confirm that market power is positively related to bank efficiency. Increased activity restrictions, strong official supervisory power, and low capital requirements are associated with high bank efficiency. Furthermore, market power has a stronger efficiency-increasing effect in a banking system characterized by the activity restrictions, supervisory power, and capital requirements described above. Foreign banks operating under strict activity restrictions in a host country with strong official supervisory power are highly efficient. The rest of this book proceeds as follows: Chap. 2 addresses the relationship between the quality of bank capital, competition and banks’ risk taking, Chap. 3 analyses the impact of bank sector reform and competition on bank stability in transition countries, Chap. 4 investigates the relationship between bank regulation and supervision on competition in emerging economies, Chap. 5 investigates the impact of bank competition and regulation on bank efficiency. Chapter 6 concludes the analysis.

References Acharya, V. V., & Richardson, M. (2009). Causes of the financial crisis. Critical Review, 21(2–3), 195–210. Andrie¸s, A. M., & C˘apraru, B. (2014). The nexus between competition and efficiency: The European banking industries experience. International Business Review, 23(3), 566–579. Barth, J. R., Caprio, G. J., & Levine, R. (2013). Bank regulation and supervision in 180 countries from 1999 to 2011. Journal of Financial Economic Policy, 5(2), 111–219. Basel Committee on Banking Supervision. (2010). Basel III: A global regulatory framework for more resilient banks and banking systems. http://www.bis.org/press/p100912.htm

6

1 Introduction

Beck, T., De Jonghe, O., & Schepens, G. (2013). Bank competition and stability: Cross-country heterogeneity. Journal of Financial Intermediation, 22(2), 218–244. Beck, T., Levine, R., & Levkov, A. (2010). Big bad banks? The winners and losers from bank deregulation in the United States. The Journal of Finance, 65(5), 1637–1667. Ben Naceur, S., Marton, K., & Roulet, C. (2018). Basel III and bank-lending: Evidence from the United States and Europe. Journal of Financial Stability, 39, 1–27. Boone, J. (2008). A new way to measure competition. The Economic Journal, 118(531), 1245–1261. Brown, C. O., & Dinç, I. S. (2011). Too many to fail? Evidence of regulatory forbearance when the banking sector is weak. The Review of Financial Studies, 24(4), 1378–1405. Brunnermeier, M. K. (2009). Deciphering the liquidity and credit crunch 2007–2008. Journal of Economic Perspectives, 23(1), 77–100. Carletti, E. (2008). Competition and regulation in banking. Handbook of Financial Intermediation and Banking, 126(5), 449–482. Coccorese, P. (2009). Market power in local banking monopolies. Journal of Banking & Finance, 33(7), 1196–1210. Fernández, A. I., González, F., & Suárez, N. (2013). The real effect of banking crises: Finance or asset allocation effects? Some international evidence. Journal of Banking & Finance, 37(7), 2419–2433. Financial Stability Board. (2012). Progress Report to the G20 on Strengthening the Oversight and Regulation of Shadow Banking. Gelos, R. G., & Roldós, J. (2004). Consolidation and market structure in emerging market banking systems. Emerging Markets Review, 5(1), 39–59. Huizinga, H., & Laeven, L. (2012). Bank valuation and accounting discretion during a financial crisis. Journal of Financial Economics, 106(3), 614–634. Jokipii, T., & Milne, A. (2011). Bank capital buffer and risk adjustment decisions. Journal of Financial Stability, 7(3), 165–178. Keeley, M. C. (1990). Deposit insurance, risk, and market power in banking. The American Economic Review, 80(5), 1183–1200. Kelly, B., Lustig, H., & Van Nieuwerburgh, S. (2016). Too-systemic-to-fail: What option markets imply about sector-wide government guarantees. American Economic Review, 106(6), 1278– 1319. Koetter, M., Kolari, J. W., & Spierdijk, L. (2012). Enjoying the quiet life under deregulation? Evidence from adjusted Lerner indices for US banks. Review of Economics and Statistics, 94(2), 462–480. Košak, M., Li, S., Lonˇcarski, I., & Marinˇc, M. (2015). Quality of bank capital and bank lending behavior during the global financial crisis. International Review of Financial Analysis, 37, 168– 183. Llewellyn, D. T. (2008). The Northern Rock crisis: A multi-dimensional problem waiting to happen. Journal of Financial Regulation and Compliance, 16(1), 35–58. Luo, Y., Tanna, S., & De Vita, G. (2016). Financial openness, risk and bank efficiency: Cross-country evidence. Journal of Financial Stability, 24, 132–148. Noman, A. H. M., Gee, C. S., & Isa, C. R. (2018). Does bank regulation matter on the relationship between competition and financial stability? Evidence from Southeast Asian countries. PacificBasin Finance Journal, 48, 144–161. Panzar, J. C., & Rosse, J. N. (1987). Testing for “monopoly” equilibrium. The Journal of Industrial Economics, 35(4), 443–456. Schaeck, K., & Cihák, M. (2008). How does competition affect efficiency and soundness in banking? New empirical evidence. European Central Bank Working Paper Series No. 932.

References

7

Schaeck, K., & Cihák, M. (2014). Competition, efficiency, and stability in banking. Financial Management, 43(1), 215–241. Williams, J., & Nguyen, N. (2005). Financial liberalisation, crisis, and restructuring: A comparative study of bank performance and bank governance in South East Asia. Journal of Banking & Finance, 29(8), 2119–2154. Zhao, T., Casu, B., & Ferrari, A. (2010). The impact of regulatory reforms on cost structure, ownership and competition in Indian banking. Journal of Banking & Finance, 34(1), 246–254.

Chapter 2

Quality of Bank Capital, Competition, and Risk-Taking

2.1 Overview This chapter empirically investigates how bank capital and competitive conditions affect bank risk-taking. Using financial data of 7620 banks on 118 countries from 2001 to 2016, we show that banks with a higher Tier 1 ratio and a lower Tier 2 ratio are lower risk-takers. A bank with greater market power in a banking system tends to reduce its risk-taking activities. Our findings also highlight that the negative relationship between Tier 1 ratio and bank risk are more pronounced in more competitive conditions. During the financial crisis, Tier 1 capital acted as a stable funding source and reduced bank risk, but the evidence on Tier 2 capital shows that a higher Tier 2 ratio results in a higher level of risk and increases bank instability.

2.2 Introduction The recent global financial crisis (GFC) of 2007–2009 was transmitted through the banking system across different countries and led to unprecedented consequences for the world economy. A series of studies indicate that deregulation and excessive competition in the banking system are the main factors that led to the GFC (Brunnermeier, 2009; Carletti, 2008; Fernández et al., 2013; Llewellyn, 2008). In 2010, the Basel Committee on Banking Supervision (BCBS) incorporated the revised capital regulatory framework, known as the Basel III Accord, to urge for enhanced bank capital regulation. On December 7, 2017, BCBS published a document finalizing

This chapter was published in Emerging Markets Finance and Trade. © Shanghai Jiao Tong University Press 2021 S. Li, Financial Regulation and Bank Performance, Contributions to Finance and Accounting, https://doi.org/10.1007/978-981-16-3509-0_2

9

10

2 Quality of Bank Capital, Competition, and Risk-Taking

the Basel III reforms, also known informally as Basel IV, and banks face significantly higher minimum capital requirements as a result of the new Basel Committee standards.1 Bank capital plays an important role in the financial system, because it acts as a financial cushion to absorb a firm’s losses and also reduces the incentive for banks to take excessive risks (Jokipii & Milne, 2011), but higher capital requirements would lead to a decrease in lending and would undermine banks’ profitability (Ben Naceur et al., 2018; Košak et al., 2015). Previous analyses also indicate that more competitive conditions among banks result in decreasing monopoly rents, which in turn leads to lower profits and capital adequacy. The latter renders the banks less able to absorb the losses and incentivizes excessive risk-taking (Keeley, 1990). However, les competition among banks may also lead to more risk-taking activities if large banks are deemed “too systemically important to fail” (Kelly et al., 2016). This study empirically investigates how bank capital and competitive conditions affect bank risk-taking. Particularly, based on the Basel Accords, we classify bank capital based on capital quality: Tier 1 capital with higher loss-absorbing capacity and Tier 2 capital with lower loss-absorbing capacity. We then examine the effects of bank capital structure on bank risk-taking. We test whether bank risk-taking behaviour varies according to the quality of bank capital, and whether this relationship changes significantly in different competitive conditions and during the financial crisis. Using the unbalanced financial information of 7620 banks from 118 countries between 2001 and 2016, we find that banks with a higher Tier 1 ratio and a lower Tier 2 ratio exhibit lower risk-taking. Further, a bank that has higher market power in a banking system tends to reduce its risk-taking activities. The results also confirm that banks with higher profitability, more funding, and higher quality of bank assets are exposed to lower risk. During a financial crisis, banks exhibit a higher level of risk-taking than in normal times. Our findings also highlight that the negative relationship between the Tier 1 ratio and bank risk are more pronounced in more competitive conditions. During the financial crisis, Tier 1 capital acted as a stable funding source and reduced bank risk, but the evidence on Tier 2 capital shows that a higher Tier 2 ratio results in a higher level of risk and increases bank instability. Our analysis extends the results of previous studies in three ways. First, although there are some studies focusing on the effect of capital requirements on risk-taking in the banking system, we focus on the capital structure and distinguish between Tier 1 and Tier 2 capital, thus considering the quality of bank capital, and examine the different effects of Tier 1 capital and Tier 2 capital on bank risk-taking. Second, we also consider the impact of competition on the relationship between bank capital and risk-taking while controlling for bank-specific characteristics and macroeconomic factors that may impact bank stability. Both the theoretical and empirical analyses show that competition affects both bank capital holding (Allen et al., 2011; Schaeck & Cihák, 2012; Chen, 2016) and bank risk-taking (Agoraki et al., 2011; Jiménez 1 KPMG, Impact of Basel 4 on EU banks, 2018. https://home.kpmg.com/xx/en/home/insights/2018/ 10/impact-of-basel-4-on-eu-banks-fs.html.

2.2 Introduction

11

et al., 2013). This indicates that the effects of Tier 1 and Tier 2 capital on bank risk-taking may change under different competitive conditions. Finally, our analysis extends the results of previous studies as we also evaluate the relationship between Tier 1 capital and Tier 2 capital and bank risk-taking during the financial crisis, which could provide further empirical evidence for policy makers regarding bank capital regulation, especially during the financial crisis. The rest of the analysis is structured as follows: Sect. 2.3 summarizes the relevant literature. Section 2.3 mainly addresses the data collection, variable definition, and data statistics. Section 2.4 presents the research design. Section 2.5 presents the empirical analysis and robustness checks, and Sect. 2.6 concludes the paper.

2.3 Literature Review Previous literature that studies the relationship between bank capital and bank risk-taking shows different findings. Koehn and Santomero (1980) and Kim and Santomero (1988) show that higher capital requirement incentives banks to choose riskier assets (see also Altunbas et al., 2007; Shrieves & Dahl, 1992). However, Jokipii and Milne (2011) indicate that bank capital is used to absorb banks’ losses and insulate banks from potential insolvency, and holding additional capital is to protect banks against large losses during a cyclical downturn and reduce the risk of insolvency (see also Rajan, 1994; Ayuso et al., 2004; Jokipii & Milne, 2008). Bank capital also acts as an incentive device, which encourages banks to behave prudently and to reduce risk-taking (Ashraf et al., 2016; Bui et al., 2017), but because of the existence of deposit insurance schemes and implicit government bailout guarantees, bank managers and shareholders tend to expand their lending and increase the potential risk (Goodhart, 2013). In the previous studies, only few studies consider the difference between the quality of Tier 1 and Tier 2 capital. Demirgüç-Kunt et al. (2013) find that Tier 1 capital has a stronger and more positive effect on stock returns than Tier 2 capital. Barrell et al. (2011) suggest that an increase in the overall capital adequacy ratio reduces banks’ risk appetite in both ex-ante and ex-post terms and that increasing Tier 2 capital based on a given capital adequacy structure increases banks’ risk appetite (see also Ashcraft, 2008). Košak et al. (2015) analyze the determinants of bank lending behavior during the GFC and find that Tier 1 capital acted as a stable source of funding during that crisis but that, although Tier 2 capital spurs bank lending during normal times, it did not do so during the crisis (Koivu, 2012). Since the GFC, regulators have already acknowledged the need to readjust and recalibrate their regulatory measures. The intention of the Basel III Accord is to significantly increase the role of Tier 1 bank capital relative to Tier 2 bank capital (Basel Committee on Banking Supervision, 2010). Similarly, the EBA, the institution charged with setting harmonized supervisory standards for banks in EU Member States, announced that major European banking groups would have to increase their Core Tier 1 capital ratios to 9% of their risk-weighted assets by June 2012 (EBA,

12

2 Quality of Bank Capital, Competition, and Risk-Taking

2011). Bank capital regulatory measures are very likely to affect the credit activity of banks (Gropp et al., 2018; Mésonnier & Monks, 2015). As regards the impact of competition on bank risking-taking, previous analysis indicates that competitive banking systems are more fragile, as competition facilitates banks’ undertaking of more risk-taking activities (Keeley, 1990). In a competitive banking system, banks benefit less from a lending relationship and this discourages their motivation to monitor borrowers’ risk properly, thus introducing higher risk (Allen & Gale, 2000, 2004). Boyd et al. (2004) also suggest that large banks in concentrated banking sectors would increase profits and decrease financial instability by providing more “capital buffers” that absorb external macroeconomic and liquidity shocks. Caminal and Matutes (2002) argue that less competition leads to reduced credit rationing and higher loans, increasing the probability of bank insolvency. However, Berger et al. (2009) indicate that by using alternative methods to allocate the risk exposure, bank risks may not increase even if market power encourages riskier asset portfolios. Tabak et al. (2012) conclude that competition affects risk-taking behavior in a non-linear way as both high and low competition levels enhance financial stability, whereas we find the opposite effect for average competition levels (see also Jiménez et al., 2013). In addition, bank size and capitalization are essential factors in explaining this relationship. Agoraki et al. (2011) suggest that banks with market power tend to undertake lower credit risk and have a lower probability of default, and capital requirements reduce risk in general, but, for banks with market power, this effect significantly weakens or can even be reversed. Laeven et al. (2016) analyze the role of bank capital, size, funding, and activities in explaining bank instability during the recent GFC and find that a higher capital ratio increases stock return, especially for large banks. Bank capital may have diverse effects on systemic stability, depending on specific policy environments: institutional frameworks enabling private market discipline and regulatory environments that foster information transparency and lower information asymmetries can substitute for capital in containing bank risk-taking. Some studies also investigate the impact of institutional environment on competition and also on risk-taking. The theoretical studies find conflicting conclusions on the relationship between bank regulation and bank performance, and a consensus on what regulatory practices are more effective for reducing bank risk has still not been reached (Barth et al., 2004, 2008). Acharya et al. (2006) claim that, although engaging in nonlending activities helps banks to achieve risk diversification, lack of experience in new services may also introduce higher volatility in incomes and higher instability. Stringent capital requirements may help improve banks’ corporate governance and reduce risk-taking behavior, but they are not necessarily related to lower loan losses (Kopecky & VanHoose, 2006; Pasiouras et al., 2006). Higher supervisory power also introduces more corruption in lending, which also increases bank risk-taking (Beck et al., 2006). Hellmann et al. (2000) claim that activities’ restrictions might incentivize bank managers to accept risky projects and Barth et al. (2004) and Laeven and Levine (2009) provide empirical findings supporting this statement. Anginer et al. (2018) show that a country’s institutional setting has a significant impact on the

2.3 Literature Review

13

relationship between bank capital and bank stability and that the impact of capital on bank risk is less pronounced for banks located in countries with better public and private monitoring of financial institutions and in countries with higher levels of information availability.

2.4 Data and Variable Definition 2.4.1 Data Sources We collect our data from different sources. The financial data for the sampled banks were collected from Bankscope. To avoid the potential bias introduced by doublecounting, this analysis employs the consolidated report; the unconsolidated report is used if consolidated information is unavailable. The country-level variables included to account for the impact of macroeconomic factors are collected from the World Bank Development Indicator Database. Additionally, we include a set of bank regulation and supervision indicators to control for the impact of institutional environment at the country-level, and the data are obtained from Barth et al. (2013).2 We keep or drop observations based on the following selection criteria: (1) for the estimation of Z-scores, we employ the mean of ROA and E/A (equity to total assets ratio) and the standard deviation of ROA (sd (ROA)) over 3 years, only keeping the banks with at least three consecutive observations; (2) we drop observations where the data on one of the variables used to estimate the Lerner index are missing; (3) for the estimation of the Lerner index, we set the minimum number of observations to 10 for each year in each country, thus deleting countries with fewer than 10 bank-year observations; and, (4) we also delete the observations where the data on one of the bank-specific and country-level control variables is missing. Using the above selection criteria, we finally have 7620 banks, accounting for 49,074 observations between 2001 and 2016. Our sample consists of commercial, savings, and co-operative banks from 118 countries. All data are adjusted based on the inflation rate and expressed in USD. To avoid the potential bias introduced by outliers, all bank-level variables employed in this study are winsorized at the 1st and 99th percentile levels.

2 In

the bank regulation and supervision information provided by Barth et al. (2013), however, the surveys were conducted in 1999, 2002, 2006, and 2011. Following the study of Anginer et al. (2014a), (2018), we employ the data from previous surveys until the newer survey data became available for matching the bank regulation and supervision variables with bank-specific variables and country control variables, that is, the data from the surveys conducted in 1999, 2002, 2006, and 2011 for the periods from 1998 to 2001, 2002 to 2005, 2006 to 2010, and 2011 to 2016, respectively. For a detailed description of the bank regulation and supervision information, see Barth et al. (2013).

14

2 Quality of Bank Capital, Competition, and Risk-Taking

2.4.2 Variable Definition (1)

Bank risk-taking measures

This study uses two measures of bank risk-taking. First, we follow the study of Jiménez et al. (2013) and employ the ratio of non-performing loans to total loans (NPL) as a measure of bank risk-taking. Second, based on the more recent study by Ashraf (2017) (see also Agoraki et al., 2011; Gaganis et al., 2018; Tabak et al., 2012), we also use a market version of the Z-score as an alternative measure of bank risk-taking. Based on the recent study by Mare et al. (2017), we use the novel non-stationary estimator and define the Z-score as in Eq. (2.1): Zit =

 R OAit + (E/A)it   ,  sd R OA

(2.1)

it

 where R OA is the moving average ROA, computed based on past and current information for   each period; E/A is the ratio of equity to total assets for period t; and,  sd ROA indicates the standard deviation of ROA computed based on past and current information for each period. A higher Z-score means a lower probability of insolvency and lower risk-taking. We also follow Anginer et al. (2014b) and employ the logarithm of the Z-score (log Zscore) to avoid the bias introduced by extreme values. (2)

Bank competition indicators

We first use the Lerner index as a measure of bank competition to evaluate banks’ competitive behavior. The higher the Lerner index, the higher the market power is. The Lerner index is defined in Eq. (2.2): Lerner Indexit = (Pit − MCit )/Pit

(2.2)

where Pit is the output price, measured by the ratio of total revenue to total assets for bank i in year t, and MCit is the marginal cost of bank i in year t.3 MCit is calculated based on Eq. (2.3): MCit =

TCit (α1 + α2 ln Qit + ϕ1 ln w1it + ϕ2 ln w 2it + ϕ3 ln w3it ) Qit

(2.3)

and the translog cost function is 3 Based

on earlier studies, we define the output price Pit as the ratio of total revenue to total assets; see also Angelini and Cetorelli (2003), Koetter et al. (2012), Fu et al. (2014), and Li (2019).

2.4 Data and Variable Definition

15

 α2 βj ln wjit (ln Qit )2 + 2 j=1 3

ln TEit = α0 + α1 ln Qit +

2  1  γj ln wjit + ϕj ln wjit ∗ ln Qit 2 j=1 j=1 3

+ +

3 3  

3

ρjk ln wjit ∗ ln wkit + αi + εit

(2.4)

j=1 k=1, k=j

where TEit is the total expenses and Qit denotes the output and is proxied by total assets of bank i in year t. Variables w1 , w2 , and w3 are input price variables: w1 is the interest expenses to total funding ratio and used as a measure of the average funding rate, w2 indicates the ratio of personnel expenses to total assets and used as the input price of personnel expenses, and w3 is the ratio of other non-interest expenses to fixed assets and employed as a proxy for the price of physical capital. Following Fu et al. (2014) and Koetter et al. (2012), we employ a stochastic cost frontier analysis to estimate Eq. (2.4) for each country in each year. As a robustness check, this analysis also constructs an alternative competitive measurement by using the Panzar-Ross model (Panzar & Rosse, 1987). Panzar and Rosse (1987) define the H-statistic to measure the market power, calculated as the sum of the elasticity of revenue with respect to three input prices. Similarly to the analysis of Anginer et al. (2014b), we use the reduced-form revenue equation and estimate the H-statistic for each country in each year4 : ln TRit = α +

3  j=1

ωj ln wjit +



πk Controlskit + αi + εit ,

(2.5)

k

where TRit is the total revenue of bank i in year t. Variables w1 , w2 , and w3 the same three input price variables used in the Lerner index estimation. A set of control variables are also included to control for bank-specific characteristics: Commercial loan ratio, the ratio of commercial loans to total assets, used to control for credit risk; Liquidity assets ratio, the ratio of liquidity assets to total assets, used to control for liquidity risk; Customer deposit ratio, the ratio of customer deposits to the sum of customer deposits and short-term funding, used to measure the bank’s funding structure; and, Equity ratio, the ratio of the equity to total assets, used to account for the leverage. Similarly to Coccorese (2009), all bank-specific factors that may have a potential effect on bank revenues but are not included in Eq. (2.5) are captured by including the bank fixed effects term (denoted by αi ). The H-statistic is defined as ω1 + ω2 + ω3 , according to Eq. (2.5). 4 Bikker

et al. (2012) indicate that a scaled revenue function leads to a significant upward bias and incorrectly measures the degree of competition; therefore, in our analysis, we employ the unscaled revenue equation to reduce the estimation bias.

16

(3)

2 Quality of Bank Capital, Competition, and Risk-Taking

Capital structure measurement and control variables

To measure each bank’s capital structure, we follow Košak et al. (2015) and define Tier 1 ratio as the Tier 1 capital to total risk-weighted assets ratio, and Tier 2 ratio as the difference between the total capital ratio and Tier 1 ratio. We include several variables to control for bank-specific time-varying effects on bank risk-taking behavior. We define Customer deposit ratio as the ratio of total customer deposit to total assets as a proxy for bank funding, Commercial loan ratio as the ratio of corporate and commercial loans to gross loans to control for the bank asset structure, Liquidity asset ratio as a control of bank liquidity risk, Foreign dummy as an indicator of bank’s ownership, and also define Size, the logarithm of total assets, to control for bank size. We also include GDP growth, Inflation, and Interest rate at the country level to account for the impact of macroeconomic factors; the variable Market capitalization, which equals to the ratio of stock market capitalization to GDP, is also included to control for the stock market size in a given country. As our sample period covers the recent GFC and the European sovereign debt, we use the bank-crisis information for individual countries from a database compiled by Laeven and Valencia (2018) to define a dummy variable Crisis dummy, which equals one if the country experienced a financial crisis in a given year, and zero otherwise. (4)

Bank regulatory variables

According to Anginer et al. (2014b), bank supervision and regulation are important factors in determining banks’ risk-taking. We use the bank regulatory information provided in Barth et al. (2013) and include several bank regulatory variables to control for bank regulation changes. We first consider the effect of capital regulation and include the following variables: Overall capital stringency, measuring the amount of capital a bank must maintain, with higher values indicating greater stringency; Official supervisory power indicating whether the supervisory authorities have the authority to take specific actions to prevent and correct problems; Banking activities restriction, related to bank activity restrictions and measuring the degree to which banks are allowed to engage in securities, insurance, and real estate activities, with higher values indicating more restrictiveness; and, Explicit deposit insurance, a binary variable that equals one if the country has explicit deposit insurance in a given year, and zero otherwise. Table 2.1 summarizes the variable definitions and the data sources employed in the analysis.

2.4.3 Descriptive Statistics and Correlation In Table 2.1, we provide descriptive statistics for the variables included in our analysis. First, we present the results for the bank risk-taking variables. The average

2.4 Data and Variable Definition

17

Table 2.1 Variable definition and data sources Variables

Definitions

Data sources

Risk-taking variables NPL

Non-performing loans divided by Bankscope total loans

log Zscore

Calculated as the sum of ROA and equity-to-assets divided by the standard deviation of ROA of each bank over past 3 years

Own calculation

Standard deviation of ROA for each bank, computed over past 3 years

Bankscope

Standard deviation of ROE for each bank, computed over past 3 years

Bankscope

   sd R OA    OE sd R

Capital structure variables Tier 1 ratio

The ratio of Tier 1 capital to total Bankscope risk-weighted assets

Tier 2 ratio

Tier 2 capital ratio, which equals to total capital ratio minus Tier 1 capital ratio

Bankscope

Lerner index

Defined as the ratio of the difference between asset price and marginal cost to asset price

Own calculation

H-statistic

Defined and calculated based on the Panzar and Rosse (1987) model for each country at each year between 2001 and 2016

Own calculation

Customer deposit ratio

The ratio of total customer deposit to total assets

Bankscope

Commercial loan ratio

The ratio of commercial loans to gross loans

Bankscope

Liquidity assets ratio

The ratio of liquidity assets to total assets

Bankscope

Size

The logarithm of total asset in $ thousand

Bankscope

Foreign dummy

A dummy variable which equals Bankscope one if the bank is foreign owned, and zero otherwise

Competition measures

Bank-control variables

Country control variables (continued)

18

2 Quality of Bank Capital, Competition, and Risk-Taking

Table 2.1 (continued) Variables

Definitions

Data sources

GDP growth

Annual growth rate of GDP at market prices based on constant local currency

World Bank Development Indicator Database

Inflation

Inflation rate

World Bank Development Indicator Database

Interest rate

The real interest rate on loans

World Bank Development Indicator Database

Crisis dummy

A dummy variable which equals one if the country experienced a financial crisis on the given year, and zero otherwise

Laeven and Valencia (2018)

Market capitalization

Stock market capitalization to GDP

World Bank Development Indicator Database

Bank regulatory measures Overall capital stringency

A variable that captures both the Barth et al. (2013) overall capital stringency and the initial capital stringency based on answers to eight questions. It ranges from zero to eight, with a higher value indicating higher capital stringency

Official supervisory power

A variable indicates whether the supervisory authorities have the authority to take specific actions to prevent and correct problems. It ranges from zero to fourteen, with fourteen indicating the highest power of the supervisory authorities

Barth et al. (2013)

Banking activities restriction

A variable measures a bank’s ability to engage in securities, insurance, and real estate activities. The ranges from 0 to 12, and a higher score indicates more restrictions on banks to engage in such activities

Barth et al. (2013)

Explicit deposit insurance

A binary variable which equals one if the country have explicit deposit insurance on the given year, and zero otherwise

Barth et al. (2013)

This table reports the definitions and data sources of the variables included in our analysis

2.4 Data and Variable Definition

19

non-performing loan ratio (NPL) is 2.285, ranging from 0 to 21.02, with a standard deviation of 4.437; the average log Zscore is 3.882, ranging from −0.321 to 6.894, with a standard deviation of 1.264. As regards bank capital structure, the average values of Tier 1 capital per risk-weighted assets (Tier 1 ratio) and Tier 2 capital per risk-weighted assets (Tier 2 ratio) are 13.577% and 1.057%, respectively. Concerning the competition variables, the Lerner index has a mean of 0.177 and a standard deviation 0.421, H-statistic has a mean of 0.613 and ranges from 0.171 to 0.916. Then, we report some descriptive statistics of the bank-specific variables. The average the average Customer deposits ratio is 0.932, and the Commercial loan ratio has a mean of 0.199, and the liquidity assets take 8.2% of the total asset on average. The value of Size varies quite substantially, ranges from 4.615 to 21.843; this confirms the importance of the size effect in the banking system. In our sample, 2.9% of banks are foreign owned. We include five country-level variables, namely GDP growth, Inflation, Interest rate, Crisis dummy, and Market capitalization, to control for the changes in macroeconomic factors and stock market size. The average country in the sample has a market capitalization to GDP ratio of 25.49, a GDP growth of 1.85%, inflation of 2.55%, and an interest rate of 5.489%. The mean value of Crisis dummy is 0.375. Finally, we report descriptive statistics on the bank regulation variables. The average country in our sample has an overall capital stringency of 5.31, official supervisory power of 12.64, banking activities restriction of 8.33, and explicit deposit insurance of 0.591. The Pearson correlations among the variables are reported in Table 2.3. The correlation coefficients between NPL and the bank-level variables suggest that the bank would have a lower non-performing loan ratio if they have a higher Tier 1 ratio and operate in a less competitive market; a higher Tier 2 ratio incentivizes the bank to engage in higher risk-taking activities. The correlations between NPL and the country-level variables indicate that a higher GDP growth, larger stock market, and lower inflation and interest rate lead to lower amounts of non-performing loans. The alternative measure of bank risk-taking, log Zscore provides results similar to those of NPL (Table 2.2).

2.5 Research Design The empirical methodology is designed to examine the effects of capital structure and competition on bank risk-taking. The analysis is performed on the bank level and the model is Risk takingijt = α0 + α1 Tier1 ratioijt + α2 Tier2 ratioijt + α3 competitionijt +  × bank controlsmijt +  × country controlsnjt + × bank regulatorykjt + εit

(2.6)

20

2 Quality of Bank Capital, Competition, and Risk-Taking

Table 2.2 Summary statistics Variable

N

Min

Median

Max

NPL

48,636

Mean 2.285

STD 4.437

0.000

0.900

21.02

log Zscore    sd R OA    sd R OE

49,074

3.882

1.264

−0.321

3.994

6.894

49,128

0.483

3.942

0.000

0.196

30.684

49,002

5.533

39.903

0.006

2.060

45.119

Tier 1 ratio

49,287

13.577

5.757

5.180

12.260

147.5

Tier 2 ratio

49,287

1.057

2.132

−3.600

1.150

4.6

Lerner index

49,287

0.177

0.421

−0.623

0.127

35.083

H-statistic

47,807

0.613

0.528

0.171

0.669

0.916

Customer deposit ratio

49,287

0.932

0.114

0.000

0.967

1.000

Commercial loan ratio

49,287

0.199

0.247

0.000

0.142

33.675

Liquidity asset ratio

49,287

0.082

0.079

−0.051

0.058

0.977

Size

49,287

12.957

1.643

4.615

12.553

21.843

Foreign dummy

49,287

0.029

0.167

0

0

1

GDP growth

49,121

1.850

2.196

– 14.814

1.786

34.500

Inflation

49,121

2.548

2.943

– 27.633

1.994

103.823

Interest rate

46,619

5.489

3.930

0.500

4.675

55.383

Crisis dummy

49,287

0.375

0.484

Market capitalization

47,820

25.494

43.312

0

0

1

0.010

123.939

1254.465

Overall capital stringency

30,102

5.307

1.922

1

7

7

Official supervisory power

30,149

12.638

Banking activities restriction

29,842

8.329

1.249

5

13

16

1.134

3

9

Explicit deposit insurance

30,100

0.591

12

0.165

0

1

1

This table reports summary statistics of the variables. Our sample period is 2001–2016. All data are inflation-adjusted, and bank-specific variables are winsorized at the 1st and 99th percentile level to reduce the influence of outliers

where Risk takingijt indicates the risk-taking of bank i in country j and year t and is represented by the non-performing loan ratio (NPL) or the logarithm of the Z-score (log Zscore) as defined in Sect. 2.4.2. The competition in the banking system is measured by the Lerner index (Lerner index). The bank control variables controlsmijt include Customer deposit ratio, Commercial loan ratio, Liquidity asset ratio, Size, and Foreign dummy. The country control variables country controlsnjt are GDP growth, Inflation, Interest rate, Market capitalization, and Crisis dummy. Bank regulatorykjt represents the bank regulatory variables Overall capital stringency, Official supervisory power, Banking activities restriction, and Explicit deposit insurance. We are also concerned with whether competition changes the relationship between capital structure and bank risk-taking. We employ Eq. (2.7) below to examine how

(2)

−0.031*** −0.036*** 0.062*** −0.054*** −0.054*** −0.027***

0.198***

−0.009**

0.058***

−0.162*** 0.189***

−0.140*** 0.106***

Tier 1 ratio (5)

Tier 2 ratio (6)

Lerner index (7)

Customer deposit ratio (8)

0.092***

0.159***

Interest rate (15)

Crisis dummy (16)

0.044***

0.044***

0.028***

0.037***

0.009**

0.009**

−0.103***

0.132***

−0.011**

−0.087***

0.196*** 0.033***

−0.138***

−0.230*** 0.050***

0.114***

0.022*** −0.015***

0.058***

0.046***

−0.033*** −0.039*** −0.026***

0.046***

−0.021*** −0.045*** 0.023***

−0.097***

Market −0.195*** 0.083*** capitalization (17)

0.142***

−0.096*** 0.118***

Inflation (14)

GDP growth (13)

0.034***

−0.047***

0.128***

Foreign dummy (12)

0.034***

−0.048***

0.127***

Size (11)

0.032***

−0.115***

0.225***

Liquidity asset ratio (10)

0.037***

0.162***

Commercial loan ratio (9)

−0.073***

−0.0135*** 0.008*

−0.042*** 1

0.721***

−0.216***

0.135*** −0.006

1

(5)

−0.175*** 1

(4)

0.117***

(3)

log Zscore (2)    sd R OA (3)    OE (4) sd R

−0.362*** 1

NPL (1)

Table 2.3 Pearson correlations

1

(7)

0.022***

0.019***

0.113***

0.016***

0.083***

−0.058***

0.032***

−0.025*** 0.0480***

−0.060*** −0.106***

0.042***

0.092***

0.065***

0.084***

0.271***

0.079***

0.121***

−0.136*** 0.008*

0.101***

1

(6)

(9)

0.173***

0.079***

0.257***

0.256***

0.156***

0.222***

0.227***

1

(10)

(continued)

−0.142*** −0.128***

−0.073*** −0.065***

−0.429*** 0.354***

−0.307*** 0.318***

−0.103*** 0.158***

−0.277*** 0.232***

−0.402*** 0.258***

−0.285*** 0.381***

−0.346*** 1

1

(8)

2.5 Research Design 21

(4)

(5)

0.143***

0.126***

−0.068*** −0.065***

Inflation (14)

Interest rate (15)

Crisis dummy (16)

−0.347*** −0.203***

Official supervisory power (19)

Banking activities −0.244*** −0.282*** restriction (20)

0.005

Overall capital stringency (18)

−0.070***

Market −0.089*** −0.143*** capitalization (17)

0.235***

0.158***

0.067***

0.103***

GDP growth (13)

1

0.287***

0.462***

1

(14)

−0.004

1

(15)

0.0533***

(6)

(7)

1

(16)

(17)

−0.091*** −0.061***

−0.103*** −0.033***

−0.152*** −0.011**

−0.032*** −0.107***

−0.058*** −0.302*** −0.518***

0.319***

−0.143*** −0.421*** −0.0947*** 0.0980***

0.609***

(8)

(18)

0.126***

0.443***

0.373***

0.125***

0.0391***

−0.0204*** 0.464***

0.154***

−0.0280*** 1

−0.144*** −0.0521*** −0.178*** 1

−0.228*** −0.113*** −0.327***

0.143***

−0.493*** −0.161*** −0.156***

0.222***

0.383***

1

(13)

(11)

Foreign dummy (12)

(12)

−0.001

−0.026*** 0.039***

−0.041*** −0.051*** −0.042***

Banking activities −0.093*** 0.087*** restriction (20)

Explicit deposit insurance (21)

−0.042*** −0.037*** −0.022***

0.075***

−0.153*** 0.127***

0.0028

Official supervisory power (19)

(3) 0.0003

(2)

−0.078***

0.229***

Overall capital stringency (18)

NPL (1)

Table 2.3 (continued) (9)

(10)

0.112**

0.549***

1

(19)

(continued)

1

(20)

−0.244*** −0.249***

−0.275*** −0.263***

−0.31***

−0.058*** −0.018***

22 2 Quality of Bank Capital, Competition, and Risk-Taking

(2)

−0.289*** −0.075***

NPL (1)

(3)

(4)

(5)

(6)

−0.264*** −0.140*** −0.0824*** 0.0925***

(7) 0.138***

This table provides the information of correlation between variables. All variables are defined as in Table 2.1 ***, **, *indicate correlation between variables are significance 1%, 5%, and 10% respectively

Explicit deposit insurance (21)

Table 2.3 (continued) (8) 0.0621***

(9) 0.208***

(10) 0.041***

2.5 Research Design 23

24

2 Quality of Bank Capital, Competition, and Risk-Taking

bank competition affects bank risk-taking and whether different qualities of bank capital strengthen or mitigate risk-taking. Risk takingijt = α0 + α1 Tier 1 ratioijt + α2 Tier 2 ratioijt + α3 competitionijt + α4 Tier 1 ratioijt ∗ competitionijt + α5 Tier 2 ratioijt ∗ competitionijt +  × controlsmijt +  × controlsnjt + × bank regulatorykjt + εijt

(2.7)

The variables here are the same as defined in Eq. (2.6), except we include the interaction term between the bank capital variables and the Lerner index. Equations (2.6) and (2.7) are estimated based on the unbalanced panel regressions, as some banks exited the market through bankruptcy, liquidity, or through M & A. We also include different types of banks, according to the main activities. There is considerable heterogeneity of the sample banks in terms of size and activities. To control for the potential heterogeneity across banks, we employ two types of estimation. First, we follow the study of Williams (2016) and use the feasible generalized least square (FGLS) regression, to account for general heteroskedasticity and crosssectional correlation in the model errors. The FGLS estimator allow for heterogeneity across banks in terms of autoregressive characteristics due to the differences in bank activities, as well as difference in variance across the banks.5 The alternative is the dynamic generalized method of moments (GMM) estimator to account for potential endogeneity of our risk-taking variable. If lagged dependent variable is correlated with the panel-level effects, the estimator may become inconsistent. Our sample also has a short time dimension and large bank dimension. Following Naceur and Omran (2011) and Noman et al. (2018), we employ the twostep system GMM procedure of Arellano and Bond (1991).6 The Arellano–Bond estimator is particularly useful in obtaining unbiased and efficient estimates in short dynamic panels with lagged endogenous variables as an explanatory variable. We employ a cluster-robust estimator to account for potentially non-i.i.d. errors and to obtain consistent standard error estimates even in the presence of autocorrelation or heteroskedasticity within panels. The bank capital structure decision is endogenous and may depend on bank-specific variables and regulatory conditions in a country (see Byoun, 2008; Flannery & Rangan, 2008; Gropp & Heider, 2010; Memmel & Raupach, 2010). We use charter value and risk-weighted assets as instrument variables based on economic arguments of Berger et al. (2019) and Gropp et al. (2018).7 5 For

detail information of feasible generalized least square (FGLS) regression, see Wooldridge (2010) and Bikker et al. (2012). 6 See Baum et al. (2007) and Roodman (2009) for detail discussion of system GMM. 7 Gropp et al. (2018) confirm that banks increase their capital ratios by reducing their risk weighted assets, not by raising their levels of equity, and this finding is consistent with debt overhang. Berger et al. (2019) find that charter value is the most important factor in explaining target capital results, and also help explain the speed of capital adjustment results.

2.5 Research Design

25

We test for the relevance of these instruments or the endogeneity of bank capital ratios using the First Stage F-test, and we use the Hansen’s J test of overidentification to check their validity.

2.6 Empirical Analysis 2.6.1 Basic Analysis In this section, we investigate the impact of capital structure and competition on bank risk-taking, by using non-performing loan ratio (NPL) as a measure of bank risk-taking. The results for FGLS regressions (Columns (1)–(4)) and the GMM estimations (Columns (5)–(8)) are reported in Table 2.4. Columns (1) and (4) report the results including on bank-level variables, while controlling for both bank fixed effects and year fixed effects. Columns (2)–(4) and (6)–(8) include country-level control variables. In columns (2) and (6), we control for both country fixed effects and year fixed effects. To avoid the estimation bias introduced the potential correlation between country-level control variables and country fixed effects, we follow Anginer et al. (2018) and include the country-year fixed effects (Country × Year FE) to control for all time-varying country factors in Columns (3)–(4) and (7)–(8). For the GMM estimation, the results of First Stage F-test and Hansen’s J statistic show that the instrument variables are both relevant and valid. The results show that Tier 1 ratio is significant and negatively related to the non-performing loan ratio at the 1% level under two estimation methods, indicating that a higher Tier 1 ratio implies lower risk-taking and a higher stability of the banking system. However, Tier 2 ratio shows different results: it is positively and significantly related to the non-performing loan ratio at the 10% level or above, implying that a higher Tier 2 ratio increases bank risk-taking. These findings provide further empirical support for Ashcraft (2008), who asserts that increasing the overall capital adequacy ratio leads to a reduction in banks’ risk appetite, but increasing the proportion of Tier 2 capital in a given capital adequacy structure increases banks’ risk appetite. Concerning the effect of competition on bank risk-taking, we find that the Lerner index is significantly and negatively related to the non-performing loan ratio, indicating that banks with higher market power have fewer non-performing loans and lower risk. This finding is consistent with those of previous studies that focus on banks in different regions (e.gAgoraki et al., 2011; Noman et al., 2018). The results for the bank-specific variables show that banks with higher funding (Customer deposit ratio), and lower commercial loans (Commercial loan ratio) and liquidity assets (Liquidity assert ratio) have a lower non-performing loan ratio. We also account for the effect of bank size (Size) and ownership and find that larger banks and foreign banks tend to exhibit higher risk-taking.

