Government intervention and banking sector stability international evidence an independent thesis submitted in fulilment of the requirements for the degree of doctor of philosophy

151 Table 4.5: Robustness test: Alternative bank stability measure – Government ownership, regulatory governance, control of corruption, political influence and bank distance to default

GOVERNMENT INTERVENTION AND

BANKING SECTOR STABILITY – INTERNATIONAL EVIDENCE

Nguyen Thi Thuy Linh

B.Sc (First Class Honours) – University of Economics Ho Chi Minh City, Viet Nam M.Sc (First Class Honours) – University of Economics Ho Chi Minh City, Viet Nam

An independent thesis submitted in fulfilment of the requirements for the degree of

Doctor of Philosophy

Department of Banking and Finance Faculty of Business and Economics

Monash University

COPYRIGHT NOTICES

Notice 1

Under the Copyright Act 1968, this thesis must be used only under the normal conditions of scholarly fair dealing In particular no results or conclusions should be extracted from it, nor should it be copied or closely paraphrased in whole or in part without the written consent of the author Proper written acknowledgement should be made for any assistance obtained from this thesis

Notice 2

I certify that I have made all reasonable efforts to secure copyright permissions for third-party content included in this thesis and have not knowingly added copyright content to my work without the owner's permission

TABLE OF CONTENTS

TABLE OF CONTENTS i

LIST OF TABLES vi

LIST OF FIGURES ix

LIST OF ABBREVIATIONS x

ABSTRACT .xi

STATEMENT OF AUTHORSHIP xiii

THESIS-RELATED RESEARCH OUTCOME xiv

ACKNOWLEDGEMENTS xv

CHAPTER 1: INTRODUCTION 1

1.1 INTRODUCTION 1

1.2 GOVERNMENT INTERVENTION IN BANKING SECTOR 4

1.3 MOTIVATION AND RESEARCH QUESTIONS 6

1.3.1.Government ownership and bank stability 7

1.3.2.Regulatory restrictions and bank stability 15

1.3.3.Government intervention in banking crises 18

1.4 RESEARCH OBJECTIVES AND HYPOTHESES 21

1.5 DATA, METHODOLOGY AND FINDINGS 23

1.6 CONTRIBUTIONS 27

1.7 IMPLICATIONS OF THE FINDINGS 30

1.8 ORGANIZATION OF THE THESIS 32

CHAPTER 2: LITERATURE REVIEW 33

2.1 INTRODUCTION 33

2.2 MARKET FAILURES AND GOVERNMENT INTERVENTION 33

2.2.1.Rationales for government intervention – Public interest theory 34

2.2.2.Limitations of government intervention – Private interest theory 36

2.3 GOVERNMENT OWNERSHIP AND BANK STABILITY 36

2.3.1 Government ownership and bank stability (H1.1) 37

2.3.2.Government ownership, political influence and bank stability (H1.2) 42 2.3.3 Government ownership, control of corruption, and bank stability (H1.3)44