Inflation

GDP growth

Foreign dummy

Size

Liquidity asset ratio

Commercial loan ratio

Customer deposit ratio

Lerner index

Tier 2 ratio

Tier 1 ratio

L.NPL

(−14.59)

(−29.03)

−0.0753*** (−20.35)

−0.0160 (−0.82) 0.0881***

(12.49)

(−0.25)

(9.76)

0.0611***

(25.83) 0.877***

(10.59) −0.0138

(1.83)

0.0899***

(69.35)

6.095***

(58.17)

1.367***

(−5.60)

−0.406***

(−101.50)

−1.225***

(39.93)

0.0621***

(−6.34)

−0.00467***

(3)

0.776***

0.0372***

(33.88)

(78.33)

0.00492*

3.035***

5.682***

(71.56)

−0.927***

−1.759***

(16.80)

(−61.26)

(−76.57)

2.743***

−0.759***

−0.956***

0.449***

(18.10)

(31.97)

(−46.21) 0.0357***

(−49.27)

0.0530***

−0.0421***

(2)

−0.0240***

(1)

0.250***

(−7.30)

−0.0366***

(8.35)

0.528***

(12.13)

0.0455***

(48.31)

4.850***

(57.25)

2.395***

(−2.88)

−0.134***

(−86.08)

−1.064***

(30.62)

0.0568***

(−46.80)

−0.0471***

(4)

(−3.09)

(1.11)

0.227

(2.00)

0.0431**

(3.65)

1.908***

(4.97)

0.966***

(−1.03)

−0.415

(−1.67)

−0.0945*

(3.55)

0.00546***

(8)

(−16.92) −0.0306

−0.124

−0.188***

(1.01)

0.162

(0.98)

0.0195

(4.97)

2.174***

(5.31)

1.342***

(0.15)

0.0376

(−3.46)

−0.172***

(−0.08)

(continued)

−0.00294

(−6.45)

−0.116***

(1.44)

0.363

(3.86)

0.138***

(6.10)

4.022***

(6.76)

2.885***

(−0.11)

−0.0517

(−2.15)

−0.0750**

(2.60)

(−4.95) 0.230***

(−3.62)

−0.245***

(32.02)

0.835***

−0.000756

−0.107***

(54.80)

0.954***

(7)

(0.02)

0.275

(0.06)

1822.5

(0.12)

7.430

(0.49)

14.09

(0.09)

24.60

(0.07)

24.06

(−0.12)

−5.828

(2.08)

0.0136**

(−0.06)

−0.476

(0.02)

−0.101***

(21.21)

(6) 0.137

0.898***

(5)

Table 2.4 Impact of capital structure and competition on bank risk-taking (non-performing loan ratio)

26 2 Quality of Bank Capital, Competition, and Risk-Taking

Yes

No

No

Year FE

Country FE

Country × Year FE

Yes

No

Yes

Yes

No

Yes

No

No

No

(33.03)

4.397***

(−33.29)

(−5.57)

(−0.28)

−0.00693***

−0.00381***

−0.0779

(80.99)

(26.84)

1.054***

(2.43)

(24.89)

0.212**

0.0750***

(2.84)

(11.11)

(4.25) 0.0553***

(3)

(2)

2.236***

(1)

Bank FE

Constant

Explicit deposit insurance

Banking activities restriction

Official supervisory power

Overall capital stringency

Market capitalization

Crisis dummy

Interest rate

Table 2.4 (continued) (4)

(6)

Yes

No

No

No

(16.72)

No

No

Yes

Yes

(3.15)

No

Yes

Yes

No

(0.05)

3.407***

Yes

No

No

No

(3.75)

Yes

No

No

No

(continued)

(5.21)

7.578***

(−2.10)

(−7.63) 3.710***

(−1.38) −0.528**

(−22.86)

(−2.50) −0.117

−0.631*** −0.631***

−0.135**

(−4.76)

(−2.77)

−0.00143***

(11.65)

0.848***

(5.90)

0.105***

−0.0957***

(−4.76)

−0.00110***

(16.87)

0.568***

(2.33)

0.0173**

(8) (−0.12)

(3.62)

3,442,862.3

(7) (−1.63)

0.182***

(0.11)

0.0141

(0.11)

4.549

(0.93)

0.511

(−0.05)

(85.37)

2.909***

(5)

0.721***

(−24.92)

−0.00522***

(33.42)

0.762***

(18.98)

0.0971***

(21.44)

2.6 Empirical Analysis 27

(6)

GMM

0.302

1.148

24.326***

41,403

(7)

GMM

0.163

2.114

15.881***

41,403

(8)

GMM

0.495

0.426

19.357***

23,010

The dependent variable is non-performing loan ratio (NPL). Our sample period is 2001–2016. Bank-level controls include Deposit ratio, Commercial loan ratio, Liquidity ratio, Size, and Foreign dummy. Country control variables include GDP growth, Inflation, Interest rate, Crisis dummy, and Market capitalization. Bank regulatory variables Overall capital stringency, Official supervisory power, Bank activities restriction, and Explicit Deposit insurance. All variables are defined in Table 2.1. Reported t-statistics (in parentheses) are based on standard errors that are heteroskedasticity consistent and clustered at the bank level. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

GMM

Estimation method

FGLS

0.927 0.295 FGLS

354.27***

(5) 43,675

Hansen J statistic (p-value) FGLS

249.41***

(4) 27,006

Hansen J statistic

FGLS

338.12***

(3) 45,608 25.495***

236.74***

(2) 45,608

First Stage F-test

Wald Chi-square

(1)

48,636

Number of observations

Table 2.4 (continued)

28 2 Quality of Bank Capital, Competition, and Risk-Taking

2.6 Empirical Analysis

29

For the country-level variables, the results for GDP growth suggest that higher GDP growth corresponds to lower risk-taking and that the non-performing loan ratio is significantly higher during a financial crisis than during normal times. We also consider the impact of bank regulation on banks’ risk-taking behavior. When including the bank regulatory variables as control variables, we find that, if the regulators have lower stringency on capital requirements, higher official supervisory power and more restrictive on bank activities, and banking system has explicit deposit insurance, banks tend to have lower risk-taking activities. We also conduct the analysis using log Zscore as an alternative measure of bank risk-taking and the results are reported in Table 2.5. The findings in Table 2.5 are largely consistent with our main findings from using non-performing loan ratio as a measure of bank risk-taking: Tier 1 ratio is significantly and positively related to log Zscore and Tier 2 ratio is negatively and significantly related to log Zscore, confirming that banks with a higher Tier 1 ratio have lower probability of insolvency but higher Tier 2 ratio corresponds with higher risk-taking. The Lerner index is positively and significantly correlated to log Zscore, confirming that higher market power results in a lower probability of insolvency and risk, similarly to the findings of Leroy and Lucotte (2017) and Fu et al. (2014).

2.6.2 Impact of Competition on the Relationship Between Capital Structure and Risk-Taking A series of theoretical and empirical studies demonstrate that competition is an important factor affecting banks’ capital holding; based on the theoretical work of Allen et al. (2011), these studies find that competition motivates banks to hold higher levels of capital because this indicates their commitment to monitoring and attracts creditworthy borrowers. Schaeck and Cihák (2012) provide empirical support for the findings of Allen et al. (2011) and confirm that competition incentivizes banks to maintain higher capital ratios. On the other hand, Chen (2016) shows that credit market competition reduces banks’ incentive to hold capital. Although the findings on the relationship between competition and capital ratio are mixed, all these results confirm the impact of competition on bank capital holding. In this section, we are concerned with whether the relationship between Tier 1 ratio and Tier 2 ratio and bank risk-taking shows a different pattern under different competitive conditions. We include the interaction terms of Tier 1 ratio and Tier 2 ratio with the Lerner index in the analysis and the results are reported in Table 2.6.8 The results show that the interaction term between Tier 1 ratio and the Lerner index is statistically significant and positively related to the non-performing loan ratio (as shown in Columns (1)–(4)) and negatively correlated to log Zscore (as reported in Columns (5)–(8)). This indicates that, if a bank has a higher market 8 For

brevity, we only report the results of FGLS estimations in Tables 2.6 and 2.7. The results for the GMM estimations are consistent with those findings and are available upon request.

Inflation

GDP growth

Foreign dummy

Size

Liquidity asset ratio

Commercial loan ratio

Customer deposit ratio

Lerner index

Tier 2 ratio

Tier 1 ratio

L. log Zscore

−0.0174***

−0.0351***

0.000908

(25.60)

(18.57)

0.0415***

0.0355***

(−2.88)

−0.0654***

(−18.79)

−0.0408***

(−53.94)

−2.311***

(−23.47)

(1.50)

(−1.18)

−0.0289

(−54.43)

−1.797***

(−26.56)

−0.478***

(13.81)

0.537***

(70.10)

0.490***

(−1.87)

−0.00227*

(98.68)

0.0558***

(4)

0.0106

(0.35)

(−2.52)

(−4.60)

(−3.83)

−0.00840***

−0.0212*** 0.0101

(−70.78)

(−81.82)

(−14.07)

−1.922***

−1.795***

−0.0433**

−0.00780***

(−20.69)

(−2.65)

(4.25) −0.465***

(10.06) −0.478***

(38.32)

−0.0273***

0.140***

(50.62)

0.393***

(−2.41)

0.369***

(49.08)

(73.01)

0.846***

0.390***

0.417***

(−2.72)

(−1.71)

−0.000489**

(86.55)

(84.27) −0.000860***

(86.89)

0.0542***

(3)

0.0521***

(2)

−0.000706*

0.0434***

(1)

Table 2.5 Impact of capital structure and competition on bank risk-taking (log Zscore)

(3.70)

(−0.37)

−0.0137

(−2.66)

−0.0206***

(−6.54)

−0.876***

(−2.37)

−0.119**

(6.16)

0.447***

(3.15)

0.0555***

(−2.09)

−0.000275**

(7)

0.0708

−0.0177**

(6.02)

0.0225***

−0.247 (−0.99)

(0.38)

0.0200

(1.17)

0.00971

(−4.21)

−0.687***

(−5.19)

−0.485***

(0.24)

61.65

(−2.89)

−1.254***

(−4.25)

−10.94***

(−1.28)

−2.485

(−0.91)

−0.0742

−1.206 (−0.42)

(2.13)

0.0329**

(−0.07)

−0.000241

(5.29)

0.0590***

(70.94)

0.748***

(0.77)

0.0767

(−1.88)

−0.00451*

(0.44)

0.00852

(12.20)

0.0444***

(71.23)

(6) 0.565***

0.748***

(5)

(8)

(continued)

0.00348

(6.07)

0.0240***

(−1.10)

−0.0459

(−3.39)

−0.0291***

(−6.09)

−0.736***

(−3.68)

−0.241***

(4.33)

0.353***

(2.64)

0.0945***

(−2.47)

−0.0106**

(2.66)

0.0194***

(48.06)

0.786***

30 2 Quality of Bank Capital, Competition, and Risk-Taking

Yes

No

No

Year FE

Country FE

Country × Year FE No

Yes

Yes

No

(22.49)

Yes

3.160***

(74.21)

Yes

No

No

No

(67.82)

3.974***

(4.46)

(−7.69) (3.17)

(−87.45)

−0.261*** 0.000294***

−0.524***

(0.40)

0.00120***

(−42.88)

0.00226

(6)

(7)

Yes

No

No

No

(17.21)

1.567***

No

No

Yes

Yes

(−0.75)

No

Yes

Yes

No

(−0.18)

Yes

No

No

No

(2.52)

(continued)

Yes

No

No

No

(−1.54)

−0.377

(0.83)

(5.55)

(1.63) 0.0494

0.127***

(5.88)

(−0.34) 0.0262

(4.54) 0.0362***

−0.00624

0.0303***

0.809**

(0.36)

0.0000541

(−2.80)

−0.0502***

(−1.32)

−0.00438

(−0.51)

−14.80

(8) (0.68)

(−2.94)

(3.93)

0.000836***

−0.00526 (−0.73)

(−8.96)

−0.109***

(−14.66)

−0.0433***

(−2.54)

(−0.16)

−0.0957

(−0.91)

−0.0732

(0.82)

−0.00344

−0.216

(5)

−0.00721***

(4.99)

0.000300***

(−51.02)

−0.498***

(−16.88)

−0.0270***

−0.0516***

(0.20)

(4) (−6.40)

(3) (−14.32)

(2)

2.591***

(1)

Bank FE

Constant

Explicit deposit insurance

Banking activities restriction

Official supervisory power

Overall capital stringency

Market capitalization

Crisis dummy

Interest rate

Table 2.5 (continued)

2.6 Empirical Analysis 31

(6)

GMM

0.121

4.824

56.05***

41,686

(7)

GMM

0.158

3.974

44.82***

41,686

(8)

GMM

0.214

2.879

28.45***

23,193

The dependent variable is log Zscore. Our sample period is 2001–2016. Bank-level controls include Deposit ratio, Commercial loan ratio, Liquidity ratio, Size, and Foreign dummy. Country control variables include GDP growth, Inflation, Interest rate, Crisis dummy, and Market capitalization. Bank regulatory variables Overall capital stringency, Official supervisory power, Bank activities restriction, and Explicit Deposit insurance. All variables are defined in Table 2.1. Reported t-statistics (in parentheses) are based on standard errors that are heteroskedasticity consistent and clustered at the bank level. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

GMM

Estimation method

FGLS

3.593 0.187 FGLS

318.93***

(5) 44,280

Hansen J statistic (p-value) FGLS

415.90***

(4) 27,142

Hansen J statistic

FGLS

368.11***

(3) 45,798 38.51***

436.12***

(2) 45,798

First Stage F-test

Wald Chi-square

(1)

49,074

Number of observations

Table 2.5 (continued)

32 2 Quality of Bank Capital, Competition, and Risk-Taking

Foreign dummy

Size

Liquidity asset ratio

Commercial loan ratio

Customer deposit ratio

Lerner index

Tier 2 ratio * Lerner index

Tier 2 ratio

Tier 1 ratio * Lerner index

Tier 1 ratio

(−15.07)

(−18.77)

(−0.71)

(10.77)

(12.32)

(19.47) 0.860***

(12.59) −0.0387

(3.18)

0.0657***

(150.41)

8.322***

(39.62)

1.137***

(−3.76)

−0.287***

(−57.02)

−1.892***

(1.52)

0.104

(22.57)

0.0445***

(34.45)

0.0687***

0.859***

0.0453***

(35.10)

(69.01)

0.00896***

3.130***

5.780***

(69.38)

−0.860***

−1.237***

(13.15)

(−53.01)

(−61.36)

2.671***

−1.653***

−1.431***

0.446***

(1.25)

(1.50)

(18.95) 0.0720

(28.32)

0.0789

0.0411***

(33.29)

(39.79)

0.0598***

0.0621***

0.0427***

(−14.75)

(6.67)

0.408***

(18.57)

0.0655***

(58.74)

5.524***

(59.32)

2.636***

(−2.46)

−0.121**

(−50.82)

−1.988***

(1.00)

0.00645

(19.38)

0.0442***

(26.60)

0.0682***

(−47.59)

(88.47)

0.0601***

(9.46)

0.357***

(50.22)

0.817***

(0.96)

(−1.08)

(−64.69)

−1.783***

(28.43)

0.475***

(5.56)

0.171***

(43.65)

0.755***

(0.83)

0.0151

(−64.48)

−2.358***

(26.52)

0.534***

(15.29)

0.575***

(53.72)

1.044***

(0.47)

0.00164

(3.08)

0.00376***

(−29.17)

−0.0399***

(91.95)

0.0680***

(8)

(−1.89)

−0.0338*

(−14.86)

(0.35)

0.0101

(−4.58)

(−1.74)

−0.0423*

(−2.87)

(continued)

(−3.43)

−0.0788***

(−28.29)

−0.0249*** −0.00981*** −0.00507*** −0.0411***

(−65.10)

−1.993***

−1.928*** (−84.56)

(21.68)

(−4.23)

−0.0489*** 0.486***

(38.84)

0.863***

(46.70)

0.794***

(−0.42)

(1.68) 0.00957

(−2.16)

−0.00119

−0.00250** 0.00189* −0.0190

(−23.05)

−0.0279***

(98.31)

0.0622***

(7)

(−27.31)

(−25.72)

−0.0278*** −0.0308***

(82.88)

0.0511***

(−59.16)

(−50.75)

−0.0628***

−0.0574***

−0.0266***

−0.0161***

(5)

(6)

Panel B: log Zscore (4)

(1)

(3)

(2)

Panel A: NPL

Table 2.6 Impact of competition on the relationship between capital structure and bank risk-taking

2.6 Empirical Analysis 33

Constant

Explicit deposit insurance

Banking activities restriction

Official supervisory power

Overall capital stringency

Market capitalization

Crisis dummy

Interest rate

Inflation

GDP growth

Table 2.6 (continued)

1.700***

0.543**

4.048***

2.432***

3.559***

(5.96) (continued)

1.343***

(−7.23) 4.312***

(7.02) 0.141***

0.0410***

−0.712*** −0.563***

(3.46)

(−2.87) (−27.30)

(−4.35) 0.0236***

−0.0103***

(4.84)

0.000309***

(−57.58)

(85.01)

3.050***

(−1.14)

−0.0000765

(−103.44)

(−16.92) −0.495***

−0.0554***

0.710***

(−31.52)

(3.28)

(−45.48)

0.00128***

(−6.73)

(−45.27) −0.542***

(0.33)

(−5.60) −0.0257***

−0.0515***

−0.0153***

−0.0387*** (−19.24)

(39.87)

0.0422***

(8)

(23.02)

0.0358***

(7)

−0.231***

0.00189

(0.40)

0.00188

(1.36)

0.00994

(−6.21)

(33.46)

0.845***

(16.46)

0.0815***

(22.13)

0.247***

(−9.40)

−0.0476***

−0.00431*** −0.00786*** −0.00633***

(83.80)

(0.65)

(27.25) 1.159***

(2.72) 0.0540

0.0795***

(16.52)

0.0520***

0.118***

(5.89)

(−22.33)

(−4.41) 0.0835***

−0.0761***

−0.0792***

(6)

(5)

(4)

Panel B: log Zscore (3)

(1)

(2)

Panel A: NPL

34 2 Quality of Bank Capital, Competition, and Risk-Taking

No

48,636

481.39***

FGLS

Country FE

Country × Year FE

Number of Observations

Wald Chi-square

Estimation method

FGLS

302.78***

45,608

No

Yes

Yes

No

FGLS

549.62***

45,608

Yes

No

FGLS

417.81***

27,006

Yes

No

No

No

(20.95)

FGLS

424.4***

49,074

No

No

Yes

Yes

(64.03)

FGLS

390.41***

45,798

No

Yes

Yes

No

(21.13)

FGLS

347.31***

45,798

Yes

No

No

No

(60.92)

(7)

FGLS

421.07***

27,142

Yes

No

No

No

(16.21)

(8)

The dependent variables are NPL and log Zscore in Panel A and Panel B, respectively. Our sample period is 2001–2016. Bank-level controls include Deposit ratio, Commercial loan ratio, Liquidity ratio, Size, and Foreign dummy. Country control variables include GDP growth, Inflation, Interest rate, Crisis dummy, and Market capitalization. Bank regulatory variables Overall capital stringency, Official supervisory power, Bank activities restriction, and Explicit Deposit insurance. All variables are defined in Table 2.1. Reported t-statistics (in parentheses) are based on standard errors that are heteroskedasticity consistent and clustered at the bank level. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

Yes

No

Year FE

No

(29.31)

(1.98) No

(19.22)

Yes

(5)

(6)

Panel B: log Zscore (4)

(1)

(3)

(2)

Panel A: NPL

Bank FE

Table 2.6 (continued)

2.6 Empirical Analysis 35

36

2 Quality of Bank Capital, Competition, and Risk-Taking

power, the negative relationship between the Tier 1 ratio and risk-taking becomes less pronounced. In other words, in a more competitive market, the effect of the Tier 1 ratio in reducing bank risk will be stronger. The interaction term between Tier 2 ratio and Lerner index is insignificant under different specifications. Based on Allen et al. (2011) and Schaeck and Cihák (2012), one potential explanation is that the competition in the banking system drives compels to hold a higher level and higher quality of bank capital, which in turn reduces bank risk.

2.6.3 Impact of the Financial Crisis on the Relationship Between Capital Structure and Risk-Taking The results of our basic analysis confirm that Tier 1 ratio and Tier 2 ratio exhibit different effects on bank risk-taking. After the GFC, the results of studies on the causes of the GFC strengthen the importance of bank capital and urge for enhanced bank capital regulation (see Anginer et al., 2018; Bui et al., 2017). We are interested in whether Tier 1 ratio and Tier 2 ratio function differently during the financial crisis. In this section, we use financial crisis information on individual countries obtained from Laeven and Valencia (2018) and include the interaction terms of Tier 1 ratio and Tier 2 ratio with Crisis dummy as additional explanatory variables. Crisis dummy is a binary variable that equals one if, in a given year, a country is experiencing a systemic crisis, and zero otherwise. Table 2.7 presents the results of the analysis with these bank crisis interactions. The coefficient of the interaction variable with Tier 1 ratio is significantly and negatively related to the non-performing loan ratio (Columns (1)–(4)) and significantly and positively related to log Zscore (Columns (5)–(8)), suggesting that the effect of the Tier 1 ratio in reducing bank risk-taking is more pronounced during a financial crisis. We also find a positive (negative) relationship between the interaction term with Tier 2 ratio and the non-performing loan ratio (log Zscore), and the coefficient of the interaction term is significant in most specifications (it is only insignificant for the non-performing loan ratio in column (3)), confirming that the positive relationship of Tier 2 ratio in increasing bank risk-taking also become pronounced during financial crisis. Our findings confirm that the negative relationship for Tier 1 ratio and positive relationship for Tier 2 ratio become more pronounced during a financial crisis.

Foreign dummy

Size

Liquidity asset ratio

Commercial loan ratio

Customer deposit ratio

Lerner index

Tier 2 ratio * Crisis dummy

Tier 2 ratio

Tier 1 ratio * Crisis dummy

Tier 1 ratio

(−16.35)

(−26.63)

(−1.40)

(2.56)

(13.36)

(20.83) 0.943***

(22.73)

−0.0743

(1.86)

0.0840***

(73.03)

6.182***

(31.78)

1.067***

(−5.30)

−0.367***

(−81.78)

−0.933***

(1.37)

0.00735

(2.58)

0.00614***

(−55.41)

−0.100***

(−73.26)

0.213**

0.0469***

(36.95)

(80.82)

0.00492*

3.093***

6.306***

(75.13)

−0.808***

−1.555***

(35.99)

(−68.22)

(−92.49)

2.606***

−0.770***

−1.029***

0.856***

(14.87)

(15.83)

(10.90)

0.0412***

0.0571***

(8.91)

(−49.63)

(−49.43)

0.0197***

−0.0901***

−0.0779***

0.0144***

(−13.24)

(−8.66)

(10.78)

0.574***

(14.69)

0.0558***

(45.11)

4.496***

(85.07)

2.963***

(−0.72)

−0.0353

(−107.39)

−1.070***

(2.05)

0.0131**

(16.69)

0.0356***

(−40.83)

−0.116***

(−12.08)

(−3.77)

−0.0482***

(0.31)

0.00943

(−5.53)

−0.0117***

−0.0277*** (−14.75)

(−51.34)

−1.822***

−1.714*** (−59.54)

(−31.36)

(−1.53)

(10.75) −0.521***

(49.24)

0.382***

(47.90)

0.370***

(−4.63)

−0.0179

0.838***

(59.34)

0.402***

(−13.14)

(5.00) −0.00982***

(9.80)

0.00763***

(33.93)

0.0435***

(52.87)

0.0386***

(6)

−0.0293***

0.0127***

(27.29)

0.0181***

(70.01)

0.0368***

−0.0118***

−0.00816***

−0.0125***

(5)

−0.0513***

Panel B: log Zscore (4)

(1)

(3)

(2)

Panel A: NPL

Table 2.7 Impact of capital structure on bank risk-taking during the financial crisis

(−2.18)

−0.0485**

(0.90)

0.00137

(−48.95)

−1.621***

(−35.04)

−0.490***

(7.47)

0.230***

(51.54)

0.353***

(−2.53)

−0.00641**

(2.72)

0.00355***

(36.38)

0.0453***

(64.02)

0.0398***

(7)

(continued)

(−2.05)

−0.0467**

(−15.62)

−0.0344***

(−138.05)

−2.293***

(−28.78)

−0.516***

(15.09)

0.560***

(137.58)

0.469***

(−2.93)

−0.00668***

(2.83)

0.00292***

(30.69)

0.0464***

(51.63)

0.0384***

(8)

2.6 Empirical Analysis 37

Constant

Explicit deposit insurance

Banking activities restriction

Official supervisory power

Overall capital stringency

Market capitalization

Crisis dummy

Interest rate

Inflation

GDP growth

1.532***

(−21.81)

(−21.24)

(−5.08)

3.959***

−0.00428***

−0.00360***

−0.316

(73.43)

(18.05)

2.578***

(7.02) (continued)

1.880***

(−7.11) 3.797***

(8.61) 0.149***

0.0422***

−0.698*** −0.509***

(3.25)

(−3.79) (−25.14)

(0.73) 0.0224***

−0.0696***

3.819***

(4.06)

0.000252***

(−48.76)

(100.97)

3.118***

(−1.58)

−0.000105

(−65.48)

−1.138***

(−15.93)

−0.0241***

0.00130

(3.06)

0.00119***

(−22.06)

(−45.61) −1.169***

(0.51)

−0.0485***

(−9.49)

−0.0190***

−0.0399*** (−17.35)

(17.12)

0.0333***

(8)

(21.58)

0.0359***

(7)

−0.868***

0.00274

(0.04)

0.000193

(1.20)

0.00878

(6)

0.740***

(−27.05)

−0.00575***

(55.22)

2.445***

(24.28)

(−25.24)

(3.68) 2.272***

0.106***

−0.0756***

0.0662***

1.539***

(19.58)

(7.75)

(4.23)

0.213***

0.0627***

0.0598***

(−17.33)

−0.0839***

−0.0899***

(1.15)

(5)

0.0222

Panel B: log Zscore (4)

(3)

(1)

(2)

Panel A: NPL

Table 2.7 (continued)

38 2 Quality of Bank Capital, Competition, and Risk-Taking

No

48,636

591.09***

FGLS

Country FE

Country × Year FE

Number of observations

Wald Chi-square

Estimation method

Yes

FGLS

501.9***

45,608

No

Yes

FGLS

426.2***

45,608

Yes

No

No

FGLS

399.6***

27,006

Yes

No

No

No

(19.17)

FGLS

340.2***

49,074

No

No

Yes

Yes

(71.85)

FGLS

468.3***

45,798

No

Yes

Yes

No

(23.38)

(6)

FGLS

434.57***

45,798

Yes

No

No

No

(70.50)

(7)

FGLS

326.22***

27,142

Yes

No

No

No

(19.87)

(8)

The dependent variables are NPL and log Zscore in Panel A and Panel B, respectively. Our sample period is 2001–2016. Bank-level controls include Deposit ratio, Commercial loan ratio, Liquidity ratio, Size, and Foreign dummy. Country control variables include GDP growth, Inflation, Interest rate, Crisis dummy, and Market capitalization. Bank regulatory variables Overall capital stringency, Official supervisory power, Bank activities restriction, and Explicit Deposit insurance. All variables are defined in Table 2.1. Reported t-statistics (in parentheses) are based on standard errors that are heteroskedasticity consistent and clustered at the bank level. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

Yes

No

Year FE

No

(29.69)

(−1.29)

No

(18.90)

Yes

(5)

Bank FE

Panel B: log Zscore (4)

(1)

(3)

(2)

Panel A: NPL

Table 2.7 (continued)

2.6 Empirical Analysis 39

40

2 Quality of Bank Capital, Competition, and Risk-Taking

2.6.4 Robustness Checks To check whether our findings are consistent, we also perform robustness checks based alternative measures of our main variables and also different subsamples.9 (1)

Alternative measures of competition and bank risk

First, we consider whether our results are robust across different measures of competition, and employ two other alternative measures of bank competition: (1) the Hstatistics based on the Panzar-Rosse model as defined in Sect. 2.4.2 and (2) the efficiency-adjusted Lerner index according to Koetter et al. (2012) (see also Fu et al., 2014). The results when using H-statistic (Columns (1)–(3) in Table 2.8) and efficiency-adjusted Lerner index (Columns (4)–(6) in Table 2.9) are consistent with our main findings.10 Second, the bank risk measure Z-score, which indicates the insolvency risk of the banks, is constructed based on ROA and equity to total assets ratio, which may highly correlated to Tier 1 ratio and Tier 2 ratio, and not really a good measure of bank risky in our analysis. To avoid the bias introduced by thepotential correlation between    Z-score and capital ratio variables, we use sd ROA and sd ROE as alternative measures of bank risk. We  find a negative  (positive) relationship between Tier 1   ratio (Tier 2 ratio) and sd ROA (sd ROE ), and this relationship becomes more pronounced during a financial crisis and in a more competitive market, which are consistent with our main findings. (2)

Robustness check on subsamples

In this section, we perform our analysis based on different subsamples. First, we examine the relationship when considering only commercial banks. The results for commercial banks (Columns (1)–(3)) show that only Tier 1 ratio (Tier 2 ratio) has a negative (positive) impact on bank risk-taking and, during the financial crisis, a higher Tier 1 ratio is related with lower risk. Second, we perform our analysis based on subsamples of banks from different countries. As in our entire sample, 9661 out of 48,636 bank-year observations are from the U.S. banking system (accounting for 19.9% of the sample). We perform our analysis by excluding the observations from the United States and check whether our findings are consistent (Columns (4)–(6)). We also report the results including only countries with more than 200 observations (Columns (7)–(9)). Finally, as our sample period covers the Basel I, Basel II, and Basel III regulatory sub-periods, and the implementation of Basel regulatory varies by country. Thus, the definitions of the Tier 1 and Tier 2 capital ratios in our sample were changing 9 For brevity, we only report the results under FGLS, the results for the GMM estimations are largely

similar and are available upon request. bank-level control variables, country control variables, and bank regulatory variables are included in the regression but not report for brevity. 10 The

Tier 1 ratio * Lerner index

Lerner index

Tier 2 ratio * Efficiency−adjusted Lerner index

Tier 1 ratio * Efficiency−adjusted Lerner index

Efficiency-adjusted Lerner index

Tier 2 ratio * Crisis dummy

Tier 1 ratio * Crisis dummy

Tier 2 ratio * H−statistic

Tier 1 ratio * H−statistic

(1.61)

0.0550

(−1.85)

−0.00142*

(2.77)

(9.16)

(−5.02)

0.109***

(13.29)

H−statistic

0.242***

−0.00580***

−0.0250***

0.0269***

Tier 2 ratio

(−27.70)

(−34.77)

(−55.69)

NPL

−0.0337***

(−1.94)

(−1.42)

(1.01)

−0.979

(1.71)

0.0658*

−2.584* (2.61)

1.005***

(21.81)

(17.80) −0.509

0.0952***

0.0792***

(−34.58)

(−10.45)

−0.0213***

(−26.41)

−0.0158***

NPL

(6)

−0.0490***

(3.82)

0.00772***

(−71.25)

−0.0355***

NPL

(5)

(−34.05)

(18.19)

0.0347***

(−55.90)

−0.0299***

NPL

(4)

−0.0477***

(9.85)

0.260***

(−2.93)

−0.0175***

NPL

−0.0376***

NPL

(3)

Tier 1 ratio

(2)

Variables

(1)

Table 2.8 Robustness checks: alternative measures of competition and bank risk

(−55.33) 0.0141***

−0.361*** (−87.28)

(15.95)

0.00606***

(−64.03)

−0.0128***

(8)   sd  ROA

−0.171***

(11.29)

0.00460***

(−44.84)

−0.00736***

(7)   sd  ROA

(−76.41)

−0.155***

(1.42)

0.00113

(−39.91)

−0.0177***

(1.47)

0.000375

(−3.24)

−0.000651***

(9)   sd  ROA

(−65.47)

−1.594***

(23.32)

0.105***

(−96.15)

−0.271***

(10)   sd  ROE

0.265***

(−61.08)

−5.594***

(29.97)

0.127***

(−115.20)

−0.326***

(11)   sd  ROE

(continued)

(−86.55)

−1.823***

(43.21)

0.299***

(−45.87)

−0.241***

(−5.29)

−0.0257***

(−79.97)

−0.147***

(12)   sd  ROE

2.6 Empirical Analysis 41

27,006

383.5***

FGLS

Number of observations

Wald Chi-square

Estimation method

FGLS

440.8***

27,006

Yes

FGLS

264.6***

27,006

Yes

Yes

(7.31)

2.321***

(3)

FGLS

346.6***

27,006

Yes

Yes

(14.10)

3.749***

(4)

FGLS

416.27***

27,006

Yes

Yes

(−11.55)

−40.37***

(5)

FGLS

327.96***

27,006

Yes

Yes

(16.41)

4.348***

(6)

FGLS

147.6***

27,006

Yes

Yes

(35.32)

1.178***

(7)

FGLS

168.9***

27,006

Yes

Yes

(33.03)

FGLS

226.2***

27,006

Yes

Yes

(35.09)

1.237***

FGLS

353.33***

26,973

Yes

Yes

(41.02)

18.38***

FGLS

365.93***

26,973

Yes

Yes

(52.71)

18.45***

(−1.28)

1.125***

−0.138

(−1.21)

(11)

−0.00784

(10) (49.85)

(9)

(34.83)

(8)

FGLS

222.6***

26,973

Yes

Yes

(36.52)

17.30***

(12)

The dependent variable is Non-performing loan ratio. Our sample period is 2001–2016. Bank-level controls include Deposit ratio, Commercial loan ratio, Liquidity ratio, Size, and Foreign dummy. Country control variables include GDP growth, Inflation, Interest rate, Crisis dummy, and Market capitalization. Bank regulatory variables Overall capital stringency, Official supervisory power, Bank activities restriction, and Explicit Deposit insurance. The bank-level control variables, country control variables, and bank regulatory variables are included in the regression but not report for brevity. All variables are defined in Table 2.1. Reported t-statistics (in parentheses) are based on standard errors that are heteroskedasticity consistent and clustered at the bank level. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

Yes

Country × Year FE

Yes

(−6.14)

(4.19)

Yes

−30.37***

(2)

1.361***

Bank and country variables

Constant

Tier 2 ratio * Lerner index

(1)

Table 2.8 (continued)

42 2 Quality of Bank Capital, Competition, and Risk-Taking

(8)

(9)

(10)

(11)

(12)

−1.301***

−1.093***

−1.377*** (−5.73)

Yes

24,616

Country × Year Yes FE

24,616

225.8***

Number of observations

Wald Chi-square

209.7***

Yes

319.8***

24,616

Yes

363.6***

21,642

Yes

332.41***

21,642

Yes

319.9***

21,642

Yes

462.15***

25,911

Yes

382.38***

25,911

Yes

Yes

(0.74) Yes

0.0733***

−2.095*** (−5.28)

0.00469

278.4***

25,911

Yes

Yes

(12.59)

(−42.98)

Yes

Yes

0.309***

(−5.22)

169.4***

5124

Yes

205.34***

5124

Yes

Yes

(−1.74)

−0.186*

(−1.31)

(9.36)

0.624***

(−8.14)

(continued)

361.3***

5124

Yes

Yes

(−6.03)

−4.924*** −1.406***

(10.46)

0.759***

(5.03)

(0.80)

Yes

(−6.03)

−1.406***

(9.36)

0.624***

(−8.14)

−0.0327 −0.123***

(−74.32)

−1.042***

(10.40)

0.0244***

(−18.02)

(30.63)

0.0815***

(−51.69)

−2.172***

(17.77)

0.0382***

(−59.21)

0.0550

Yes

(−68.39)

−0.956***

(14.37)

0.0338***

(−50.25)

(−41.37)

(1.58)

(1.17)

(−8.18)

−0.927***

(2.19)

0.0325**

(−3.21)

−0.113***

0.0927

0.0533

Yes

(−10.38)

(4.23)

(0.45)

(−92.36)

(2.12)

0.0421***

(−5.01)

(46.23)

−2.652***

(−88.02)

−1.207***

(−88.31)

(19.96)

0.00305**

(−4.13)

0.00369

(12.23)

(12.40)

0.0326***

(−18.12)

0.0950***

0.0308***

0.0316***

(−70.34)

(−11.82)

(7)

(−38.67)

(6)

Subperiod after Implementation of Basel III −0.307*** −0.297***

(5)

(4)

(3)

Countries with more than 200 observations

−0.0496*** −0.0710*** −0.0201*** −0.0336*** −0.0442*** −0.0261*** −0.0576*** −0.0778*** −0.0217*** −0.297***

(2)

(1)

Yes

Bank and country variables

Tier 2 ratio * Crisis dummy

Tier 1 ratio * Crisis dummy

Tier 2 ratio * Lerner index

Tier 1 ratio * Lerner index

Lerner index

Tier 2 ratio

Tier 1 ratio

Countries without US

Commercial Banks

Table 2.9 Robustness checks: subsamples

2.6 Empirical Analysis 43

FGLS

(2)

(3)

FGLS

FGLS

FGLS

(5)

(4)

FGLS

(1) FGLS

(6) FGLS

(7) FGLS

(8)

Countries with more than 200 observations

FGLS

(9) FGLS

(10) FGLS

(11)

FGLS

(12)

Subperiod after Implementation of Basel III

The dependent variable is Non-performing loan ratio (NPL). Our sample period is 2001–2016. Bank-level controls include Customer deposit ratio, Commercial loan ratio, Liquidity asset ratio, Size, and Foreign dummy. Country control variables include GDP growth, Inflation, Interest rate, Crisis dummy, and Market capitalization. Bank regulatory variables Overall capital stringency, Official supervisory power, Bank activities restriction, and Explicit Deposit insurance. The bank-level control variables, country control variables, and bank regulatory variables are included in the regression but not report for brevity. All variables are defined in Table 2.1. Reported t-statistics (in parentheses) are based on standard errors that are heteroskedasticity consistent and clustered at the bank level. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

Estimation method

Countries without US

Commercial Banks

Table 2.9 (continued)

44 2 Quality of Bank Capital, Competition, and Risk-Taking

2.6 Empirical Analysis

45

substantially between 2001 and 2016. Based on the implementation timeline of Basel III for different countries, we focus our analysis on the most recent subperiod and perform our analysis on the countries that complete the implementation of Basel III regulatory.11,12 The results show that our main findings are consistent under these alternative sample selection criteria’s.

2.7 Conclusion In this analysis, we examine the relationship between capital structure, competition, and bank risk-taking using cross-country data from 118 countries in the period between 2001 and 2016. We consider the difference between Tier 1 and Tier 2 capital quality of bank capital. The results indicate that banks with a higher Tier 1 ratio and a lower Tier 2 ratio are lower risk-takers. A bank with higher market power in its banking system tends to reduce its risk-taking activities. The results also confirm that banks with more customer deposit and higher quality of bank assets are exposed to lower risk. During a financial crisis, banks exhibit a higher level of risk-taking than during normal times. Our findings also highlight a sharp contrast between the impact of different types of bank capital quality during the financial crisis and in different competitive conditions. During the financial crisis, Tier 1 capital acted as a stable funding source and reduced the bank risk, but the evidence on Tier 2 capital shows that a higher Tier 2 ratio results in a higher level of risk and increases bank instability. In a banking system, competition compels banks to hold more capital, which in turn reduces bank risk, but only in the case of Tier 1 capital. Our findings provide some policy implications for bank regulation as follows: The negative (positive) relationship between Tier 1 ratio (Tier 2 ratio) and bank risk-taking confirms the importance of the quality of bank capital in the banking system—more enhanced regulation of high-quality capital should be implemented to reduce bank risk-taking behavior and enhance effective risk management, and this is more important during a financial crisis. Although our results suggest that higher market power reduces bank risk-taking, increases in market power make the banking system less contestable, which makes the banking system less competitive. This finding emphasizes the importance of a regulatory framework that achieves a balance between reducing bank risk-taking and avoiding potentially excessive competition.

11 The

timeline for the adoption of Basel III regulatory framework by Basel Committee members can be find in the progress report progress report on adoption of the Basel regulatory framework provided by Basel Committee on Banking Supervision. See also https://www.bis.org/bcbs/implem entation/rcap_reports.htm. 12 The results of the interaction terms for Crisis dummy are dropped because of collinearity.

46

2 Quality of Bank Capital, Competition, and Risk-Taking

References Acharya, V. V., Hasan, I., & Saunders, A. (2006). Should banks be diversified? Evidence from individual bank loan portfolios. The Journal of Business, 79(3), 1355–1412. https://doi.org/10. 1086/500679 Agoraki, M.-E.K., Delis, M. D., & Pasiouras, F. (2011). Regulations, competition and bank risktaking in transition countries. Journal of Financial Stability, 7(1), 38–48. Allen, F., Carletti, E., & Marquez, R. (2011). Credit market competition and capital regulation. The Review of Financial Studies, 24(4), 983–1018. https://doi.org/10.1093/rfs/hhp089 Allen, F., & Gale, D. (2000). Bubbles and crises. The Economic Journal, 110(460), 236–255. https:// doi.org/10.1111/1468-0297.00499 Allen, F., & Gale, D. (2004). Competition and financial stability. Journal of Money, Credit and Banking, 453–490. Altunbas, Y., Carbo, S., Gardener, E. P., & Molyneux, P. (2007). Examining the relationships between capital, risk and efficiency in European banking. European Financial Management, 13(1), 49–70. Angelini, P., & Cetorelli, N. (2003). The effects of regulatory reform on competition in the banking industry. Journal of Money, Credit and Banking, 35(5), 663–684. Anginer, D., Demirgüç-Kunt, A., & Mare, D. S. (2018). Bank capital, institutional environment and systemic stability. Journal of Financial Stability, 37, 97–106. https://doi.org/10.1016/j.jfs.2018. 06.001 Anginer, D., Demirguc-Kunt, A., & Zhu, M. (2014). How does competition affect bank systemic risk? Journal of Financial Intermediation, 23(1), 1–26. Anginer, D., Demirguc-Kunt, A., & Zhu, M. (2014). How does deposit insurance affect bank risk? Evidence from the recent crisis. Journal of Banking & Finance, 48, 312–321. Arellano, M., & Bond, S. (1991). Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. The Review of Economic Studies, 58(2), 277–297. Ashcraft, A. B. (2008). Does the market discipline banks? New evidence from regulatory capital mix. Journal of Financial Intermediation, 17(4), 543–561. Ashraf, B., Arshad, S., & Hu, Y. (2016). Capital regulation and bank risk-taking behavior: Evidence from Pakistan. International Journal of Financial Studies, 4(3), 16. Ashraf, B. N. (2017). Political institutions and bank risk-taking behavior. Journal of Financial Stability, 29, 13–35. https://doi.org/10.1016/j.jfs.2017.01.004 Ayuso, J., Pérez, D., & Saurina, J. (2004). Are capital buffers pro-cyclical? Evidence from Spanish panel data. Journal of Financial Intermediation, 13(2), 249–264. Barrell, R., Davis, E. P., Fic, T., & Karim, D. (2011). Bank capital composition, regulation and risk taking. NIESR and Brunel University London. Barth, J. R., Caprio, G., & Levine, R. (2004). Bank regulation and supervision: What works best? Journal of Financial Intermediation, 13(2), 205–248. Barth, J. R., Caprio, G., & Levine, R. (2008). Bank regulations are changing: For better or worse? Comparative Economic Studies, 50(4), 537–563. Barth, J. R., Caprio, G. J., & Levine, R. (2013). Bank regulation and supervision in 180 countries from 1999 to 2011. Journal of Financial Economic Policy, 5(2), 111–219. Basel Committee on Banking Supervision. (2010). Basel III: A global regulatory framework for more resilient banks and banking systems. From http://www.bis.org/press/p100912.htm Baum, C. F., Schaffer, M. E., & Stillman, S. (2007). Enhanced routines for instrumental variables/GMM estimation and testing. Stata Journal, 7(4), 465–506. Beck, T., Demirguc-Kunt, A., & Levine, R. (2006). Bank concentration, competition, and crises: First results. Journal of Banking & Finance, 30(5), 1581–1603. Ben Naceur, S., Marton, K., & Roulet, C. (2018). Basel III and bank-lending: Evidence from the United States and Europe. Journal of Financial Stability, 39, 1 https://doi.org/10.1016/j.jfs.2018. 08.002

References

47

Berger, A. N., Klapper, L. F., & Turk-Ariss, R. (2009). Bank competition and financial stability. Journal of Financial Services Research, 35(2), 99–118. Berger, A. N., Öztekin, Ö., & Roman, R. A. (2019). How does competition affect bank capital structure? Evidence from a natural experiment. ECB Working Paper. Bikker, J. A., Shaffer, S., & Spierdijk, L. (2012). Assessing competition with the Panzar-Rosse model: The role of scale, costs, and equilibrium. Review of Economics and Statistics, 94(4), 1025–1044. Boyd, J. H., De Nicoló, G., & Smith, B. D. (2004). Crises in competitive versus monopolistic banking systems. Journal of Money, Credit and Banking, 487–506. Brunnermeier, M. K. (2009). Deciphering the liquidity and credit crunch 2007–2008. Journal of Economic Perspectives, 23(1), 77–100. https://doi.org/10.1257/jep.23.1.77 Bui, C., Scheule, H., & Wu, E. (2017). The value of bank capital buffers in maintaining financial system resilience. Journal of Financial Stability, 33, 23–40. https://doi.org/10.1016/j.jfs.2017. 10.006 Byoun, S. (2008). How and when do firms adjust their capital structures toward targets? The Journal of Finance, 63(6), 3069–3096. Caminal, R., & Matutes, C. (2002). Market power and banking failures. International Journal of Industrial Organization, 20(9), 1341–1361. https://doi.org/10.1016/S0167-7187(01)00092-3 Carletti, E. (2008). Competition and regulation in banking. Handbook of Financial Intermediation and Banking, 126(5), 449–482. Chen, Y. (2016). Bank capital and credit market competition: Will competitive pressure lead to higher capital levels? Journal of International Money and Finance, 69, 247–263. https://doi.org/ 10.1016/j.jimonfin.2016.07.006 Coccorese, P. (2009). Market power in local banking monopolies. Journal of Banking & Finance, 33(7), 1196–1210. Demirgüç-Kunt, A., Detragiache, E., & Merrouche, O. (2013). Bank capital: Lessons from the financial crisis. Journal of Money, Credit and Banking, 45(6), 1147–1164. EBA. (2011). EBA recommendation on the creation and supervisory oversight of temporary capital buffers to restore market confidence (EBA/REC/2011/1), December 8, 2011. Fernández, A. I., González, F., & Suárez, N. (2013). How do bank competition, regulation, and institutions shape the real effect of banking crises? International evidence. Journal of International Money and Finance, 33, 19–40. Flannery, M. J., & Rangan, K. P. (2008). What caused the bank capital build-up of the 1990s? Review of Finance, 12(2), 391–429. Fu, X. M., Lin, Y. R., & Molyneux, P. (2014). Bank competition and financial stability in Asia Pacific. Journal of Banking & Finance, 38, 64–77. Gaganis, C., Hasan, I., Papadimitri, P., & Tasiou, M. (2018). National culture and risk-taking: Evidence from the insurance industry. Journal of Business Research, 97, 104–116. Goodhart, C. A. E. (2013). Ratio controls need reconsideration. Journal of Financial Stability, 9(3), 445–450. Gropp, R., & Heider, F. (2010). The determinants of bank capital structure. Review of Finance, rfp030. Gropp, R., Mosk, T., Ongena, S., & Wix, C. (2018). Banks response to higher capital requirements: evidence from a quasi-natural experiment. The Review of Financial Studies, hhy052-hhy052. https://doi.org/10.1093/rfs/hhy052 Hellmann, T. F., Murdock, K. C., & Stiglitz, J. E. (2000). Liberalization, moral hazard in banking, and prudential regulation: Are capital requirements enough? American Economic Review, 90(1), 147–165. Jiménez, G., Lopez, J. A., & Saurina, J. (2013). How does competition affect bank risk-taking? Journal of Financial Stability, 9(2), 185–195. https://doi.org/10.1016/j.jfs.2013.02.004 Jokipii, T., & Milne, A. (2008). The cyclical behaviour of European bank capital buffers. Journal of Banking & Finance, 32(8), 1440–1451.

48

2 Quality of Bank Capital, Competition, and Risk-Taking

Jokipii, T., & Milne, A. (2011). Bank capital buffer and risk adjustment decisions. Journal of Financial Stability, 7(3), 165–178. https://doi.org/10.1016/j.jfs.2010.02.002 Keeley, M. C. (1990). Deposit insurance, risk, and market power in banking. The American Economic Review, 80(5), 1183–1200. Kelly, B., Lustig, H., & Van Nieuwerburgh, S. (2016). Too-systemic-to-fail: What option markets imply about sector-wide government guarantees. American Economic Review, 106(6), 1278– 1319. Kim, D., & Santomero, A. M. (1988). Risk in banking and capital regulation. The Journal of Finance, 43(5), 1219–1233. Koehn, M., & Santomero, A. M. (1980). Regulation of bank capital and portfolio risk. The Journal of Finance, 35(5), 1235–1244. https://doi.org/10.1111/j.1540-6261.1980.tb02206.x Koetter, M., Kolari, J. W., & Spierdijk, L. (2012). Enjoying the quiet life under deregulation? Evidence from adjusted Lerner indices for US banks. Review of Economics and Statistics, 94(2), 462–480. Koivu, T. (2012). Monetary policy, asset prices and consumption in China. Economic Systems, 36(2), 307–325. Kopecky, K. J., & VanHoose, D. (2006). Capital regulation, heterogeneous monitoring costs, and aggregate loan quality. Journal of Banking & Finance, 30(8), 2235–2255. Košak, M., Li, S., Lonˇcarski, I., & Marinˇc, M. (2015). Quality of bank capital and bank lending behavior during the global financial crisis. International Review of Financial Analysis, 37, 168– 183. Laeven, L., & Levine, R. (2009). Bank governance, regulation and risk taking. Journal of Financial Economics, 93(2), 259–275. Laeven, L., Ratnovski, L., & Tong, H. (2016). Bank size, capital, and systemic risk: Some international evidence. Journal of Banking & Finance, 69(Supplement 1), S25–S34. https://doi.org/10. 1016/j.jbankfin.2015.06.022 Laeven, L., & Valencia, F. (2018). Systemic banking crises revisited. IMF Working Paper No. 206. International Monetary Fund. Leroy, A., & Lucotte, Y. (2017). Is there a competition-stability trade-off in European banking? Journal of International Financial Markets, Institutions and Money, 46, 199–215. https://doi.org/ 10.1016/j.intfin.2016.08.009 Li, S. (2019). The impact of bank regulation and supervision on competition: Evidence from emerging economies. Emerging Markets Finance and Trade, 55(10), 2334–2364. https://doi. org/10.1080/1540496X.2018.1547191 Llewellyn, D. T. (2008). The Northern Rock crisis: A multi-dimensional problem waiting to happen. Journal of Financial Regulation and Compliance, 16(1), 35–58. Mésonnier, J.-S., & Monks, A. (2015). Did the EBA capital exercise cause a credit crunch in the Euro Area? International Journal of Central Banking, 11, 75–117. Mare, D. S., Moreira, F., & Rossi, R. (2017). Nonstationary Z-Score measures. European Journal of Operational Research, 260(1), 348–358. https://doi.org/10.1016/j.ejor.2016.12.001 Memmel, C., & Raupach, P. (2010). How do banks adjust their capital ratios? Journal of Financial Intermediation, 19(4), 509–528. Naceur, S. B., & Omran, M. (2011). The effects of bank regulations, competition, and financial reforms on banks’ performance. Emerging Markets Review, 12(1), 1–20. Noman, A. H. M., Gee, C. S., & Isa, C. R. (2018). Does bank regulation matter on the relationship between competition and financial stability? Evidence from Southeast Asian countries. PacificBasin Finance Journal, 48, 144–161. Panzar, J. C., & Rosse, J. N. (1987). Testing for “monopoly” equilibrium. The Journal of Industrial Economics, 35(4), 443–456. Pasiouras, F., Gaganis, C., & Zopounidis, C. (2006). The impact of bank regulations, supervision, market structure, and bank characteristics on individual bank ratings: A cross-country analysis. Review of Quantitative Finance and Accounting, 27(4), 403–438.

References

49

Rajan, R. G. (1994). Why bank credit policies fluctuate: A theory and some evidence. The Quarterly Journal of Economics, 109(2), 399–441. Roodman, D. (2009). How to do xtabond2: An introduction to difference and system GMM in Stata. The Stata Journal, 9(1), 86–136. Schaeck, K., & Cihák, M. (2012). Banking competition and capital ratios. European Financial Management, 18(5), 836–866. https://doi.org/10.1111/j.1468-036X.2010.00551.x Shrieves, R. E., & Dahl, D. (1992). The relationship between risk and capital in commercial banks. Journal of Banking & Finance, 16(2), 439–457. Tabak, B. M., Fazio, D. M., & Cajueiro, D. O. (2012). The relationship between banking market competition and risk-taking: Do size and capitalization matter? Journal of Banking & Finance, 36(12), 3366–3381. https://doi.org/10.1016/j.jbankfin.2012.07.022 Williams, B. (2016). The impact of non-interest income on bank risk in Australia. Journal of Banking & Finance, 73, 16–37. https://doi.org/10.1016/j.jbankfin.2016.07.019 Wooldridge, J. M. (2010). Econometric analysis of cross section and panel data. MIT Press.

Chapter 3

Banking Sector Reform, Competition, and Bank Stability

3.1 Overview This study tests the impact of banking sector reform and competition on bank stability based on unbalanced data from 22 transition countries from 1998 to 2016. The initial results not only highlight the positive relationship between market power and bank fragility, but also confirm the positive relationship between bank reform and stability. Our findings also show that both higher activity restrictions and more explicit guidelines for asset diversification increase bank stability, but this positive effect significantly weakens for banks with higher market power. More stringent capital requirements in combination with higher market power increase the risk of bank insolvency.

3.2 Introduction The recent global financial crisis between 2008 and 2010 influenced the global economy through banking systems across countries and led to unprecedented consequences. Some studies highlight that deregulation and excessive competition in financial sectors were important factors that drove the financial crisis (Brunnermeier, 2009; Carletti, 2008; Fernández et al., 2013; Llewellyn, 2008). The impact of bank regulation and competition on financial stability has thus attracted more attention from academics and policymakers since then (Acharya & Richardson, 2009; Beck et al., 2010, 2013; Schaeck & Cihák, 2014). With the development of information technology (Marinˇc, 2013) as well as the liberalization, deregulation, and integration of capital markets (Das & Ghosh, 2006; Koetter et al., 2012), the number of financial reforms has increased markedly. A number of studies demonstrate that financial reforms develop the financial market, This chapter was published in Emerging Markets Finance and Trade. © Shanghai Jiao Tong University Press 2021 S. Li, Financial Regulation and Bank Performance, Contributions to Finance and Accounting, https://doi.org/10.1007/978-981-16-3509-0_3

51

52

3 Banking Sector Reform, Competition, and Bank Stability

increase the market share of private credit, reduce the financing cost, and raise risktaking (Cubillas & González, 2014). The banking systems in transition countries have undergone extensive reforms in the past two decades and this provides us with a unique setting in which to test the relationships among banking sector reform, competitive conditions, and financial stability. In this study, we investigate the impact of both bank reform and competition on stability by analyzing the empirical relationships among banking sector reform, competition, and stability in transition economies. By using unbalanced data on 1121 banks from 22 transition countries between 1998 and 2016, our findings confirm the negative relationship between market power and stability, supporting the view that competition increases bank stability. The results also confirm that bank reform increases banking stability, supporting the view that better financial development leads to a more stable banking system. Both higher activity restrictions and more explicit guidelines for asset diversification increase bank stability, but this positive effect significantly weakens for banks with higher market power. More stringent capital requirements in combination with higher market power increase the risk of bank insolvency. Declaring insolvency power, private monitoring, financial statement transparency, and deposit insurance have only a direct impact on bank stability. Our analysis extends previous studies in two ways. First, banking systems in transition countries are particularly interesting because banks play no economic role in planned economies; by contrast, the financial systems in most transition countries are now dominated by banks rather than the equity market. Since the 1990s, financial reforms have been implemented across transition countries, including financial liberalization, bank privatization, widening access to foreign banks, and restructuring national banking systems, which have changed the regulation system of the banking sector as well as reshaped the competitive conditions of banks. The recent financial crisis and global recession also enable us to test the resilience of new institutions and regulatory structures. Second, transition countries exhibit a number of unique characteristics, and government policies and regulations can differ from those in developed and developing economies. Therefore, we evaluate the impact of competition on bank stability while considering the impact of banking sector reforms and bank regulations as well as controlling for bank-specific characteristics and macroeconomic factors that may impact bank stability. We also perform robustness checks by using different estimation methods and subsamples. The rest of the paper is structured as follows. The third section summarizes the banking sector reform in transition countries and the related literature. The fourth section mainly addresses the sample selection, variable definition, and data statistics. The fifth section describes the empirical methodology employed in this study. The sixth section presents the empirical analysis and robustness checks and the last section concludes.

3.3 Banking Sector Reform in Transition Countries and Literature Review

53

3.3 Banking Sector Reform in Transition Countries and Literature Review 3.3.1 Banking Sector Reforms in Transition Countries Before the transition phase, banking sectors in transition countries were largely controlled by the state and lending activities were showing a marked preference toward state-owned firms, practically the only form of non-monetary financing of the budget deficit. Because of the implicit government guarantee, banks had no incentives to monitor their risk. As a result, they were saddled with a large number of bad loans concentrated in the state-owned banking sector (Perotti, 1993). Financial reforms have been implemented across transition countries since the 1990s, including resolving non-performing loans, privatizing state-owned banks, introducing foreign investors, and restructuring national banking systems. As discussed in the Introduction, this has changed the regulation system of the banking sector and reshaped the competitive conditions of the banking system. Transition countries have enacted strict bankruptcy laws to control the level of non-performing loans. For example, non-performing loans in Hungary were replaced by government securities and transferred to a government collection agency, and the stakes of large state-owned banks owned by the government were sold to foreign investors. Bad loans in the banking system in the Czech Republic were replaced by government bonds and the bad assets were transferred to a new bank (Bonin et al., 2014). The privatization programs of government-owned banks have experienced several stages (Bonin et al., 2005b). After the bank crises in the 1990s, transition countries started to introduce foreign investors into financial markets. To privatize banks in Poland, both domestic initial public offerings and foreign investment were used. The bank privatization programs in Romania, Bulgaria, Croatia, and the Czech Republic involved negotiations between governments and foreign banks (Bonin et al., 2014). With the introduction of foreign investors, the majority of government-owned banks were taken over by privately owned and well-capitalized financial institutions (Bonin et al., 2005a). Indeed, foreign investors now hold a large share of the banking sectors in transition countries (Fang et al., 2014). The initial bank regulation reforms focused on opening the financial market, facilitating banking system competition, and creating diversities of financial institutions. Financial reforms including liberalizing interest rates, reducing directed credit programs, and lowering bank activity restrictions were implemented (Bonin et al., 2014). New and prudential supervision guidelines concerning capital requirements, official supervisory power, loan classifications, information disclosure, and private sector monitoring were established following the rapid adoption of financial reforms. For example, bank lending in Croatia is now strictly constrained by central bank regulation, which mitigates the effects of booms. Moreover, formal Financial Stability Reports and macro-prudential policy responses have been introduced by Romania and several other transition countries to control systemic risk in the banking system (Bonin et al., 2014).

54

3 Banking Sector Reform, Competition, and Bank Stability

3.3.2 Literature Review Previous analyses indicate that competitive banking systems are fragile. In a competitive banking system, banks benefit less from the lending relationship, and this discourages them from monitoring borrowers’ risk properly and thus raises risk (Allen & Gale, 2004). Boyd et al. (2004) also suggest that large banks in concentrated banking sectors can enhance profits and decrease financial instability by providing more “capital buffers” that absorb external macroeconomic and liquidity shocks. Further, concerning the competition–fragility nexus, although deposit insurance can mitigate bank instability by preventing banks from insolvency, it also leads to moral hazard by providing incentives for banks to undertake risk-taking activities. Thus, deposit insurance may increase bank instability in a more competitive market (Matutes & Vives, 1996) and thus constraints on deposit interest rates are needed to prevent excessive risk-taking in a competitive banking system (Hellmann et al., 2000). Some studies indicate that competitive banking systems are more stable. Less competition leads to reduced credit rationing and higher loans, increasing the probability of bank insolvency (Caminal & Matutes, 2002). However, Berger et al. (2009) indicate that by using alternative methods to allocate the risk exposure, bank risks may not increase, even if market power encourages riskier asset portfolios. On the impact of banking sector reform on risk-taking behavior, previous studies have mixed findings. From the traditional viewpoint, financial liberalization is generally associated with increased market competition, which in turn introduces greater risk-taking activities and intensified moral hazard issues (Grossman, 1992); Boyd et al., 1998). Dick (2006) also provides empirical evidence for this view based on bank information from the United States after the Riegle–Neal Act in the 1990s. However, other researchers have different views, stating that banks are more able to pursue economies of scale and scope and diversify income flows, and thus become less risky, if they are liberalized (Claessens & Klingebiel, 2001). Hellmann et al. (2000) claim that activity restrictions might incentivize bank managers to accept risky projects and Barth et al. (2004) and Laeven and Levine (2009) provide empirical evidence for this statement. Theoretical studies find conflicting conclusions on the relationship between bank regulation and bank performance, and no consensus on which regulatory practices are more effective in reducing bank risk has been reached (Barth et al., 2004, 2007). Acharya et al. (2006) claim that although engaging in non-lending activities helps banks diversify risk, the lack of experience in new services may also introduce higher volatility in incomes and higher instability. Stringent capital requirements may help improve banks’ corporate governance and reduce risk-taking behavior, but they are not necessarily related to lower loan losses (Kopecky & VanHoose, 2006). Greater supervisory power also introduces more corruption in lending and this raises banking instability (Beck et al., 2006).

3.3 Banking Sector Reform in Transition Countries and Literature Review

55

Only a few recent studies provide direct findings on the relationships among banking sector reform, competition, and bank performance by focusing on transition countries. Brissimis et al. (2008) confirms that banking sector reforms, by increasing competition and introducing more risk-taking activities, improve bank efficiency. Agoraki et al. (2011) confirm that banks with higher market power tend to have a lower probability of default and that higher activity restrictions and capital requirements increase bank stability. Fungáˇcová and Weill (2013) explore the role of bank competition on the occurrence of bank failures based on a sample in Russia and confirm that increasing bank competition could undermine bank stability. Fang et al. (2014) suggest that financial stability has improved substantially since the banking sector reform.

3.4 Data and Variables 3.4.1 Sample We collect banks’ financial data from Bankscope. To avoid the bias introduced by double counting in the sample, we use the consolidated financial report of each bank, and only the unconsolidated report if the consolidated report is not available. Information on banking sector reform is collected from the transition report provided by the European Bank for Reconstruction and Development (EBRD). To control for the impact of bank regulation, we also employ the bank regulation and supervision data provided by Barth et al. (2013).1 The country-level variables used to control for the change in macroeconomic factors and capital market size are collected from the World Development Indicators database. We drop observations based on the following selection criteria: (1) observations where the data on one of the variables employed to estimate the Lerner index are missing; (2) to estimate the Lerner index, we set the minimum number of observations to 10 for each year in each country; thus, we delete countries with fewer than 10 bankyear observations; (3) to compute the Z-score, we employ the moving average ROA (return on assets) and moving average standard deviation of ROA (sd(ROA)) over three years, and we only keep banks that have at least three consecutive observations; and (4) we delete observations where the data on one of the bank-specific and countrylevel control variables are missing.

1 For

the bank regulation and supervision information provided by Barth et al. (2013), however, the surveys were conducted in 1999, 2002, 2006, and 2011. Following the study of Anginer et al. (2014a), we employ the previous survey data until the new survey data become available for matching the bank regulation and supervision variables with the bank-specific variables and country control variables: the survey data found in 1999 for 1998–2001, the survey data found in 2002 for 2002– 2005, the survey data found in 2006 for 2006–2010, and the survey data found in 2011 for 2011– 2016.

56

3 Banking Sector Reform, Competition, and Bank Stability

After applying the selection criteria, we finally have 1121 banks, which account for 4445 observations between 1998 and 2016. Our sample consists of commercial banks, saving banks, cooperative banks, and bank holding companies from 22 transition countries.2 All data are adjusted based on the inflation rate and expressed in USD.

3.4.2 Variables (1)

Bank stability measures

We follow Laeven and Levine (2009) and employ the Z-score to measure bank stability. The Z-score is an indicator of bank solvency extensively used in banking studies (e.g., Demirgüç-Kunt & Huizinga, 2010, Beck et al., 2013). It indicates the number of standard deviations by which returns would have to fall from the mean to wipe out all equity in the bank (Boyd & Runkle, 1993). Based on the recent study of Mare et al. (2017), we employ the novel non-stationary estimator and calculate the Z-score as follows: Z it =

 R OAit + (E/A)it    sd R OA

(3.1)

it

 where R OA is defined as moving average ROA computed based on past and current information for each period;  E/A  is defined as the ratio of equity to total assets for the  current period (t), and sd R OA indicates the standard deviation of ROA computed based on past and current information for each period.3 A higher Z-score indicates a lower probability of insolvency and higher bank stability. We also follow Laeven and Levine (2009) and employ the logarithm of the Z-score to avoid the bias  introduced   by extreme values. As a robustness check, we use ROA volatility (sd R OA ), ROE    volatility (sd R OE ), and the non-performing loan ratio (NPL, defined as the ratio of non-performing loans to total loans) as alternative measures of bank stability. (2)

Bank competition indicators

This study first employs the Lerner index to measure bank competition and evaluates banks’ competitive behavior. The Lerner index is defined as follows:

2 Focusing only on European cooperative banks between 2006 and 2014, Clark et al. (2018) study the

relationship between competition and financial stability and show that a hump-shaped relationship exists between market power and stability. 3 Mare et al. (2017) also summarize the methods for computing the Z-score and outline some important limitations.

3.4 Data and Variables

57

Lerner Indexit = (Pit − MCit )/Pit

(3.2)

In our analysis, Pit indicates the output price, measured as the total revenue to total assets ratio for bank i in year t, and MCit indicates the marginal cost of bank i in year t.4 MCit is defined as   3  TCit MCit = ρk ln wkit + ϕ3 ln Tt α1 + α2 ln Qit + Qit k=1

(3.3)

and the translog total cost function is 3 

1 2 ln TCit = βk ln wkit + γk ln wkit + α1 ln Qit 2 k=1 k=1 3

 1 1 + α2 ln Q it2 + ρk ln Qit ln wkit + ϕ1 ln Tt + ϕ1 ln T2t 2 2 k=1 3

+ ϕ3 ln Tt ln Qit +

3 

ϕk+3 ln Tt ln wkit + α0 + αi + εit

(3.4)

k=1

where T Cit is defined as total costs and Qit denotes output, proxied by the total assets of bank i in year t. We define w1 as the interest expenses to total funding ratio and employ this to measure the average funding rate; w2 indicates the ratio of personnel expenses to total assets and this is used as the input price of personnel expenses; and w3 is the ratio of other non-interest expenses to fixed assets and this is employed as a proxy of the price of physical capital. T is a time trend variable used to capture the impact of technological development. Equation (3.4) is estimated based on stochastic cost frontier analysis for each country in each year (see also Fu et al., 2014; Koetter et al., 2012). We also estimate the efficiency-adjusted Lerner index based on Koetter et al. (2012) as a robustness check. For another robustness check, we define an alternative competition indicator based on the Panzar–Rosse model (Panzar & Rosse, 1987). The Panzar–Rosse model constructs an H-statistic to measure market power, where the H-statistic is calculated as the sum of the elasticity of revenue with respect to three input prices. Based on Anginer et al. (2014b), we employ the reduced-form revenue equation and estimate the H-statistic for each country in each year5 :

4 We

define the output price Pit as the ratio of total revenue to total assets following Angelini and Cetorelli (2003), Koetter et al. (2012), and Fu et al. (2014). 5 Bikker et al. (2012) indicate that a scaled revenue function leads to a significant upward bias and incorrectly measures the degree of competition. Hence, in our analysis, we employ the unscaled revenue equation to reduce estimation bias.

58

3 Banking Sector Reform, Competition, and Bank Stability

ln OPIit = α +

3 

β j ln w jit +

j=1



πk Controlskit + αi + εit

(3.5)

k

where OPI it indicates the operating revenue of bank i in year t and w1 , w2 , and w3 are the bank input price variables as defined in the Lerner index estimation. We also add a set of control variables to control for the impact of bank-specific characteristics.6 All bank-specific factors that may affect operating revenue but are not included in Eq. (3.5) are captured by including bank fixed effects (FE; denoted by αi ). The Hstatistic, which is defined as the sum of the elasticities of revenues with respect to three input prices, is defined as β1 + β2 + β3 .7 Finally, we define the Herfindahl– Hirschman index (HHI), which is calculated as the sum of the square of the market share of each bank in the banking system to measure bank concentration. (3)

Banking sector reform measure

To measure banking sector reform in transition countries, we collect bank reform information from the transition reports provided by the EBRD. The EBRD follows the transition progress of former socialist countries in different sectors since the 1990s. The bank reform indicator for transition countries considers the degree of interest rate liberalization, bank credit allocation, presence of private banks, and whether bank supervision and regulation are prudent (see also Brissimis et al., 2008; Fang et al., 2014). This indicator varies between 1 and 4.33, with a higher value indicating greater financial development. (4)

Bank regulation variables

As we are also interested in the effect of bank regulation on the relationship between competition and bank stability, we employ the set of regulation variables provided by Barth et al. (2013) to examine the impact of bank regulation from different aspects. The first regulatory variable Activity restriction is related to banks’ activity restrictions and this measures the degree to which banks are allowed to engage in

6 The control variables include the following: Customer Loan, defined as the ratio of customer loans

to total assets, to control for credit risk; NEA, defined as the ratio of other non-earning assets to total assets, to control for the composition of the asset; Customer Deposit, defined as the ratio of customer deposits to the sum of customer deposits and short-term funding, to measure the funding structure of the bank; and Equity Ratio, defined as the equity to total assets ratio, to account for leverage. 7 To deal with the potential heteroskedasticity problem, we estimate the unscaled PR revenue model in Eq. (3.5) by using pooled feasible generalized least squares, and clustered standard errors are also used to account for general heteroskedasticity and cross-sectional correlation in the model errors (Arellano 1987). We also follow the analysis of Bikker et al. (2012) and employ an equilibrium ROE test to confirm whether banks operate in a long-run equilibrium. By using ROE as the dependent variable in Eq. (3.5), the H-statistic under the ROE test equals zero if the banking system operates in a long-run equilibrium.

3.4 Data and Variables

59

securities, insurance, and real estate activities (higher values indicate higher restrictions). Second, we consider the effect of capital regulation and include Capital stringency index, which measures the amount of capital a bank must maintain (higher values indicate greater stringency). We use Declaring insolvency power to measure official supervisory power (higher values indicate greater power). Diversification index captures whether there are explicit, verifiable, and quantifiable guidelines for banks’ asset diversifications and whether banks are allowed to make loans outside national borders (higher values indicate more diversification). Private monitoring index measures whether there are incentives for private monitoring by firms (higher values indicate more private monitoring). Finally, Financial statement transparency indicates the transparency of banks’ financial statement practices to examine the impact of external governance and Deposit insurance, which equals one if the country has deposit insurance and zero otherwise, is used to investigate the impact of deposit insurance on bank stability.8 (5)

Bank-specific and country-level control variables

We use several variables to control for bank-specific time-varying effects on bank stability. We define log EQ as the logarithm of the total equity to total assets ratio to control for leverage, log LLP as the logarithm of the ratio of loan loss provision to gross loans to control for loan quality, log DEP as the ratio of total customer deposits to total assets as a proxy of bank funding, FA as the ratio of fixed assets to total assets to measure a bank’s asset structure, and log (TA), defined as the logarithm of total assets, to control for bank size. All the bank-specific variables employed in this analysis are winsorized at the 1st and 99th percentiles to avoid the effects of outliers. We also use GDP growth and Inflation at the country level to account for the impact of macroeconomic factors. The binary variable Crisis equals one if the country was experiencing a systemic crisis in a given year and zero otherwise, following Laeven and Valencia (2018), to control for the economic cycle in a country. Table 3.1 summarizes the definition and data sources of the variables employed in this analysis.

3.4.3 Summary Statistics Panel A in Table 3.2 provides the statistics of the variables included in our analysis. First, we present the results for the bank stability measurements. The average log Zscore is 2.42, ranging from −5.85 to 12.13; sd(ROA) and sd(ROE) have means 8 We also include Foreign bank limitation, which measures whether foreign banks own domestic banks and may enter a country’s banking industry (lower values indicate greater stringency), and Entry barriers, which indicates whether various types of legal submissions are required to obtain a banking license (higher values indicate greater stringency). However, the results show no significant impact on bank stability.

60

3 Banking Sector Reform, Competition, and Bank Stability

Table 3.1 Variable definition and data sources Variables

Definitions

Data sources

Calculated as the sum of ROA and equity-to-assets divided by the standard deviation of ROA of each bank over past 3 years

Bankscope

Standard deviation of ROA for each bank, computed over past 3 years

Bankscope

Standard deviation of ROE for each bank, computed over past 3 years

Bankscope

Nonperforming loans divided by total loans

Bankscope

Dependent variables log Zscore

   OA sd R    sd R OE NPL

Competitive condition indicators Lerner index

Lerner index is equal to the difference Own calculation between asset price and marginal cost, normalized by asset price

H-statistic

H-statistic is calculated as the sum of Own calculation the elasticity of revenue with respect to three input prices for each country at each year between 1998 and 2016 based on the Panzar and Rosse (1987) model

HHI

Hirschmann-Herfindahl index (HHI) Own calculation of concentration of total assets, which is the sum of the squares of the market shares (assets) of each bank in each country at each year

Bank-specific variables log TA

The logarithm of total asset in $ thousand

Bankscope

log EQ

The logarithm of total equity to total assets ratio

Bankscope

log LLP

The logarithm of ratio of loan loss provision to gross loans

Bankscope

NIM

The ratio of net interest income to total earning asset

Bankscope

log DPS

The logarithm of ratio of total customer deposit to total asset

Bankscope

FA

The ratio of fixed assets to total assets Bankscope (continued)

3.4 Data and Variables

61

Table 3.1 (continued) Variables

Definitions

Data sources

Country control variables GDP growth

Annual growth rate of GDP at market World Bank Development prices based on constant local Indicator Database currency

Inflation

Inflation rate

Crisis

A binary variable which equals one if Laeven and Valencia (2010) the country was experiencing a systemic crisis in a given year and zero othewise

World Bank Development Indicator Database

Bank reform index Bank reform index

An indicator which considers the European Bank for degree of liberalization of interest Reconstruction and rate, the allocation of bank credit, Development (EBRD) whether significant lending to private enterprise exits, whether private banks have a significant presence, and whether bank supervision and regulation are prudent. The indicator goes from 1 to 4.33, with higher numbers indicating higher stages of development

Bank regulation variables Activity restriction

A variable measures a bank’s ability to engage in securities, insurance, and real estate activities. The ranges from 0 to 12, and a higher score indicates more restrictions on banks to engage in such activities

Barth et al. (2013)

Capital regulatory index

A variable that captures both the overall capital stringency and the initial capital stringency based on answers to eight questions. It ranges from zero to eight, with a higher value indicating higher capital stringency

Barth et al. (2013)

Declaring insolvency power

A variable measure whether the supervisory authorities have the power to declare a deeply troubled bank insolvent. It ranges from 0 to 4 and higher values indicate greater power

Barth et al. (2013)

Diversification index

A variable that ranges from zero to two, with higher values indicating more diversification

Barth et al. (2013)

(continued)

62

3 Banking Sector Reform, Competition, and Bank Stability

Table 3.1 (continued) Variables

Definitions

Private monitoring index

A variable that measures whether Barth et al. (2013) there incentives/ability for the private monitoring of firm, with higher values indicating more private monitoring

Data sources

Financial statement transparency

The transparency of bank financial Barth et al. (2013) statements practices, ranges from 0 to 6 and higher values indicate better transparency

Deposit insurance

A binary variable which equals one if Barth et al. (2013) the country has explicit deposit insurance scheme and zero otherwise

This table reports the definitions and data sources of the variables included in our analysis

of 1.82 and 15.4 with standard deviations of 2.78 and 24.01, respectively; and the average non-performing loans ratio (NPL) is 9.09, ranging from 0 to 114.8. We also summarize the competition variables. The Lerner index has a mean of 0.295 with a standard deviation 0.207; its H-statistic is 0.311, ranging from 0.153 to 0.848, and the average HHI is 0.158, ranging from 0.068 to 0.986. For the bank-specific variables, the average log TA is 14.39 and its value varies substantially, which confirms the importance of the size effect in a banking system. The average logarithm of customer deposits to total assets (log DPS) is −0.553, while log LLP has a mean of −4.244, and the average fixed asset ratio is 0.032. We include three country-level variables, namely GDP growth, Inflation, and Crisis, to control for the changes in macroeconomic factors and the economic cycle. An average transition country in the sample has GDP growth of 4.03% and inflation of 10.34%, and 18.8% of countries have experienced a crisis. Finally, we report the descriptive statistics of the bank reform indicator and bank regulation variables. An average country in our sample has a bank reform indicator of 2.87, activity restriction of 6.33, capital regulatory index of 7.07, declaring insolvency power of 1.61, diversification index of 1.35, private monitoring index of 6.96, and financial statement transparency of 4.85. In total, 91.6% of the observations have an explicit deposit insurance scheme. Panel B in Table 3.2 reports the sample distribution by calendar year and country. The number of bank-year observations in our sample ranges from 45 to 648, with the largest number in 2008 and the lowest number in 1998. Russia accounts for 38.5% of our sample observations.

3.5 Empirical Methodology The adopted empirical methodology is designed to examine the effects of banking sector reform and competition on bank stability. The analysis is performed at the bank level and the model is defined as follows:

4445 0.188

4445 2.866

3874 6.331

3886 7.072

3854 1.614

Bank reform index

Activity restriction

Capital regulatory index

Declaring insolvency power

4445 10.340

Inflation

Crisis

4445 0.032

4445 4.026

0.795

4445 −0.553

log DPS

GDP growth

1.118

4445 −4.244

log LLP

FA

0.520

4445 2.473

log EQ

0.091

0.658

1.572

1.517

0.616

0.390

9.098

4.835

0.033

3.496

4445 0.158

4445 14.389

0.530

log TA

4042 0.311

H−statistic

13.249 0.207

HHI

3492 9.092

4445 0.295

24.013

4438 15.400

Lerner index

2.784

4367 1.823

NPL

1.505

4445 2.419

STD

log Zscore    sd R OA    sd R OE

Mean

N

Variable

Panel A: Data statistics

Table 3.2 Summary statistics

0

3

3

1.000

4

10

11

4.000

1.000

75.201

−18.930 0.000

0.365 34.500

0.000

0.000

−14.800

−9.531

1.099

4.416

−1.139

0.986

0.848

2.446

13.663

0.135

0.481

0.253

4.330

7.596

0.932

2.581

2.769

15.562

0.162

0.725

0.430

11.145

16.118

1.998

3.343

Median P75

1

6

5

2.670

0.000

3.367

1.785

0.010

2

8

6

2.670

0.000

9.075

5.017

0.022

(continued)

2

8

7

3.330

0.000

15.170

7.296

0.043

−0.705 −0.250 −0.085

−4.883 −4.213 −3.520

2.146

12.079

0.124

0.003

0.124

114.810 1.315 0.678

27.451

−10.331

0.439

1.696

P25

164.821 3.872

5.152

0.068

0.153

0.000

0.000

0.338

19.010

12.130

−5.848 0.000

Max

Min

3.5 Empirical Methodology 63

2521 4.851

3252 0.916

Financial statement transparency

Deposit insurance

66

114

82

138

160

271

58

46

88

138

90

Armenia

Azerbaijan

Belarus

Bosnia and Herzegovina

Bulgaria

Croatia

Czech Republic

Estonia

Georgia

Hungary

Kazakhstan

1.48

2.02

3.10

1.98

1.03

1.30

6.10

3.60

3.10

1.84

2.56

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

260

648

564

438

184

170

123

108

94

75

58

45

Frequency

Year

48

Albania

1.08

Sample distribution by year

Frequency Percentage

Country

0.277

0.855

1.065

0.481

STD

Sample distribution by country

Panel B: Sample distribution by year and by country

3980 1.352

3712 6.957

Private monitoring index

Mean

Diversification index

N

Variable

Panel A: Data statistics

Table 3.2 (continued)

0

2

5

0

Min

5.85

14.58

12.69

9.85

4.14

3.82

2.77

2.43

2.11

1.69

1.30

1.01

Percentage

1

6

9

2

Max

1

5

6

1

P25

1

5

7

1

(continued)

1

5

8

2

Median P75

64 3 Banking Sector Reform, Competition, and Bank Stability

4445

Uzbekistan

Total

100

2.05

8.01

4.43

2.88

38.47

4.45

5.62

1.71

2.41

0.52

N

Mean

Total

2016

2015

2014

2013

2012

2011

2010

STD

4445

178

261

208

232

259

268

272

Min

100

4.00

5.87

4.68

5.22

5.83

6.03

6.12

Max

P25

Median P75

This table reports summary statistics of the variables for the full sample, and sample distribution according to year and country. Our sample period is 1998–2016. All data are inflation-adjusted and bank-specific variables are winsorized at the 1st and 99th percentile level to reduce the influence of outliers

356

91

Ukraine

197

Slovenia

198

Romania

1710

250

Poland

128

76

Macedonia (FYROM)

Slovakia

107

Latvia

Russia

23

Kyrgyz Republic

Variable

Panel A: Data statistics

Table 3.2 (continued)

3.5 Empirical Methodology 65

66

3 Banking Sector Reform, Competition, and Bank Stability

bank riski jt = α0 + α1 competition i jt + α2 bank re f orm indexi jt +  × bank_contr olsmi jt +  × countr y_contr olsn jt + αi + μt + εi jt

(3.6)

where bank riski jt indicates bank i’s stability in country j in year t and is measured as the logarithm of the Z-score (log Zscore) as defined in Eq. (3.1). The main explanatory variables of interest are competition, measured by the Lerner index (Lerner index), and banking sector reform, measured by the bank reform indicator (Bank reform index) provided in the EBRD’s transition reports. The bank-specific control variables include bank size (log TA), the equity ratio (log EQ), the loan loss provision ratio (log LLP), the net interest income ratio (NIM), the total customer deposits ratio (log DPS), and fixed assets (FA). The country-level control variables include GDP growth (GDP growth), the inflation rate (Inflation), and Crisis. We also include bank FE (αi ) to control for the unobserved time-invariant heterogeneity across banks and yearly FE (μt ) to control for the macroeconomic factors and monetary policy, which may vary over time. We are also concerned about whether bank regulation changes the relationships among bank competition, reform, and stability. We thus employ Eq. (3.7) to examine how bank regulation affects risk and whether the relationship between competition and stability is strengthened or mitigated: bank riski jt = α0 + α1 competition i jt + α2 bank re f orm indexi jt + α3 bank r egulaiton jt + α4 competition i jt ∗ bank re f orm index jt + α5 competition i jt + bank r egulaiton jt +  × bank_contr olsmi jt + × countr y_contr olsn jt + αi + μt + εijt

(3.7)

All the variables are as defined in Eq. (3.6) except that we include the interaction term between the bank reform index and Lerner index, bank regulatory variables, and interaction terms between the bank regulatory variables and Lerner index. Equations (3.6) and (3.7) are estimated based on two types of estimation methods. First, we employ the bank FE model with robust standard errors and include the year dummy variables to control for the heterogeneity and changes in unobservable features. Second, we also consider potential endogeneity bias: the level of bank stability could affect bank competition and the degree of banking sector reform may affect not only bank stability, but also the competitiveness of banks. To address this potential endogeneity bias, we Anginer et al. (2014b) and use the two-stage least squares (2SLS) estimator. Based on previous studies (e.g., Leroy & Lucotte, 2017; Schaeck & Cihák, 2008), we use the first lag of the Lerner index, the first lag of the bank reform index, and personnel expenses (w2 ), which is a proxy of a bank’s efficiency, as instrumental variables (IVs). The first stage F-test and Hansen’s J test are respectively employed to test for the relevance and validity of the instruments of the Lerner index.

3.6 Empirical Analysis

67

3.6 Empirical Analysis 3.6.1 Basic Results In this section, we investigate the impact of bank competition and reform on bank stability in transition countries, after controlling for the effects of the bank-specific and country-level characteristics. Table 3.3 presents the results for both the FE regressions (columns (1)–(3)) and the IV estimations (columns (4)–(6)). Columns (1) and (4) of Table 3.3 present the results including only the Lerner index and country-level explanatory variables. The results confirm that the Lerner index is negatively and statistically significant at 5% for the different estimation methods. After including the bank-specific variables in columns (2) and (5), the results remain unchanged. This finding is consistent with those of previous studies that focus on banks from different regions and also suggest that competition enhances bank stability (e.g., Anginer et al., 2014a; Schaeck et al., 2009; Yeyati & Micco, 2007).9 Columns (3) and (6) of Table 3.3 report the results for the bank reform index. The results confirm the significantly positive correlation for the bank reform index, suggesting that higher banking sector development is positively related to bank stability in transition countries. The results for the bank-specific variables show that banks with higher equity capital exhibit higher stability, while a higher loan loss provision leads to bank fragility. We also account for the effect of bank funding (log DPS), size (log TA), and fixed assets (FA), but none of these factors has a significant effect on bank stability. For the country-level variables, the results for GDP growth and Inflation suggest that higher GDP growth and lower inflation correspond to higher bank stability. We also conduct ananalysis  based  onthe different measures of risk and compe  tition. First, we use sd R OA , sd R OE , and NPL as alternative measures of bank risk. Panel A of Table 3.4presents  the results. We  find a positive relationship between   the Lerner index and sd R OA as well as sd R OE , suggesting that higher market power increases the volatility of bank profit and leads to bank fragility. Higher market power also corresponds to higher non-performing loans. The results for the bank reform  index  confirm that financial development leads to a significant reduction  in sd R OA and NPL, providing further evidence that bank reform is positively associated with bank stability. In Panel B of Table 3.4, we test whether our findings are consistent under alternative measures of bank competition. In columns (1) and (2), we first use the H-statistic as an alternative measure and find that it is positively and statistically significantly related to the Z-score, which confirms that competition as measured by the H-statistic 9 Yeyati

and Micco (2007) find a positive link between bank risk (as measured by the Z-score) and competition (as captured by the H-statistic) based on a sample of commercial banks from eight Latin American countries between 1993 and 2002. Schaeck et al. (2009) use the data from 31 systemic banking crises in 45 countries in 1980–2005 and find that competition reduces the likelihood of a crisis. Anginer et al. (2@014a) use a sample of publicly traded banks from 63 countries between 1997 and 2009 and find a positive relationship between competition and systemic stability.

68

3 Banking Sector Reform, Competition, and Bank Stability

Table 3.3 Bank reform, competition and stability: basic analysis Variables

(1)

(2)

(3)

(4)

(5)

(6)

Lerner index

−0.314**

−0.335***

−0.335***

−0.659***

−0.648***

−0.644***

(−2.34)

(−2.69)

(−2.69)

(−2.65)

(−2.76)

(−2.76)

Bank reform index

0.0726*

0.217*

(1.85)

(1.72)

Bank-specific variables log TA

0.0342

0.0366

0.118

0.0934

(0.90)

(0.95)

(0.97)

(0.77)

1.170***

1.172***

1.208***

1.220***

(15.38)

(15.38)

(6.00)

(6.11)

log LLP

−0.180***

−0.180***

−0.243***

−0.239***

(−6.84)

(−6.84)

(−4.46)

(−4.48)

log DPS

−0.0433

−0.0425

0.0444

0.0431

(−0.73)

(−0.72)

(0.33)

(0.32)

−0.887

−0.833

1.179

1.276

(−0.77)

(−0.73)

(0.56)

(0.60) −0.0182

log EQ

FA Country control variables GDP growth

0.0322***

0.0142*

0.0147*

0.00329

−0.0168

(3.92)

(1.88)

(1.95)

(0.21)

(−1.07)

(−1.15)

Inflation

−0.00421

−0.00675*

−0.00702*

−0.0152**

−0.0155**

−0.0169***

(−1.06)

(−1.81)

(−1.86)

(−2.27)

(−2.58)

(−2.79)

Crisis

−0.0402

0.00486

0.0102

−0.203

−0.243

−0.233

(−0.43)

(0.06)

(0.12)

(−1.21)

(−1.59)

(−1.53)

2.216***

−2.029**

−2.265**

(9.26)

(−2.36)

(−2.45)

Year fixed effects

Yes

Yes

Yes

Yes

Yes

Yes

Bank fixed effects

Yes

Yes

Yes

Yes

Yes

Yes

N

4445

4445

4445

1387

1387

1387

Adjusted R-squared

0.0248

0.159

0.160

0.293

0.148

0.151

F-statistic of first stage

12.11

15.31

14.98

Hansen’s J statistic

8.363

6.694

7.036

Constant

(continued)

3.6 Empirical Analysis

69

Table 3.3 (continued) Variables

(1)

(2)

(3)

Hansen’s J statistic (p-value) Estimation method

FE

FE

FE

(4)

(5)

(6)

0.193

0.213

0.350

IV

IV

IV

This table examines the impact of bank reform and competition on bank stability. The dependent variable is the logarithm of Z-score as defined in Eq. (3.1). Our sample period is 1998–2016. Lerner index is estimated based on Eq. (3.2) and employed as a measure of bank competitive indicator. Bank reform index is a bank reform indicator. Bank-level controls include log TA, log TA * log TA, log EQ, log LLP, log DPS, and FA. Country control variables include GDP growth, and Inflation. All variables are defined in Table 3.1. Reported t-statistics (in parentheses) are based on standard errors that are heteroskedasticity consistent and clustered at the bank level. The Hansen’s J test evaluates the joint validity of instruments used. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively.

increases bank stability. We also use the concentration ratio (i.e., the HHI) as another measure of competition, but find no significant impact on the Z-score. This is consistent with the findings of previous studies (e.g., Anginer et al., 2014a; Claessens & Laeven, 2004) that the level of competition is not necessarily related to market structure.