2.3.4.Government ownership, regulatory governance quality and bank stability

(H1.4) 45

2.4 REGULATORY RESTRICTIONS AND BANK STABILITY 46

2.4.1.Theories on regulatory restrictions 46

2.4.2.Interest rate controls and bank stability 48

2.4.3.Credit controls and bank stability 50

2.4.4.Entry restrictions and bank stability 51

2.5 THE TIMING OF GOVERNMENT INTERVENTION IN SYSTEMIC BANKING CRISES AND CRISIS DURATION AND COSTS 53

2.5.1.Theories of government intervention in banking crises 53

2.5.2.Timing of government intervention and crisis duration 55

2.5.3.The timing of government intervention and crisis costs 57

2.6 CHAPTER SUMMARY 59

CHAPTER 3: DATA AND METHODOLOGY 63

3.1 INTRODUCTION 63

3.2 SAMPLE AND DATA 64

3.2.1 Data sources 64

3.2.2.Sample units and sampling procedure for RQ1 and RQ2 66

3.2.3.Sample units and sampling procedure for RQ3 71

3.3 METHODOLOGY FOR RQ1 AND RQ2 74

3.3.1 Variables for bank stability, government ownership and regulatory restrictions (RQ1 and RQ2) 74

3.3.2 Empirical models and estimation methods (RQ1 and RQ2) 89

3.4 METHODOLOGY FOR RQ3 102

3.4.1.Variable definitions 103

3.4.2.Empirical models and estimation methods (RQ3) 118

3.5 CHAPTER SUMMARY 133

CHAPTER 4: GOVERNMENT OWNERSHIP, REGULATORY RESTRICTIONS AND BANK STABILITY RESULTS 135

4.1 INTRODUCTION 135

4.2 DESCRIPTIVE STATISTICS AND CORRELATION ANALYSES FOR RQ1 AND RQ2 135

4.2.1.Descriptive statistics 135

4.2.2.Correlation analysis 142

4.3 EMPIRICAL RESULTS FOR RQ1 145

4.3.1.Results for H1.1 through H1.4 145

4.3.2.Control variables for H1.1 through H1.4 156

4.3.3.Robustness tests (H1.1 through H1.4) 159

4.3.4.Summary of RQ1 182

4.4 EMPIRICAL RESULTS FOR RQ2 186

4.4.1.Results for H2.1 through H2.3 186

4.4.2.Control variables for H2.1 through H2.3 195

4.4.3.Robustness tests (H2.1 through H2.3) 197

4.4.4.Summary of RQ2 204

4.5 SUMMARY FOR CHAPTER 4 206

CHAPTER 5: THE TIMING OF INTERVENTION, CRISIS DURATION AND CRISIS COSTS 211

5.1 INTRODUCTION 211

5.2 DESCRIPTIVE STATISTICS AND CORRELATION ANALYSIS FOR RQ3 211

5.2.1.Descriptive statistics 212

5.2.2 Correlation analysis 219

5.3 EMPIRICAL RESULTS – THE TIMING OF GOVERNMENT INTERVENTION AND THE CRISIS DURATION (H3.1) 221

5.3.1 Plots of hazard rate h(t) (crisis duration) 221

5.3.2.The timing of government intervention and the crisis duration 223

5.3.3.Government intervention – Containment measures 225

5.3.4.Government intervention – Resolution measures 226

5.3.5.Other control variables 227

5.3.6.Robustness tests 228

5.4 EMPIRICAL RESULTS – THE TIMING OF GOVERNMENT INTERVENTION AND THE CRISIS OUTPUT LOSSES (H3.2) 235

5.4.1.The timing of government intervention and the crisis output losses 235

5.4.2.Government intervention - Containment measures 238

5.4.3.Government intervention – Resolution measures 239

5.4.4.Other control variables 239

5.4.5.Robustness tests 240

5.5 EMPIRICAL RESULTS – THE TIMING OF GOVERNMENT INTERVENTION AND THE CRISIS FISCAL COSTS (H3.3) 246

5.5.1.The timing of government intervention and the crisis fiscal costs 246

5.5.2.Government intervention - Containment measures 248

5.5.3.Government intervention – Resolution measures 249

5.5.4.Other control variables 249

5.5.5.Robustness tests 250

5.6 SUMMARY OF CHAPTER 5 255

CHAPTER 6: CONCLUSION 258

6.1 INTRODUCTION 258

6.2 REVIEW OF RESEARCH QUESTIONS, OBJECTIVES, HYPOTHESES, AND FINDINGS 259

6.2.1.RQ1: Is bank stability related to level of government ownership? 260

6.2.2 RQ2: Is bank stability related to regulatory restrictions on banks? 264

6.2.3.RQ3: Are the crisis duration and crisis costs related to the timing of government intervention? 266