3.6.2 Impact of Bank Regulation on the Relationship Between Competition and Stability The results of our basic analysis confirm that competition and bank reform increase bank stability in transition countries. Further, studies of the causes of the global financial crisis argue that deregulation and excessive competition are the main factors that lead to financial system fragility (e.g., Brunnermeier, 2009; Llewellyn, 2008; Milne, 2009). Moreover, previous analyses of the relationship between bank regulation and competition have also highlighted the importance of the former in determining a bank’s competitive conditions (Angelini & Cetorelli, 2003; Claessens & Laeven, 2004; Demirguc-Kunt et al., 2004). In this section, we extend our analysis to examine whether the relationship between bank competition and stability varies considering the difference in national bank regulations. In Table 3.5, we investigate the impact of the bank regulatory variables on risk. Both the bank-specific variables and the country-level control variables are included as explanatory variables but not reported for brevity. We test the impact of the following bank regulation variables: Activity restriction, Capital regulatory index, Declaring insolvency power, Diversification index, Private monitoring index, Financial statement transparency, and Deposit insurance.

70

3 Banking Sector Reform, Competition, and Bank Stability

Table 3.4 Bank reform, competition and stability: alternative measures Panel A: Alternative measures of bank stability (3)

Variables

(1) (2)       sd R OA sd R OE

NPL

(4) (5) (6)       sd R OA sd R OE NPL

Lerner index

0.566**

5.843***

1.885**

0.751*

8.721**

(2.29)

(2.94)

(2.42)

(1.66)

(2.57)

(2.73)

Bank reform index

−0.106*

−0.0238

−1.301**

−0.419**

−1.083

−2.057**

(−1.93)

(−0.03)

(−2.35)

(−2.03)

(−0.58)

(−2.18)

Constant

8.352***

86.66***

30.82***

5.893***

(5.22)

(4.87)

(3.58)

Bank-specific variables

Yes

Yes

Yes

Yes

Yes

Yes

Country control variables

Yes

Yes

Yes

Yes

Yes

Yes

Year fixed effects

Yes

Yes

Yes

Yes

Yes

Yes

Bank fixed effects

Yes

Yes

Yes

Yes

Yes

Yes

N

4367

4438

3492

1351

1387

1306

Adjusted R-squared

0.342

0.707

0.233

0.510

0.727

0.253

F-statistic of first stage

12.73

14.40

10.67

Hansen’s J statistic

5.492

5.302

5.86

Hansen’s J statistic (p-value)

0.482

0.506

0.293

IV

IV

IV

Estimation method

FE

FE

FE

Panel B: Alternative measures of bank competition (1)

(2)

(3)

(4)

Variables

log Zscore log Zscore log Zscore log Zscore

H-statistic

0.00581**

0.0110**

(2.28) HHI Bank reform index Constant

(2.33) 0.301

−0.464

(0.92)

(−0.35)

0.0718**

0.0714

0.196*

0.197**

(2.15)

(1.23)

(1.74)

(2.54)

−3.345**

−2.644***

(−2.31)

(−2.59)

Bank-specific variables

Yes

Yes

Yes

Yes

Country control variables

Yes

Yes

Yes

Yes

Year fixed effects

Yes

Yes

Yes

Yes (continued)

3.6 Empirical Analysis

71

Table 3.4 (continued) Panel A: Alternative measures of bank stability (1)

(2)

(3)

(4)

Bank fixed effects

Yes

Yes

Yes

Yes

N

4042

4445

1297

1387

Adjusted R-squared

0.157

0.158

0.141

0.141

F-statistic of first stage

12.94

14.39

Hansen’s J statistic

9.019

6.835

Hansen’s J statistic (p-value)

0.173

0.336

IV

IV

Estimation method

FE

FE

(5)

(6)

This table examines the impact of bank reform and competition on bank stability. Our sample period is 1998–2016. Bank Reform Index is a bank reform indicator. Bank-level controls include log TA, log EQ, log LLP, log DPS, and FA. Country control variables include GDP growth, Inflation, and Crisis. All variables are defined in Table 3.1. Bank-level control variables and country-level variables are also included in the regressions but not reported for brevity. Reported t-statistics (in parentheses) are based on standard errors that are heteroskedasticity consistent and clustered at the bank level. The Hansen’s J test evaluates the joint validity of instruments used. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

The results for Activity restriction indicate that activity restrictions increase bank stability, as they reduce the potential channels of credit risk contagion (Anginer & Demirguc-Kunt, 2014). More stringent capital requirements can provide banks with a higher capital buffer to absorb losses, minimize the risk contagion among banks, and provide more incentives for investors to monitor the bank’s risk-taking (Allen et al., 2011; Repullo, 2004). Thus, higher capital regulatory stringency can reduce the probability of bank insolvency and increase bank stability, with the results for Capital regulatory index providing empirical support for this conclusion. Further, supervisory authorities that have greater power to declare a deeply troubled bank insolvent can reduce the risk contagion, while explicit guidelines for asset diversification can help banks increase the efficiency of asset management, reduce the potential risk introduced by non-performing loans, and diversify potential risks. The results for Declaring insolvency power and Diversification index confirm that if the supervisory authority in a country has greater power to declare insolvency and provides more explicit guidelines for diversification, the banking system would tend to be more stable. We also expect more private monitoring to be positively correlated with bank stability, as this provides more incentives for the private monitoring of firms. The results in columns (5) and (10) are consistent with our expectation. Considering the impact of external governance, we expect better financial statement transparency to be associated with lower bank fragility because information asymmetry may serve as a channel through which risks and shocks can be propagated in the banking sector (Hong & Stein, 2003). Moreover, better financial statement transparency

Constant

Deposit insurance

Financial statement transparency

Private monitoring index

Diversification index

Declaring insolvency power

Capital regulatory index

Activity restriction

(8)

(1.59)

(1.67)

0.120*

(−2.78) (1.22)

0.136

(−2.47)

(9)

(10)

(11)

(12)

(13)

(−3.39)

(−3.01)

(−3.21)

(−3.71)

(−3.77)

(−3.30)

(−3.73) (continued)

(−1.88)

−3.246*** −2.911*** −2.963*** −2.214*** −3.970*** −4.195*** −2.240***

−0.152*

(1.24)

0.138

(−2.64)

(−0.20)

(2.02)

(14) −0.522***

−0.0238

0.249**

(1.39)

(3.46)

(1.49)

0.243

(−2.56)

0.136

0.232***

(2.99)

(2.15)

(1.63)

0.185

(−2.65)

0.146***

0.124**

(2.20)

(1.69)

(1.20)

0.133

(−2.64)

0.178**

0.1000*

(0.99)

(2.07)

(1.50)

0.167

(−2.41)

0.0590

0.00183**

(2.08)

(1.46)

0.162

(−2.72)

0.00188**

(2.41)

(1.73)

0.185

(−2.83)

0.0913**

(1.31)

0.131*

(−2.75)

(2.32)

(1.78)

0.0915

(7)

0.0697**

Bank regulation variables

(1.57)

(1.20)

0.115*

(−2.37)

(6)

(−2.78)

0.106

(−2.91)

0.0835

(5)

(−2.84)

Bank reform index

(4) −0.368*** −0.375*** −0.506*** −0.369*** −0.486** −0.537*** −0.472** −0.532*** −0.518*** −0.612**

(3)

Lerner index

(2)

(1)

−0.376*** −0.386*** −0.316**

Variables

Table 3.5 Relationship between bank competition, reform, regulation and stability

72 3 Banking Sector Reform, Competition, and Bank Stability

FE

3980

FE

0.157

3712

FE

0.153

FE

0.172

2521

Yes

FE

0.154

3252

Yes

Yes

0.137

1868

Yes

Yes

Yes

IV

0.454

5.436

12.32

0.134

1871

Yes

Yes

Yes

Yes

(9)

IV

0.245

1.562

11.76

0.122

1827

Yes

Yes

Yes

Yes

(10)

IV

0.816

7.021

12.43

0.135

1871

Yes

Yes

Yes

Yes

(11)

IV

0.135

2.619

12.36

0.146

1807

Yes

Yes

Yes

Yes

(12)

IV

0.624

8.219

11.37

0.150

1441

Yes

Yes

Yes

Yes

(13)

IV

0.420

3.901

8.959

0.134

1870

Yes

Yes

Yes

Yes

(14)

This table examines the impact of bank reform and competition on bank stability. The dependent variable is the logarithm of Z-score as defined in Eq. (3.1). Our sample period is 1998–2016. Lerner index is estimated based on Eq. (3.2) and employed as a measure of bank competitive indicator. Bank reform index is a bank reform. Bank-level controls include log TA, log EQ, log LLP, log DPS, and FA. Country control variables include GDP growth, Inflation, and Crisis. Bank regulation variables include Activity restriction, Capital regulatory index, Declaring insolvency power, Diversification index, Private monitoring index, Financial statement transparency, and Deposit insurance. All variables are defined in Table 3.1. Bank-level control variables and country-level variables are also included in the regressions but not reported for brevity. Reported t-statistics (in parentheses) are based on standard errors that are heteroskedasticity consistent and clustered at the bank level. The Hansen’s J test evaluates the joint validity of instruments used. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

FE

0.137

3854

Yes

Yes

Yes

Yes

(8)

IV

FE

0.151

3886

Yes

Yes

Yes

Yes

(7)

Estimation method

0.154

Adjusted R-squared

Yes

Yes

Yes

Yes

(6)

0.355

3874

N

Yes

Yes

Yes

Yes

(5)

Hansen’s J statistic (p-value)

Yes

Bank fixed effects

Yes

Yes

Yes

(4)

3.660

Yes

Year fixed effects

Yes

Yes

(3)

Hansen’s J statistic

Yes

Country Control variables

Yes

(2)

12.76

Yes

Bank-specific variables

F-statistic of first stage

(1)

Variables

Table 3.5 (continued)

3.6 Empirical Analysis 73

74

3 Banking Sector Reform, Competition, and Bank Stability

can significantly lower the information asymmetry between a bank’s stakeholders. The coefficient of Financial statement transparency in the IV estimation is significantly positive, confirming that better financial statement transparency increases bank stability, which is consistent with our expectation. Deposit insurance is negatively correlated with the Z-score, which is consistent with the findings of DeLong and Saunders (2011) and Fu et al. (2014), confirming that deposit insurance coverage makes depositors less likely to enforce market discipline and intensifies moral hazard problems. Table 3.6 examines the effect of bank regulation on the relationship between bank competition and stability, after controlling for the impact of bank reform. Columns (1) and (9) in Table 3.6 include the interaction term between the bank reform index and Lerner index and test whether the relationship between competition and stability varies when considering banking sector reform. The coefficients of the interaction term are significantly negative, which indicates that the effect of bank competition in increasing bank stability is more pronounced when transition countries experienced higher bank reforms. Columns (2)–(8) and (10)–(16) in Table 3.6 present the results of the interaction terms between the bank regulation variables and Lerner index, after controlling for the impact of bank reform. The coefficients of the interaction terms for Activity restrictions and Diversification index are significantly positive (columns (2) and (9) and columns (5) and (13), respectively), indicating that the effect of competition in reducing bank risk is less pronounced if the bank is exposed to higher activity restrictions and provided with more explicit guidelines for diversification. The negative coefficients of the interaction term for Capital regulatory index in columns (3) and (11) provide evidence that the benefit of bank competition in reducing bank instability is greater in a banking system that has more stringent capital requirements.

3.6.3 Robustness Checks We also perform robustness checks based on different specifications and subsamples from our main models.10 First, following Koetter et al. (2012), we employ the efficiency-adjusted Lerner index as an alternative measure of bank competition. Our main findings are consistent (see columns (1) and (2) of Table 3.7). Second, we introduce country FE into the regression to control for the heterogeneity between countries instead of bank FE, while year FE are also considered. The results remain unchanged (see columns (3) and (4) of Table 3.7). Third, we adopt different sample selection criteria. Columns (5) and (6) of Table 3.7 report the results including only countries with more than 100 observations and columns (7) and (8) report the findings excluding banks in Russia. The results show that our findings are consistent under these alternative sample selection criteria. 10 For brevity, we only report the results for the FE estimations. The results for the IV estimations are consistent with those findings and are available upon request.

Diversification index * Lerner index

Diversification index

Declaring insolvency power * Lerner index

Declaring insolvency power

Capital regulatory index * Lerner index

Capital regulatory index

Activity restriction * Lerner index

Activity restriction

Lerner index * Bank reform index

(1.81)

0.0677*

(1.88)

0.0673*

(−2.04)

−0.00319**

(0.01)

0.000297

(0.36)

(−0.13)

(−2.27)

(2.51)

(0.03)

0.00527

(0.69)

0.0527

(0.39)

0.0439

(2.24)

(−1.03) 0.0993**

(−1.74)

(4) −0.486

0.0862** 0.0424

(0.75)

(3) −0.532*

−0.0298** −0.0152

(2.09)

(−2.34) 0.0525

(−1.80) 0.0637**

Lerner index

Bank reform index

(2) −0.801**

(1) −0.268*

Variables

Table 3.6 The impact of bank reform and regulation on stability: through the channel of competition (5)

0.425*

(0.39)

0.0449

(0.17)

0.0204

(1.03)

0.0731

(−2.12)

−1.049**

(6)

(0.58)

0.0724

(1.47)

0.112

(−2.29)

−0.252**

(7)

(0.13)

0.0259

(1.22)

0.145

(−1.93)

−1.774*

(8)

(continued)

(0.41)

0.0486

(1.32)

0.0971

(−0.73)

−0.380

3.6 Empirical Analysis 75

Yes Yes Yes Yes 4313 0.161

Country control variables

Year fixed effects

Bank fixed effects

N

Adjusted R-squared

0.154

3765

Yes

Yes

Yes

Yes

0.150

3777

Yes

Yes

Yes

Yes

(−2.85)

(−3.23)

(−2.26)

(3)

−4.918*** −4.468***

(2)

−3.277**

(1)

Bank-specific variables

Constant

Deposit insurance * Lerner index

Deposit insurance

Financial statement transparency * Lerner index

Financial statement transparency

Private monitoring index * Lerner index

Private monitoring index

Variables

Table 3.6 (continued) (1.74)

(5)

(−0.44)

−0.0531

(2.69)

0.170***

(6)

(1.02)

0.228

(0.93)

0.104

(7)

(−0.48)

−0.186

(0.06)

0.0105

(8)

0.136

3732

Yes

Yes

Yes

Yes

(−3.49)

0.157

3870

Yes

Yes

Yes

Yes

(−2.72)

0.153

3620

Yes

Yes

Yes

Yes

(−3.43)

0.175

2434

Yes

Yes

Yes

Yes

(−3.09)

(continued)

0.154

3252

Yes

Yes

Yes

Yes

(−3.36)

−3.234*** −3.878*** −5.741*** −6.481*** −4.662***

(4)

76 3 Banking Sector Reform, Competition, and Bank Stability

Declaring insolvency power

Capital regulatory index * Lerner index

Capital regulatory index

Activity restriction * Lerner index

Activity restriction

Lerner index * Bank reform index (−0.72)

(−2.12)

(2.60)

0.199***

(1.03)

0.0475

(1.07) −0.127

(1.27) −0.0215**

(−1.89)

−0.103*

(0.52)

0.0214

(−0.16)

−0.0280

(1.39)

(0.28)

0.272

(11)

0.154

FE

(3)

0.120

(−1.62)

(−1.84) 0.142

−1.353

−0.463*

Lerner index

Bank reform index

(10)

FE

(2)

(9)

FE

(1)

Variables

Estimation method

Hansen’s J statistic (p-value)

Hansen’s J statistic

F-statistic of first stage

Variables

Table 3.6 (continued)

(0.69)

0.0824

(0.31)

0.0519

(1.53)

0.169

(−1.99)

−0.738**

(12)

FE

(4)

(−0.33)

−0.0568

(1.16)

0.128

(−1.74)

−1.338*

(13)

FE

(5)

(0.02)

0.00349

(1.73)

0.186*

(−1.34)

−1.782

(14)

FE

(6)

(−0.07)

−0.0188

(1.43)

0.233

(−2.32)

−2.041**

(15)

FE

(7)

(continued)

(0.01)

0.00165

(1.24)

0.138

(−0.17)

−0.120

(16)

FE

(8)

3.6 Empirical Analysis 77

Yes Yes

Country control variables

(9)

Bank-specific variables

Constant

Deposit insurance * Lerner index

Deposit insurance

Financial statement transparency * Lerner index

Financial statement transparency

Private monitoring index * Lerner index

Private monitoring index

Diversification index * Lerner index

Diversification index

Declaring insolvency power * Lerner index

Variables

Table 3.6 (continued)

Yes

Yes

(10)

Yes

Yes

(11)

(12)

Yes

Yes

(0.25)

0.0704

Yes

Yes

(1.94)

0.683*

(−0.30)

−0.0447

(13)

Yes

Yes

(0.97)

0.173

(2.21)

0.190**

(14)

Yes

Yes

(0.95)

0.292

(1.27)

0.169

(15)

(continued)

Yes

Yes

(−0.86)

−0.445

(−0.24)

−0.0577

(16)

78 3 Banking Sector Reform, Competition, and Bank Stability

Yes 1871 0.134 14.45 4.742 0.315 IV

Year fixed effects

Bank fixed effects

N

Adjusted R-squared

F-statistic of first stage

Hansen’s J statistic

Hansen’s J statistic (p-value)

Estimation method

(10)

IV

0.424

3.867

12.52

0.138

1868

Yes

Yes

(11)

IV

0.208

5.882

11.68

0.134

1871

Yes

Yes

(12)

IV

0.831

1.478

11.18

0.122

1827

Yes

Yes

(13)

IV

0.0963

7.875

11.83

0.139

1871

Yes

Yes

(14)

IV

0.559

2.992

11.95

0.146

1807

Yes

Yes

(15)

IV

0.0579

9.133

11.13

0.151

1441

Yes

Yes

(16)

IV

0.424

3.868

11.70

0.134

1870

Yes

Yes

This table examines the impact of bank reform and competition on bank stability. The dependent variable is the logarithm of Z-score as defined in Eq. (3.1). Our sample period is 1998–2016. Lerner index is estimated based on Eq. (3.2) and employed as a measure of bank competitive indicator. Bank reform index is a bank reform indicator. Bank-level controls include log TA, log TA * log TA, log EQ, log LLP, log DPS, and FA. Country control variables include GDP growth, Inflation, and Crisis. Bank regulation variables include Activity restriction, Capital regulatory index, Declaring insolvency power, Diversification index, Private monitoring index, Financial statement transparency, and Deposit insurance. All variables are defined in Table 3.1. Bank-level control variables and country-level variables are also included in the regressions but not reported for brevity. Reported t-statistics (in parentheses) are based on standard errors that are heteroskedasticity consistent and clustered at the bank level. The Hansen’s J test evaluates the joint validity of instruments used. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

(9) Yes

Variables

Table 3.6 (continued)

3.6 Empirical Analysis 79

Bank reform index * foreign dummy

Lerner index * foreign dummy

Lerner index * bank reform index * Commercial dummy

Bank reform index * commercial dummy

Lerner index * commercial dummy

Commercial dummy

Lerner index * bank reform index

Bank reform index

Lerner index

Efficiency-adjusted Lerner index * Bank reform index

(−2.59)

0.545*** (3.25)

(1.84)

(−2.72)

−0.965***

0.125*

(−2.19)

Efficiency-adjusted Lerner index

(2)

(1) −0.0510** −2.752***

Variables

Efficiency-adjusted Lerner index

(1.04)

0.0729

(−2.38)

(−2.40)

(0.43)

0.0433 −0.0603**

(0.46)

0.0462

(−1.72)

−0.0298**

(1.97)

0.0637***

(−1.98)

(−2.43)

(6)

−0.228**** −0.0590*

(5)

(−2.41)

(4)

Countries with more than 100 observations

−0.336** −0.268**

(3)

Country fixed effects

Table 3.7 Impact of bank competition and reform on bank stability: robustness checks

(continued)

(1.90)

0.124**

(−2.69)

−0.453***

(7)

Russia excluded

80 3 Banking Sector Reform, Competition, and Bank Stability

Yes No 3738 0.138 FE

Country fixed effects

N

Adjusted R-squared

Estimation method

Yes

Year fixed effects

Bank fixed effects

Yes Yes

Country control variables

FE

0.142

3738

No

Yes

Yes

Yes

Yes

(−3.28)

FE

0.647

4440

Yes

No

Yes

Yes

Yes

(−1.93)

−3.339*** −2.046*

(3)

(−2.15)

(2)

FE

0.651

4313

Yes

No

Yes

Yes

Yes

(−1.88)

−3.056*

(4)

Country fixed effects

−2.051**

(1)

Efficiency-adjusted Lerner index

Bank-specific variables

Constant

Lerner index * bank reform index * EU entry

Bank reform index * EU entry

Lerner index * EU entry

EU entry

Lerner index * bank reform index * foreign dummy

Variables

Table 3.7 (continued)

FE

0.150

2567

No

Yes

Yes

Yes

Yes

(−1.44)

−2.283

(5)

FE

0.145

2524

No

Yes

Yes

Yes

Yes

(−1.65)

−2.650*

(6)

Countries with more than 100 observations

FE

0.181

2735

No

Yes

Yes

Yes

Yes

(continued)

(−2.04)

−2.196**

(7)

Russia excluded

3.6 Empirical Analysis 81

Lerner index * commercial dummy

Commercial dummy

Lerner index * Bank reform index

Bank reform index

Lerner index

Efficiency-adjusted Lerner index * Bank reform index

Efficiency-adjusted Lerner index

Variables

Table 3.7 (continued)

(2.40)

(0.44) 0.00626

2.936**

0.154

(1.76)

(−2.34)

(2.22)

(1.24) −0.304*

0.819**

0.0719

−0.0575**

(2.46)

0.123**

(−2.70)

(−0.16)

−0.336*** −0.260

(−1.76)

(10)

−0.302*

(9)

Russia excluded Commercial banks (8)

(12)

(2.25)

0.0726**

(−2.69)

(−2.02)

−0.0029**1

(0.53)

0.0335

(−1.67)

−0.335*** −0.233*

(11)

Foreign banks (14)

(1.56)

0.0886

(−2.62)

(continued)

(−0.15)

−0.0184

(0.97)

0.0621

(−0.66)

−0.327*** −0.234

(13)

EU entry

82 3 Banking Sector Reform, Competition, and Bank Stability

Bank reform index * EU entry

Lerner index * EU entry

EU entry

Lerner index * bank reform index * foreign dummy

Bank reform index * Foreign dummy

Lerner index * foreign dummy

Lerner index * bank reform index * commercial dummy

Bank reform index * commercial dummy

Variables

Table 3.7 (continued) (9)

(−0.57)

−0.351

(−2.12)

−0.791**

(0.00)

(10)

Russia excluded Commercial banks (8)

(11)

(−0.43)

−0.109

(1.28)

0.153

(−0.38)

−0.321

(12)

Foreign banks

(−0.02)

(continued)

0.0936

(−1.76)

−2.412*

−0.0111 (1.44)

(14)

0.196

(13)

EU entry

3.6 Empirical Analysis 83

2623 0.189 FE

N

Adjusted R-squared

Estimation method

FE

0.160

4445

No

Yes

Yes

Yes

Yes

(11)

FE

0.165

4313

No

Yes

Yes

Yes

Yes

(−3.30)

FE

0.160

4445

No

Yes

Yes

Yes

Yes

(−2.45)

FE

0.162

4313

No

Yes

Yes

Yes

Yes

(−2.28)

−3.291**

(12)

Foreign banks

−5.980*** −2.265**

(10)

(1.64)

0.641

(0.64)

(14)

FE

0.161

4445

No

Yes

Yes

Yes

Yes

(−2.71)

FE

0.163

4313

No

Yes

Yes

Yes

Yes

(−2.27)

−2.470*** −3.303**

(13)

EU entry

This table examines the impact of bank reform and competition on bank stability according to different estimation specifications and subsamples. The dependent variable is the logarithm of Z-score as defined in Eq. (3.1). Our sample period is 1998–2016. Lerner index is estimated based on Eq. (3.2) and employed as a measure of bank competitive indicator. Bank reform index is a bank reform indicator. Bank-level controls include log TA, log EQ, log LLP, log DPS, and FA. Country control variables include GDP growth, Inflation, and Crisis. All variables are defined in Table 3.1. Reported t-statistics (in parentheses) are based on standard errors that are heteroskedasticity consistent and clustered at the bank level. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

Yes No

Country fixed effects

Yes

Year fixed effects

Bank fixed effects

Yes Yes

(−2.40)

(−1.68)

Country control variables

−2.437**

−3.074*

(9)

Russia excluded Commercial banks (8)

Bank-specific variables

Constant

Lerner index * bank reform index * EU entry

Variables

Table 3.7 (continued)

84 3 Banking Sector Reform, Competition, and Bank Stability

3.6 Empirical Analysis

85

We then examine the relationship when considering different types of banks. The results for commercial banks (columns (9) and (10)) show that they enjoy more stability than other types of banks and that the relationships among bank reform, competition, and stability do not significantly change, while the results for foreign banks (columns (11) and (12)) show no significant differences between foreign banks and domestic banks, although the former are regulated not only by the home country but also by host countries.11 Finally, we test whether transition countries’ decision to join the European Union (EU) significantly changes the relationships among bank reform, competition, and stability.12 We include the binary variable EU entry, which equals one after the country joined the EU, to examine the impact of EU entry. The results in columns (13) and (14) suggest that EU entry increases bank stability and that the positive relationship between bank competition and stability becomes more pronounced after joining the EU.13

3.7 Conclusion This study tests the influence of banking sector reform and competition on bank stability based on unbalanced data from 22 transition countries between 1998 and 2016. The initial results confirm the negative relationship between market power and bank stability and the positive relationship between bank reform and stability. These findings indicate that competition incentivizes banks to diversify risk and thus reduces bank instability and that banks experience an improvement in financial stability after banking sector reforms. In terms of bank regulations, both higher activity restrictions and more explicit guidelines for asset diversification increase bank stability, but this positive effect significantly weakens for banks with higher market power. More stringent capital requirements in combination with higher market power increase the risk of bank insolvency. The results also show that higher declaring insolvency power, private 11 Previous

studies such as De Haas and Van Lelyveld (2006), Havrylchyk and Jurzyk (2011), and Fang et al. (2014) argue that foreign banks are not significantly sensitive to the bank reforms in host countries but are more sensitive to the wishes of parent banks in home countries and that foreign banks’ lending criteria are highly associated with the parent country’s regulation rather than that of the host country (Ongena et al., 2013). These conclusions may explain our findings: foreign banks are not sensitive to bank reform and changes in competitive conditions in host countries as well as the relationships among bank reform, competition, and stability do not vary significantly for domestic banks or foreign banks. 12 During our sample period, some transition countries joined the EU including the Czech Republic, Hungary, Poland, Slovakia, and Slovenia (2004), Bulgaria and Romania (2007), and Croatia (2013). The banks in transition countries that joined the EU are subject to changes in the macroeconomic environment, financial markets, and bank regulation. 13 We are also concerned about how the bank reform changes affect bank stability. As a robustness check, we use the change in the bank reform indicator as an alternative measure of bank reform. The results are consistent with our main findings and are available upon request.

86

3 Banking Sector Reform, Competition, and Bank Stability

monitoring, and financial statement transparency enhance bank stability, whereas having a deposit insurance scheme is negatively related to financial soundness. Our findings suggest several policy implications for transition countries. First, the negative relationship between market power and stability confirms the importance of competitive conditions in the banking system, while the positive relationship between bank reform and stability confirms the importance of bank reforms such as liberalizing interest rates, increasing lending to private enterprises, and increasing the market share of private banks. Second, the findings that activity restrictions, capital regulation, official supervisory actions, private monitoring, and higher financial statement transparency are correlated with higher bank stability provide empirical support that an appropriate bank regulation framework is crucial for ensuring bank stability. Finally, while our results suggest that higher activity restrictions reduce bank fragility, increases in activity restrictions make the financial market less contestable, which makes the banking system less competitive. This finding emphasizes the importance for a regulatory framework to strike the right balance between increasing bank stability and avoiding potential competition.

References Acharya, V. V., Hasan, I., & Saunders, A. (2006). Should banks be diversified? Evidence from individual bank loan portfolios. The Journal of Business, 79(3), 1355–1412. Acharya, V. V., & Richardson, M. (2009). Causes of the financial crisis. Critical Review, 21(2–3), 195–210. Agoraki, M.-E.K., Delis, M. D., & Pasiouras, F. (2011). Regulations, competition and bank risktaking in transition countries. Journal of Financial Stability, 7(1), 38–48. Allen, F., Carletti, E., & Marquez, R. (2011). Credit market competition and capital regulation. The Review of Financial Studies, 24(4), 983–1018. Allen, F., & Gale, D. (2004). Competition and financial stability. Journal of Money, Credit and Banking, 453–490. Angelini, P., & Cetorelli, N. (2003). The effects of regulatory reform on competition in the banking industry. Journal of Money, Credit and Banking, 35(5), 663–684. Anginer, D., & Demirguc-Kunt, A. (2014). Has the global banking system become more fragile over time? Journal of Financial Stability, 13, 202–213. Anginer, D., Demirguc-Kunt, A., & Zhu, M. (2014a). How does competition affect bank systemic risk? Journal of Financial Intermediation, 23(1), 1–26. Anginer, D., Demirguc-Kunt, A., & Zhu, M. (2014b). How does deposit insurance affect bank risk? Evidence from the recent crisis. Journal of Banking & Finance, 48, 312–321. Barth, J. R., Caprio, G., & Levine, R. (2004). Bank regulation and supervision: What works best? Journal of Financial Intermediation, 13(2), 205–248. Barth, J. R., Caprio, G., & Levine, R. (2007). Bank regulation is changing: But for better or worse. Comparative Economic Studies. Barth, J. R., Caprio, G. J., & Levine, R. (2013). Bank regulation and supervision in 180 countries from 1999 to 2011. Journal of Financial Economic Policy, 5(2), 111–219. Beck, T., De Jonghe, O., & Schepens, G. (2013). Bank competition and stability: Cross-country heterogeneity. Journal of Financial Intermediation, 22(2), 218–244. Beck, T., Demirguc-Kunt, A., & Levine, R. (2006). Bank concentration, competition, and crises: First results. Journal of Banking & Finance, 30(5), 1581–1603.

References

87

Beck, T., Levine, R., & Levkov, A. (2010). big bad banks? The winners and losers from bank deregulation in the United States. The Journal of Finance, 65(5), 1637–1667. Berger, A. N., Klapper, L. F., & Turk-Ariss, R. (2009). Bank competition and financial stability. Journal of Financial Services Research, 35(2), 99–118. Bikker, J. A., Shaffer, S., & Spierdijk, L. (2012). Assessing competition with the Panzar-Rosse model: The role of scale, costs, and equilibrium. Review of Economics and Statistics, 94(4), 1025–1044. Bonin, J. P., Hasan, I., & Wachtel, P. (2005a). Bank performance, efficiency and ownership in transition countries. Journal of Banking & Finance, 29(1), 31–53. Bonin, J. P., Hasan, I., & Wachtel, P. (2005b). Privatization matters: Bank efficiency in transition countries. Journal of Banking & Finance, 29(8), 2155–2178. Bonin, J. P., Hasan, I., & Wachtel, P. (2014). Banking in transition countries. BOFIT Discussion Paper No. 8/2014. Available at SSRN: https://ssrn.com/abstract=2416826. Boyd, J. H., & Runkle, D. E. (1993). Size and performance of banking firms: Testing the predictions of theory. Journal of Monetary Economics, 31(1), 47–67. https://doi.org/10.1016/0304-393 2(93)90016-9. Boyd, J. H., Chang, C., & Smith, B. D. (1998). Moral hazard under commercial and universal banking. Journal of Money, Credit and Banking, 30(3), 426–468. Boyd, J. H., De Nicoló, G., & Smith, B. D. (2004). Crises in competitive versus monopolistic banking systems. Journal of Money, Credit and Banking, 487–506. Brissimis, S. N., Delis, M. D., & Papanikolaou, N. I. (2008). Exploring the nexus between banking sector reform and performance: Evidence from newly acceded EU countries. Journal of Banking and Finance, 32(12), 2674–2683. Brunnermeier, M. K. (2009). Deciphering the liquidity and credit crunch 2007–2008. Journal of Economic Perspectives, 23(1), 77–100. Caminal, R., & Matutes, C. (2002). Market power and banking failures. International Journal of Industrial Organization, 20(9), 1341–1361. Carletti, E. (2008). Competition and regulation in banking. Handbook of Financial Intermediation and Banking, 126(5), 449–482. Claessens, S., & Klingebiel, D. (2001). Competition and scope of activities in financial services. The World Bank Research Observer, 16(1), 19–40. Claessens, S., & Laeven, L. (2004). What drives bank competition? Some international evidence. Journal of Money, Credit and Banking, 36(3), 563–583. Clark, E., Mare, D. S., & Radi´c, N. (2018). Cooperative banks: What do we know about competition and risk preferences? Journal of International Financial Markets, Institutions and Money, 52, 90–101. Cubillas, E., & González, F. (2014). Financial liberalization and bank risk-taking: International evidence. Journal of Financial Stability, 11, 32–48. Das, A., & Ghosh, S. (2006). Financial deregulation and efficiency: An empirical analysis of Indian banks during the post reform period. Review of Financial Economics, 15(3), 193–221. De Haas, R., & Van Lelyveld, I. (2006). Foreign banks and credit stability in Central and Eastern Europe. A panel data analysis. Journal of Banking and Finance, 30(7), 1927–1952. DeLong, G., & Saunders, A. (2011). Did the introduction of fixed-rate federal deposit insurance increase long-term bank risk-taking? Journal of Financial Stability, 7(1), 19–25. Demirgüç-Kunt, A., & Huizinga, H. (2010). Bank activity and funding strategies: The impact on risk and returns. Journal of Financial Economics, 98(3), 626–650. Demirguc-Kunt, A., Laeven, L., & Levine, R. (2004). Regulations, market structure, institutions, and the cost of financial intermediation. Journal of Money, Credit and Banking, 36(3), 593–622. Dick, A. A. (2006). Nationwide branching and it’s impact on market structure, quality, and bank performance. The Journal of Business, 79(2), 567–592. Fang, Y., Hasan, I., & Marton, K. (2014). Institutional development and bank stability: Evidence from transition countries. Journal of Banking & Finance, 39, 160–176.

88

3 Banking Sector Reform, Competition, and Bank Stability

Fernández, A. I., González, F., & Suárez, N. (2013). How do bank competition, regulation, and institutions shape the real effect of banking crises? International evidence. Journal of International Money and Finance, 33, 19–40. Fu, X. M., Lin, Y. R., & Molyneux, P. (2014). Bank competition and financial stability in Asia Pacific. Journal of Banking & Finance, 38, 64–77. Fungáˇcová, Z., & Weill, L. (2013). Does competition influence bank failures? Economics of Transition, 21(2), 301–322. Grossman, R. S. (1992). Deposit insurance, regulation, and moral hazard in the thrift industry: Evidence from the 1930s. The American Economic Review, 800–821. Havrylchyk, O., & Jurzyk, E. (2011). Profitability of foreign banks in Central and Eastern Europe. Economics of Transition, 19(3), 443–472. https://doi.org/10.1111/j.1468-0351.2010.00406.x Hellmann, T. F., Murdock, K. C., & Stiglitz, J. E. (2000). Liberalization, moral hazard in banking, and prudential regulation: Are capital requirements enough? American Economic Review, 90(1), 147–165. Hong, H., & Stein, J. C. (2003). Differences of opinion, short-sales constraints, and market crashes. The Review of Financial Studies, 16(2), 487–525. Koetter, M., Kolari, J. W., & Spierdijk, L. (2012). Enjoying the quiet life under deregulation? Evidence from adjusted Lerner indices for US banks. Review of Economics and Statistics, 94(2), 462–480. Kopecky, K. J., & VanHoose, D. (2006). Capital regulation, heterogeneous monitoring costs, and aggregate loan quality. Journal of Banking & Finance, 30(8), 2235–2255. Laeven, L., & Levine, R. (2009). Bank governance, regulation and risk taking. Journal of Financial Economics, 93(2), 259–275. Laeven, L., & Valencia, F. (2010). Resolution of banking crises: The good, the bad and the ugly. International Monetary Fund Working Paper. Laeven, L., & Valencia, F. (2018). Systemic banking crises revisited. IMF Working Paper No. 206. International Monetary Fund. Leroy, A., & Lucotte, Y. (2017). Is there a competition-stability trade-off in European banking? Journal of International Financial Markets, Institutions and Money, 46, 199–215. Llewellyn, D. T. (2008). The Northern Rock crisis: A multi-dimensional problem waiting to happen. Journal of Financial Regulation and Compliance, 16(1), 35–58. Mare, D. S., Moreira, F., & Rossi, R. (2017). Nonstationary Z-Score measures. European Journal of Operational Research, 260(1), 348–358. Marinˇc, M. (2013). Banks and information technology: Marketability vs. relationships. Electronic Commerce Research, 13(1), 71–101. Matutes, C., & Vives, X. (1996). Competition for deposits, fragility, and insurance. Journal of Financial Intermediation, 5(2), 184–216. Milne, A. (2009). The fall of the house of credit: What went wrong in banking and what can be done to repair the damage? Cambridge University Press. Ongena, S., Popov, A., & Udell, G. F. (2013). “When the cat’s away the mice will play”: Does regulation at home affect bank risk-taking abroad? Journal of Financial Economics, 108(3), 727–750. Panzar, J. C., & Rosse, J. N. (1987). Testing for “monopoly” equilibrium. The Journal of Industrial Economics, 35(4), 443–456. Perotti, E. C. (1993). Bank lending in transition economies. Journal of Banking & Finance, 17(5), 1021–1032. Repullo, R. (2004). Capital requirements, market power, and risk-taking in banking. Journal of Financial Intermediation, 13(2), 156–182. Schaeck, K., & Cihák, M. (2014). Competition, efficiency, and stability in banking. Financial Management, 43(1), 215–241. Schaeck, K., & Cihák, M. (2008). How does competition affect efficiency and soundness in banking? New empirical evidence. European Central Bank Working Paper Series No. 932.

References

89

Schaeck, K., Cihak, M., & Wolfe, S. (2009). Are competitive banking systems more stable? Journal of Money, Credit and Banking, 41(4), 711–734. Yeyati, E. L., & Micco, A. (2007). Concentration and foreign penetration in Latin American banking sectors: Impact on competition and risk. Journal of Banking & Finance, 31(6), 1633–1647.

Chapter 4

The Impact of Bank Regulation and Supervision on Competition

4.1 Overview This article empirically investigates the influence of bank regulation and supervision on the competitive landscape in banking systems. Using the information on 23 emerging economies from 1996 to 2016, we confirm that banking systems with fewer activity restrictions and (foreign) bank entry barriers are more competitive. Greater capital strictness and official supervision enhance competition in the banking industry. Our findings also highlight that greater explicit guidelines on asset diversification and deposit insurance coverage and lower private-sector monitoring are associated with more intensive bank competition. A further examination reveals that, during a bank crisis, the relationship between activity restrictions, entry barriers, diversification guidelines, and competition become more pronounced, and the positive effect of foreign bank limitations, capital strictness, official supervision, and private monitoring on competitive conditions become less effective. Finally, we divide our sample into foreign banks and domestic banks and find that foreign banks are more sensitive to official supervision and private monitoring, and less sensitive to activity restrictions, foreign bank limitations, and diversification guidelines.

4.2 Introduction Over the past two decades, because of the development of information technology, globalization, and deregulation, the banking system has undergone dramatic changes. Emerging economies have experienced considerable economic development and financial reforms, including financial liberalization, bank privatization, widening access to foreign banks, and restructuring national banking systems, which aimed to reshape competitive conditions in the banking sector. For instance, the This chapter was published in Emerging Markets Finance and Trade. © Shanghai Jiao Tong University Press 2021 S. Li, Financial Regulation and Bank Performance, Contributions to Finance and Accounting, https://doi.org/10.1007/978-981-16-3509-0_4

91

92

4 The Impact of Bank Regulation and Supervision on Competition

Chinese government recapitalized the state-owned commercial banks and reduced the number of non-performing loans (NPLs) by injecting new capital; additional changes included interest rate liberalization, fewer activity restrictions, and privatization of the banks through selling shares on the market and introducing minority foreign ownership stakes (Hasan et al., 2009). In India, the deregulation of the banking system includes interest rate liberalization, the removal of bank entry restrictions, private ownership restrictions, and activity restrictions; these structural deregulations were aimed at promoting competition. Prudential norms on NPLs and capital requirements were also implemented to strengthen financial stability (Zhao et al., 2010). These changes decreased profitability in traditional bank activities and led to massive mergers and acquisitions (M&A) among financial institutions in emerging economies. While changes in the financial market are the main driving forces for bank consolidation in developed economies, financial supervisory authorities play an important role in the bank consolidation process in emerging economies (Gelos & Roldós, 2004). The overall differences in economic development, bank consolidation, and regulation between developed economies and emerging economies created distinctive features in competitive conditions in the banking industry in emerging economies. Following the recent global financial crisis, policy makers reshaped bank regulations substantially, especially in emerging economies. The goal of this study is to examine the impact of bank regulation and supervision on competitive conditions in the banking sector in emerging economies and examine whether the unique characteristics in emerging markets shed light on the relationship between bank regulation and competition from different perspectives. Using a sample of 1629 banks in 23 emerging economies between 1996 and 2016, this study employs the Panzar and Rosse (1987) methodology and constructs the competitive variable H-statistic as a measure of competition in the banking system. We also compute the Lerner index (see also Coccorese, 2009; Koetter et al., 2012) and the Boone indicator (2008) as alternative measures of competition. We investigate how competition evolved under different types of bank regulation and supervision. As our sample period covers several banking crises in some emerging economies as well as the recent global financial crisis, we also examine whether the relationship between bank regulation and competition changed during the banking crisis. Considering the different roles played by domestic banks and foreign banks, we also study whether the impact of bank regulation on foreign banks exhibits different patterns. Our analysis shows that banking systems with higher concentration and fewer activity restrictions and entry barriers are more competitive. We also find that reducing foreign bank limitations and increasing capital strictness and official supervisory power also enhances competition in the banking sector. The results also provide evidence that competition in banking systems with fewer government-owned banks and fewer diversification guidelines tends to be more intensive. The results also confirm that higher private monitoring of banks and deposit insurance coverage significantly contribute to an increase in bank competition.

4.2 Introduction

93

During a banking crisis, the impact of activity restrictions, entry barriers, and diversification guidelines on competitive conditions become more effective, whereas foreign bank limitations, capital strictness, supervisory power, and private monitoring become less effective. To further investigate whether the relationship between bank regulation and competition varies according to the type of ownership, we divide the sample banks into two groups: foreign banks and domestic banks. Our findings suggest that foreign banks are more sensitive to official supervisory power and private monitoring and less sensitive to activity restrictions, foreign bank limitations, and diversification guidelines. Our analysis extends previous studies in several respects. First, we focus our analysis on 23 emerging economies. In emerging economies, cross-border M&As are the main reason for consolidation in the banking system, and financial authorities are deeply involved in the banking system and serve an important role in its restructuring. These factors indicate that bank consolidation and competitive conditions in emerging economies exhibit different characteristics. During our sample period, financial reforms have been implemented across the emerging economies, including financial liberalization, bank privatization, widening access to foreign banks, and restructuring the national banking systems, so as to reshape competitiveness in the banking sector. The unique institutional setting of the emerging economies can provide further insight into previous research. We also employ three different measures of competition (H-statistic, Lerner index, and Boone indicator) and incorporate bank regulatory and supervisory factors to examine the impact of bank regulation and supervision on competition in emerging economies, while controlling for a variety of bank-level and country-level characteristics that may affect bank competition. The rest of the paper is structured as follows. The third section provides a review of the previous studies on competition and then summarizes the literature on bank regulation and competition. The fourth section mainly addresses the estimation of competition variables, sample selection, data, and research methodology. The fifth section offers empirical analysis and robustness checks, and last section draws the final conclusions.

4.3 Literature Review 4.3.1 Previous Studies on Competition The previous literature includes empirical studies that measure industrial competition using structural and non-structural approaches. The structural approach is based on the traditional structure-conduct-performance (SCP) paradigm and links concentration, competition, and firm performance. SCP assumes that the market structure, which is reflected in the level of market concentration, affects firm behavior and then in turn determines firm performance (Bain, 1951). The problem is that the SCP

94

4 The Impact of Bank Regulation and Supervision on Competition

analysis does not explicitly incorporate the effect of regulation and examine whether regulatory changes increase or decrease the relationship between market structure and competitive behavior in the banking system. One non-structural approach to measuring competition is the new empirical industrial organization (NEIO) approach. Unlike the SCP paradigm, which tries to determine competition from the market structure in a given industry, the NEIO models directly analyze firm conduct to detect the market power of firms. The NEIO models rely on a comparative statics analysis, as in the Panzar and Rosse (1987) model. The PR model measures market power using the H-statistic, which is calculated as the sum of revenue elasticities with respect to input prices. It measures how much a change in factor prices affects the firm’s equilibrium revenue. The PR model has been widely used to measure competition in the banking industry. Vesala (1995) investigates how deregulation in the 1980s affected competition among Finnish banks. Coccorese (2004, 2009) analyzes competitive conditions in the Italian banking industry. Matthews et al. (2007) and Maudos and Solís (2011) employ both the PR model and the Lerner index to analyze competition in the British and Mexican banking industries, respectively. These findings mostly indicate that banks operate under monopolistic competition. An alternative non-structural technique to the PR model is to estimate a parameter that directly measures firms’ competitive behavior from information on firm costs and demand. For example, the Lerner index is a relative markup of the price over marginal cost and measures firm market power (Lerner, 1934). The higher the markup is, the greater the market power. The Lerner index ranges from 0 in the case of perfect competition to 1 in the case of monopoly. A number of studies (e.g., Bikker & Haaf, 2002) show empirically that the H-statistic and the Lerner index are negatively correlated. That is, the relative price–cost markup (smaller Lerner index) decreases with higher competition (higher H-statistic). The Lerner index is widely used to estimate competition in the banking sector. Coccorese (2009) points out that the Lerner index is a true reflection of the bank’s degree of market power. Angelini and Cetorelli (2003) assess the behavior of Italian regional banks and find that deregulation fostered a reduction in the price-costs margin. Fu et al. (2014) use the Lerner index as a measure of bank competition and investigate the influence of bank competition on individual bank fragility, and Anginer et al. (2014a) use both the H-statistic and the Lerner index to measure bank competition and find a robust negative relationship between bank competition and systemic risk. Based on the hypothesis that more efficient banks are rewarded more in more competitive markets, Boone (2008) developed a new indicator. The basic intuition underlying the Boone indicator is that competition improves efficiency in bank performance with regard to profitability and market share and weakens the performance of inefficient banks. The Boone indicator has two major advantages compared to other measures: (1) it is based on strong theoretical foundations and captures competition due to both a fall in entry barriers and to more aggressive behavior on the part of incumbents, and (2) it captures the dynamics and non-price strategy in the market, while the other measures of competition are based on static price competition. The Boone indicator has been used in considerable literature in recent banking studies,

4.3 Literature Review

95

including Delis (2012), Schaeck and Cihák (2014), Khan et al. (2017), Clark, et al. (2018). The empirical evidence on bank competition in emerging market economies is growing. For example, Gelos and Roldós (2004) examine the evolution of the market structure in the banking system in emerging markets based on the PR model and confirm that markets have not become less competitive there. Jeon et al. (2011) show that an increase in foreign bank penetration enhances competition in these host countries’ banking sector, and the positive foreign bank penetration and banking competition link has a spillover effect from foreign banks to their domestic counterparts. Apergis (2015) employs the PR methodology and assesses competition across the banking systems in emerging market economies, while emphasizing the impact of the recent financial crisis on the degree of competition in these banking systems. Other papers examine bank competitive conditions focusing on one or a few emerging economies: Tan (2016) and Tan and Floros (2018) investigate the relationship between bank competition, risk, and profitability based on evidence in the Chinese banking industry; Soedarmono et al. (2011, 2013) and Khan et al. (2017) examine bank competition based on emerging markets in Asia.

4.3.2 Bank Regulation and Effects on Bank Competition The traditional theory indicates that bank deregulation is positively associated with efficiency because of a reduction in regulatory costs imposed on the banking system. Based on both theoretical and empirical evidence, Keeley (1990) confirms that deregulation reduces bank market power in the US market, and Dick (2006) suggests that higher loan loss provisions after the deregulation. Matutes and Vives (2000) study the links between competition for deposits and risk-taking incentives and conclude that the welfare performance of the market and the appropriateness of alternative regulatory measures depend on the degree of rivalry and the deposit insurance regime. Hellmann et al. (2000) analyze the relationship between competition for deposits and capital regulation in a dynamic framework in which banks choose their asset risk privately and compete for deposits and find that capital requirement regulation is not an optimal choice for controlling risk-taking incentives. A number of studies have investigated the effect of regulations and factors that presumably are related to competition in the banking system. Based on a survey of bank regulation and supervision in the banking system, Barth et al. (2004) find that higher entry barriers reduce bank efficiency and lead to an increase in interest rate margins and personnel expenses, which empirical confirms that entry restrictions reduce bank competition. Using a cross-country sample, Claessens et al. (2001) examine the impact of foreign banks on domestic banking and confirm that the introduction of foreign banks increases efficiency at domestic banks. In a cross-country study based on bank financial data in 72 countries, DemirgucKunt et al. (2004) confirm that tighter bank entry restrictions and activity restrictions are negatively related to banking efficiency, especially at foreign banks. Gelos and

96

4 The Impact of Bank Regulation and Supervision on Competition

Roldós (2004) examine the evolution of market structure in emerging markets and argue that lower entry barriers mitigated a decline in competition driven by consolidation. Claessens and Laeven (2004) investigate competitive conditions across 50 banking systems, finding that systems with greater foreign bank entry and few entry and activity restrictions are more competitive. Delis et al. (2011) examine the relationship between bank productivity and regulatory framework in 22 transition countries and find that private monitoring and activity restrictions have a significant impact on productivity. Only a few recent studies provide direct findings on the relationship between banking regulation and competition by focusing only on emerging economies. For example, Beck et al. (2005) analyze the impact of liquidation, federalization, privatization, and reconstruction on the performance of state banks in Brazil; Fu and Heffernan (2009) investigate the effect of reforms in China’s banking sector on market structure and performance; Brissimis et al. (2008) examine the impact of banking sector reform on bank performance in ten emerging economies in Europe; Zhao et al. (2010) evaluate the impact of financial sector reforms on cost structure, ownership, and competition in the Indian banking system; Tompson (2004) gives a detailed analysis of banking reform in Russia; and Williams and Nguyen (2005) and Noman et al. (2018) examine the impact of financial liberalization, bank restructuring, and bank regulation on bank performance in countries in Southeast Asia.

4.4 Estimation Methodology and Data 4.4.1 Estimation of Competition Variables (1)

Panzar-Rosse model

In our analysis, we first employ the PR model to construct the H-statistic to measure market power. Panzar and Rosse (1987) show that the H-statistic, which is equal to or less than zero, indicates a collusive or joint monopoly equilibrium or monopolistic competition without the threat of entry; a value between zero and one means monopolistic competition, and a value equal to one indicates perfect competition. Based on Demirguc-Kunt et al. (2004) and Anginer et al. (2014b), we employ the following reduced-form revenue equation and estimate the H-statistic for each country in each year1 : ln OPIit = α +

3  j=1

1 Bikker

βj ln wjit +



πk Controlskit + αi + εit

(4.1)

k

et al. (2012) indicate that a scaled revenue function leads to a significant upward bias and incorrectly measures the degree of competition, so in our analysis we employ the unscaled revenue equation to reduce the estimation bias.

4.4 Estimation Methodology and Data

97

where OPI it is the operating income (as a measure of revenue).2 We define w1 as the ratio of annual interest expenses to total funding and use it as a proxy for the average funding rate; w2 is the ratio of personnel expenses to total assets used as an proxy for personnel expenses; and w3 is the ratio of other non-interest expenses to fixed assets used as a proxy for the price of physical capital. We also add a set of control variables: Customer Loan, which is defined as the ratio of customer loans to total assets to control for credit risk; NEA, which is defined as the ratio of other nonearning assets to total assets to control for asset composition; Customer Deposit, which is defined as the ratio of customer deposits to the sum of customer deposits and short-term funding to capture the funding structure of the bank; and Equity Ratio, which is defined as the ratio of equity to total assets to account for bank leverage. As in Coccorese (2009), all bank-specific factors that have potential effects on the level of operating income but are not included in Eq. (4.1) are captured through the insertion of dummy variables associated with banks (denoted by αi ). We winsorize all variables at the 1st and 99th percentile levels to reduce the influence of outliers. The H-statistic, which is defined as the sum of the elasticities of revenues with respect to input prices, is then given by β1 + β2 + β3 . As we use unscaled revenue in Eq. (4.1), one potential caveat is that the scale differences across banks may introduce heteroskedastic standard errors into the coefficients, which in turn inflates the H-statistic. In our analysis, we estimate the unscaled PR revenue model using pooled feasible generalized least squares (FGLS) to address the heteroskedasticity problem and clustered standard errors to account for general heteroskedasticity and cross-sectional correlation in the model errors (Arellano, 1987). An important feature of the H-statistic is that the PR model must be based on firms that operate in a long-run equilibrium (Panzar & Rosse, 1987). We follow the analysis of Bikker et al. (2012) and employ an equilibrium return on equity (ROE) test to check whether banks operate in a long-run equilibrium. By using ROE instead of operating income (OPI) as an independent variable in Eq. (4.1), the H-statistic under the ROE test equals zero if the banking system operates in a long-run equilibrium. (2)

Lerner index

We also employ the Lerner index as an alternative measure of competition and measure banks’ competitive behavior based on information on costs and demand. The Lerner index is calculated as follows: Lerner Indexit = (Pit − MCit )/Pit

(4.2)

In our analysis, Pit is the output price proxied by the ratio of total revenue to total assets, and MC it is the marginal cost.3 The marginal cost is derived from the total 2 For

a robustness check, we also use total revenue as another measure of revenue, and the findings are consistent. 3 Based on earlier studies, we define the output price P as the ratio of total revenues to total assets; it see also Angelini and Cetorelli (2003) and Fu et al. (2014).

98

4 The Impact of Bank Regulation and Supervision on Competition

cost function—that is: MCit =

TCit (α1 + α2 ln Qit + α9 ln w1it + α10 ln w 2it + α11 ln w3it ) Qit

(4.3)

where the translog total cost function is α  2 (ln Qit )2 + α3 ln w1it 2 α  α  6 7 + α5 ln w3it + (ln w1it )2 + (ln w2it )2 2 2

ln TCit = α0 + α1 ln Qit +

+ α4 ln w2it α  8 + (ln w3it )2 +α9 ln w1it ∗ ln Qit + α10 ln w2it ∗ ln Qit 2 +α11 ln w3it ∗ ln Qit +α12 ln w1it ∗ ln w2it + α13 ln w2it ∗ ln w3it + α14 ln w3it ∗ ln w1it + αi + εit

(4.4)

TC it denotes the total operating expenses, and Qit represents output, measured by total assets; w1 , w2 , and w3 represent the bank’s input prices as defined previously in the PR model. Following Koetter et al. (2012) and Fu et al. (2014), we use the stochastic cost frontier analysis to estimate Eq. (4.4) for each country in each year. (3)

Boone indicator

Finally, we also use Boone indicator as defined in Boone (2008). The Boone indicator evaluates the relationship between bank competition and efficiency, which indicates that banks with higher efficiency also perform better and have higher profits. The idea behind the Boone indicator is that as the relationship between efficiency and profits increases, so does the degree of competition. The Boone model is defined as follows: ln Profitit = α + β ln MCit + εit

(4.5)

where Profit it is the profit before tax of bank i in year t, MCit is the marginal cost as defined in Eq. (4.3). The Boone indicator β is negative and decreases with increases in competition. The log–log specification is employed to deal with potential heteroskedasticity.

4.4 Estimation Methodology and Data

99

4.4.2 Data Sources, Sample Selection, and Variable Definitions (1)

Data Sources

Based on the International Monetary Fund’s definition of emerging economies, we focus our analysis on 23 emerging markets.4 We obtained the data from different sources. The financial data for each bank is collected from Bankscope. To avoid double counting, we use the consolidated financial information on each bank if available and unconsolidated reports otherwise. We obtained the bank regulation and supervision data from Barth et al. (2013), and the country economy development variables come from the World Bank Development Indicator Database. (2)

Sample selection

Based on a complete sample of banks obtained from Bankscope, we apply the following selection criteria: (1) we delete observations in which data on one of the variables employed in estimating the H-statistic, Lerner index, and Boone indicator are missing; (2) for the estimation of competition measures, we set the minimum number of observations to 20 for each year in each country, thus we delete countries with less than 20 bank-year observations; (3) we also delete observations in which the data on one of the bank-level and country-level control variables are missing. We use unbalanced data between 1996 and 2016, and our sample consists of commercial banks, saving banks, co-op banks, and bank holding companies. Our analysis has a total of 1629 banks and a total of 12,856 bank-year observations, with the largest number of observations in 2010 (914) and the lowest number in 1996 (236),5,6 All the data are inflation adjusted and expressed in USD. (3)

Variable Definitions

As we are interested in the impact of bank regulation and supervision on competition in the banking industry, we collect information of bank regulation and supervision from Barth et al. (2013) and define a set of variables to measure bank regulation and supervision in the banking system from different perspectives. 4 See detailed information on emerging economies in World Economic Outlook: Adjusting to Lower

Commodity Prices, International Monetary Fund, October 2015. our database, banks in Bangladesh and South Africa changed their accounting standard from local GAAP to international accounting standards during 2005–2006. In order to include as long a sample period as possible, we also include all bank observations (including those before the implementation of international accounting standards). Thus, for those two countries, we have two different accounting standards in our sample period: local GAAP (before 2005 or 2006) and international accounting standards (after 2005 or 2006). We use a fixed-effect model and include the year dummy variables to control for the effect of the changes in accounting standards. 6 In our sample period, especially during the global financial crisis in 2008–2009, some banks were bailed out. As information on the bailed-out banks in our sample is not available, we include both bank fixed-effect and year-fixed effects to avoid potential bias introduced by the bailed-out banks. 5 In

100

4 The Impact of Bank Regulation and Supervision on Competition

We define Bank concentration as the degree of concentration of deposits at the five largest banks in a given country and Government-owned banks as the extent to which the banking system’s assets are government owned, to examine the impact of market structure on competition in a given country. To indicate restrictions on bank activity in a given country, we define Activity restriction as a bank’s ability to engage in securities, insurance, and real estate activities; it ranges from 0 to 12, and a higher score indicates more restrictions on banks engaging in such activities. To measure the regulations on bank competition, we define two variables: Foreign bank limitation, which measures whether foreign banks may own domestic banks and whether foreign banks may enter a country’s banking industry, in which lower values indicate greater strictness; and Entry barriers, which indicates that various types of legal documentation are required to obtain a banking license, in which higher values indicate more documentation. We use Capital regulatory to indicate the capital regulatory strictness, in which a higher value indicates greater strictness. Official supervisory power and Diversification index measure official actions by supervisors, and Private monitoring index is used to as a proxy for private monitoring. We also employ Deposit insurance ratio, which is defined as the size of the deposit insurance fund relative to total bank assets, as a measure of the deposit insurance coverage in a given country. We use several variables to control for bank-specific time-varying effects on competition. We define TCD as the ratio of total customer deposits to total assets to control for the deposit size, LLP as the ratio of loan loss provision to gross loans to control for loan quality, NIIC as the ratio of non-interest income to total operating income, ROA as a proxy for profitability, and log (TA), defined as the logarithm of gross total assets, to control for bank size. As country control variables, we include GDP per capita, GDP growth, and Inflation in a country to account for economic cycles. We also add Interest rate and Market capitalization to GDP to control for changes in monetary policy and capital market size in a given country. In our analysis, competition variables, including the H-statistic and the Lerner index, are estimated for each country in each year, and bank-level financial data and country-level control variables are also collected for each calendar year. The bank regulation and supervision information comes from Barth et al. (2013); however, their surveys were conducted in 1999, 2002, 2006, and 2011. Following Anginer et al. (2014a), we employ the previous survey data until new survey data become available to match the bank regulation and supervision variables with bank-specific variables and country control variables.7 Table 4.1 summarizes the definitions of the variables and data sources employed in this analysis.

7 Specifically, the survey data conducted in 1999 for 1996–2001, the survey data conducted in 2002

for 2002–2005, the survey data conducted in 2006 for 2006–2010, and the survey data conducted in 2011 for 2011–2016.

4.4 Estimation Methodology and Data

101

Table 4.1 Variable definition and data sources Variables

Definitions

Data sources

Dependent variables H-statistic

H-statistic is calculated as the sum of the Own calculation elasticity of revenue with respect to three input prices for each country at each year between 1996 and 2016 based on the Panzar and Rosse (1987) model

Lerner index

Lerner index is equal to the difference between asset price and marginal cost, normalized by asset price

Own calculation

ln profit

Logarithm of the profit before tax in $ thousand

Bankscope

TCD

The ratio of total customer deposits to total assets

Bankscope

LLP

The ratio of loan loss provision to gross loans

Bankscope

NIIC

Non-interest income divided by total operating income

Bankscope

Bank control variables

ROA

Net income divided by total assets

Bankscope

log TA

Logarithm of total asset in $ thousand

Bankscope

Bank crisis

A dummy variable that equals one if the Laeven and Valencia (2010) country is going through a systemic crisis in a given year, and zero otherwise

Foreign dummy

A binary variable which equals one if the Bankscope bank is a foreign subsidiary and zero otherwise

Country control variables Market capitalization to GDP

Stock market capitalization to GDP

World Bank Development Indicator Database

GDP per capita

GDP per capital in $ thousand

World Bank Development Indicator Database

GDP growth

Annual growth rate of GDP at market prices based on constant local currency

World Bank Development Indicator Database

Inflation

Inflation rate

World Bank Development Indicator Database

Interest rate

The interest rate charged by banks on loans to prime customers

World Bank Development Indicator Database

Regulation and supervision variables Bank concentration

The degree of concentration of deposits in the 5 largest banks

Barth et al. (2013)

Government-owned banks

The extent to which the banking system’s assets are government owned

Barth et al. (2013) (continued)

102

4 The Impact of Bank Regulation and Supervision on Competition

Table 4.1 (continued) Variables

Definitions

Activity restriction

A variable measures a bank’s ability to Barth et al. (2013) engage in securities, insurance, and real estate activities. The ranges from 0 to 12, and a higher score indicates more restrictions on banks to engage in such activities

Data sources

Foreign bank limitation

A variable measures whether foreign Barth et al. (2013) banks may own domestic banks and whether foreign banks may enter a country’s banking industry. The variable ranges from 0 to 4, and lower values indicate greater stringency

Entry barriers

A variable which indicates that whether Barth et al. (2013) various types of legal submissions are required to obtain a banking license. The variable ranges from 0 to 8, and higher values indicate greater stringency

Capital regulatory index

A variable that captures both the overall capital stringency and the initial capital stringency based on answers to eight questions. It ranges from zero to eight, with a higher value indicating higher capital stringency

Barth et al. (2013)

Official supervisory power

A variable that ranges from zero to fourteen, with fourteen indicating the highest power of the supervisory authorities

Barth et al. (2013)

Diversification index

A variable that ranges from zero to two, with higher values indicating more diversification

Barth et al. (2013)

Private monitoring index

A variable that measures whether there incentives/ability for the private monitoring of firm, with higher values indicating more private monitoring

Barth et al. (2013)

Deposit insurance ratio

The size of the deposit insurance fund relative to total bank assets

Barth et al. (2013)

This table reports the definitions and data sources of the variables included in our analysis

4.4.3 Statistical Results We provide the results in Panel A of Table 4.2, beginning with the competition variables. The average H-statistic is 0.575, ranging from 0.165 to 0.816, and the Lerner index has a mean of 0.225 with a standard deviation of 0.374.

4.4 Estimation Methodology and Data

103

Table 4.2 Summary statistics Panel A: Data statistics Variable

N

Mean

STD

Min

Max

P25

Median

H-statistic

12,856 0.575

0.534

0.165

0.816

0.269

0.527

Lerner index

8792

0.374

0.192

0.901

0.287

0.258

ln Profit

20,782 10.389 3.757

0.943

18.265

1.647

9.733

TCD

12,856 0.535

0.256

0.031

0.816

0.335

0.613

LLP

12,856 0.016

0.018

0.000

0.057

0.002

0.009

NIIC

12,856 0.311

0.233

0.045

0.882

0.122

0.260

ROA

12,856 0.014

0.013

−0.003

0.043

0.005

0.011

log TA

12,856 14.862 2.660

10.284

19.537

12.831 14.581

Bank crisis

12,856 0.081

0.273

0

1

0

0

Foreign dummy

12,856 0.179

0.383

0

1

0

0

0.225

Market capitalization to GDP 12,856 62.929 53.275 0.850 (%)

321.984 26.753 46.601

GDP per capita ($ thousand)

12,856 5.280

3.587

0.395

16.007

2.400

4.596

GDP growth (%)

12,856 4.580

4.172

−13.127 14.231

3.005

5.094

Inflation (%)

12,856 8.320

8.824

−5.016

3.525

6.859 12.134

75.271

Interest rate (%)

12,856 18.469 15.716 3.536

67.083

7.470

Bank concentration (%)

10,677 56.683 18.385 4.000

93.000

52.000 57.000

Government-owned banks (%)

11,333 30.787 23.439 0.000

75.000

11.000 32.000

Activity restriction

12,714 7.702

2.013

4

12

6

8

Foreign bank limitation

12,693 3.670

0.674

0

4

3

4

Entry barriers

12,856 7.511

0.905

3

8

7

8

Capital regulatory index

12,645 6.455

2.263

2

10

5

6

Official supervisory power

12,856 11.578 2.356

4

16

10

12

Diversification index

11,559 1.269

0.757

0

2

1

1

Private monitoring index

12,465 8.565

1.239

6

11

8

9

Deposit insurance ratio

7441

0.325

0

1

0

0

0.120

Sample distribution by year Year

Frequency

Percentage

Country

Frequency

1996

236

1.84

Argentina

819

Percentage 6.37

1997

263

2.05

Bangladesh

201

1.56

1998

419

3.26

Brazil

1,490

11.59

1999

476

3.70

Bulgaria

203

1.58

2000

574

4.46

Chile

384

2.99

2001

537

4.18

China

1,307

10.17

2002

520

4.04

Colombia

354

2.75 (continued)

104

4 The Impact of Bank Regulation and Supervision on Competition

Table 4.2 (continued) Sample distribution by year 2003

482

3.75

Hungary

2004

608

4.73

India

853

6.64

2005

710

5.52

Indonesia

897

6.98

2006

741

5.76

Malaysia

975

7.58

2007

755

5.87

Mexico

164

1.28

2008

763

5.93

Pakistan

435

3.38

2009

817

6.36

Peru

346

2.69

2010

914

7.11

Philippines

468

3.64

2011

884

6.88

Poland

470

3.66

2012

743

5.78

Romania

247

1.92

2013

666

5.18

Russia

1,247

9.70

2014

591

4.60

South Africa

532

4.14

2015

585

4.55

Thailand

359

2.79

2016

572

4.45

Turkey

451

3.51

Ukraine

120

0.93

Venezuela

234

1.82

Total

12,856

100

Total

12,856

100

300

2.33

This table reports summary statistics of the variables for the full sample, and sample distribution according to year and country. Our sample period iss 2000–2016. All data are inflation-adjusted and bank-specific variables are winsorized at the 1st and 99th percentile level to reduce the influence of outliers

Then we report the results for bank-specific variables. The average ratio of customer deposits to total assets (TCD) is 54%, and LLP has an average of approximately 1.6% of gross loans. NIIC accounts for roughly 31.1% of the total operating income, and the average ROA is 1.4%. The average log TA value is 14.86, but the size of the banks in our sample varies substantially. This implies that any analysis needs to account for the size effect. In our sample, 17.9% of the banks are foreign owned. We include five country-level variables (Market capitalization to GDP, GDP per capita, GDP growth, Inflation, and Interest rates) to control for the changes of stock market and macroeconomic factors. The average market capitalization to GDP ratio is 62.9, GDP per capita is $5,280, GDP growth is 4.58%, inflation is 8.32%, and interest rate is 18.47%. Finally, we report descriptive statistics of bank regulation and supervision variables. The average for bank concentration is 56.68%, government-owned banks is 30.79%, activity restrictions 7.7, foreign bank limitations 3.67, entry barriers 7.51, capital regulatory index 6.46, official supervisory power 11.58, diversification index 1.27, private monitoring index 8.57, and deposit insurance coverage ratio is 12%.

4.4 Estimation Methodology and Data

105

In Panel B of Table 4.2, we report the sample distribution by calendar year and by country. The number of banks ranges from 236 to 914, with the largest number in 2010 and the lowest number in 1996.

4.4.4 Determinants of Bank Competition To examine the impact of bank regulation and supervision on competition, we regress the competition variables on a set of bank-specific variables, country variables, and bank regulation and supervision variables. The regression model is as follows: Competitionit = β0 +

N 

βn bank_controlsnit

n=1

+

J 

αj country_controlsjkt + γ bank_regulationkt

j=1

+ νi + μt + εi,t

(4.6)

The dependent variable Competition is one of the bank competition variables: the H-statistic or the Lerner index. bank_controls are a set of bank-specific control variables: TCD, LLP, NIIC, ROA, and log TA. country_controls are the country-level control variables: Market capitalization to GDP, GDP per capita, GDP growth, Inflation, and Interest rate. bank_regulation are bank regulation and supervision variables: Bank concentration, Government-owned banks, Activity restriction, Foreign bank limitation, Entry barriers, Capital regulatory index, Official supervisory power, Diversification index, Private monitoring index, and Deposit insurance ratio. ε is an error term. Standard errors are heteroskedasticity consistent and are clustered at the bank level. We follow Gormley and Matsa (2014) in employing bank-specific fixed effects to control for unobserved time-invariant heterogeneity across banks that might affect competition. Regressions include year-fixed effects to control for macroeconomic factors and monetary policy that may vary over time.

4.5 Empirical Results 4.5.1 Bank Regulation, Supervision, and the H-Statistic In this section, we investigate how bank regulation and supervision affect competition in the banking industry from emerging markets, after controlling for the effects of bank-level and country-level characteristics. As the H-statistic and the Lerner index with standard errors are generated from the first-stage regressions as defined in the

106

4 The Impact of Bank Regulation and Supervision on Competition

third section, to increase the accuracy of our second-stage estimation, we follow Doidge et al. (2006) and Chue and Cook (2008) and weigh each observation by the inverse of the standard error of the H-statistic and the Lerner index for each country in each year obtained at the first stage. With this procedure, the H-statistic and the Lerner index of observations that are estimated more precisely in the first-stage regressions have a higher weight in the second-stage analysis. We first examine the impact of bank regulation and supervision on the H-statistic, and the results are reported in Table 4.3. Considering the impact of market structure on competition, the result in column 1 in Table 4.3 indicates that bank concentration is positively and statistically correlated with the H-statistic, which indicates that banks in a more concentrated system are exposed to a higher level of competition. This positive relationship between the H-statistic, which measures bank competition based on the new empirical industrial organization (NEIO) approach, and bank concentration, which defines market structure based on the structure-conduct-performance (SCP) paradigm, suggests that the level of competition is not necessarily related to market structure (see also Fernández De Guevara et al., 2005; Ryan et al., 2014). Two possible explanations exist for this finding: first, the banking sector is often observed to be simultaneously concentrated and competitive, and so concentration may be a poor proxy for underlying market power; moreover, a more serious issue is that market structure and concentration may proxy for a whole range of conduct that determines bank and market characteristics, including average bank size, bank complexity in terms of product variety and activities, the ease of information flow in the market, and the overall size of the market. We also include the degree of government bank ownership, to measure the level of banking system’s assets owned by the government. The result indicates that higher government bank ownership seems to reduce competition in the banking industry. Column 2 in Table 4.3 tests the impact that regulations on bank activity have on competition. Higher Activity restriction indicates more restrictions on banks engaging in securities, insurance, and real estate activities, and the result indicates that higher activity restrictions result in lower competition. This confirms that, although activity restrictions reduce the potential risks in the banking industry (Anginer & Demirguc-Kunt, 2014), but this also reduces competition in the banking system. The positively significant coefficient of Foreign bank limitation and negatively significant coefficient of Entry barriers confirm that less foreign bank entry or ownership limitations and lower banking entry requirements positively affect competition in the banking system. These findings suggest that lower foreign bank limitations and banking entry requirements increase competition in the banking system. To examine the impact of capital regulation, we include the variable Capital regulatory index. A higher Capital regulatory index indicates greater strictness, and our results provide evidence that competition in the banking system is more intense if capital strictness is more pronounced in the country. Previous studies (e.g., Agoraki et al., 2011; Chortareas et al., 2012) indicate that strengthening capital restrictions can improve efficient bank operations, which in turn enhances competition among banks.

(1)

(2.89)

−0.0243

(−1.34)

−0.0235

(−1.29)

(−1.95)

(−1.94)

(2.88)

−0.0105*

−0.0104*

0.407***

(1.14)

(1.18)

0.397***

0.0118

0.0122

0.264***

(3.67)

0.261***

Interest rate

Inflation

GDP growth

GDP per capita

Market capitalization

(2)

(3.62)

(−1.17)

−0.0210

(2.79)

0.377***

(−1.94)

−0.0103*

(1.19)

0.0123

(3.36)

0.239***

(3)

(−1.29)

−0.0229

(2.99)

0.423***

(−1.96)

−0.0105**

(1.08)

0.0109

(3.50)

0.245***

(4)

(−1.48)

−0.0266

(2.71)

0.366***

(−1.88)

−0.00993*

(1.25)

0.0130

(3.64)

0.259***

(5)

(−1.75)

−0.0305*

(2.85)

0.386***

(−1.97)

−0.0109**

(1.22)

0.0129

(3.81)

0.273***

(6)

(−1.69)

−0.0302*

(3.26)

0.462***

(−1.83)

−0.00994*

(1.23)

0.0128

(3.67)

0.261***

(7)

(−1.87)

−0.0323*

(2.97)

0.430***

(−1.91)

−0.0102*

(1.36)

0.0146

(3.59)

0.261***

(8)

(−0.85)

−0.0158

(3.69)

0.493***

(−1.89)

−0.00917*

(1.27)

0.0102

(1.84)

0.144*

(9)

0.0224***

(11.05)

(9.89)

(4.86)

(4.73)

0.0218***

0.00763***

(5.32)

(4.71)

0.00709***

0.0106***

(7.58)

(7.92)

0.00993***

0.0679***

(−17.00)

0.0694***

(−16.96)

(10.25)

0.0221***

(4.37)

0.00666***

(4.50)

0.00955***

(8.68)

0.0734***

(−17.24)

(11.96)

0.0243***

(5.86)

0.00844***

(6.93)

0.0133***

(9.73)

0.0739***

(−17.72)

(9.16)

0.0200***

(4.72)

0.00691***

(3.69)

0.00721***

(8.11)

0.0740***

(−16.75)

(9.78)

0.0222***

(4.13)

0.00585***

(4.59)

0.00967***

(7.45)

0.0679***

(−15.33)

(10.17)

0.0212***

(5.52)

0.00837***

(5.53)

0.0120***

(7.80)

0.0641***

(−17.48)

(9.41)

0.0220***

(3.81)

0.00551***

(4.20)

0.00865***

(7.81)

0.0772***

(−14.78)

(continued)

(17.39)

0.0446***

(2.32)

0.00681**

(16.03)

0.0280***

(32.20)

0.161***

(−15.73)

−0.00784*** −0.00794*** −0.00772*** −0.00831*** −0.00766*** −0.00741*** −0.00822*** −0.00728*** −0.0124***

Country control variables

log TA

ROA

NIIC

LLP

TCD

Bank control variables

Variables

Table 4.3 Impact of bank regulation and supervision on H-statistic

4.5 Empirical Results 107

(1)

Deposit insurance ratio

Private monitoring index

Diversification index

Official supervisory power

Capital regulatory index

Entry barriers

Foreign bank limitation

Activity restriction

Government-owned −0.00522*** banks (−4.29)

(13.35)

Bank concentration 0.00371***

Regulation and supervision variables

Variables

Table 4.3 (continued)

(−2.22)

−0.0161**

(2)

(3.11)

0.0528***

(3)

(−10.00)

−0.0937***

(4)

(3.69)

0.0336***

(5)

(4.49)

0.0360***

(6)

(−6.68)

−0.137***

(7)

(4.74)

0.0827***

(8)

(continued)

(1.83)

0.105*

(9)

108 4 The Impact of Bank Regulation and Supervision on Competition

9365

0.186

Number of observations

R-squared

0.130

12,714

Yes

0.131

12,693

Yes

Yes

(0.17)

0.0435

(3)

0.134

12,856

Yes

Yes

(3.32)

0.911***

(4)

0.130

12,645

Yes

Yes

0.139

12,856

Yes

Yes

(−0.33)

−0.0929

−0.0345 (−0.13)

(6)

(5)

0.141

11,559

Yes

Yes

(2.54)

0.649**

(7)

0.168

12,465

Yes

Yes

(−1.66)

−0.571*

(8)

0.292

7441

Yes

Yes

(−1.97)

−0.583**

(9)

The dependent variable is H-statistic which is estimated based on Panzar and Rosse (1987) model as defined in Eq. (4.1). Our sample period is 1996–2016. Bank-level controls include TCD, LLP, NIIC, ROA, and log TA. Country control variables include Market capitalization, GDP per capita, GDP growth, Inflation, and Interest rate. Regulation and supervision variables include Bank concentration, Government-owned banks, Activity restriction, Foreign bank limitation, Entry barriers, Capital regulatory index, Official supervisory power, Diversification index, Private monitoring index, and Deposit insurance ratio. All variables are defined in Table 4.1. Reported t-statistics (in parentheses) are based on standard errors that are heteroskedasticity consistent and clustered at the bank level. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

Yes

Year fixed effects

Yes

(1.36)

Yes

(0.99)

Bank fixed effects

0.369

0.250

Constant

(2)

(1)

Variables

Table 4.3 (continued)

4.5 Empirical Results 109

110

4 The Impact of Bank Regulation and Supervision on Competition

Results for official supervisory power are in columns 6 and 7 in Table 4.3. We find that the cross-country difference in bank competition can be explained by the differences in official supervisory power and explicit guidelines for asset diversification, in which higher official supervisory power increases competition and greater explicit guidelines for asset diversification reduce competition. Next we explore the relationship between private monitoring and bank competition. If the banks have greater incentives or ability to engage in private monitoring, then we would expect that that a higher private monitoring index is associated with higher competition. The result for private monitoring is in column 8 in Table 4.3. Consistent with our expectations, a higher private monitoring index significantly increases bank competition in a given country. Finally, we test the impact of deposit insurance on bank competition. While deposit insurance provides banks with a guarantee when they fail and prevents banks runs, it also leads to moral hazard and excessive risk-taking (Anginer et al., 2014b), which may in turn enhance competition among banks. In column 9 in Table 4.3, we include the variable Deposit insurance ratio, which is defined as the size of the deposit insurance fund relative to total bank assets, to examine the impact of deposit insurance on competition in the banking system. The positive and significant coefficient of Deposit insurance ratio confirms that deposit insurance increases competition in the banking system. Among bank-level control variables, we find that customer deposits (TCD) and profitability (ROA) are positively and significantly related to bank competition in all specifications, and higher non-interest income (NIIC) indicates lower competition in the banking system. Among country-level variables, we find that the development of the stock market has a negatively significant impact on competitiveness in the banking system, and higher GDP per capita, GDP growth, inflation, and interest rate are statistically and positively significant. This suggests that in emerging economies, the development of the economy and the stock market show the same patterns and are the main factors driving competition in the banking system.

4.5.2 Alternative Measures of Competition We also consider whether our results are robust across different measures of competition, and the regression results are in Table 4.4. First, we use the Lerner index as a measure of competition. The results in Panel A in Table 4.4 show that Activity restriction is positively and significantly related to the Lerner index, indicating that higher activity restrictions reduce bank competition; Foreign bank limitation is negatively and significantly related to the Lerner index, which suggests that lower foreign bank limitations lead to higher competition in the banking system. The negatively significant coefficients of Capital regulatory index and Official supervisory power provide evidence that higher capital strictness and supervisory power result in higher competition among banks in a given country. Diversification index is positively and statistically significantly related to the Lerner index, which confirms that greater

(1)

(2)

(3)

(4)

Deposit insurance ratio

Private monitoring index

Diversification index

Official supervisory power

Capital regulatory index

Entry barriers

Foreign bank limitation

Activity restriction

Government-owned 0.00212 banks (1.52)

(1.15)

Bank concentration 0.000498

(1.91)

0.00735*

(−1.85)

−0.0188*

(−0.50)

−0.00534

Panel A: Impact of bank regulation and supervision on Lerner index

Variable

(−2.29)

−0.0140**

(5)

(−1.95)

−0.00555*

(6)

(2.43)

0.0257**

(7)

(−3.25)

−0.0239***

(8)

(−2.56)

−0.188**

(9)

Table 4.4 Impact of bank regulation and supervision on competition. Alternative measure of competition: Lerner index and Boone indicator

(continued)

(10)

4.5 Empirical Results 111

6456

0.201

Number of observations

R-squared

0.199

8773

Yes

0.193

8816

Yes

Yes

(1.93)

0.456*

(3)

0.195

8907

Yes

Yes

(2.36)

0.556**

(4)

0.201

8725

Yes

Yes

(1.72)

0.413*

(5)

Entry barriers * ln MC

Foreign bank limitation * ln MC

Activity restriction * ln MC

Government-owned banks * ln MC

Bank concentration * ln MC

ln MC

(12.69)

0.00527***

(−3.71)

(12.00)

0.0423***

(−0.08)

−0.00179***

(−9.91)

(−11.47)

(3) −0.00242

(2)

−0.145*** −0.368***

(1)

(−12.83)

−0.182***

(6.17)

0.302***

(4)

(6)

0.195

8907

Yes

Yes

(2.47)

0.574**

(6)

(0.86)

0.0118

(−7.57)

(−5.60)

−0.146*** −0.150***

(5)

Panel B: Estimation of bank regulation and supervision affecting the Boone indicator

Yes

Bank and Year fixed effects

Yes

(2.46)

Yes

(2.23)

Bank and country control variables

0.567**

0.525**

Constant

(2)

(1)

Variable

Table 4.4 (continued)

(2.81)

0.0941***

(7)

0.198

8218

Yes

Yes

(2.68)

0.608***

(7)

(−11.67)

−0.242***

(8)

0.200

8631

Yes

Yes

(2.95)

0.706***

(8)

(12.91)

0.538***

(9)

0.275

5361

Yes

Yes

(0.10)

0.0290

(9)

(continued)

(−12.42)

−0.210***

(10)

(10)

112 4 The Impact of Bank Regulation and Supervision on Competition

17,059

20,782

131.5

Number of observations

Chi-squared

465.4

20,033

Yes

(204.44)

9.486***

(3)

490.7

20,701

Yes

(203.31)

9.361***

(4)

253.9

20,782

Yes

(329.33)

9.996***

(5)

(6)

1037.4

19,971

Yes

(228.95)

9.267***

(−15.10)

−0.0447***

3513.0

20,599

Yes

(190.54)

8.394***

(−34.98)

−0.0858***

(7)

339.4

19,828

Yes

(219.38)

9.449***

(2.40)

0.0242**

(8)

962.1

19,898

(211.14)

9.340***

(−22.95)

−0.119***

(9)

200.7

15,483

Yes

(201.58)

9.556***

(−1.17)

−0.285

(10)

The dependent variables are Lerner index and log of profit in Panel A and Panel B, respectively. Our sample period is 1996–2016. Bank-level controls include TCD, LLP, NIIC, ROA, and log TA. Country control variables include Market capitalization, GDP per capita, GDP growth, Inflation, and Interest rate. Regulation and supervision variables include Bank concentration, Government-owned banks, Activity restriction, Foreign bank limitation, Entry barriers, Capital regulatory index, Official supervisory power, Diversification index, Private monitoring index, and Deposit insurance ratio. All variables are defined in Table 4.1. Bank-level control variables and country-level variables are also included in the regressions but not reported for brevity. Reported t-statistics (in parentheses) are based on standard errors that are heteroskedasticity consistent and clustered at the bank level. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

524.5

Yes

9.601***

(190.89)

9.581***

(2)

(157.27)

(1)

Bank and year fixed Yes effects

Constant

Deposit insurance ratio * ln MC

Private monitoring index * ln MC

Diversification index * ln MC

Official supervisory power * ln MC

Capital regulatory index * ln MC

Variable

Table 4.4 (continued)

4.5 Empirical Results 113

114

4 The Impact of Bank Regulation and Supervision on Competition

explicit guidelines for asset diversification increase the market power of banks and thus reduces competition in the banking system. We also find that Private monitoring index and Deposit insurance ratio are negatively and significantly associated with the Lerner index. The negative relationship indicates that a higher private monitoring index and deposit insurance coverage reduce the market power of banks and increase bank competition. For a robustness check, we also use the Boone indicator as another alternative measure of competition to see the effect on it of bank regulations and supervision. We include the interaction terms of marginal cost (lnMC) and bank regulation and supervision variables bank_regulation to analyze the change in the Boone indicator due to interaction with bank_regulation (captured in the coefficient of the interaction term). The regression results of the Boone indicator are reported in Panel B in Table 4.4. As expected, the regression in column 1 in Panel B shows that marginal cost is negatively and significantly associated with profit. We find that the interaction term of bank concentration and marginal cost, Bank concentration * ln MC, is negatively and significantly associated with profit. This indicates that the negative Boone indicator is lower for a banking industry with higher bank concentration and provides empirical support that competition is higher in banking systems with higher concentration. The interaction term between Government-owned banks and marginal cost is positively and significantly related to profit. This indicates that the Boone indicator is lower for government-owned banks than for other banks. The interaction terms for Activity restriction and Diversification index are positively and significantly related to profit. Hence, higher activity restrictions and greater explicit guidelines for asset diversification lead to a decrease in bank competition. The coefficients of the interaction terms for Foreign bank limitation and Capital regulatory index are significantly negative, which further confirms that higher foreign bank limitations and greater capital strictness corresponds with an increase in bank competition. The interaction terms Official supervisory power and Private monitoring index are negatively and significantly related to profit. This confirms that Official supervisory power and Private monitoring index are negatively and significantly related to the Boone indicator. Hence, higher official supervisory power and private monitoring are associated with higher competition in the banking industry.

4.5.3 Impact of the Financial Crisis Following the recent global financial crisis, policy makers reshaped bank regulation substantially. We are concerned with how the crisis affects competition in the banking system and whether the relationship between bank regulation and competition shows a different pattern during the crisis. In this section, we use bank-crisis information for individual countries from a database compiled by Laeven and Valencia (2018) and include interaction terms between country-level regulation variables and Bank Crisis

4.5 Empirical Results

115

as additional explanatory variables. Bank Crisis is a dummy variable that equals one if the country is going through a systemic crisis in a given year, and zero otherwise. The results with the bank crisis interactions are in Table 4.5. The coefficient on the interaction variable for bank concentration is negative (column 1 in Table 4.5), suggesting that competition in a more highly concentrated banking system is lower during a bank crisis; in banking systems with more government-owned banks, competition among banks is much higher during a bank crisis. We also see that the negative relationship between Activity restrictions and bank competition is more pronounced during a bank crisis, and the results of Entry barriers and Diversification index have similar findings, indicating that the effect of Activity restrictions, Entry barriers, and Diversification index in reducing bank competition is more pronounced during a bank crisis. The coefficients of the interaction terms for Foreign bank limitation, Capital regulatory index, Official supervisory power, Private monitoring index, and Deposit insurance ratio are significantly negative, indicating that a reduction in foreign bank limitations and increase in capital strictness, supervisory power, and private monitoring are more effective in enhancing bank competition in normal times than during a bank crisis. Our findings confirm that during a bank crisis, the impact of activity restrictions, entry barriers, and diversification guidelines on competitive conditions become more effective while foreign bank limitations, capital strictness, supervisory power, and private monitoring become less effective.