6.3 CONTRIBUTION 268

6.4 IMPLICATIONS 272

6.5 LIMITATIONS AND FUTURE RESEARCH DIRECTIONS 274

6.5.1 Limitations 274

6.5.2.Future research directions 276

APPENDIX 1: SAMPLING PROCEDURES AND SAMPLE 278

APPENDIX 2: DESCRIPTIVE STATISTICS 282

APPENDIX 3: CORRELATION ANALYSIS 293

REFERENCES 294

LIST OF TABLES

Table 1.1: Summary of the relevant literature to identify the research gaps addressed

in this thesis 9

Table 2.1 Summary of Research Questions and Hypotheses 60

Table 3.1: Data sources 64

Table 3.2: Sample selection for RQ1 and RQ2 68

Table 3.3: Sample description classified by country for RQ1 and RQ2 69

Table 3.4: Sample selection for RQ3 71

Table 3.5: Sample description for RQ3 72

Table 3.6: Definitions of Variables (Equations 3.6 to 3.10) 92

Table 3.7: Testing procedure for Hypotheses H1.1 to H2.3 98

Table 3.8: Definitions of Variables (Equations 3.16 to 3.18) 121

Table 3.9: Testing procedure for Hypotheses H3.1 127

Table 3.10: Testing procedure for hypotheses H3.2 and H3.3 130

Table 4.1: Descriptive statistics for variables used in Equations 3.6 to 3.10 137

Table 4.2: Correlation matrix for main independent variables used in Equations 3.6 to 3.10 143

Table 4.3: Level of government ownership and bank stability (H1.1) 147

Table 4.4: Government ownership, regulatory governance, control of corruption, political influence and bank Z-SCORE (H1.2 through H1.4) 151

Table 4.5: Robustness test: Alternative bank stability measure – Government ownership, regulatory governance, control of corruption, political influence and bank distance to default BDD (H1.1 through H1.4) 161

Table 4.6: Robustness tests: Alternative bank stability measure – Government ownership, regulatory governance, control of corruption, political influence and bank distance to capital (H1.1 through H1.4) 163

Table 4.7: Robustness tests: Alternative bank stability measure – Government ownership, regulatory governance quality, control of corruption, political influence and bank non-performing loan ratios (H1.1 through H1.4) 165 Table 4.8: Robustness test: Alternative bank stability measures – Government ownership, regulatory governance quality, control of corruption, political influence

Table 4.9: Robustness test: Alternative measures of government ownership –

Government ownership and bank Z-SCORE 170

Table 4.10: Robustness test: Generalized least squares fixed effect estimation method – Level of government ownership and bank stability (H1.1) 174Table 4.11: Robustness test: Fixed-effect two-stage least squares estimation method – Level of government ownership and bank stability (H1.1) 176Table 4.12: Robustness test: Generalized least squares fixed effect estimation method – Level of government ownership, regulatory governance, control of corruption, political influence and bank stability (H1.2 to H1.4) 178Table 4.13: Robustness test: Fixed-effect two-stage least squares estimation method – Level of government ownership, regulatory governance, control of corruption, political influence and bank stability (H1.2 to H1.4) 180Table 4.14: Summary of the results of testing hypotheses related to RQ1 183Table 4.15: Regulatory restrictions and bank stability (H2.1 through H2.3) 188Table 4.16: Robustness test–- Regulatory restrictions and bank stability: Alternative bank stability measures 198Table 4.17: Robustness test: Regulatory restrictions and bank stability – Alternative estimation methods 202Table 4.18: Summary of the results of testing hypotheses associated with RQ2 205Table 4.19: Major findings for RQ1 and RQ2 207 Table 5.1: Descriptive statistics for variables used in Equations 3.16 to 3.18 213 Table 5.2: Summary information of systemic banking crises 216Table 5.3: Correlation matrix main independent variables in Equations 3.16 and 3.18 220Table 5.4: Estimation of Cox proportional hazard model for H3.1 (Question 3.16) – The timing of government intervention and crisis duration 224Table 5.5: Estimation of Weibull proportional hazard model for H3.1 – Robustness tests (Equation 3.16) 230Table 5.6: Estimation of proportional hazard model for H3.1 using alternative approach to define crisis dates – Robustness tests (Equation 3.16) 232Table 5.7: Robustness tests: Alternative control variables – The timing of government intervention and the crisis duration (H3.1) 234