4.5.4 Further Tests: Foreign Banks Versus Domestic Banks In this subsection, we further investigate whether the relationship between bank regulation and competition differs between domestic banks and foreign banks. Foreign banks have better profitability performance than domestic banks in developing countries (Claessens et al., 2001), and foreign banks are subject to regulations not only in the host country but also in the parent country. De Haas and Van Lelyveld (2006) and Ongena et al. (2013) also find that foreign banks are more sensitive to the strictness of parent-country regulation, rather than to the institutional conditions in the host countries, and the impact of institutional reforms can have a less significant impact on foreign banks (Fang et al., 2014). To test these matters, we introduce a binary variable Foreign dummy, which equals one if the bank is a foreign subsidiary and zero otherwise. The results, including the interaction terms between bank regulation variables and Foreign dummy, are in Table 4.6. The results of Activity restriction, Foreign bank limitation, and Diversification index indicate that the changes in competition in the banking industry with activity restrictions, foreign bank limitations, and diversification guidelines are more pronounced at domestic banks than at foreign banks, and the impact of official supervisory power and private monitoring on competition are greater for foreign banks than domestic banks. Our findings suggest that foreign banks are more sensitive

(1)

(2)

(10.94)

0.00321***

Entry barriers * bank crisis

Entry barriers

Foreign bank limitation * bank crisis

Foreign bank limitation

Activity restriction * bank crisis

Activity restriction

Government-owned banks 0.00802*** * bank crisis (3.94)

(−5.41)

Government-owned banks −0.00628***

Bank concentration * bank −0.00988*** crisis (−8.92)

Bank concentration

(−6.18)

−0.0288***

(−2.17)

−0.0157**

Bank crisis, regulation, and supervision variables

Variables

(−3.15)

−0.0354***

(2.81)

0.0470***

(3)

Table 4.5 Bank regulation, supervision, bank crisis, and H-statistic

(−5.11)

−0.0227***

(−9.77)

−0.0937***

(4)

(5)

(6)

(7)

(8)

(continued)

(9)

116 4 The Impact of Bank Regulation and Supervision on Competition

Deposit insurance ratio * bank crisis

Deposit insurance ratio

Private monitoring index * bank crisis

Private monitoring index

Diversification index * bank crisis

Diversification index

Official supervisory power * bank crisis

Official supervisory power

Capital regulatory index * bank crisis

Capital regulatory index

Variables

Table 4.5 (continued)

(1)

(2)

(3)

(4)

(5)

(−4.16)

−0.0196***

(3.61)

0.0330***

(−4.08)

−0.0146***

(4.37)

0.0352***

(6)

(−6.96)

−0.198***

(−7.21)

−0.152***

(7)

(−5.75)

−0.0220***

(4.78)

0.0829***

(8)

(continued)

(−6.56)

−0.195***

(1.95)

0.113*

(9)

4.5 Empirical Results 117

9365

0.191

Number of observations

R-squared

0.135

12,714

Yes

(3)

0.134

12,693

Yes

Yes

(0.13)

0.0325

(4)

0.136

12,856

Yes

Yes

(3.23)

0.882***

0.133

12,645

Yes

Yes

(−0.23)

0.141

12,856

Yes

Yes

(−0.36)

(6) −0.101

(5) −0.0636

(7)

0.141

11,559

Yes

Yes

(2.72)

0.685***

(8)

0.172

12,465

Yes

Yes

(−1.81)

−0.617*

(9)

0.292

7441

Yes

Yes

(−1.85)

−0.549*

The dependent variable is H-statistic which is estimated based on Panzar and Rosse (1987) model as defined in Eq. (4.1). Our sample period is 1996–2016. Bank-level controls include TCD, LLP, NIIC, ROA, and log TA. Country control variables include Market capitalization, GDP per capita, GDP growth, Inflation, and Interest rate. Regulation and supervision variables include Bank concentration, Government-owned banks, Activity restriction, Foreign bank limitation, Entry barriers, Capital regulatory index, Official supervisory power, Diversification index, Private monitoring index, and Deposit insurance ratio. All variables are defined in Table 4.1. Reported t-statistics (in parentheses) are based on standard errors that are heteroskedasticity consistent and clustered at the bank level. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

Yes

Bank and year fixed effects

Yes

(1.13)

Yes

(0.89)

Bank and country control variables

0.225

Constant

(2)

0.305

(1)

Variables

Table 4.5 (continued)

118 4 The Impact of Bank Regulation and Supervision on Competition

(1)

(−3.53)

−0.00495***

(0.23)

0.0000592

(12.55)

0.00370***

Entry barriers * foreign dummy

Entry barriers

Foreign bank limitation * foreign dummy

Foreign bank limitation

Activity restriction * foreign dummy

Activity restriction

Government-owned banks * foreign −0.000912 dummy (−0.48)

Government-owned banks

Bank concentration * foreign dummy

Bank concentration

Bank crisis, regulation, and supervision variables

Variables

(1.98)

0.0215**

(−2.78)

−0.0232***

(2)

(−3.28)

−0.108***

(3.75)

0.0701***

(3)

Table 4.6 Bank regulation, supervision, foreign ownership, and H-statistic

(−1.11)

−0.0358

(−9.16)

−0.0918***

(4)

(5)

(6)

(7)

(8)

(continued)

(9)

4.5 Empirical Results 119

Deposit insurance ratio * foreign dummy

Deposit insurance ratio

Private monitoring index * foreign dummy

Private monitoring index

Diversification index * foreign dummy

Diversification index

Official supervisory power * foreign dummy

Official supervisory power

Capital regulatory index * foreign dummy

Capital regulatory index

Variables

Table 4.6 (continued)

(1)

(2)

(3)

(4)

(5)

(0.05)

0.000396

(3.59)

0.0335***

(2.47)

0.0455**

(3.02)

0.0263***

(6)

(2.80)

0.0765***

(−7.35)

−0.150***

(7)

(2.30)

0.0754**

(3.86)

0.0690***

(8)

(continued)

(0.04)

0.000341

(1.83)

0.105*

(9)

120 4 The Impact of Bank Regulation and Supervision on Competition

9365

0.186

Number of observations

R-squared

0.130

12,714

Yes

(3)

0.132

12,693

Yes

Yes

(0.15)

0.0385

(4)

0.134

12,856

Yes

Yes

(3.45)

0.941***

0.130

12,645

Yes

Yes

(−0.13)

0.140

12,856

Yes

Yes

(−0.38)

(6) −0.107

(5) −0.0347

(7)

0.142

11,559

Yes

Yes

(2.47)

0.635**

(8)

0.168

12,465

Yes

Yes

(−1.68)

−0.574*

(9)

0.292

7441

Yes

Yes

(−1.97)

−0.583**

The dependent variable is H-statistic which is estimated based on Panzar and Rosse (1987) model as defined in Eq. (4.1). Our sample period is 1996–2016. Bank-level controls include TCD, LLP, NIIC, ROA, and log TA. Country control variables include Market capitalization, GDP per capita, GDP growth, Inflation, and Interest rate. Regulation and supervision variables include Bank concentration, Government-owned banks, Activity restriction, Foreign bank limitation, Entry barriers, Capital regulatory index, Official supervisory power, Diversification index, Private monitoring index, and Deposit insurance ratio. All variables are defined in Table 4.1. Reported t-statistics (in parentheses) are based on standard errors that are heteroskedasticity consistent and clustered at the bank level. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

Yes

Bank and year fixed effects

(1.48) Yes

(0.96)

Yes

Bank and country control variables

0.404

0.244

Constant

(2)

(1)

Variables

Table 4.6 (continued)

4.5 Empirical Results 121

122

4 The Impact of Bank Regulation and Supervision on Competition

to official supervisory power and private monitoring and less sensitive to activity restrictions, foreign bank limitations, and diversification guidelines than domestic banks.

4.5.5 Robustness Checks To assess the reliability of our results, in this section we conduct a series of robustness checks with an alternative econometric model and different subsamples. We consider an alternative regression specification to assess whether our findings are consistent under different estimations. Specifically, instead of controlling for bank-fixed effects, we include country-fixed effects in the regression to control for country differences and year-fixed effects to control for time-varying effects. The results are in Panel A in Table 4.7. The relationships between bank regulation variables and H-statistics are consistent with our findings, except that Governmentowned banks and Deposit insurance ratio become insignificant. Thus, the results confirm that our results on bank regulation and competition are robust to alternative regression specifications. As a robustness check, we also performed our analysis across different subsamples. First, we replicate the analysis on the subsample limited to commercial banks. The result in Panel B in Table 4.7 shows that our main findings still hold for commercial banks. Second, as our sample countries cover the BRICS (Brazil, Russia, India, China, South Africa) countries, it is therefore interesting and warranted to perform our analysis on subsamples of banks in the BRICS countries and check whether the impact of bank regulation on competition changes in these five biggest and most rapidly growing emerging economies. The results in Panel C in Table 4.7 show that the findings of bank regulation variables are consistent with our previous conclusions. Third, Bonin et al. (2005) show that data from Bankscope suffer from several problems and are less accurate, especially for transition countries among the former Soviet republics. To check that our results are not driven by potential data problems, we perform an analysis on the subsample without banks in the former Soviet republics (see Panel D in Table 4.7). The results were largely unchanged. Based on the subsample analysis, we can conclude that our results are robust based on alternative sample selection criteria.

4.6 Conclusion This study examines the influence of bank regulation and supervision on competitive conditions using cross-country data from 23 emerging economies between 1996 and 2016. Both structural and nonstructural approaches are employed to measure competition in the banking system. The main results indicate that banking systems with higher concentration and fewer activity restrictions and entry barriers are more

(15.33)

0.00414***

Deposit insurance ratio

Private monitoring index

Diversification index

Official supervisory power

Capital regulatory index

Entry barriers

Foreign bank limitation

Activity restriction

(−1.44)

Government-owned banks −0.00168

Bank concentration

Panel A: Alternative specification

(1)

(−2.48)

−0.0148**

(2)

(4.72)

0.0757***

(3)

(−12.45)

−0.0987***

(4)

Table 4.7 Impact of bank regulation and supervision on competition: robustness check

(4.43)

0.0377***

(5)

(4.88)

0.0347***

(6)

(−8.25)

−0.153***

(7)

(5.18)

0.0781***

(8)

(continued)

(0.48)

0.0236

(9)

4.6 Conclusion 123

0.307

R-squared

(11.07)

Official supervisory power

Capital regulatory index

Entry barriers

Foreign bank limitation

Activity restriction

(−1.65)

Government-owned banks −0.00226*

Bank concentration

0.00346***

9365

Number of observations

Panel B: Commercial banks

Yes

Country and year fixed effects

(−2.61)

−0.0201***

0.194

12,714

Yes

Yes

(−1.61)

(−2.87)

Yes

(2)

−0.192

(1)

−0.281***

Bank and country control variables

Constant

Table 4.7 (continued) (3)

(3.46)

0.0684***

0.193

12,693

Yes

Yes

(−5.15)

−0.571***

(4)

(−11.00)

−0.109***

0.199

12,856

Yes

Yes

(3.30)

0.333***

(5)

(2.73)

0.0316***

0.194

12,645

Yes

Yes

(−4.76)

−0.704***

(6)

(2.57)

0.0261**

0.201

12,856

Yes

Yes

(−4.25)

−0.559***

(7)

0.267

11,559

Yes

Yes

(0.68)

0.0737

(8)

0.237

12,465

Yes

Yes

(−4.92)

−1.123***

(9)

(continued)

0.335

7441

Yes

Yes

(−7.07)

−1.084***

124 4 The Impact of Bank Regulation and Supervision on Competition

0.196

R-squared

(1.71)

Activity restriction

(−15.33)

Government-owned banks −0.0961***

Bank concentration

0.000580*

5712

Number of observations

Panel C: BRICS

Yes

Bank and year fixed effects

(−4.74)

−0.0838***

0.168

8100

Yes

Yes

(2.33)

Yes

0.667**

(1.92)

(2)

0.521*

(1)

Bank and country control variables

Constant

Deposit insurance ratio

Private monitoring index

Diversification index

Table 4.7 (continued)

0.170

8112

Yes

Yes

(0.89)

0.236

(3)

0.168

8236

Yes

Yes

(4.42)

1.265***

(4)

0.168

8043

Yes

Yes

(0.80)

0.246

(5)

0.173

8236

Yes

Yes

(0.80)

0.257

(6)

(7)

0.129

7378

Yes

Yes

(3.21)

0.875***

(−4.87)

−0.122***

0.183

7934

Yes

Yes

(−0.12)

−0.0517

(2.28)

0.0547**

(8)

(continued)

0.334

4391

Yes

Yes

(−2.35)

−0.731**

(0.46)

0.0305

(9)

4.6 Conclusion 125

Yes

3947

0.684

Bank and year fixed effects

Number of observations

R-squared

0.550

5429

Yes

Yes

(18.68)

(15.66)

Yes

6.853***

(2)

6.139***

(1)

Bank and country control variables

Constant

Deposit insurance ratio

Private monitoring index

Diversification index

Official supervisory power

Capital regulatory index

Entry barriers

Foreign bank limitation

Table 4.7 (continued) (3)

0.545

5429

Yes

Yes

(13.97)

6.004***

(4.74)

0.216***

0.543

5429

Yes

Yes

(15.91)

6.326***

(−4.74)

−0.0775***

(4)

0.539

5429

Yes

Yes

(14.35)

6.067***

(4.74)

0.0368***

(5)

0.539

5429

Yes

Yes

(11.58)

5.587***

(4.74)

0.0366***

(6)

0.578

4343

Yes

Yes

(5.66)

4.111***

(−4.74)

−0.195***

(7)

0.538

5429

Yes

Yes

(18.89)

8.230***

(4.74)

0.129***

(8)

(continued)

0.805

3044

Yes

Yes

(−217.35)

−128.5***

(−4.14)

0.442***

(9)

126 4 The Impact of Bank Regulation and Supervision on Competition

(1)

(2)

(9.29)

0.00375***

Deposit insurance ratio

Private monitoring index

Diversification index

Official supervisory power

Capital regulatory index

Entry barriers

Foreign bank limitation

Activity restriction

(−1.52)

Government-owned banks −0.00474

Bank concentration

(−2.69)

−0.0194***

Panel D: Subsample without transition countries

Table 4.7 (continued)

(3.09)

0.0510***

(3)

(−10.54)

−0.0971***

(4)

(3.40)

0.0347***

(5)

(4.94)

0.0397***

(6)

(−6.80)

−0.144***

(7)

(5.18)

0.0916***

(8)

(continued)

(5.36)

0.271***

(9)

4.6 Conclusion 127

8810

0.197

Number of observations

R-squared

0.136

11,967

Yes

(3)

0.137

11,946

Yes

Yes

(0.18)

0.0522

(4)

0.139

12,109

Yes

Yes

(3.10)

0.979***

0.136

11,898

Yes

Yes

(−0.34)

0.145

12,109

Yes

Yes

(0.48)

(6) 0.145

(5) −0.115

(7)

0.153

10,812

Yes

Yes

(2.09)

0.607**

(8)

0.172

11,718

Yes

Yes

(−0.81)

−0.289

(9)

0.322

6756

Yes

Yes

(3.35)

0.955***

The dependent variable is H-statistic which is estimated based on Eq. (4.1). Our sample period is 1996–2016. Bank-level controls include TCD, LLP, NIIC, ROA, and log TA. Country control variables include Market capitalization, GDP per capita, GDP growth, Inflation, and Interest rate. Regulation and supervision variables include Bank concentration, Government-owned banks, Activity restriction, Foreign bank limitation, Entry barriers, Capital regulatory index, Official supervisory power, Diversification index, Private monitoring index, and Deposit insurance ratio. All variables are defined in Table 4.1. Bank-level control variables and country-level variables are also included in the regressions but not reported for brevity. Reported t-statistics (in parentheses) are based on standard errors that are heteroskedasticity consistent and clustered at the bank level. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

Yes

Bank and year fixed effects

Yes

(1.37)

Yes

(0.96)

(2)

0.425

(1)

0.284

Bank and country control variables

Constant

Table 4.7 (continued)

128 4 The Impact of Bank Regulation and Supervision on Competition

4.6 Conclusion

129

competitive. The results also confirm that lowering foreign bank limitations and increasing capital strictness and official supervisory power also enhances competition in the banking sector. Our evidence also highlights that banking systems with fewer government-owned banks and diversification guidelines and higher private monitoring of banks and deposit insurance coverage tend to more competitive. In addition, our analyses indicate that during a bank crisis the relationship between activity restrictions, entry barriers, diversification guidelines, and competition is intensified while the positive effect of foreign bank limitation, capital stringency, official supervisory power, and private monitoring on competitive condition diminishes. Moreover, by examining the impact of bank regulation and supervision on foreign vs. domestic banks, our findings show that foreign banks are more sensitive to official supervisory power and private monitoring and less sensitive to activity restrictions, foreign bank limitations, and diversification guidelines. Our findings raise several important policy implications for financial regulators in emerging economies. First, to increase competition in the banking system, regulators need to be more cautious about evaluating and approving consolidation between banks in order to prevent excessive concentration in the banking system at the country level. Financial authorities should also reduce policy interventions in the banking system. Second, to improve the efficiency of capital allocation in the banking system, regulators also should seek to reduce activity restrictions and lower (foreign) bank entry barriers. Third, a certain level of capital strictness and official supervisory power should be maintained to reduce bank risk-taking behavior and enhance effective risk management, and this is more important for foreign banks that seek to take advantage of differences in regulations between the parent country and host countries. Finally, deposit insurance fosters competition among banks, and increasing deposit insurance coverage should be encouraged to promote competition and efficiency in the banking system. This research has a number of limitations. For instance, the data on bank regulation were obtained from the database in Barth et al. (2013), which only covers data until 2011. After 2010, new regulations and supervision were designed to respond to the global financial crisis (Claessens & Kodres, 2014), which, in turn, could reflect different findings across our sample countries. During our sample period, especially during the global financial crisis, some banks were bailed out by their government. Therefore, this raises concern regarding potential bias introduced by the bailed-out banks. Therefore, future research should consider the changes in bank regulation and supervision in emerging economies in recent years, and detailed information on the bailed-out banks in emerging economies also should be collected and examined in future studies.

130

4 The Impact of Bank Regulation and Supervision on Competition

References Agoraki, M.-E. K., Delis, M. D., & Pasiouras, F. (2011). Regulations, competition and bank risktaking in transition countries. Journal of Financial Stability, 7(1), 38–48. Angelini, P., & Cetorelli, N. (2003). The effects of regulatory reform on competition in the banking industry. Journal of Money, Credit and Banking, 35(5), 663–684. Anginer, D., & Demirguc-Kunt, A. (2014). Has the global banking system become more fragile over time? Journal of Financial Stability, 13, 202–213. Anginer, D., Demirguc-Kunt, A., & Zhu, M. (2014a). How does competition affect bank systemic risk? Journal of Financial Intermediation, 23(1), 1–26. Anginer, D., Demirguc-Kunt, A., & Zhu, M. (2014b). How does deposit insurance affect bank risk? Evidence from the recent crisis. Journal of Banking & Finance, 48, 312–321. Apergis, N. (2015). Competition in the banking sector: New evidence from a panel of emerging market economies and the financial crisis. Emerging Markets Review, 25, 154–162. Arellano, M. (1987). PRACTITIONERS’ CORNER: Computing robust standard errors for withingroups estimators. Oxford Bulletin of Economics and Statistics, 49(4), 431–434. Bain, J. S. (1951). Relation of profit rate to industry concentration: American manufacturing, 1936– 1940. The Quarterly Journal of Economics, 65(3), 293–324. Barth, J. R., Caprio, G., & Levine, R. (2004). Bank regulation and supervision: What works best? Journal of Financial Intermediation, 13(2), 205–248. Barth, J. R., Caprio, G. J., & Levine, R. (2013). Bank regulation and supervision in 180 countries from 1999 to 2011. Journal of Financial Economic Policy, 5(2), 111–219. Beck, T., Crivelli, J. M., & Summerhill, W. (2005). State bank transformation in Brazil—choices and consequences. Journal of Banking & Finance, 29(8), 2223–2257. Bikker, J. A., & Haaf, K. (2002). Competition, concentration and their relationship: An empirical analysis of the banking industry. Journal of Banking & Finance, 26(11), 2191–2214. Bikker, J. A., Shaffer, S., & Spierdijk, L. (2012). Assessing competition with the Panzar-Rosse model: The role of scale, costs, and equilibrium. Review of Economics and Statistics, 94(4), 1025–1044. Bonin, J. P., Hasan, I., & Wachtel, P. (2005). Privatization matters: Bank efficiency in transition countries. Journal of Banking & Finance, 29(8), 2155–2178. Boone, J. (2008). A new way to measure competition. The Economic Journal, 118(531), 1245–1261. Brissimis, S. N., Delis, M. D., & Papanikolaou, N. I. (2008). Exploring the nexus between banking sector reform and performance: Evidence from newly acceded EU countries. Journal of Banking & Finance, 32(12), 2674–2683. Chortareas, G. E., Girardone, C., & Ventouri, A. (2012). Bank supervision, regulation, and efficiency: Evidence from the European Union. Journal of Financial Stability, 8(4), 292–302. Chue, T. K., & Cook, D. (2008). Emerging market exchange rate exposure. Journal of Banking & Finance, 32(7), 1349–1362. Claessens, S., Demirgüç-Kunt, A., & Huizinga, H. (2001). How does foreign entry affect domestic banking markets? Journal of Banking & Finance, 25(5), 891–911. Claessens, S., & Kodres, L. (2014). The regulatory responses to the global financial crisis: Some uncomfortable questions (No. 14–46). International Monetary Fund Working Paper. Claessens, S., & Laeven, L. (2004). What drives bank competition? Some international evidence. Journal of Money, Credit and Banking, 36(3), 563–583. Clark, E., Radi´c, N., & Sharipova, A. (2018). Bank competition and stability in the CIS markets. Journal of International Financial Markets, Institutions and Money, 54, 190–203. Coccorese, P. (2004). Banking competition and macroeconomic conditions: A disaggregate analysis. Journal of International Financial Markets, Institutions and Money, 14(3), 203–219. Coccorese, P. (2009). Market power in local banking monopolies. Journal of Banking & Finance, 33(7), 1196–1210. De Haas, R., & Van Lelyveld, I. (2006). Foreign banks and credit stability in Central and Eastern Europe. A panel data analysis. Journal of Banking & Finance, 30(7), 1927–1952.

References

131

Delis, M. D. (2012). Bank competition, financial reform, and institutions: The importance of being developed. Journal of Development Economics, 97(2), 450–465. Delis, M. D., Molyneux, P., & Pasiouras, F. (2011). Regulations and productivity growth in banking: Evidence from transition economies. Journal of Money, Credit and Banking, 43(4), 735–764. Demirguc-Kunt, A., Laeven, L., & Levine, R. (2004). Regulations, market structure, institutions, and the cost of financial intermediation. Journal of Money, Credit and Banking, 36(3), 593–622. Dick, A. A. (2006). Nationwide branching and it’s impact on market structure, quality, and bank performance. The Journal of Business, 79(2), 567–592. Doidge, C., Griffin, J., & Williamson, R. (2006). Measuring the economic importance of exchange rate exposure. Journal of Empirical Finance, 13(4), 550–576. Fang, Y., Hasan, I., & Marton, K. (2014). Institutional development and bank stability: Evidence from transition countries. Journal of Banking & Finance, 39, 160–176. Fernández De Guevara, J., Maudos, J., & Perez, F. (2005). Market power in European banking sectors. Journal of Financial Services Research, 27(2), 109–137. Fu, X., & Heffernan, S. (2009). The effects of reform on China’s bank structure and performance. Journal of Banking & Finance, 33(1), 39–52. Fu, X. M., Lin, Y. R., & Molyneux, P. (2014). Bank competition and financial stability in Asia Pacific. Journal of Banking & Finance, 38, 64–77. Gelos, R. G., & Roldós, J. (2004). Consolidation and market structure in emerging market banking systems. Emerging Markets Review, 5(1), 39–59. Gormley, T. A., & Matsa, D. A. (2014). Common errors: How to (and not to) control for unobserved heterogeneity. Review of Financial Studies, 27(2), 617–661. Hasan, I., Wachtel, P., & Zhou, M. (2009). Institutional development, financial deepening and economic growth: Evidence from China. Journal of Banking & Finance, 33(1), 157–170. Hellmann, T. F., Murdock, K. C., & Stiglitz, J. E. (2000). Liberalization, moral hazard in banking, and prudential regulation: Are capital requirements enough? American Economic Review, 90(1), 147–165. Jeon, B. N., Olivero, M. P., & Wu, J. (2011). Do foreign banks increase competition? Evidence from emerging Asian and Latin American banking markets. Journal of Banking & Finance, 35(4), 856–875. Keeley, M. C. (1990). Deposit insurance, risk, and market power in banking. The American Economic Review, 80(5), 1183–1200. Khan, H. H., Kutan, A. M., Ahmad, R. B., & Gee, C. S. (2017). Does higher bank concentration reduce the level of competition in the banking industry? Further evidence from South East Asian economies. International Review of Economics & Finance, 52, 91–106. Koetter, M., Kolari, J. W., & Spierdijk, L. (2012). Enjoying the quiet life under deregulation? Evidence from adjusted Lerner indices for US banks. Review of Economics and Statistics, 94(2), 462–480. Laeven, L., & Valencia, F. (2010). Resolution of banking crises: the good, the bad and the ugly. International Monetary Fund Working Paper, WP/10/146. Laeven, L., & Valencia, F. (2018). Systemic Banking Crises Revisited. IMF Working Paper No. 206. International Monetary Fund. Lerner, A. P. (1934). The concept of monopoly and the measurement of monopoly power. The Review of Economic Studies, 1(3), 157–175. Matthews, K., Murinde, V., & Zhao, T. (2007). Competitive conditions among the major British banks. Journal of Banking & Finance, 31(7), 2025–2042. Matutes, C., & Vives, X. (2000). Imperfect competition, risk taking, and regulation in banking. European Economic Review, 44(1), 1–34. Maudos, J., & Solís, L. (2011). Deregulation, liberalization and consolidation of the Mexican banking system: Effects on competition. Journal of International Money and Finance, 30(2), 337–353.

132

4 The Impact of Bank Regulation and Supervision on Competition

Noman, A. H. M., Gee, C. S., & Isa, C. R. (2018). Does bank regulation matter on the relationship between competition and financial stability? Evidence from Southeast Asian countries. PacificBasin Finance Journal, 48, 144–161. Ongena, S., Popov, A., & Udell, G. F. (2013). “When the cat’s away the mice will play”: Does regulation at home affect bank risk-taking abroad? Journal of Financial Economics, 108(3), 727–750. Panzar, J. C., & Rosse, J. N. (1987). Testing for “monopoly” equilibrium. The Journal of Industrial Economics, 35(4), 443–456. Ryan, R. M., O’Toole, C. M., & McCann, F. (2014). Does bank market power affect SME financing constraints? Journal of Banking and Finance, 49, 495–505. https://doi.org/10.1016/j.jbankfin. 2013.12.024. Schaeck, K., & Cihák, M. (2014). Competition, efficiency, and stability in banking. Financial Management, 43(1), 215–241. Soedarmono, W., Machrouh, F., & Tarazi, A. (2011). Bank market power, economic growth and financial stability: Evidence from Asian banks. Journal of Asian Economics, 22(6), 460–470. Soedarmono, W., Machrouh, F., & Tarazi, A. (2013). Bank competition, crisis and risk taking: Evidence from emerging markets in Asia. Journal of International Financial Markets, Institutions and Money, 23, 196–221. Tan, Y. (2016). The impacts of risk and competition on bank profitability in China. Journal of International Financial Markets, Institutions and Money, 40, 85–110. Tan, Y., & Floros, C. (2018). Risk, competition and efficiency in banking: Evidence from China. Global Finance Journal, 35, 223–236. Tompson, W. (2004). Banking reform in Russia: Problems and prospects. OECD Economics Department Working Papers, No. 410, OECD publishing. Vesala, J. (1995). Testing for competition in banking: Behavioral evidence from Finland. Bank of Finland Studies Working Paper. Bank of Finland. Helsinki. Williams, J., & Nguyen, N. (2005). Financial liberalisation, crisis, and restructuring: A comparative study of bank performance and bank governance in South East Asia. Journal of Banking & Finance, 29(8), 2119–2154. Zhao, T., Casu, B., & Ferrari, A. (2010). The impact of regulatory reforms on cost structure, ownership and competition in Indian banking. Journal of Banking & Finance, 34(1), 246–254.

Chapter 5

Bank Competition, Regulation, and Efficiency

5.1 Overview This study investigates the impact of bank competition and regulation on bank efficiency in the Asia–Pacific region during 2001–2016. The result reveals that market power is positively related to bank efficiency. We also find that stringent activity restrictions, strong official supervisory power, and low capital requirements are associated with high bank efficiency. Furthermore, market power has a stronger efficiency-increasing effect in a banking system characterized by the activity restrictions, supervisory power, and capital requirements described above. Foreign banks operating under increased activity restrictions in a host country with strong official supervisory power have relatively high efficiency.

5.2 Introduction Banks, a special type of enterprise highly regulated by the government, constitute an important part of the financial system. Since the 1980s, the global financial market has become increasingly integrated, and both the developed and developing countries have embarked on financial liberalization. In the process of financial liberalization, deregulation, and increased integration, banks have expanded their services and engaged in greater risk-taking activities to improve their productive efficiency (Luo et al., 2016). With increased competition following financial liberalization and deregulation, banks have to improve their efficiency (Andrie¸s & C˘apraru, 2014; Schaeck & Cihák, 2008). In this paper, we examine how bank competition and regulation affect bank efficiency, using a sample of 1261 banks across 28 Asia Pacific countries for the period This chapter is co-authored with Xiaolin Li and was published in Asia-Pacific Journal of Accounting & Economics. © Shanghai Jiao Tong University Press 2021 S. Li, Financial Regulation and Bank Performance, Contributions to Finance and Accounting, https://doi.org/10.1007/978-981-16-3509-0_5

133

134

5 Bank Competition, Regulation, and Efficiency

2001–2016. We employ a stochastic frontier analysis (SFA) model to estimate bank efficiency. Our findings confirm that market power is positively related to bank efficiency. Increased activity restrictions, strong official supervisory power, and low capital requirements are associated with high bank efficiency. Furthermore, market power has a stronger efficiency-increasing effect in a banking system characterized by the activity restrictions, supervisory power, and capital requirements described above. Foreign banks operating under strict activity restrictions in a host country with strong official supervisory power are highly efficient. We choose the Asia–Pacific region for this sample study, considering the limited research on the above relationship so far in this part of the world.1 In addition, since the 1990s, governments across the region have implemented financial reforms, including financial liberalization, bank privatization, widening access to foreign banks, and restructuring of the national banking systems, to enhance the competitive strength of the banking sector. For instance, the Chinese government has recapitalized state-owned commercial banks and replaced non-performing loans by injecting new capital, in addition to undertaking interest rate liberalization, reducing activity restrictions, and privatization of banks by selling shares on the market and introducing minority foreign ownership stakes (Fang et al., 2019; Hasan et al., 2009). In India, the deregulation of the banking system includes interest rate liberalization and removal of bank entry restrictions, private ownership restrictions, and activity restrictions aimed at promoting competition. Prudential norms on non-performing loans and capital requirements are also implemented to strengthen financial stability (Zhao et al., 2010). The banking sector in Asia–Pacific region underwent a restructuring process after the 1997–1998 Asian financial crisis and the recent global financial crisis (Noman et al., 2018; Williams & Nguyen, 2005). Our study contributes to the literature in several ways. First, we focus on the Asia– Pacific region. Since the 1990s, the development of information technology alongside financial reforms have reshaped the banking systems in the Asia–Pacific region (see Das & Ghosh, 2006; Koetter et al., 2012; Marinˇc, 2013). Technology development and financial reforms have facilitated financial market development, increased the market share of private credit, reduced the financing cost of private credit (Bae & Goyal, 2009; Djankov et al., 2007), and encouraged banks to undertake risky but more value-added activities (John et al., 2008; Laeven & Levine, 2009). Second, we employ SFA to estimate bank efficiency and evaluate the impact of competition and regulation on bank efficiency while accounting for bank-specific characteristics, market structure, financial development, and macroeconomic factors that may impact bank performance. We use an alternative definition of competition, perform different robustness analyses, and use different estimation methods for different subsamples. 1 Previous

studies examine the national banking systems in the Asia–Pacific region from different perspectives. Soedarmono et al. (2011) and Liu et al. (2012) analyze the competition and financial stability relationship based on a sample of 12 Asian and 4 South East Asian countries. Fu et al. (2014) consider the tradeoff between competition and financial stability based on 14 Asia–Pacific countries, whereas Noman et al. (2018) examine the impact of bank regulation on the relationship between bank competition and financial stability based on five South East Asian countries.

5.2 Introduction

135

Third, our analysis enhances previous studies by evaluating the impact of bank regulation on the relationship between competition and efficiency, which could provide implications for policymakers. The rest of the paper is structured as follows. Section 5.3 reviews the related literature. Section 5.4 presents the variable definitions and research design. Section 5.5 provides an empirical analysis and robustness checks. Section 5.6 concludes the paper.

5.3 Related Literature Review 5.3.1 The Impact of Competition on Bank Efficiency Under the traditional competition-efficiency hypothesis, competition improves resource allocation and transfers profit to more efficient firms and, thus, increases the efficiency of firms (e.g., Olley & Pakes, 1996; Stiroh, 2000; Tirole, 1988). As competition increases, banks are forced to specialize, providing services at lower prices and minimizing costs (Zarutskie, 2013). Competition also forces banks to improve lending efficiency and reduce credit risk (Dick & Lehnert, 2010). Schaeck and Cihák (2008) examine how competition among banks impacts efficiency, and indicate that competition has a positive effect on cost and profit efficiency. Efficiency improvement would lead to bank expansion, growth in market share, and in turn a more concentrated market (Andrie¸s & C˘apraru, 2014). However, under the alternative competition-inefficiency hypothesis, greater competition leads to higher inefficiency. In a more competitive banking system, banks may respond to an increased proportion of low-quality borrowers by reducing their lending standards and expanding credit, and borrower-specific information becomes more dispersed, as each bank is informed about a smaller pool of borrowers (Dell’Ariccia & Marquez, 2006). This reduces banks’ screening ability, increasing information asymmetry between banks and borrowers, and creating inefficiency as more low-quality borrowers obtain financing (Marquez, 2002). Weill (2004) and Casu and Girardone (2009a, b) test the effect of competition on bank efficiency from the EU and confirm a negative relationship. Maudos and Fernández de Guevara (2007) and Koetter et al. (2012) analyze the relationship between market power in loan and deposit markets and efficiency and confirm a positive relationship between market power and cost X-efficiency, and reject the so-called quiet life hypothesis proposed by Berger and Hannan (1998).

5.3.2 Literature Review on Bank Regulation and Bank Efficiency The traditional theory also confirms that bank regulation affects bank efficiency. Banking regulation is a combination of supervisory and restrictive policies aiming

136

5 Bank Competition, Regulation, and Efficiency

to both protect the banking sector from excessive risk-taking, and minimize moral hazard (Barth et al., 2004). In this section, we review the literature pertaining to regulations that influence bank competition and efficiency, focusing on activity restrictions, capital requirements, and supervisor power. These are the main concerns of regulators and supervisors for policy implications designed to promote competition and efficiency in the banking industry.2 Activity restrictions refer to regulatory restrictions on a bank’s ability to engage in securities, insurance, and real estate activities. These are the key determinants of the scope of a bank’s operations, which may influence both bank competition and efficiency (Barth et al. 2013a, b). Given the influence of activity restrictions, the proponents claim that activity restrictions prevent the formation of large banks and complex structures, which are difficult to monitor or discipline. By forcing banks to focus only on their specializations, activity restrictions would lead to improved bank efficiency. Pasiouras et al. (2009) provide empirical support for the proponents and show that activity restrictions improve profit efficiency. However, bank activity restrictions also prevent banks from diversifying income sources and achieving economies of scope and scale, which in turn hinder the bank’s efficiency and performance (Claessens & Klingebiel, 2001). Chortareas et al. (2012) and Barth et al. (2013a, b) empirically confirm that tighter activity restrictions are negatively associated with bank efficiency. Capital requirements indicate the minimum capital requirements that bank needs to hold against its risk-weighted assets, and are considered as a regulatory tool for building a capital buffer and improving financial stability (Laeven & Levine, 2009). Concerning the theories of capital requirements, Allen et al. (2011) identified a positive relationship between capital requirements and bank efficiency, as capital requirements incentivize bank managers to improve risk management and force bank investors to control risk-taking activities. In support of this theoretical view, Ben Naceur and Kandil (2009) and Chortareas et al. (2012) confirm that stringent capital requirements lead to improved bank efficiency. On the other hand, some studies, including Pasiouras et al. (2009), argue that because of the agency cost from the shareholder–debtholder relationship, higher capital requirements may hinder bank efficiency. The importance of supervisory power has been stressed for maintaining a stable and well-performing banking system (Basel Committee on Banking Supervision, 2012), but there are conflicting views about the importance of supervisory power. Some theoretical studies (e.g., Beck et al., 2006) suggest that supervisors require higher power to ensure their independence, and more resources to prevent banks from engaging in undesirable activities and taking excessive risks. Following this view, higher official supervisory power leads to improved bank efficiency. Conversely, other analyses claim that higher supervisory power leads to higher corruption and larger unofficial economies (Djankov et al., 2002), which significantly contributed to deepening several recent systemic banking crises (Quintyn & Taylor, 2003). Given this view, banks are forced to make sub-optimal decisions under higher supervisory 2 See

Barth et al. (2004) for the definitions and measurement of bank regulation variables.

5.3 Related Literature Review

137

power, which in turn leads to reduced bank efficiency. Empirical analyses on supervisory power indicate conflicting views, similar to those in the theoretical studies. Pasiouras et al. (2009) and Chortareas et al. (2012) confirm that supervisory power is positively associated with bank efficiency, while Barth et al. (2004) and DemirgüçKunt and Detragiache (2011) show that greater supervisory power leads to higher risks and has a weak effect on bank efficiency. In sum, previous studies show that bank regulation might influence bank efficiency in different ways. However, it is yet to be empirically examined how regulations influence bank efficiency in a competitive environment in a disaggregated manner, especially in the Asia–Pacific region.

5.4 Variables Definition and Research Design 5.4.1 Bank Efficiency Estimation and Main Variables Definition (1)

Estimation of bank efficiency

We use stochastic frontier analysis (SFA) to estimate bank profit and cost efficiency, following Bonin et al. (2005) and Luo et al. (2016). The SFA model is defined as follows: TPit ln W3it



TCit W3it



    2 Qmit Wmit = α0 + + α1m ln β1m ln W3it W3it m=1 m=1   2      2 Qmit Q1it Q2it 1 ln α2m ln + α3 ln + 2 m=1 W3it W3it W3it   2     2 Wmit Wmit W2it 1 ln β2m ln + β3 ln + 2 m=1 W3it W3it W3it      2  2  Qmit Wnit ln + α41 ln(EQit ) γmn ln + W3it W3it m=1 n=1 α52 2 α42 T + (ln(EQit ))2 + α51 Tt + 2 2 t  2       Qmit Wmit μm ln + πm ln + Tt W3it W3it m=1  2       Qmit Wmit ρm ln + τm ln + ln EQit W3it W3it m=1 2 



+ α6 Tt ln EQit + uit + νit

(5.1)

138

5 Bank Competition, Regulation, and Efficiency

In Eq. (5.1), TPit (TCit ) indicates the total profit (total cost) of bank i in year t. Qm indicates the bank output variables, which are measured by (1) the ratio of interest income to gross loans (denoted by Q1 ) and (2) the ratio of non-interest income to other earning assets (denoted by Q2 ). Wn indicates input price variables, including (1) the ratio of interest expenses to total deposits (denoted by W1 ), (2) the ratio of other operating expenses to fixed assets (denoted by W2 ), and (3) the price of labor, which is defined as the ratio of personnel expenses to total assets (denoted by W3 ). To control for scale bias and potential heteroskedasticity, the dependent variables (TPit and TCit ), input prices (W1 and W2 ), and output variables (Q1 and Q2 ) are normalized by the price of labor (W3 ). Following Pasiouras et al. (2009), we also include the variable ln(EQ), which is defined as the logarithm of equity as a quasi-fixed input to control for the risk preference difference between banks. As our sample period covers 2001–2016, in order to control for the impact of technological development on bank efficiency, we introduce a time trend variable, T . The error term in the SFA model is decomposed into two parts: uit and νit . The error term uit indicates the profit (cost) inefficiency and is distributed as a truncated normal variable, and νit indicates random error and is distributed as a normal variable.3 The inefficiency for each bank is defined as Profit Ineffit = exp(uit ) for profit inefficiency and Cost Ineffit = exp(−uit ) for cost inefficiency. (2)

Measure of bank competition

In our analysis, to measure the competition in banking industry, we estimate the Lerner index and evaluate banks’ competitive behavior. The Lerner index is defined as follows: Lerner Indexit = (Pit − MCit )/Pit

(5.2)

The Lerner index is a relative mark-up of price over marginal cost and measures firm market power (Lerner, 1934). The higher the mark-up, the greater is the market power. The Lerner index ranges from 0 in the case of perfect competition to 1 in the case of monopoly. In our analysis, Pit indicates output price, which is measured by the ratio of total revenue to total assets for bank i in year t, and MCit indicates the marginal cost of bank i in year t.4 MCit is defined as in Eq. (5.3): MCit =

TCit (β1 + β2 lnQ3it + β9 lnW1it + β10 lnW2it + β11 lnW3it ) Q3it

(5.3)

The translog total cost function is

3 See

the detailed explanation of the SFA model in Battese and Coelli (1995). it as the ratio of total revenues to total assets. See also Koetter et al. (2012) and Fu et al. (2014).

4 Based on earlier studies, we define the output price P

5.4 Variables Definition and Research Design

139

3  β2 β2+m lnWmit (lnQ3it )2 + 2 m=1  3  3  1 2 + β5+m (lnWmit ) + lnQ3it β8+m lnWmit 2 m=1 m=1

lnTCit = β0 + β1 lnQ3it +

+ β12 (lnW1it )(lnW2it ) + β13 (lnW1it )(lnW3it ) + β14 (lnW2it )(lnW3it ) + εit

(5.4)

where TCit is defined as the total cost of bank i in year t. W1 , W2 , and W3 are input price variables as defined in Sect. 5.4.1. Q3 is the output variable, which is a proxy for total assets. Equation (5.4) is estimated based on the stochastic cost frontier analysis for each country in each year (see also Koetter et al., 2012; Fu et al., 2014). We alternatively employ the recent competition measure, Boone indicator, defined in Boone (2001, 2008). The Boone indicator captures the link between firm performance and efficiency; firms with higher efficiency have superior performance, which results in higher profits. The hypothesis of the Boone indicator is that the relationship between efficiency and profits is increasing in the degree of competition. The Boone model is defined as follows: ln TPit = α + β ln MCit + εit

(5.5)

where TPit indicates the total profit, MCit is the marginal cost of bank i in year t, and β is defined as the Boone indicator. The Boone indicator β is negative and decreasing in the level of competition. We use the log–log specification to deal with potential heteroskedasticity. (3)

Bank regulatory variables

Given the impact of bank regulations on bank efficiency, we employ three indexbased measures to control for country-specific differences in banking regulations, following Chortareas et al. (2012), Barth et al. (2013a, b), and Luo et al. (2016). The variable Activity restrictions is related to bank activity restrictions and measures the degree to which banks are allowed to engage in securities, insurance, and real estate activities. A higher value indicates greater restrictions. The variable Capital stringency index measures both initial and overall capital stringency and estimates whether the capital requirements reflect certain risk elements and whether allowances are made for certain market value losses before the minimum capital adequacy is determined. The variable Official supervisory power indicates whether the supervisory authorities have the authority to take specific actions to prevent and correct problems, with a higher value indicating greater powers of supervision.