Table 5.8: Ordinary Least Squares estimates for H3.2 – The timing of government

intervention and the crisis output losses 236

Table 5.9: Robustness test: Alternative crisis dating approaches – Timing of government intervention and the crisis output losses (H3.2) 242

Table 5.10: Robustness test: Alternative output losses estimates – The timing of government intervention and the crisis output losses (H3.2) 244

Table 5.11: Ordinary Least Squares estimates for H3.3 – The timing of government intervention and the crisis fiscal costs 247

Table 5.12: Robustness test: Alternative crisis dating approach – The timing of government intervention and the crisis fiscal costs (H3.3) 252

Table 5.13: Robustness test for H3.3: Alternative fiscal costs estimates 254

Table 5.14: Major findings of RQ3 255

Table 5.15: Summary of the hypothesis testing results associated with RQ3 256

Table 6.1: Research questions, related hypotheses and results 261

Table A1.1: Sample description classified by year for RQ1 and RQ2 278

Table A2.1: Government ownership of banks in developed and developing countries 282

Table A2.2: Descriptive statistics for other main variables used in Equation 3.6 to 3.10 in developed and developing countries 286

Table A2.3: Summary information of the types of government intervention in banking crises 290

Table A3.1: VIF values for independent variables in Equation 3.16 to 3.18 293

LIST OF FIGURES

Figure 1.1: Government intervention in banking sector 5Figure 5.1 Non-parametrically estimated survivor functions for crisis duration with and without early government intervention 222

LIST OF ABBREVIATIONS

EFW Economic Freedom of the World

FE-2SLS Fixed-Effect Two-Stage Least Squares

GFC Global Financial Crisis

GLSFE Generalized Least Squares Fixed Effects

GMM Generalized Method of Moments

IFS International Financial Statistics

IFES International Foundation for Electoral Systems

IMF International Monetary Fund

OLS Ordinary Least Squares

PCR Principal Components Regression

WDI World Development Indicator

WEO World Economic Outlook

WGI Worldwide Governance Indicators

Two samples are used to examine these matters Government ownership and regulatory restrictions in relation to bank stability are investigated by applying the generalized method of moments (GMM) procedure on a sample of 2,898 commercial banks from 81 countries over the period 1995–2011 The timing of government intervention associated with the crisis duration and crisis costs is investigated using survival analysis and ordinary least squares (OLS) regression on a smaller sample of

65 systemic banking crises in 56 countries during 1980–2012

This thesis shows that the level of government ownership is associated with bank stability This relation, however, differs among country groups While it is negatively associated in developing countries, the association is positive in developed countries Furthermore, a higher quality of regulatory governance and control of corruption appeared to reduce the negative association in developing countries while strengthening the positive effect in developed ones In contrast, political influence during election years exacerbates the negative impact of government ownership on bank stability in developing countries as well as reducing the positive effect in

developed ones In terms of regulatory restrictions, each type of restriction was found

to impact differently impact on bank stability Interest rate and entry restrictions are positively associated with bank stability in developing countries but not developed countries, whereas credit controls are negatively related in both developed and developing countries In regard to the timing of government intervention and the crisis duration and costs, early intervention is associated with shorter crisis duration and less output losses but does not impact on the crisis fiscal costs

While most relevant studies focus on the difference in bank stability between government-owned and privately-owned banks, this thesis contributes to the bank government ownership literature by providing new evidence for the relation between the actual level of government ownership and bank stability The difference found between developed and developing countries further adds to the literature by showing that the same level of government ownership has a different effect on bank stability depending on the quality of regulatory governance, the control of corruption and political influence This thesis also extends the literature of bank regulation by providing new evidence of the relation between different dimensions of regulatory restrictions, interest rate controls, credit controls and sector entry restrictions, in relation to bank stability Finally, the thesis contributes to the crisis management literature by providing new evidence of the association between the timing of intervention and the crisis duration and costs

These findings should be of interest to academics, government authorities, policy makers, regulators, central bankers, taxpayers and other bank stakeholders

STATEMENT OF AUTHORSHIP

I declare that this thesis is my own work and contains no material previously published or written by another person(s) except where due reference is made

This thesis has not been submitted or does not contain material extracted in whole or

in part from a thesis or report presented by me for another degree, diploma or award

at any university

NGUYEN Thi Thuy Linh

THESIS-RELATED RESEARCH OUTCOME

Nguyen, T.T.L., Skully, M.T., and Perera, G.H.S (2014, May), Government

interventions in systemic banking crises: Timing, duration and costs Paper presented

at the Department of Banking and Finance, Monash University, Australia

Nguyen, T.T.L., Skully, M.T., and Perera, G.H.S (2013), Government ownership,

economic development, regulation quality and bank stability: International evidence,

Paper being invited to be revised and resubmitted to Journal of Banking and Finance

Nguyen, T.T.L., Skully, M.T., and Perera, G.H.S (2012, November), Regulatory

restrictions and banking sector stability: International evidence from 73 countries,

Paper presented at the 25th PhD Conference in Business and Economics, Perth, Australia

Nguyen, T.T.L., Skully, M.T., and Perera, G.H.S (2012, October), Regulatory

restrictions, ownership structure and banking sector stability: Evidence from selected Asian countries Paper presented at the 27th International Conference of the American Committee for Asian Economic Studies, Melbourne, Australia; Invited to

be revised and resubmitted to the Journal of Asian Economics

Nguyen, T.T.L., Skully, M.T., and Perera, G.H.S (2012, September), Government

ownership and bank stability? Bank-level evidence during period 1997-2010 Paper

presented at the 30th SUERF Colloquium on “States, Banks, and the Financing of the Economy”, Zurich, Switzerland and at the 2012 International Conference for the

ACKNOWLEDGEMENTS

First and foremost, my deepest gratitude is to my main supervisor, Professor Michael Skully, for his constructive guidance, useful feedbacks, untiring support, and encouragement during the course of my PhD His in-depth knowledge on a broad spectrum of banking topics has been extremely beneficial for me I am also greatly indebted to my associate supervisor, Dr Shrimal Perera, for his careful reading, critical and constructive comments on the revisions of this thesis Long discussions with him helped me sort out the difficulties in my work

I gratefully acknowledge the Australian Government for the Australia Award that made my PhD work possible They have provided me with a high value scholarship and continual assistance during my candidature My work was also generously supported by the Department of Banking and Finance, Faculty of Business and Economics and Monash University Institute of Graduate Research Likewise, I thankfully acknowledge the conference funding offered by SUERF – The European Money and Finance Forum

My appreciation is also to Professor Christine Brown and Professor Philip Gray – independent assessors during confirmation, mid-candidature and pre-submission presentations Their insightful comments and constructive criticisms at different stages of my research were thought-provoking and helped me hold to a high research standard Special thanks are also due to Associate Professor Kevin Tant for his advices on different issues, teaching opportunities, and for the best interest he has shown in my education

For their comments and advice, my gratitude is given to participants at the 27thInternational Conference of the American Committee for Asian Economic Studies (Melbourne), the 2012 International Conference for the Financial Engineering and Banking Society (London), the 30th SUERF Colloquium on “States, Banks, and the Financing of the Economy” (Zurich), the 25th PhD Conference in Business and Economics (Perth), research seminars and Brown Bag presentations at Monash University I feel particularly grateful to the insightful feedback from Professor Ike Mathur (Southern Illinois University), Professor Michael Plummer (John Hopkins University), Professor Terry Walter (SIRCA), Professor Jeffrey Sheen (Macquarie University), Professor Kostas Andriosopoulos (ESCP Europe Business School), Associate Professor David Tripe (Massey University), Dr Patrick McColgan (University of Strathclyde), Dr Leonard Nakamura (Federal Reserve Bank of Philadelphia), Dr Linden Luo (Australian Treasury) and anonymous referees