140

(4)

5 Bank Competition, Regulation, and Efficiency

Control variables

Bank-specific variables: We use a set of bank-specific variables to control for timevarying effects on bank performance. The probability of bank insolvency is proxied by Z-score (e.g., Demirgüç-Kunt & Huizinga, 2010; Laeven & Levine, 2009), which indicates the number of standard deviations by which returns would have to fall from the mean to wipe out all equity in the bank. A higher Z-score indicates a lower probability of bank failure.5 We define the variable TCD as the ratio of total customer deposits to total assets to control for the funding of the bank, LLP as the ratio of loan loss provision to gross loans to control for the loan quality, FA as the ratio of fixed assets to total assets to measure bank asset structure, and log (TA) as the logarithm of total assets to control for bank size. To control for the difference introduced by banks’ specializations, we also define two dummy variables, Commercial dummy and Savings dummy, which take the value of one for a commercial bank and a savings bank, respectively. Country control variables: We first consider the effect of market structure and define the variable Bank concentration, which indicates the degree of concentration of deposits in the five largest banks in a given country, and the variable Governmentowned banks, which measures the extent to which the banking system’s assets are government owned, to examine the impact of market structure on bank efficiency in a given country. During our sample period, most economies experienced financial reforms. Thus, we include different measures to account for the financial development of the banking system in the Asia–Pacific region. We first include the variable Financial openness, an index defined by Chinn and Ito (2008).6 The normalized index Financial openness ranges from 0 to 1, with a higher value indicating a higher level of financial openness. Second, we include the variable Financial freedom, which is obtained from the annual report of the Heritage Foundation and indicates the financial freedom of a given country. Financial freedom ranges from 0 to 100 and measures the extent of government regulation of financial services, the extent of state intervention in banks and other financial services, the difficulty of opening and operating financial services firms, and government influence on the allocation of credit.7 A higher value of Financial freedom indicates greater financial freedom. 5 The Z-score is calculated as follows: Z-score = (ROA

it +ETAit )/sd (ROA)it , where ROAit is defined as return on assets for bank i at year t, ETAit is the ratio of equity to assets, and sd (ROA) indicates the standard deviation of ROA. We employ a three-year rolling window to calculate sd (ROA) instead of the whole sample period, allowing for the time variation of sd (ROA). See Demirgüç-Kunt and Huizinga (2010) for a detailed discussion. 6 The Financial openness index is constructed according to information on four categories: (1) the presence of multiple exchange rates, (2) restrictions on current account transactions, (3) restrictions on capital account transactions, and (4) the requirement to surrender export proceeds. Information on each category is provided in the Annual Report on Exchange Arrangements and Exchange Restrictions, published by the International Monetary Fund (IMF); a category is assigned 1 if the restriction does not exist and 0 otherwise. 7 For detailed information about the index of economic freedom, defined by the Heritage Foundation, see https://www.heritage.org/index/about.

5.4 Variables Definition and Research Design

141

Finally, we include GDP growth and Inflation at country level to account for the impact of macroeconomic factors, and define the variable Crisis, which equals one if the country experienced a systemic financial crisis in a given year, to control for the impact of financial crisis on bank efficiency. Table 5.1 summarizes the definition and data sources of variables employed in this analysis.

5.4.2 Model Specification Based on the profit (cost) inefficiency estimated in Eq. (5.1), we investigate the determinants of profit (cost) inefficiency of banks. The model is specified as follows: Ineffit = α0 + β1 Compeitionit + γRegulationjt + × Bank_controlsit +  × Country_controlsjt + αi + ϕt + εit

(5.6)

where Ineff it represents the profit (cost) inefficiency for bank i at year t estimated based on Eq. (5.1), Compeitionit is the competition indicator, which is a proxy for the Lerner index, or Boone indicator, and Regulationjt denotes one of the bank regulatory characteristics of country j, including Activity restrictions, Capital stringency index, and Official supervisory power. Bank_controlsit indicates bank-specific variables, including Z-score, TCD, LLP, FA, log TA, Commercial dummy and Savings dummy. Country_controlsjt is a matrix of country-level control variables, including Bank concentration, Government-owned banks, Financial openness, Financial freedom; GDP growth, Inflation, and Crisis. Both bank fixed effects (αi ) and year fixed effects (ϕt ) are considered to control for unobserved time-invariant heterogeneity across banks, as well as macroeconomic factors and monetary policy, which may vary over time. We use the method proposed by Vogelsang (2012), who extend Driscoll and Kraay’s (1998) study with bank fixed effects to a setting with bank and year fixed effects, to estimate standard errors and hypothesis tests robust to heteroskedasticity, serial correlation, and spatial correlation.

5.4.3 Data Selection Our analysis focuses on 28 countries from Asia Pacific between 2001 and 2016. The financial data of each bank are collected from Bankscope. To avoid double counting, we use the consolidated financial information of each bank, if available, and unconsolidated reports otherwise. Data on financial openness are obtained from

142

5 Bank Competition, Regulation, and Efficiency

Table 5.1 Variables definition and data sources Variables

Definitions

Data sources

Variables for efficiency estimation TP

Total profit of bank in a given year, expressed in $1000

Bankscope

TC

Total cost of bank in a given year, expressed in $1000

Bankscope

Q1

The ratio of interest income to gross loans

Bankscope

Q2

The ratio of non-interest income Bankscope to other earning assets

W1

The ratio of interest expense to total deposits

Bankscope

W2

The ratio of other operating expenses to fixed assets

Bankscope

W3

The ratio of personnel expense to total assets

Bankscope

ln EQ

The logarithm of equity capital Bankscope of banking institution in a given year

T

The time trend variable

Self-definition

Lerner index

Lerner index is equal to the difference between asset price and marginal cost, normalized by asset price

Own calculation

Boone indicator

A measure of degree of competition, calculated as the elasticity of profits to marginal costs

Own calculation

Competition indicators

Bank-specific variables Z-score

Calculated as the sum of ROA Bankscope and equity-to-assets divided by the standard deviation of ROA of each bank over past 3 years

TCD

The ratio of total customer deposits to total assets

Bankscope

LLP

The ratio of loan loss provision to gross loans

Bankscope

FA

The ratio of fixed assets to total assets

Bankscope

log TA

The logarithm of total asset in $ Bankscope 1000 (continued)

5.4 Variables Definition and Research Design

143

Table 5.1 (continued) Variables

Definitions

Data sources

Foreign dummy

Takes a value of 1 if the bank is a foreign-owned bank, and 0 otherwise

Bankscope

Commercial dummy

Takes a value of 1 if the bank is a commercial bank and 0 otherwise

Bankscope

Saving dummy

Takes a value of 1 if the bank is a saving bank and 0 otherwise

Bankscope

The degree of concentration of deposits in the 5 largest banks

Barth et al. (2013a)

Market structure variables Bank concentration

Government-owned banks The extent to which the banking Barth et al. (2013a) system’s assets are government owned Financial development variables Financial openness

The Chinn-Ito index measures the degree of capital account openness. Construction is based on transactions from the IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions (AREAER)

Chinn and Ito (2008). This paper uses the 2015 version of the database, available from: http:// web.pdx.edu/ ito/Chinn-Ito website.htm

Financial freedom

Financial freedom component of the Heritage index of Economic Freedom

The Heritage Foundation. Available from: http://www.her itage.org/index/

Country control variables GDP growth

Annual growth rate of GDP at World Bank Development market prices based on constant Indicator Database local currency

Inflation

Inflation rate

Crisis

A binary variable which equals Laeven and Valencia (2018) one if the country was experiencing a systemic crisis in a given year, and zero otherwise

World Bank Development Indicator Database

Bank regulatory variables Activity restrictions

A variable measures a bank’s ability to engage in securities, insurance, and real estate activities. The ranges from 0 to 12, and a higher score indicates more restrictions on banks to engage in such activities

Barth et al. (2013a)

(continued)

144

5 Bank Competition, Regulation, and Efficiency

Table 5.1 (continued) Variables

Definitions

Capital regulatory index

A variable that captures both the Barth et al. (2013a) overall capital stringency and the initial capital stringency based on answers to eight questions. It ranges from 0 to 8, with a higher value indicating higher capital stringency

Data sources

Official supervisory power A variable indicates whether the Barth et al. (2013a) supervisory authorities have the authority to take specific actions to prevent and correct problems. It ranges from 0 to 14, with fourteen indicating the highest power of the supervisory authorities This table reports the definitions and data sources of the variables included in our analysis

Chinn and Ito (2008), and we use the 2015 version of the database.8 Data on the financial freedom index are collected from the annual report of the Heritage Foundation. We obtain the market structure variables and bank regulation data from Barth et al. (2013a, b),9 the country economy development variables are collected from the World Bank Development Indicator Database, and information on bank crises is obtained from Laeven and Valencia (2018). Based on the complete sample of banks obtained from Bankscope, we apply the following selection criteria: (1) we delete observations for which data on one of the variables employed in estimating profit (cost) inefficiency are missing; (2) we delete observations for which data on one of the variables employed in estimating the Lerner index are missing; (3) to estimate the Z-score, we employ the mean of return on assets (ROA) and equity to total asset ratio (ETA) and the standard deviation of ROA (sd (ROA)) over three years, retaining only those banks that have at least three consecutive years’ observations; (4) we delete the observations for which data on one of the bank-specific and country-level control variables are missing or drop out of the 5th and 95th percentiles of bank-specific variables.

8 As the information of financial openness for 2016 is not available, we use interpolation and extrap-

olation techniques, following Goetz et al. (2016), to fill in the missing observation of financial openness of 2016 for each country. 9 Information on bank regulation is provided by Barth et al. (2013a, b), but the surveys were conducted in 1999, 2002, 2006, and 2011. Following Anginer et al. (2014), we employ the previous survey data until the new survey data become available for matching the bank regulation and supervision variables with bank-specific variables and country control variables. Surveys were conducted in 1999 for the years 1998–2001, in 2002 for the years 2002–2005, in 2006 for the years 2006–2010, and in 2011 for the years 2011–2016.

5.4 Variables Definition and Research Design

145

We finally have 1261 banks and 9193 bank-year observations between 2000 and 2016. Our sample consists of commercial, savings, and co-operative banks from 28 Asia–Pacific countries. The data are inflation adjusted and expressed in USD.

5.4.4 Summary Statistics We first report the summary of profit inefficiency and cost inefficiency for 28 Asia– Pacific economies between 2000 and 2016 by country (Panel A), by year (Panel B), and group difference analysis (Panel C) in Table 5.2. Considering bank efficiency by country, we find that on average, banks in Korea have the highest profit and cost efficiency. Comparing profit and cost inefficiency by year, we confirm an increasing trend for both profit and cost efficiency from 2000 to 2016. Profit and cost efficiencies during the global financial crisis (2008–2010) are slightly higher than before the global financial crisis (2005–2007), but the difference is insignificant. We then provide the summary statistics in Table 5.3. Profit inefficiency ranges from 0.238 to 0.632, with an average of 0.467; cost inefficiency has a mean of 0.492 with a standard deviation of 0.120, and varies from 0.249 to 0.662. The mean of the Lerner index is 0.685 with a standard deviation 0.589, and the average Boone indicator is − 1.286, ranging from −10.39 to −0.166. For the bank regulation variables, an average country in our sample has activity restrictions of 8.02, a capital regulatory index of 6.81, and official supervisory power of 11.56.

5.5 Empirical Results 5.5.1 Basic Results We examine the impact of competition and regulation on bank efficiency, and Table 5.4 reports the basic results. The dependent variables, profit inefficiency and cost inefficiency, are reported in columns (1)–(5) and columns (6)–(10), respectively. The negative and statistically significant coefficients of the Lerner index for both profit and cost inefficiency in different specifications indicate that increased market power is associated with higher bank efficiency, which is consistent with the competitioninefficiency view of Maudos and Fernández de Guevara (2007). In examining the role of regulation on bank efficiency, first, the coefficient of Activity restrictions in columns (2) and (7) is negative and significant, suggesting that activity restrictions appear to be an effective regulatory tool to improve bank efficiency, supporting the findings of Pasiouras et al. (2009) and Djalilov and Piesse (2019). This is because strict bank activity restrictions drive banks to acquire greater expertise and specialization, and in turn become more profitable and cost efficient in specific market segments.

146

5 Bank Competition, Regulation, and Efficiency

Table 5.2 Summary of profit inefficiency and cost inefficiency according to countries and years Panel A: Mean by country Country

Profit inefficiency

Cost inefficiency

Observations

Australia

0.453

0.469

316

Azerbaijan

0.561

0.585

128

Bangladesh

0.504

0.526

406

Bhutan

0.536

0.562

13

Cambodia

0.533

0.562

64

China (Mainland)

0.383

0.409

740

Fiji

0.628

0.659

6

Hong Kong SAR (China)

0.451

0.479

310

India

0.465

0.485

882

Indonesia

0.485

0.506

674

Japan

0.467

0.500

2825

Kazakhstan

0.512

0.537

253

Korea

0.260

0.272

57

Kyrgyzstan

0.580

0.609

56

Malaysia

0.448

0.469

702

Maldives

0.310

0.340

12

Nepal

0.504

0.525

216

New Zealand

0.475

0.495

148

Pakistan

0.502

0.526

383

Papua new Guinea

0.506

0.551

9

Philippines

0.492

0.521

326

Samoa

0.599

0.634

6

Singapore

0.370

0.388

61

Sir Lanka

0.497

0.518

266

Tajikistan

0.601

0.625

40

Thailand

0.440

0.465

277

Tonga

0.614

0.646

7

Turkmenistan

0.538

0.561

10

Panel B: Mean by year Year

Profit inefficiency

Cost inefficiency

Observations

2001

0.541

0.566

348

2002

0.542

0.570

353

2003

0.540

0.570

340

2004

0.543

0.573

456 (continued)

5.5 Empirical Results

147

Table 5.2 (continued) Panel A: Mean by country Country

Profit inefficiency

Cost inefficiency

Observations

2005

0.542

0.572

489

2006

0.533

0.563

510

2007

0.524

0.554

538

2008

0.522

0.551

557

2009

0.524

0.552

556

2010

0.523

0.550

600

2011

0.516

0.544

648

2012

0.510

0.537

678

2013

0.511

0.537

664

2014

0.311

0.329

779

2015

0.306

0.323

874

2016

0.302

0.318

803

Total

0.467

0.492

9193

Panel C: Group difference Periods

Before the global financial crisis (2005–2007)

During the global financial crisis (2008–2010)

Group difference (t-statistic)

Group mean (Profit inefficiency)

0.532

0.523

0.009 (1.06)

Group mean (Cost inefficiency)

0.562

0.551

0.011 (1.35)

This table gives a summary of the mean of profit inefficiency and cost inefficiency according to country and year, and group difference analysis

Second, the positive coefficient of Capital regulatory index in columns (3) and (8) suggests that higher capital requirements are associated with lower bank efficiency in the Asia–Pacific region. With higher bank capital requirements, bank owners may pursue a costly financing policy and prioritize equity over deposits (Djalilov & Piesse, 2019), which in turn reduces the banks’ incentives to screen and monitor lending when equity capital becomes more expensive to raise than deposits, ultimately leading to higher risks and lower efficiency (Barth et al., 2004). Finally, columns (4) and (9) show that the effect of official supervisory power is negative and significant, suggesting that higher official supervisory power increases bank efficiency. This finding supports the argument that powerful official supervision drives banks to improve corporate governance, reduce risk-taking activities, and improve banks’ functioning, which potentially leads to higher efficiency (Beck et al., 2006). For the bank and country control variables, the negative and significant coefficient of Z-score is significantly negative for both profit and cost inefficiency, suggesting that

148

5 Bank Competition, Regulation, and Efficiency

Table 5.3 Summary statistics Variables

N

Mean

Standard Min deviation

Max

P10

Median P90

Dependent variables Profit inefficiency

9193

0.467

0.115

0.238

0.632

0.279

0.497

0.600

Cost inefficiency

9193

0.492

0.120

0.249

0.662

0.294

0.524

0.627

−0.078 0.636

1.527

Competition indicators Lerner index

9193

0.685

0.589

−0.258 2.024

Boone indicator

9045 −1.286

1.935

−10.39 −0.166 −4.123 −0.731 −0.271

Bank-specific variables Z-score

9193

3.630

3.518

−0.515 13.216

0.299

2.622

8.921

TCD

9193

0.759

0.171

0.195

0.951

0.500

0.803

0.933

LLP

9193

0.007

0.008

−0.001 0.031

0.000

0.005

0.019

FA

9193

0.014

0.011

0.001

0.044

0.003

0.011

0.029

log TA

9193

16.969

3.927

2.966

28.865

12.615

16.025

22.978

Foreign dummy

9193

0.143

0.350

0

1

0

0

1

Commercial dummy

9193

0.478

0.500

0

1

0

0

1

Saving dummy

9193

0.010

0.098

0

1

0

0

0

Country control variables Bank concentration 9193

54.173 13.801

36

100

40

52

76

Government-owned 9193 banks

22.904 27.927

0

96

0

3

69

Financial openness

9193

46.811 17.086

10

90

30

50

70

Financial freedom

9193

54.173 13.801

36

100

40

52

76

GDP growth

9193

4.281

3.619

−5.417 34.500

0.336

4.220

8.480

Inflation

9193

3.729

5.100

−18.93 28.164

−1.101 2.463

8.984

Crisis

9193

0.022

0.148

0

1

0

0

0

Activity restrictions 7192

8.018

1.950

3

12

5

8

10

Capital regulatory index

7220

6.809

2.238

2

10

4

7

10

Official supervisory 7358 power

11.556

1.960

6

16

9

12

14

Regulatory variables

This table reports summary statistics of the variables for the full sample. Our sample period is 2001–2016. All data are inflation-adjusted, and bank-specific variables are winsorized at the 5th and 95th percentile level to reduce the influence of outliers

Bank concentration

(5)

(−1.56)

(2.16)

(1.12)

(0.63)

0.00861**

(−4.14)

0.00351

0.0000501

(0.34)

0.000212***

(2.65)

(2.04)

0.000265**

(2.76)

0.00872***

(0.62)

0.00157

(−108.86)

(−110.34)

0.00156

(−303.55)

−0.00433***

(4.16) −0.0288***

(4.69)

−0.0288***

(7.85)

−0.0298***

(2.47) 0.252***

(2.75)

0.0906**

(5.11)

0.0221***

(−2.73)

0.271***

0.289***

(3.68)

(5.00)

0.101***

(7.89)

0.108***

0.0211***

0.0241***

(−2.76)

(2.64)

0.000207***

(3.38)

0.0102***

(1.39)

0.00197

(−177.43)

−0.0288***

(6.97)

0.264***

(2.55)

0.0730**

(6.70)

0.0205***

(−2.79)

(0.65)

0.0000636

(2.70)

0.00821***

(0.92)

0.00132

(−176.55)

−0.0289***

(6.73)

0.258***

(3.44)

0.101***

(6.69)

0.0211***

(−2.77)

(4.39)

0.000352***

(4.65)

0.00787***

(−0.74)

−0.000751

(−323.94)

−0.0312***

(9.77)

0.336***

(4.78)

0.133***

(9.74)

0.0284***

(−1.96)

(1.31)

0.000184

(5.54)

0.0122***

(1.31)

0.00317

(−120.43)

−0.0304***

(6.26)

0.323***

(3.24)

0.112***

(6.40)

0.0266***

(−0.98)

(−4.89)

(7) −0.00389

−0.0000696

(−1.51)

−0.00217***

(6)

(3.22)

0.000396***

(6.28)

0.0123***

(1.29)

0.00317

(−118.08)

−0.0304***

(5.65)

0.304***

(2.93)

0.103***

(6.38)

0.0275***

(−0.91)

−0.0000655

(−5.10)

−0.00365***

(8)

Dependent variable: cost inefficiency

(−3.61)

(−7.51)

−0.000796*** −0.000279

(4)

−0.000222*** −0.000203*** −0.000199*** −0.000201*** −0.000200*** −0.000116*

Country control variables

Saving dummy

Commercial dummy

log TA

FA

LLP

TCD

Z-score

(−1.70)

(−3.43)

(−5.25)

Bank-specific variables

−0.000334*** −0.000562*

−0.00252***

Lerner index

(3)

(2)

(1)

Variables

Dependent variable: profit inefficiency

Table 5.4 Impact of bank competition and regulation and bank efficiency (9)

(4.55)

0.000356***

(7.39)

0.0130***

(2.67)

0.00366***

(−192.78)

−0.0304***

(8.84)

0.314***

(3.32)

0.0891***

(8.68)

0.0255***

(−0.93)

−0.0000641

(−1.28)

−0.00348

(10)

(continued)

(2.17)

0.000207**

(6.68)

0.0118***

(2.14)

0.00299**

(−192.19)

−0.0305***

(8.91)

0.312***

(4.10)

0.111***

(8.83)

0.0267***

(−0.98)

−0.0000677

(−7.99)

−0.00392***

5.5 Empirical Results 149

(5.44)

(−8.48)

Official supervisory power

Capital regulatory index

Activity restrictions

(−5.45)

(−1.55)

(6.03)

(8.65)

0.0647***

(3.10)

0.000187***

(5)

(−0.38) (−1.35)

−0.00225

(−4.12)

(−2.38)

(2.66) −0.000826***

(1.78) −0.000451**

0.000633***

0.000550*

(−1.73)

(−0.55)

(−0.94) −0.0000248

−0.000459*

(−1.21)

(−0.59) −0.000490

(−8.45) −0.0000844

(−2.12)

(−1.62)

(−0.68)

(−1.05) −0.00132

−0.0000351

(−1.68)

−0.0000764

(−1.78)

(−9.00) −0.000149*

(−5.62) −0.000188*

−0.000763**

(−0.93)

−0.00171

(−4.69)

−0.000674

−0.000294*** −0.0000702

(−4.34)

−0.000386*** −0.000166

Bank regulation variables

Crisis

Inflation

GDP growth

(4.21)

0.0394***

(3.35)

0.000156***

(4)

(0.33)

0.00165

(2.28)

0.0000950**

(6)

(2.78)

0.0241***

(1.40)

0.0000982

(7)

(1.99)

0.0198**

(2.05)

0.000167**

(8)

Dependent variable: cost inefficiency (9)

(1.68)

0.0103*

(2.98)

0.000125***

(10)

(5.10)

0.0353***

(3.29)

0.000175***

(−6.63) (0.31)

0.0000313

(−1.31)

−0.00224

(−4.05)

(−2.02)

−0.000672**

(−0.42)

−0.00562

(−0.14)

−0.000241*** −0.0000101

(−2.06)

−0.000175**

(−10.85)

(1.99)

0.000574**

(−0.98)

−0.00528

(−0.24)

−0.0000167

(0.12)

0.0000114

(−6.75)

(−0.84)

−0.000152

(−1.38)

−0.00343

(−0.10)

−0.00000592

(0.42)

0.0000357

(−10.56)

(continued)

(−3.27)

−0.000602***

(2.70)

0.000604***

(−1.84)

−0.000458*

(−1.49)

−0.00610

(0.48)

0.0000295

(1.21)

0.000102

(−10.07)

−0.000223*** −0.000225*** −0.000235*** −0.000249*** −0.000244*** −0.000282*** −0.000268*** −0.000279*** −0.000284*** −0.000282***

(4.20)

Financial freedom

0.0471***

(1.72)

(1.14)

0.0525***

0.0220***

Financial openness

0.000164*

(3)

0.0000944

(2)

Dependent variable: profit inefficiency

(1)

Government−owned 0.000118** banks (2.54)

Variables

Table 5.4 (continued)

150 5 Bank Competition, Regulation, and Efficiency

Yes

9081

0.976

Year fixed effects

N

R−squared

(2.49)

0.972

7039

Yes

Yes

(5.44)

Yes

Bank fixed effects

0.487***

0.715***

Constant

(2)

(1)

Variables

0.972

7220

Yes

Yes

(6.30)

0.159***

(3)

Dependent variable: profit inefficiency

Table 5.4 (continued)

0.972

7208

Yes

Yes

(3.23)

0.192***

(4)

0.972

7039

Yes

Yes

(6.06)

0.355***

(5)

0.981

9081

Yes

Yes

(9.65)

0.599***

(6)

0.978

7039

Yes

Yes

(4.12)

0.436***

(7)

0.978

7220

Yes

Yes

(0.84)

0.871

(8)

Dependent variable: cost inefficiency (9)

0.978

7208

Yes

Yes

(7.23)

0.341***

(10)

0.978

7039

Yes

Yes

(1.23)

0.682

5.5 Empirical Results 151

152

5 Bank Competition, Regulation, and Efficiency

higher bank stability corresponds to higher bank efficiency. We find that increasing banks’ customer deposits, loan loss provisions, and fixed assets decreases profit and cost efficiency. The significantly negative coefficient of log TA indicates that larger banks enjoy higher efficiency compared with small banks. Commercial banks are more efficient than saving banks. Considering the impact of market structure, our results show that higher bank concentration and more government-owned banks are associated with lower bank efficiency.10 Higher financial openness leads to lower bank efficiency and higher financial freedom is related to higher efficiency. This table examines the impact of competition and regulation on bank efficiency. The dependent variables are Profit inefficiency and Cost inefficiency in columns (1)– (5) and columns (6)–(10), respectively as defined in Eq. (5.1). Our sample period is 2001–2016. Bank-level controls include Z-score, TCD, LLP, FA, log TA, Commercial dummy, and Saving dummy. Country control variables include Bank concentration, Government-owned banks, Financial openness, Financial freedom, GDP growth, Inflation, and Crisis. Bank regulation variables include Activity restrictions, Capital regulatory index, Official supervisory power. All variables are defined in Table 5.1. Both bank and year fixed effects are included and reported t-statistics (in parentheses) are based on standard errors that are estimated based on Vogelsang (2012) and robust to heteroskedasticity, serial correlation, and spatial correlation. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

5.5.2 Impact of Bank Regulation on the Relationship Between Competition and Efficiency The results of our basic analysis in Sect. 5.5.1 indicate that both market power and bank regulation have an impact on bank efficiency. Previous studies confirm that bank regulation is also an important factor in determining banks’ competitive conditions (e.g., Claessens & Laeven, 2004; Demirguc-Kunt et al., 2004; Li, 2019) and has a significant effect on the relationship between competition and bank performance (e.g., Agoraki et al., 2011; Noman et al., 2018). Thus, in this subsection, we test whether bank regulation affects the relationship between competition and efficiency in our sample, and the results are reported in Table 5.5. The coefficient of the interaction term of Activity restrictions and Lerner index remains negative and significant, suggesting that increased activity restrictions strengthen the positive effect of market power on bank efficiency, probably because tighter activity restrictions result in lower competition in the banking system (Anginer

10 The positive coefficient of Bank concentration is not contradictory to the main conclusion that increased market power (measured by Lerner index) is associated with higher bank efficiency, as the degree of concentration is not necessarily related to the degree of competition, and competition and concentration can coexist (Mamatzakis et al., 2005). See Claessens and Laeven (2004), Beck et al. (2006), and Olivero et al. (2011) for the detail discussion.

Government-owned banks

Bank concentration

Saving dummy

Commercial dummy

log TA

FA

LLP

TCD

(−4.21)

−0.00653*** (−0.71)

−0.00101

(6)

(2.99)

(1.88)

(3.08)

0.000169***

(0.65)

0.000100*

0.000256***

(2.84)

(2.73)

0.0000626

0.00872***

0.00869***

(0.99)

(−177.02)

(−175.21)

(1.14)

−0.0288***

−0.0288***

0.00143

(6.71)

(6.96)

0.00164

0.256***

0.265***

(3.26)

(3.40)

(6.97)

0.0953***

(6.69)

0.0996***

0.0220***

0.0212***

(−2.77)

(3.54)

0.000167***

(2.70)

0.000214***

(3.33)

0.00985***

(1.44)

0.00204

(−177.55)

−0.0288***

(6.93)

0.264***

(2.52)

0.0724**

(6.66)

0.0205***

(−2.80)

(3.00)

0.000174***

(0.61)

0.0000565

(1.67)

0.00496*

(−0.32)

(2.19)

0.000102**

(2.04)

0.000192**

(6.77)

0.0123***

(2.30)

(−190.15) 0.00322**

(−32.34)

−0.0304***

(9.15)

0.319***

(4.10)

0.112***

(8.75)

0.0266***

(−0.98)

−0.000443

−0.0229***

(7.92)

0.304***

(2.67)

0.0770***

(7.69)

0.0236***

(−3.00)

(3.49)

0.000174***

(4.61)

0.000381***

(6.83)

0.0123***

(2.12)

0.00294**

(−193.38)

−0.0304***

(8.91)

0.310***

(4.06)

0.110***

(9.05)

0.0274***

(−0.96)

(−3.73)

−0.0107***

(5)

(3.27)

0.000138***

(4.62)

0.000364***

(7.30)

0.0126***

(2.73)

0.00374***

(−192.88)

−0.0304***

(8.79)

0.313***

(3.28)

0.0884***

(8.62)

0.0255***

(−0.94)

−0.0000649

(−4.22)

−0.00969***

(7)

Dependent variable: cost inefficiency

−0.0000667

(−1.88)

−0.00450*

(4)

(−2.78)

(−1.35)

(−2.16)

(3)

−0.000201*** −0.000200*** −0.000202*** −0.000214*** −0.0000682

−0.00356

−0.00363**

Lerner index

Z-score

(2)

(1)

Variables

Dependent variable: profit inefficiency

Table 5.5 Impact of bank regulation on the relationship between competition and efficiency

(continued)

(3.58)

0.000185***

(1.93)

0.000172*

(4.76)

0.00792***

(0.69)

0.000916

(−35.15)

−0.0235***

(11.04)

0.384***

(3.34)

0.0894***

(10.02)

0.0291***

(−1.24)

−0.0000845

(−4.97)

−0.0132***

(8)

5.5 Empirical Results 153

Official supervisory power

Capital regulatory index * Lerner index

Capital regulatory index

(4.01)

0.0252***

−3.245 (−0.93)

(5) (3.03)

0.0198***

(6) (1.71)

0.0104*

(7)

Dependent variable: cost inefficiency (4)

(−1.28)

(−9.75) (0.36)

(3.31)

(2.08)

0.000455*

0.000718***

0.000448**

(−0.98)

(−1.74)

(−1.73)

(−1.49) −0.000339*

(−0.67) −0.000402*

−0.000421

(−1.10)

−0.00567

(−0.29)

−0.0000176

−0.000206

(−1.19)

−0.00618

(0.89)

−0.000308

−0.0000338

(−0.32)

−0.000288

(−3.65) 0.0000638

(1.00)

(−1.47)

(−0.89)

(−0.52)

(−5.84) −0.000390*** 0.0000309

0.000263

−0.00140

−0.00179

−0.000386

(−1.37)

(−1.25)

−0.0000312

(−1.65)

(−2.15)

−0.0000816

(−1.85)

−0.0000815

(−9.22) −0.000147*

(−8.17)

−0.000193**

(−8.05)

−0.000167*

Activities restrictions −0.000509** * Lerner index (−2.39)

Activity restrictions

Crisis

Inflation

GDP growth

(6.05)

0.0395***

(3) (0.06)

0.210

(8)

(3.50)

0.000697***

(0.51)

0.000128

(−0.81)

−0.00540

(−0.44)

−0.0000248

(0.05)

0.00000428

(−9.94)

0.000337

(−1.18)

−0.00319

(−0.02)

−0.00000136

(0.46)

0.0000387

(−10.89)

(continued)

0.000574**

(4.61)

0.000923***

(−1.33)

−0.000390

(−1.33)

−0.000284

(−0.51)

−0.000147

(−0.76)

−0.00801

(−0.08)

−0.0000053

(−1.54)

−0.000152

(−6.31)

−0.000232*** −0.000231*** −0.000260*** −0.000210*** −0.000272*** −0.000273*** −0.000297*** −0.0002***

(6.62)

(7.99)

Financial freedom

0.0471***

0.0541***

Financial openness

(2)

Dependent variable: profit inefficiency

(1)

Variables

Table 5.5 (continued)

154 5 Bank Competition, Regulation, and Efficiency

Yes

Yes

7039

0.972

N

R-squared

0.972

7066

Yes

Yes

(4.19)

(7.13)

Year fixed effects

2.951***

0.356***

(2)

(−0.97)

(−2.28)

0.972

7208

Yes

Yes

(4.51)

0.973

7039

Yes

Yes

(5.99)

0.492***

−0.000461** 0.944***

(1.69) −0.00110

(−0.13)

(4)

(3)

Dependent variable: profit inefficiency

(1)

Bank fixed effects

Constant

Official supervisory power * Lerner index

Variables

Table 5.5 (continued)

0.978

7039

Yes

Yes

(6.99)

0.746***

(5)

0.978

7066

Yes

Yes

(5.43)

0.395***

(6) (2.40)

(8)

0.978

7208

Yes

Yes

(5.21)

0.738***

(−2.81)

0.979

7039

Yes

Yes

(3.27)

0.329***

(−5.57)

−0.000541*** −0.0011***

(1.37)

(7)

Dependent variable: cost inefficiency

5.5 Empirical Results 155

156

5 Bank Competition, Regulation, and Efficiency

et al., 2014; Li, 2019). The coefficient of the interaction term for Capital regulatory index and Lerner index remains positive and significant, demonstrating that the market power effect in increasing bank efficiency is less pronounced if the bank is exposed to stringent capital requirements. The interaction term of Official supervisory power and Lerner index suggests implications similar to those of activity restrictions; the negative coefficient of the interaction term provides evidence that strong supervisory power of bank regulators intensifies the market power effect in increasing bank efficiency. This table examines the impact of bank regulation on the relationship between competition and bank efficiency. The dependent variables are Profit inefficiency and Cost inefficiency in columns (1)–(4) and columns (5)–(8), respectively as defined in Eq. (5.1). Our sample period is 2001–2016. Bank-level controls include Z-score, TCD, LLP, FA, log TA, Commercial dummy, and Saving dummy. Country control variables include Bank concentration, Government-owned banks, Financial openness, Financial freedom, GDP growth, Inflation, and Crisis. Bank regulation variables include Activity restrictions, Capital regulatory index, Official supervisory power. All variables are defined in Table 5.1. Both bank and year fixed effects are included and reported t-statistics (in parentheses) are based on standard errors that are estimated based on Vogelsang (2012) and robust to heteroskedasticity, serial correlation, and spatial correlation. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

5.5.3 Further Test: Foreign Banks Versus Domestic Banks Some previous studies indicate that foreign banks and domestic banks have different patterns: foreign banks earn more profit than domestic banks (Claessens et al., 2001), and are more sensitive to strict parent country regulation than to the institutional conditions in host countries (e.g., De Haas & Van Lelyveld, 2006; Ongena et al., 2013). In this subsection, we further test whether the determinants of efficiency have significantly changed for foreign banks compared to domestic banks.11 For the market power effect in increasing profit and cost efficiency, the results in Table 5.6 show no significant difference between foreign and domestic banks. The coefficients of interaction terms for Activity restrictions and Official supervisory power confirm that foreign banks operating in a host country with higher activity restrictions and official supervisory power have higher bank efficiency. This is probably because foreign banks outperform domestic banks in terms of profitability and cost efficiency by exploiting their advanced technologies, a highly skilled labor force, and better risk management practices in host countries with higher activity restrictions and official supervisory power (Haque & Brown, 2017).

11 We

also include the dummy variable Foreign dummy in the regression, but it is dropped from the model owing to collinearity.

Government-owned banks

Bank concentration

Saving dummy

Commercial dummy

log TA

FA

LLP

TCD

(1.77)

(2.53)

(0.53) 0.0000949*

(2.67) 0.000117**

0.0000503

(2.71)

(1.13) 0.000215***

0.00862***

0.00356

(1.08)

(−4.10)

(−32.28)

(−175.36) 0.00156

(−303.87) −0.00429***

−0.0229***

−0.0288***

−0.0298***

(2.53)

0.000132**

(2.58)

0.000203***

(2.02)

0.00591**

(−0.22)

−0.000307

(7.66)

(7.09)

0.291***

0.271***

(2.11)

(7.86)

(3.40)

(3.73)

0.0609**

(7.93)

0.0244***

(−2.91)

−0.000209***

(−0.86)

−0.000503

(3)

0.290***

0.0998***

0.110***

(6.68)

(−2.81)

(−3.60) (7.90)

−0.000203***

−0.000221*** 0.0211***

(−0.38)

(−5.37)

0.0242***

−0.000228

−0.00280***

Lerner index

Z-score

(2)

(1)

Variables

Dependent variable: pinefficiency

Table 5.6 Impact of ownership on the relationship between bank competition, regulation and bank efficiency

(3.29)

0.000153***

(2.66)

0.000210***

(3.37)

0.0101***

(1.44)

0.00204

(−177.38)

−0.0288***

(6.92)

0.263***

(2.52)

0.0722**

(6.68)

0.0205***

(−2.78)

−0.000200***

(−1.47)

−0.000843

(4)

(2.84) (continued)

0.000170***

(0.75)

0.0000738

(2.69)

0.00822***

(0.96)

0.00139

(−178.10)

−0.0289***

(6.62)

0.255***

(3.39)

0.0990***

(6.51)

0.0208***

(−2.76)

−0.000199***

(−0.39)

−0.000234

(5)

5.5 Empirical Results 157

−0.000439*** (−4.13)

−0.000226*** (−7.94) −0.000168*

−0.000221*** (−8.38) −0.000381***

Capital regulatory index * foreign dummy

Capital regulatory index

Activities restrictions * foreign dummy

Activity restrictions

Foreign dummy * Lerner index

Crisis

Inflation

GDP growth

Financial freedom

(−0.81) −0.000461

(−1.06) −0.00168 (−0.78) −0.000547 (−0.41)

(−4.73) −0.000589 (−0.60) 0.00160 (1.29)

(continued)

0.000359

(2.50) 0.000241

(−0.60)

(−3.56)

(−2.06)

0.000632**

−0.00107***

−0.0000269** −0.000160

(−2.19)

(−2.69)

(−0.33)

−0.000441

(−1.39)

−0.00230

(−0.39)

−0.0000253

(−0.81)

−0.0000732

(−8.50)

−0.000242***

−0.000637**

(0.11)

0.000153

(−0.54)

−0.000485

(−0.58)

−0.0000345

(−1.70)

−0.000152*

(−9.02)

−0.000250***

(8.80)

0.0663***

(5)

−0.000748***

(−0.36)

−0.00577

(0.81)

0.0000529

(−1.86) −0.0000686

(−4.29) −0.000296***

(−5.76)

−0.000197***

(6.04)

0.0399***

(−1.00)

(7.76)

(4.29)

−3.495

0.0522***

0.0224***

(4)

(3)

Financial openness

(2)

Dependent variable: pinefficiency (1)

Variables

Table 5.6 (continued)

158 5 Bank Competition, Regulation, and Efficiency

9081 0.976

N

R-squared

0.972

7039

Yes

(8.74)

(−1.01)

(−1.96)

TCD (9.74)

−0.0000697

−0.000116* 0.0265***

(−7.66)

(−4.30)

0.0284***

−0.00415***

−0.00210***

Lerner index

Z-score

(7)

(6)

Variables

Dependent variable: cost inefficiency

Yes

Year fixed effects

Yes

(8.75)

(4.88) Yes

4.341***

1.378***

(2)

Dependent variable: pinefficiency (1)

Bank fixed effects

Constant

Official supervisory power * foreign dummy

Official supervisory power

Variables

Table 5.6 (continued)

(10.29)

0.0299***

(−1.15)

−0.0000787

(−1.22)

−0.00388

(8)

0.973

7066

Yes

Yes

(45.30)

0.752***

(0.51)

(3)

(8.66)

0.0254***

(−0.92)

−0.0000635

(−6.80)

−0.00357***

(9)

0.972

7208

Yes

Yes

(39.55)

(8.60) (continued)

0.0264***

(−0.97)

−0.0000671

(−7.54)

−0.00410***

(10)

0.972

7039

Yes

Yes

(62.16)

0.876***

(−2.22)

(−0.35) 0.874***

(−2.77) −0.00119**

(−1.54)

−0.000598***

(0.58)

(5)

−0.000541

−0.000321

(4)

5.5 Empirical Results 159

GDP growth

Financial freedom

Financial openness

Government-owned banks

Bank concentration

Saving dummy

Commercial dummy

log TA

−0.000213*** (−6.41)

−0.000270*** (−9.80) 0.0000265

−0.000281*** (−10.81) −0.000174**

−0.000213**

0.0000292

(−10.66)

−0.000288***

(1.74)

0.0108* (−0.01)

(3.79)

−0.0437

0.0236***

(0.35)

(2.89)

0.000122***

(4.59)

0.000359***

(7.40)

0.0129***

(2.74)

0.00377***

(−192.35)

−0.0304***

(8.76)

0.311***

(3.23)

0.0869***

(9)

(3.19)

0.000148***

(4.05)

0.000317***

(5.29)

0.00884***

(0.86)

0.00113

0.00175

(2.11)

(2.28)

(1.97) 0.0000988**

(4.39) 0.0000949**

0.000183**

(6.73)

(4.65) 0.000352***

0.0122***

0.00788***

(2.27)

(−0.73)

(−34.85)

(−190.28) 0.00318**

(−323.81) −0.000740

−0.0237***

−0.0304***

−0.0312***

(10.48)

(9.23)

0.366***

0.322***

(9.77)

(2.63)

0.0709***

(8)

0.336***

(4.06)

(4.77)

FA

0.111***

0.134***

LLP

(7)

Dependent variable: cost inefficiency (6)

Variables

Table 5.6 (continued)

(continued)

0.000112

(−10.20)

−0.000280***

(5.30)

0.0370***

(2.97)

0.000157***

(2.29)

0.000220**

(6.67)

0.0119***

(2.21)

0.00309**

(−193.78)

−0.0305***

(8.78)

0.308***

(4.00)

0.109***

(10)

160 5 Bank Competition, Regulation, and Efficiency

Official supervisory power * foreign dummy

Official supervisory power

Capital regulatory index * foreign dummy

Capital regulatory index

Activities restrictions * foreign dummy

Activity restrictions

Foreign dummy * Lerner index

Crisis

Inflation

Variables

Table 5.6 (continued)

−0.00108

−0.00132 (−1.12)

0.000450 (0.40)

−0.00141 (−0.73)

−0.000812** (−2.15)

(continued)

(−1.63)

(0.23)

(0.52) −0.000321 0.0000461

(0.31)

(2.63) 0.000312

0.000136

0.000620***

(−0.10)

(−1.78)

(−2.04) −0.0000250

(−2.37)

−0.000644**

(−0.95)

−0.00115

−0.00116*

(−0.49)

−0.000599

(−1.48)

−0.00612

(0.50)

0.0000307

(1.31)

(10)

(−2.51)

(−0.95)

(−0.77)

−0.00338

(−0.04)

−0.00000252

(0.35)

(9)

−0.0000153**

−0.000651**

(−1.25)

(−0.96)

(−1.33)

(−0.12) −0.00748

(−0.10) −0.00555

(−4.06)

(−2.14) −0.00000742

(0.31) −0.00000634

(−2.05) −0.000241*** −0.00226

(8)

(7)

Dependent variable: cost inefficiency (6)

5.5 Empirical Results 161

Yes 9081 0.981

N

R-squared

0.978

7039

Yes

Yes

0.979

7066

Yes

Yes

(43.28)

0.713***

(8)

0.978

7208

Yes

Yes

(9)

0.978

7039

Yes

Yes

(10)

This table examines the impact of bank regulation on the relationship between competition and bank efficiency. The dependent variables are Profit inefficiency and Cost inefficiency in columns (1)–(5) and columns (6)–(10), respectively as defined in Eq. (5.1). Our sample period is 2001–2016. Bank-level controls include Z-score, TCD, LLP, FA, log TA, Commercial dummy, and Saving dummy. Country control variables include Bank concentration, Government-owned banks, Financial openness, Financial freedom, GDP growth, Inflation, and Crisis. Bank regulation variables include Activity restrictions, Capital regulatory index, Official supervisory power. All variables are defined in Table 5.1. Both bank and year fixed effects are included and reported t-statistics (in parentheses) are based on standard errors that are estimated based on Vogelsang (2012) and robust to heteroskedasticity, serial correlation, and spatial correlation. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

Yes

Year fixed effects

(54.09)

(56.17)

Bank fixed effects

0.874***

0.964***

Constant

(7)

Dependent variable: cost inefficiency (6)

Variables

Table 5.6 (continued)

162 5 Bank Competition, Regulation, and Efficiency

5.5 Empirical Results

163

5.5.4 Robustness Checks To address the reliability of our findings, we conduct different robustness checks based on an alternative definition of competition, different estimation methods, and different subsamples. First, we use the Boone indicator as an alternative measure of bank competition. The Boone indicator directly captures the link between bank performance and efficiency and does not impose more stringent data requirements than the Lerner index does (Boone, 2008). The results using the Boone indicator as a measure of bank competition are reported in Table 5.7. While our main results are largely unchanged, the negative and statistically significant coefficient of the Boone indicator indicates that profit and cost inefficiencies increase with greater bank competition, which is consistent with the findings based on the Lerner index. Second, we also perform our analysis according to different estimation methods. We first introduce the country fixed effects in the regression to control for heterogeneity between countries instead of bank fixed effects while the year fixed effects are also considered. Then, we perform our basic analysis based on the instrumental variable (IV) analysis to control for a potential endogeneity and heteroskedasticity problem (see also Berger et al., 2009; Fu et al., 2014). The competitive indicator Lerner index is instrumented by lagged Lerner index, Restructuring power and Financial conglomerate restrictiveness. Restructuring power indicates whether the supervisory authorities have the power to restructure and reorganize a troubled bank, and Financial conglomerate restrictiveness indicates the restrictions on financial conglomeration of the financial system in a country, and measures the extent to which banks may own and control non-financial firms, non-financial firms may own and control banks, and non-bank financial firms may own and control banks; higher values are more restrictive. To obtain consistent and efficient estimates in the presence of non-independently and identically distributed errors, we use the GMM estimator in the IV estimation analysis and estimators are clustered at bank level to account for within-cluster correlation of the disturbances (see also Driscoll & Kraay, 1998).12 The country fixed effects estimation and IV regression results are presented in Table 5.8. For brevity, we report only the results of profit inefficiency.13 The results of the Lerner index and bank regulation variables under the different estimation methods are largely unchanged. This table examines the impact of competition and risk on bank efficiency. The dependent variable is Profit inefficiency as defined in Eq. (5.1). Our sample period is 2001–2016. Bank-level controls include Z-score, TCD, LLP, FA, log TA, Commercial 12 The instrumental variables are statistically significant in the first-stage analysis. In the first-stage regression, the Lerner index is negatively related to Restructuring power and Financial conglomerate restrictiveness at 1%. In addition, we employ different tests for weak identification and underidentification, such as the Kleibergen–Paap rk LM and Wald F statistic (Kleibergen & Paap, 2006) and the Cragg–Donald Wald F statistic (Hall & Peixe, 2000), and confirm that the instruments are valid. 13 The results of cost inefficiency under the country fixed effects estimation and IV regression are consistent with our main findings and are available upon request.

Government-owned banks

Bank concentration

Saving dummy

Commercial dummy

log TA

FA

LLP

TCD

(1.70)

(2.92)

(1.17)

0.000134*

0.000134***

(1.93)

(2.22)

0.000168

0.000188*

(1.18)

(−4.27)

0.00875**

(0.35)

−0.00435***

0.00359

(−109.98)

0.000892

(−305.92)

(1.96)

0.000180*

(1.63)

0.000268

(2.57)

0.00854**

(0.35)

0.000905

(−107.73)

−0.0290***

(4.17)

(4.45)

−0.0290***

(8.27)

−0.0299***

0.251***

(2.34)

0.0875**

(5.32)

0.0227***

(−2.56)

−0.000185**

0.254***

(2.44)

0.305***

(3.04)

(5.32)

0.0905**

0.0904***

(8.26)

(−2.50)

0.0221***

(−3.92)

0.0250***

(−6.03)

−0.000184**

(−8.41)

−0.000235***

(−0.95)

Z-score

−0.00129

−0.00126***

−0.00134***

Boone indicator

(3)

(2)

(1)

Variables

Dependent variable: profit inefficiency

Table 5.7 Alternative measure of bank competition

(3.81)

0.000176***

(1.54)

0.000149

(3.28)

0.00967***

(0.90)

0.00129

(−178.54)

−0.0290***

(6.79)

0.260***

(2.52)

0.0727**

(6.78)

0.0206***

(−2.71)

−0.000192***

(−7.84)

−0.00127***

(4)

(3.84)

0.000226***

(2.23)

0.000236**

(2.74)

0.00832***

(0.47)

0.000677

(−177.08)

−0.0290***

(6.35)

0.243***

(3.08)

0.0908***

(7.12)

0.0222***

(−2.56)

−0.000182**

(−1.62)

−0.00123

(5)

(2.76)

0.000117***

(4.76)

0.000435***

(5.14)

0.00849***

(−0.55)

−0.000545

(−327.53)

−0.0312***

(9.33)

0.320***

(5.05)

0.141***

(9.72)

0.0278***

(−1.19)

−0.0000687

(−6.09)

−0.000958***

(6)

(2.12)

0.000146**

(3.14)

0.000416***

(5.21)

0.0119***

(0.84)

0.00206

(−119.77)

−0.0306***

(5.46)

0.280***

(3.51)

0.124***

(6.49)

0.0263***

(0.03)

0.00000202

(−4.45)

−0.000940***

(7)

(2.40)

0.000193**

(3.46)

0.000517***

(5.67)

0.0116***

(0.82)

0.00207

(−116.41)

−0.0307***

(5.24)

0.276***

(3.31)

0.120***

(6.41)

0.0270***

(0.00)

4.03e-08

(−1.36)

−0.000965

(8)

Dependent variable: cost inefficiency (9)

(3.49)

0.000149***

(4.54)

0.000412***

(7.02)

0.0123***

(1.85)

0.00257*

(−193.66)

−0.0307***

(7.96)

0.284***

(4.07)

0.110***

(8.42)

0.0244***

(−0.02)

−0.00000110

(−6.09)

−0.000977***

(10)

(continued)

(4.31)

0.000229***

(5.08)

0.000502***

(6.42)

0.0115***

(1.36)

0.00191

(−193.15)

−0.0307***

(7.72)

0.270***

(4.46)

0.122***

(8.98)

0.0266***

(0.04)

0.00000244

(−1.82)

−0.000929*

164 5 Bank Competition, Regulation, and Efficiency

Bank fixed effects

Constant

Official supervisory power

Capital regulatory index

Activity restrictions

Crisis

Inflation

GDP growth

Financial freedom

(−1.06)

(1.03)

(3.48)

Yes

(3.53)

Yes

5.217***

4.431*** Yes

(2.76)

1.913*** (3.22) Yes

Yes

(3.46)

3.142***

(−3.46)

4.131***

(−2.68)

(3.22) −0.000701***

(1.72) −0.000505***

0.000765***

0.000574*

(−0.22)

(−1.85)

(−1.09)

−0.00172

(−0.43)

−0.0000280

(−0.14)

−0.0000128

(−9.05)

−0.000264***

(6.16)

0.0477***

(5)

−0.0000596

(−0.75)

−0.000673

(−0.60)

−0.0000364

(−0.33)

−0.0000290

(−8.71)

−0.000250***

(4.70)

0.0319***

(4)

−0.000305*

(−1.16)

(−1.03) −0.00127

(−1.02)

−0.00113

(−4.46)

0.00103

−0.0000767

(−0.64)

(−1.00)

−0.0000759

(−3.78)

(−5.49) −0.0000618

(−5.81)

−0.0000993

(−7.98)

−0.000344***

−0.000289***

(3.40) −0.000242***

(3.36)

−0.000245***

(2.30)

−0.000221***

(3) 0.0380***

0.0125**

Financial openness

(2)

0.0341***

(1)

Variables

Dependent variable: profit inefficiency

Table 5.7 (continued)

Yes

(2.12)

0.486**

(−1.17)

−0.00309

(−4.00)

−0.000242***

(−0.75)

−0.0000653

(−11.54)

−0.000318***

(0.41)

0.00213

(6)

Yes

(0.89)

1.148

(−2.23)

−0.0000771**

(−1.34)

−0.00564

(−0.99)

−0.0000686

(1.57)

0.000145

(−7.95)

−0.000329***

(1.43)

0.0134

(7)

Yes

(0.57)

0.725

(2.20)

0.000689**

(−0.60)

−0.00597

(−0.67)

−0.0000452

(2.26)

0.000199**

(−7.58)

−0.000330***

(2.17)

0.0220**

(8)

Dependent variable: cost inefficiency (9)

Yes

(1.19)

0.317

(−0.90)

−0.000159

(−0.76)

−0.00427

(−0.87)

−0.0000491

(2.63)

0.000213***

(−11.62)

−0.000327***

(1.85)

0.0119*

(10)

Yes (continued)

(1.42)

0.531

(−2.66)

−0.000495***

(3.68)

0.000822***

(−0.33)

−0.0000822

(−1.56)

−0.00621

(−0.36)

−0.0000214

(2.78)

0.000230***

(−12.03)

−0.000344***

(3.74)

0.0272***

5.5 Empirical Results 165

Yes

8936

0.977

Year fixed effects

N

R-squared

0.973

6938

Yes

(2)

0.973

7087

Yes

(3)

0.974

7074

Yes

(4)

0.973

6938

Yes

(5)

0.982

8936

Yes

(6)

0.979

6938

Yes

(7)

0.979

7087

Yes

(8)

Dependent variable: cost inefficiency (9)

0.979

7074

Yes

(10)

0.979

6938

Yes

This table examines the impact of competition and regulation on bank efficiency. The dependent variables are Profit inefficiency and Cost inefficiency in columns (1)–(5) and columns (6)–(10), respectively as defined in Eq. (5.1). Our sample period is 2001–2016. Bank-level controls include Z-score, TCD, LLP, FA, log TA, Commercial dummy, and Saving dummy. Country control variables include Bank concentration, Government-owned banks, Financial openness, Financial freedom, GDP growth, Inflation, and Crisis. Bank regulation variables include Activity restrictions, Capital regulatory index, Official supervisory power. All variables are defined in Table 5.1. Both bank and year fixed effects are included and reported t-statistics (in parentheses) are based on standard errors that are estimated based on Vogelsang (2012) and robust to heteroskedasticity, serial correlation, and spatial correlation. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

(1)

Variables

Dependent variable: profit inefficiency

Table 5.7 (continued)

166 5 Bank Competition, Regulation, and Efficiency

(−1.47)

−0.00178

(7)

(2.75)

0.00831***

(0.83)

0.00115

(−186.63)

−0.0289***

(6.60)

0.251***

(3.13)

0.0913***

(7.13)

0.0224***

(−2.80)

(3.26)

0.00973***

(1.13)

0.00155

(−188.61)

−0.0288***

(6.93)

0.263***

(2.56)

0.0732**

(6.87)

0.0209***

(−2.84)

(2.58)

0.00778***

(0.65)

0.000910

(−186.17)

−0.0290***

(6.72)

0.258***

(3.45)

0.101***

(6.85)

0.0215***

(−2.83)

(0.51)

0.00139

(1.35)

0.00205

(−169.65)

−0.0294***

(7.59)

0.326***

(0.20)

0.00627

(8.90)

0.0278***

(−1.58) 0.0273***

(0.43)

0.00119

(1.30)

0.00199

(−170.64)

−0.0295***

(7.84)

0.340***

(0.28)

0.00888

(8.72)

(0.50)

0.00137

(1.23)

0.00187

(−168.18)

−0.0294***

(7.55)

0.325***

(0.24)

0.00746

(8.89)

0.0280***

(−1.53)

−0.000110

(−2.00)

−0.00238**

(8)

(0.57)

0.00156

(1.19)

0.00181

(−170.16)

−0.0294***

(7.68)

0.328***

(0.05)

0.00165

(8.77)

0.0274***

(−1.63)

−0.000117

(−1.09)

−0.00130

(9)

(2.24)

0.00907**

(3.20)

0.00596***

(−90.97)

−0.0286***

(6.19)

0.426***

(−1.15)

−0.0522

(8.35)

0.0362***

(−3.86)

−0.000570***

(−2.51)

−0.0219**

(10)

(−8.51)

(−7.81) −0.000155* (−1.75)

(−2.22)

(−9.01)

−0.000197**

(−8.35) (−1.06)

−0.0000947

(−8.44)

(3.45)

0.000410***

(−5.20)

(3.65)

0.000433***

(−4.76)

(3.47)

0.000416***

(−5.32)

(3.27)

0.000391***

(−5.66)

(continued)

(−1.63)

−0.000528

(−0.32)

−0.000215*** −0.00022*** −0.000227*** −0.000240*** −0.000236*** −0.000160*** −0.000148*** −0.000165*** −0.000174*** −0.0000179

0.00812***

(2.58)

(0.79)

(−4.41)

0.00326

0.00110

−0.00455***

(1.04)

−0.0289***

(−187.65)

−0.0298***

(−319.29)

0.270***

(7.06)

0.287***

(7.81)

0.101***

(3.43)

0.108***

(3.67)

0.0215***

(6.86)

0.0245***

(8.07)

(−2.86)

−0.000125*

(−2.01)

−0.00235**

(6)

Panel B: IV estimation

(−1.76)

(−0.55)

−0.000298

(5)

−0.000225*** −0.00021*** −0.000202*** −0.000204*** −0.000204*** −0.000113

(−1.67)

−0.000827*

(4)

(−3.66)

Government-owned −0.000390*** −0.000168* banks (−4.39) (−1.86)

Bank concentration

Saving dummy

Commercial dummy

log TA

FA

LLP

TCD

Z-score

(−1.09)

−0.000349** −0.000586

(−2.44)

−0.00253***

(−5.26)

Lerner index

(3)

(2)

(1)

Variables

Panel A: Country fixed effects

Table 5.8 Robustness checks based on alternative estimation methods

5.5 Empirical Results 167

Country fixed effects

Constant

Official supervisory power

Capital regulatory index

Activity restrictions

Crisis

Inflation

GDP growth

Financial freedom −0.000586

−0.000827

−0.000298

−0.00235**

−0.00178

−0.0215

Yes

4.443***

(3.74)

2.347***

5.542*** Yes

(2.81) Yes

(7.72)

5.481*** Yes

(6.15)

2.020***

−0.000768*** (−3.85)

(−2.30)

(2.29) −0.000438**

0.000562** (2.43)

0.000497**

−0.000563** (−2.21)

(−0.68)

−0.00085***

(−1.08)

0.0209

(−3.57)

(5.72)

Yes

−0.0224 (−1.13)

No

(−0.90)

−0.0278

(−1.58)

No

(−2.47)

−0.000655**

(−1.52)

−0.0273

(−0.86)

(−2.83)

0.0215

(−2.84)

−0.0245

(−2.80)

(−0.87)

(−1.47) −0.000125*

(−2.01)

(−0.01)

−0.0000437

(−1.31)

−0.0000981

(7)

(−1.76)

(−0.55)

(−0.07)

−0.000247

(−1.73)

−0.000115*

(6)

Panel B: IV estimation

(−2.86)

(−1.57)

(−2.41)

−0.00230**

(−0.40)

−0.0000260

(5)

−0.000225*** −0.00021*** −0.000202*** −0.000204*** −0.000204*** −0.000113

(−1.09)

(−0.61)

−0.000543

(−0.65)

−0.0000387

(4)

(−3.66)

−0.000349

(−0.64)

−0.00253***

(−5.26)

(−1.41)

−0.00135

−0.00183*

(−1.94)

0.000589

(0.60)

−0.0000818 (−1.38)

(−1.02)

(3)

−0.000293*** −0.0000666

Financial openness

(2)

(−4.67)

(1)

Variables

Panel A: Country fixed effects

Table 5.8 (continued)

No

(0.41)

0.000131

(−0.89)

−0.0280

(−1.53)

−0.000110

(−2.00)

−0.00238**

(−0.03)

−0.000111

(−1.77)

−0.000118*

(8)

No

(−2.36)

−0.000489**

(−0.77)

−0.0274

(−1.63)

−0.000117

(−1.09)

−0.00130

(−0.10)

−0.000324

(−0.79)

−0.0000546

(9)

No (continued)

(−2.78)

−0.000731***

(1.24)

0.000465

(−2.96)

−0.00186***

(−1.35)

−0.0362

(−3.86)

−0.000570***

(2.51)

0.0219**

(1.02)

0.00167

(1.27)

0.000160

(10)

168 5 Bank Competition, Regulation, and Efficiency

0.972

−

Yes

9081

0.976

−

-

Year fixed effects

N

R-squared

Hansen J statistic

Hansen J statistic (p-value)

−

0.972

7066

Yes

No

(3)

−

0.973

7208

Yes

No

(4)

−

0.972

7039

Yes

No

(5)

0.111

3.385

0.979

4273

Yes

Yes

(6)

0.124

2.833

0.979

4249

Yes

Yes

(7)

Panel B: IV estimation

0.086

4.903

0.979

4273

Yes

Yes

(8)

0.280

2.545

0.979

4273

Yes

Yes

(9)

0.181

1.786

0.973

5812

Yes

Yes

(10)

This table examines the impact of competition and regulation on bank efficiency. The dependent variable is Profit inefficiency as defined in Eq. (5.1). Our sample period is 2001–2016. Bank-level controls include Z-score, TCD, LLP, FA, log TA, Commercial dummy, and Saving dummy. Country control variables include Bank concentration, Government-owned banks, Financial openness, Financial freedom, GDP growth, Inflation, and Crisis. Bank regulation variables include Activity restrictions, Capital regulatory index, Official supervisory power. All variables are defined in Table 5.1. Reported t-statistics (in parentheses) are based on standard errors that are heteroskedasticity consistent and clustered at the country level. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

7039

Yes

No

No

Bank fixed effects

(2)

(1)

Variables

Panel A: Country fixed effects

Table 5.8 (continued)

5.5 Empirical Results 169

170

5 Bank Competition, Regulation, and Efficiency

dummy, and Saving dummy. Country control variables include Bank concentration, Government-owned banks, Financial openness, Financial freedom, GDP growth, Inflation, and Crisis. Bank regulation variables include Activity restrictions, Capital regulatory index, Official supervisory power. All variables are defined in Table 5.1. Both bank and year fixed effects are included and reported t-statistics (in parentheses) are based on standard errors that are estimated based on Vogelsang (2012) and robust to heteroskedasticity, serial correlation, and spatial correlation. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively Finally, we performed our analysis on different subsamples: commercial banks and subsample of countries with more than 200 observations (see Table 5.9). Similar to Table 5.8, we report only the results of profit inefficiency for brevity.14 Our main findings are still consistent.

5.6 Conclusions This study investigates the impact of competition and regulation on bank efficiency based on international evidence in the Asia–Pacific region. Our sample consists of unbalanced financial data of 9193 observations from 1261 banks operating in 28 Asia–Pacific countries between 2001 and 2016. Our results indicate that market power is positively related to bank efficiency. The findings also highlight that stringent activity restrictions, strong official supervisory power, and low capital requirements are associated with increased bank efficiency. Furthermore, the market power effect in increasing profit and cost efficiency is more pronounced in a banking system characterized by the above activity restrictions, supervisory power, and capital requirements. Foreign banks operating under strict activity restrictions in a host country with strong official supervisory power have high profit and cost efficiency. Our findings highlight several important policy implications for financial regulators in Asia–Pacific region. First, in order to increase bank efficiency, regulators need to follow a more cautious approach to evaluating and approving bank consolidation to prevent excessive competition in the banking system at the country level. Second, a certain level of activity restrictions and official supervisory power should be maintained to improve bank efficiency, and this is more important for foreign banks that seek to take advantage of regulation differences between the parent country and host countries. Finally, high capital requirements reduce bank efficiency in the Asia–Pacific region, suggesting that policy makers need to identify the weaknesses in the present capital regulations and improve their effectiveness to enhance bank efficiency.

14 The results of cost inefficiency for commercial banks and subsample of countries with more than

200 observations have the similar findings and are available upon request.

Financial openness

Government-owned banks

Bank concentration

Saving dummy

Commercial dummy

log TA

FA

LLP

TCD

(−6.34)

−0.000366***

(−0.33)

(2.68)

0.129***

0.133***

(0.30)

−0.0000284

0.000196***

(2.65)

−

−

0.0000148

0.000106

−

−

0.000125***

−

−

−

(−30.57)

0.104***

(0.15)

0.00000752

(1.27)

−

−

−

(−31.38)

−0.0255***

−

−0.0263***

−0.0223***

(5.95)

0.306***

(−13.01)

(6.35)

(−0.18)

−0.00619

−

0.323***

(6.84)

(0.03)

(−0.94) 0.404***

(10.60) 0.00120

(11.14) −0.0327

(10.63)

(−6.83)

(−6.81) 0.0392***

−0.00038***

−0.000409***

(−2.15)

−0.00168**

(3)

0.0386***

(−3.07)

(−2.17)

0.0395***

−0.00570***

−0.00146**

Lerner index

Z-score

(2)

(1)

Variables

Panel A: Commercial banks

Table 5.9 Robustness checks based on different subsamples

0.115***

(3.17)

0.000152***

(1.60)

0.000121

−

−

−

−

(−32.24)

−0.0258***

(5.91)

0.324***

(−1.11)

−0.0395

(10.30)

0.0370***

(−6.48)

−0.000375***

(−0.03)

−0.0000835

(4)

0.0347

(−1.24) (continued)

−0.0000661

(−1.91)

−0.000173*

−

−

−

−

(−29.06)

−0.0233***

(6.39)

0.315***

(−0.31)

−0.0105

(10.94)

0.0377***

(−6.73)

−0.000364***

(−2.70)

−0.00834***

(5)

5.6 Conclusions 171

Official supervisory power * Lerner index

Official supervisory power

Capital regulatory index * Lerner index

Capital regulatory index

Activities restrictions * Lerner index

Activity restrictions

Crisis

Inflation

GDP growth

Financial freedom

Variables

Panel A: Commercial banks

Table 5.9 (continued)

(−2.10)

−0.000174 (−0.74)

(continued)

(2.21) −0.000486**

(−2.23)

0.000636**

(1.72)

(−0.13) −0.000663**

(−5.89) 0.000330*

(−2.79) −0.0000263

−0.00161***

(−4.21) −0.00066***

(−0.52)

−0.00100***

(−1.40)

(−1.53) −0.00976***

(−4.24)

(−1.26)

−0.00589

−0.000132

−0.00093***

(−0.58)

(−1.21)

−0.00663

(−1.75)

−0.000120*

(1.04)

(−0.73)

−0.00794

−0.00766

(−3.59)

−0.000248***

(4.14)

(2.00) 0.000111

−0.000217

(−1.97)

(−2.91)

(3.53)

(−1.97) 0.000390***

0.0000778**

(0.39)

(5)

(−1.79)

−0.000147**

−0.000217***

(−1.02) 0.000335***

−0.0000700**

(10.83)

(4)

−0.000511*

(5.21)

(0.58)

(−0.85) 0.000481***

(−1.96) 0.0000539

−0.0000359

(9.59)

(11.75) −0.0000300

(7.35)

(3)

(2)

−0.0000696*

(1)

172 5 Bank Competition, Regulation, and Efficiency

Yes 4342 0.740

N

R-squared

Commercial dummy

log TA

FA

LLP

TCD

(0.65)

(−5.61)

(−174.14)

(−172.62) 0.000945

(−345.61) −0.00550***

−0.0290***

−0.0289***

−0.0300***

(0.52)

0.000760

(6.80)

(6.75)

0.263***

0.259***

(3.82)

(7.68)

(3.68)

(3.63)

0.113***

(6.83)

0.0216***

(−2.89)

−0.000209***

(−1.41)

−0.00213

(8)

0.779

4005

Yes

Yes

(2.81)

5.542***

(3)

0.292***

0.109***

0.101***

(6.71)

(−2.86)

(−3.39) (7.33)

−0.000207***

−0.000205*** 0.0212***

(−2.10)

(−4.05)

0.0225***

−0.00356**

−0.00193***

Lerner index

Z-score

(7)

(6)

0.789

398z7

Yes

Yes

Variables

Panel B: Banks with more than 200 observations

Yes

Year fixed effects

(3.74)

(5.72)

Bank fixed effects

4.443***

2.347***

Constant

(2)

(1)

Variables

Panel A: Commercial banks

Table 5.9 (continued)

(0.91)

0.00130

(−175.05)

−0.0289***

(6.82)

0.264***

(3.30)

0.0961***

(6.14)

0.0190***

(−2.92)

−0.000210***

(−3.64)

−0.00869***

(9)

0.782

4094

Yes

Yes

(7.72)

5.481***

(4)

(−0.48) (continued)

−0.000677

(−32.55)

−0.0239***

(7.33)

0.283***

(2.87)

0.0841***

(7.52)

0.0232***

(−3.04)

−0.000216***

(−5.01)

−0.0144***

(10)

0.803

3987

Yes

Yes

(6.15)

2.020***

(5)

5.6 Conclusions 173

Activities restrictions * Lerner index

Activity restrictions

Crisis

Inflation

GDP growth

Financial freedom

Financial openness

Government-owned banks

−0.00144 (−1.53)

−0.000393 (−0.40)

(−0.72)

(−2.19)

(continued)

(−0.36)

−0.000110

(−1.59)

−0.00529

(0.73)

0.0000542

(−3.99)

−0.000511***

(−7.51)

−0.000275***

−0.000171

(−0.41)

−0.000370

(−1.42)

−0.0000887

(−1.13)

−0.000115

(−10.35)

−0.000297***

(−1.07)

−3.906

(3.38)

0.000195***

(1.80)

0.000183*

(1.62)

0.00486

(10)

(−1.06)

(−1.39)

−0.00132

(−2.68)

−0.000165***

(−8.85)

−0.000258***

(7.23)

0.0472***

(3.64)

0.000172***

(1.29)

0.000123

(2.94)

0.00879***

(9)

−0.000470**

−0.000320

(−1.99)

(−6.10)

(−2.16)

(−2.37) −0.000132**

−0.000245**

−0.000115 (−1.40)

(−9.58)

(−8.51)

−0.000344***

−0.000220**

−0.000277***

−0.000230***

(6.87)

(7.55)

(2.68) 0.0485***

(5.91)

(1.95) 0.0511***

(2.64)

0.000151***

(2.25)

0.000232**

(2.53)

0.00797**

(8)

0.0304***

0.000103*

(1.41)

(1.18) 0.000124***

0.000147

0.000114

(2.52)

(0.68)

Bank concentration

0.00817**

0.00202

Saving dummy

(7)

(6)

Variables

Panel B: Banks with more than 200 observations

Table 5.9 (continued)

174 5 Bank Competition, Regulation, and Efficiency

8644 0.979

N

R-squared

0.973

6772

Yes

(−0.17)

0.973

6772

Yes

Yes

(0.82)

0.356

(2.57)

(1.31)

0.974

6896

Yes

Yes

(1.54)

0.974

6772

Yes

Yes

(0.77)

0.675

(−6.52)

(−3.89) 0.2952

(2.35) −0.00146***

(−0.39) −0.000786***

0.000637**

0.000556**

0.000284 −0.000104

−0.0000544

(10)

(1.16)

(9)

0.000312

(8)

This table examines the impact of competition annd risk on bank efficiency. The dependent variable is Profit inefficiency as defined in Eq. (5.1). Our sample period is 2001–2016. Bank-level controls include Z-score, TCD, LLP, FA, log TA, Commercial dummy, and Saving dummy. Country control variables include Bank concentration, Government-owned banks, Financial openness, Financial freedom, GDP growth, Inflation, and Crisis. Bank regulation variables include Activity restrictions, Capital regulatory index, Official supervisory power. All variables are defined in Table 5.1. Both bank and year fixed effects are included and reported t-statistics (in parentheses) are based on standard errors that are estimated based on Vogelsang (2012) and robust to heteroskedasticity, serial correlation, and spatial correlation. Significance levels at 10%, 5%, and 1% are indicated by *, **, and ***, respectively

Yes

Year fixed effects

Yes

(4.473)

Yes

2.176***

(5.17)

(7)

5.146***

(6)

Bank fixed effects

Constant

Official supervisory power * Lerner index

Official supervisory power

Capital regulatory index * Lerner index

Capital regulatory index

Variables

Panel B: Banks with more than 200 observations

Table 5.9 (continued)

5.6 Conclusions 175

176

5 Bank Competition, Regulation, and Efficiency

References Agoraki, M.-E.K., Delis, M. D., & Pasiouras, F. (2011). Regulations, competition and bank risktaking in transition countries. Journal of Financial Stability, 7(1), 38–48. Allen, F., Carletti, E., & Marquez, R. (2011). Credit Market competition and capital regulation. The Review of Financial Studies, 24(4), 983–1018. https://doi.org/10.1093/rfs/hhp089 Andrie¸s, A. M., & C˘apraru, B. (2014). The nexus between competition and efficiency: The European banking industries experience. International Business Review, 23(3), 566–579. Anginer, D., Demirguc-Kunt, A., & Zhu, M. (2014). How does competition affect bank systemic risk? Journal of Financial Intermediation, 23(1), 1–26. Bae, K. H., & Goyal, V. K. (2009). Creditor rights, enforcement, and bank loans. The Journal of Finance, 64(2), 823–860. Barth, J. R., Caprio, G., & Levine, R. (2004). Bank regulation and supervision: What works best? Journal of Financial Intermediation, 13(2), 205–248. Barth, J. R., Caprio, G. J., & Levine, R. (2013a). Bank regulation and supervision in 180 countries from 1999 to 2011. Journal of Financial Economic Policy, 5(2), 111–219. Barth, J. R., Lin, C., Ma, Y., Seade, J., & Song, F. M. (2013b). Do bank regulation, supervision and monitoring enhance or impede bank efficiency? Journal of Banking & Finance, 37(8), 2879–2892. Basel Committee on Banking Supervision. (2012). Core Principles for Effective Banking Supervision. Battese, G. E., & Coelli, T. J. (1995). A model for technical inefficiency effects in a stochastic frontier production function for panel data. Empirical Economics, 20(2), 325–332. Beck, T., Demirgüç-Kunt, A., & Levine, R. (2006). Bank supervision and corruption in lending. Journal of Monetary Economics, 53(8), 2131–2163. Ben Naceur, S., & Kandil, M. (2009). The impact of capital requirements on banks’ cost of intermediation and performance: The case of Egypt. Journal of Economics and Business, 61(1), 70–89. Berger, A. N., & Hannan, T. H. (1998). The efficiency cost of market power in the banking industry: A test of the “quiet life” and related hypotheses. The Review of Economics and Statistics, 80(3), 454–465. Berger, A. N., Klapper, L. F., & Turk-Ariss, R. (2009). Bank competition and financial stability. Journal of Financial Services Research, 35(2), 99–118. Bonin, J. P., Hasan, I., & Wachtel, P. (2005). Bank performance, efficiency and ownership in transition countries. Journal of Banking & Finance, 29(1), 31–53. Boone, J. (2001). Intensity of competition and the incentive to innovate. International Journal of Industrial Organization, 19(5), 705–726. Boone, J. (2008). A new way to measure competition. The Economic Journal, 118(531), 1245–1261. Casu, B., & Girardone, C. (2009a). Does competition lead to efficiency? The case of EU commercial banks. Available at SSRN: https://ssrn.com/abstract=1200362. Casu, B., & Girardone, C. (2009b). Testing the relationship between competition and efficiency in banking: A panel data analysis. Economics Letters, 105(1), 134–137. Chinn, M. D., & Ito, H. (2008). A new measure of financial openness. Journal of Comparative Policy Analysis: Research and Practice, 10(3), 309–322. Chortareas, G. E., Girardone, C., & Ventouri, A. (2012). Bank supervision, regulation, and efficiency: Evidence from the European Union. Journal of Financial Stability, 8(4), 292–302. Claessens, S., Demirgüç-Kunt, A., & Huizinga, H. (2001). How does foreign entry affect domestic banking markets? Journal of Banking & Finance, 25(5), 891–911. Claessens, S., & Klingebiel, D. (2001). Competition and scope of activities in financial services. The World Bank Research Observer, 16(1), 19–40. Claessens, S., & Laeven, L. (2004). What drives bank competition? Some international evidence. Journal of Money, Credit and Banking, 36(3), 563–583. Das, A., & Ghosh, S. (2006). Financial deregulation and efficiency: An empirical analysis of Indian banks during the post reform period. Review of Financial Economics, 15(3), 193–221.

References

177

De Haas, R., & Lelyveld, I. Van. (2006). foreign banks and credit stability in central and Eastern Europe. A panel data analysis. Journal of Banking & Finance, 30(7), 1927–1952. Dell’Ariccia, G., & Marquez, R. (2006). Lending booms and lending standards. The Journal of Finance, 61(5), 2511–2546. https://doi.org/10.1111/j.1540-6261.2006.01065.x Demirgüç-Kunt, A., & Detragiache, E. (2011). Basel core principles and bank soundness: Does compliance matter? Journal of Financial Stability, 7(4), 179–190. Demirgüç-Kunt, A., & Huizinga, H. (2010). Bank activity and funding strategies: The impact on risk and returns. Journal of Financial Economics, 98(3), 626–650. Demirguc-Kunt, A., Laeven, L., & Levine, R. (2004). Regulations, market structure, institutions, and the cost of financial intermediation. Journal of Money, Credit and Banking, 36(3), 593–622. Dick, A. A., & Lehnert, A. (2010). Personal bankruptcy and credit market competition. The Journal of Finance, 65(2), 655–686. Djalilov, K., & Piesse, J. (2019). Bank regulation and efficiency: Evidence from transition countries. International Review of Economics & Finance, 64(2019), 308–322. https://doi.org/10.1016/j.iref. 2019.07.003 Djankov, S., La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (2002). The regulation of entry. The Quarterly Journal of Economics, 117(1), 1–37. Djankov, S., McLiesh, C., & Shleifer, A. (2007). Private credit in 129 countries. Journal of Financial Economics, 84(2), 299–329. Driscoll, J. C., & Kraay, A. C. (1998). Consistent covariance matrix estimation with spatially dependent panel data. Review of Economics and Statistics, 80(4), 549–560. Fang, J., Lau, C.-K.M., Lu, Z., Tan, Y., & Zhang, H. (2019). Bank performance in China: A perspective from bank efficiency, risk-taking and market competition. Pacific-Basin Finance Journal, 56(2019), 290–309. https://doi.org/10.1016/j.pacfin.2019.06.011 Fu, X. M., Lin, Y. R., & Molyneux, P. (2014). Bank competition and financial stability in Asia Pacific. Journal of Banking & Finance, 38(2014), 64–77. Goetz, M. R., Laeven, L., & Levine, R. (2016). Does the geographic expansion of banks reduce risk? Journal of Financial Economics, 120(2), 346–362. Hall, A. R., & Peixe, F. P. M. (2000). Data mining and the selection of instruments. Journal of Economic Methodology, 7(2), 265–277. Haque, F., & Brown, K. (2017). Bank ownership, regulation and efficiency: Perspectives from the Middle East and North Africa (MENA) Region. International Review of Economics & Finance, 47(2017), 273–293. https://doi.org/10.1016/j.iref.2016.10.015 Hasan, I., Wachtel, P., & Zhou, M. (2009). Institutional development, financial deepening and economic growth: Evidence from China. Journal of Banking & Finance, 33(1), 157–170. John, K., Litov, L., & Yeung, B. (2008). Corporate governance and risk-taking. The Journal of Finance, 63(4), 1679–1728. Kleibergen, F., & Paap, R. (2006). Generalized reduced rank tests using the singular value decomposition. Journal of Econometrics, 133(1), 97–126. Koetter, M., Kolari, J. W., & Spierdijk, L. (2012). Enjoying the quiet life under deregulation? Evidence from adjusted Lerner indices for US Banks. Review of Economics and Statistics, 94(2), 462–480. Laeven, L., & Levine, R. (2009). Bank governance, regulation and risk taking. Journal of Financial Economics, 93(2), 259–275. Laeven, L., & Valencia, F. (2018). Systemic banking crises revisited. IMF Working Paper No. 206. International Monetary Fund. Lerner, A. P. (1934). The concept of monopoly and the measurement of monopoly power. The Review of Economic Studies, 1(3), 157–175. Li, S. (2019). The impact of bank regulation and supervision on competition: Evidence from emerging economies. Emerging Markets Finance and Trade, 55(10), 2334–2364. https://doi. org/10.1080/1540496X.2018.1547191 Liu, H., Molyneux, P., & Nguyen, L. H. (2012). Competition and risk in South East Asian commercial banking. Applied Economics, 44(28), 3627–3644.

178

5 Bank Competition, Regulation, and Efficiency

Luo, Y., Tanna, S., & De Vita, G. (2016). Financial openness, risk and bank efficiency: Cross-country evidence. Journal of Financial Stability, 24(2016), 132–148. Mamatzakis, E., Staikouras, C., & Koutsomanoli-Fillipaki, N. (2005). Competition and concentration in the banking sector of the South Eastern European Region. Emerging Markets Review, 6(2), 192–209. Marinˇc, M. (2013). Banks and information technology: Marketability vs relationship. Electronic Commerce Research, 13(1), 71–101. Marquez, R. (2002). Competition, adverse selection, and information dispersion in the banking industry. The Review of Financial Studies, 15(3), 901–926. https://doi.org/10.1093/rfs/15.3.901 Maudos, J., & Fernández de Guevara, J. F. (2007). The cost of market power in banking: social welfare loss vs. cost inefficiency. Journal of Banking & Finance, 31(7), 2103–2125. Noman, A. H. M., Gee, C. S., & Isa, C. R. (2018). Does bank regulation matter on the relationship between competition and financial stability? Evidence from Southeast Asian Countries. PacificBasin Finance Journal, 48(2018), 144–161. Olivero, M. P., Li, Y., & Jeon, B. N. (2011). Competition in banking and the lending channel: Evidence from bank-level data in Asia and Latin America. Journal of Banking & Finance, 35(3), 560–571. Olley, G. S., & Pakes, A. (1996). The dynamics of productivity in the telecommunications equipment industry. Econometrica, 64(6), 1263–1297. Ongena, S., Popov, A., & Udell, G. F. (2013). “When the cat’s away the mice will play”: Does regulation at home affect bank risk-taking abroad? Journal of Financial Economics, 108(3), 727–750. Pasiouras, F., Tanna, S., & Zopounidis, C. (2009). The impact of banking regulations on banks’ cost and profit efficiency: Cross-country evidence. International Review of Financial Analysis, 18(5), 294–302. Quintyn, M., & Taylor, M. W. (2003). Regulatory and supervisory independence and financial stability. Cesifo Economic Studies, 49(2), 259–294. Schaeck, K., & Cihák, M. (2008). How does competition affect efficiency and soundness in banking? New empirical evidence. European Central Bank Working Paper Series No. 932. Soedarmono, W., Machrouh, F., & Tarazi, A. (2011). Bank market power, economic growth and financial stability: Evidence from Asian Banks. Journal of Asian Economics, 22(6), 460–470. Stiroh, K. J. (2000). Compositional dynamics and the performance of the US banking industry. FRB of New York Staff Report No. 98. Tirole, J. (1988). The theory of industrial organization. MIT Press. Vogelsang, T. J. (2012). Heteroskedasticity, autocorrelation, and spatial correlation robust inference in linear panel models with fixed-effects. Journal of Econometrics, 166(2), 303–319. Weill, L. (2004). On the relationship between competition and efficiency in the EU banking sector. Kredit Und Kapital, 37(3), 329–352. Williams, J., & Nguyen, N. (2005). Financial liberalisation, crisis, and restructuring: A comparative study of bank performance and bank governance in South East Asia. Journal of Banking & Finance, 29(8), 2119–2154. Zarutskie, R. (2013). Competition, financial innovation and commercial bank loan portfolios. Journal of Financial Intermediation, 22(3), 373–396. Zhao, T., Casu, B., & Ferrari, A. (2010). The impact of regulatory reforms on cost structure, ownership and competition in Indian banking. Journal of Banking & Finance, 34(1), 246–254.

Chapter 6

Conclusion

6.1 Overview This chapter gives a general discussion on the main findings of this book. The primary objective of this study is to evaluate the impact of financial regulatory on bank performance in normal times and during the global financial crisis. This chapter briefly summarizes the findings in each chapter of the main findings and provides a systematic overview of them.

6.2 Summary of the Main Findings Chapter 2 empirically investigates how bank capital and competitive conditions affect bank risk-taking. Particularly, based on the Basel Accords, we classify bank capital based on capital quality: Tier 1 capital with higher loss-absorbing capacity and Tier 2 capital with lower loss-absorbing capacity. We then examine the effects of bank capital structure on bank risk-taking. We test whether bank risk-taking behaviour varies according to the quality of bank capital, and whether this relationship changes significantly in different competitive conditions and during the financial crisis. Using the unbalanced financial information of 7620 banks from 118 countries between 2001 and 2016, we find that banks with a higher Tier 1 ratio and a lower Tier 2 ratio exhibit lower risk-taking. Further, a bank that has higher market power in a banking system tends to reduce its risk-taking activities. The results also confirm that banks with higher profitability, more funding, and higher quality of bank assets are exposed to lower risk. During a financial crisis, banks exhibit a higher level of risk-taking than in normal times. Our findings also highlight that the negative relationship between the Tier 1 ratio and bank risk are more pronounced in more competitive conditions. During the financial crisis, Tier 1 capital acted as a stable funding source and reduced bank risk,

© Shanghai Jiao Tong University Press 2021 S. Li, Financial Regulation and Bank Performance, Contributions to Finance and Accounting, https://doi.org/10.1007/978-981-16-3509-0_6

179

180

6 Conclusion

but the evidence on Tier 2 capital shows that a higher Tier 2 ratio results in a higher level of risk and increases bank instability. Chapter 3 investigates the impact of both bank reform and competition on stability by analyzing the empirical relationships among banking sector reform, competition, and stability in transition economies. By using unbalanced data on 1121 banks from 22 transition countries between 1998 and 2016, our findings confirm the negative relationship between market power and stability, supporting the view that competition increases bank stability. The results also confirm that bank reform increases banking stability, supporting the view that better financial development leads to a more stable banking system. Both higher activity restrictions and more explicit guidelines for asset diversification increase bank stability, but this positive effect significantly weakens for banks with higher market power. More stringent capital requirements in combination with higher market power increase the risk of bank insolvency. Declaring insolvency power, private monitoring, financial statement transparency, and deposit insurance have only a direct impact on bank stability. The goal of Chap. 4 is to examine the impact of bank regulation and supervision on competitive conditions in the banking sector in emerging economies and examine whether the unique characteristics in emerging markets shed light on the relationship between bank regulation and competition from different perspectives. Using a sample of 1629 banks in 23 emerging economies between 1996 and 2016, this study employs the Panzar and Rosse (1987) methodology and constructs the competitive variable H-statistic as a measure of competition in the banking system. We also compute the Lerner index (see also Coccorese, 2009; Koetter et al., 2012) and the Boone indicator (2008) as alternative measures of competition. We investigate how competition evolved under different types of bank regulation and supervision. As our sample period covers several banking crises in some emerging economies as well as the recent global financial crisis, we also examine whether the relationship between bank regulation and competition changed during the banking crisis. Considering the different roles played by domestic banks and foreign banks, we also study whether the impact of bank regulation on foreign banks exhibits different patterns. Our analysis shows that banking systems with higher concentration and fewer activity restrictions and entry barriers are more competitive. We also find that reducing foreign bank limitations and increasing capital strictness and official supervisory power also enhances competition in the banking sector. The results also provide evidence that competition in banking systems with fewer government-owned banks and fewer diversification guidelines tends to be more intensive. The results also confirm that higher private monitoring of banks and deposit insurance coverage significantly contribute to an increase in bank competition. During a banking crisis, the impact of activity restrictions, entry barriers, and diversification guidelines on competitive conditions become more effective, whereas foreign bank limitations, capital strictness, supervisory power, and private monitoring become less effective. To further investigate whether the relationship between bank regulation and competition varies according to the type of ownership, we divide the sample banks into two groups: foreign banks and domestic banks. Our findings suggest that foreign banks are more sensitive to official supervisory power and private

6.2 Summary of the Main Findings

181

monitoring and less sensitive to activity restrictions, foreign bank limitations, and diversification guidelines. Chapter 5 examines how bank competition and regulation affect bank efficiency, using a sample of 1261 banks across 28 Asia Pacific countries for the period 2001– 2016. We employ a stochastic frontier analysis (SFA) model to estimate bank efficiency. Our findings confirm that market power is positively related to bank efficiency. Increased activity restrictions, strong official supervisory power, and low capital requirements are associated with high bank efficiency. Furthermore, market power has a stronger efficiency-increasing effect in a banking system characterized by the activity restrictions, supervisory power, and capital requirements described above. Foreign banks operating under strict activity restrictions in a host country with strong official supervisory power are highly efficient.