Thanks are also due to the following staff at the Monash University for facilitating

my learning and teaching experience and for numerous discussions on related topics that helped me improve my knowledge in the area: Associate Professors Elaine Hutson, Associate Professor Cameron Truong, Dr Kym Brown, Ms Katherine Avram, Dr Binh Do, Dr Viet Cao, Dr Minh Do, Dr Tram Vu, and Dr Maria Strydom at Monash University Their critical remarks and suggestions have always been very helpful in improving my skills

I would like to acknowledge the work of Dr Guenter A Plum who provided professional editorial service in my thesis I gratefully acknowledge the assistance offered to me by Mr Amish Haria and Ms Danielle Change from Bureau van Dijk

Electronic Publishing on the BankScope database, the CD historical disks and analysis software Professor Ugo Panizza (Graduate Institute of International and Development Studies in Geneva) also deserves a special note for his kindness in providing me the historical ownership data which is no longer available on the BankScope database The help of Dr Rasika Amarasiri, Manager Research IT Services, with different databases is also thankfully acknowledged I also thank the subject librarian, Mr David Horne, other library staff, and the Monash University Institute of Graduate Research staff for their support and advice Especially, thanks are due to Ms Kinda Say (Australia Award coordinator), Ms Judy Beart, Ms Jess Jie Han, Ms Cherie Lau, Ms Krista Constantas and Ms Connie Thalmmier

I also thank my fellow PhD students for their academically useful discussions, research inspirations and being very good friends: Anil Perera, Davin Wang, Yoni Navon, Pearpilai Jutasompakorn, Jyotimor Podder, Chloe Ho, Pham Phu Quoc, Vu Hoang Van, Nguyen Thi Ngoc My, Nguyen Yen Ngoc, Xing Yan, Xiaoting Wei, Neeru Chaudhry, and Terry Shi They have helped me stay sane through these difficult years Their support and care helped me overcome setbacks and stay focused

on my study

Very special thanks should go to my former advisors at the University of Economics

Ho Chi Minh city for their various forms of support during my study: Professor Nguyen Dong Phong, A Professor Pham Van Nang, A Professor Tran Hoang Ngan,

A Professor Nguyen Trong Hoai and A Professor Nguyen Minh Kieu Thanks are also due to my lovely colleagues at the School of Banking

Last but not least, I wish to dedicate this thesis to my family for their love, support and encouragement throughout this journey In particular, to my dear husband, Ho Nam Giang; I am grateful for his love, companionship, endless patience, and encouragement to make the completion of this thesis possible To my seven-year-old and ten-month-old sons, Ho Nam Nhat Minh and Ho Nam Nhat Huy, thanks for being my endless sources of energy and bundles of joy to give me strength to overcome difficulties encountered in my pursuit of the PhD degree I am deeply indebted to my parents, Nguyen Huu Ly and Dinh Thi Hue Binh for providing unconditional love and care, raising me with a love of research and supporting me in all my pursuits To my parents-in-law, Ho Tu Cuong and Chu Thi Hue, thanks for your generous support, understanding and care of my children when I was working

on this thesis Thanks are also due to my brother Nguyen Duy Khanh, sister-in-law Tran Thuy Tien and brother-in-law Ho Nam Chau for their good wishes and care of

my parents while I was studying

CHAPTER 1: INTRODUCTION

This thesis examines three distinct but related issues concerning government intervention in banking sectors in both developed and developing countries over the period from 1980–2012:1 (1) the level of government ownership in relation to bank stability, (2) the association between regulatory restrictions and bank stability, and (3) the impact of the timing of government intervention on the duration and costs of systemic banking crises While the first and second research issues remain widely debated and controversial issues among policymakers and academics, the third is largely unexplored

Given the banking crises of the last few decades including the mid-2007 global financial crisis (GFC), bank stability2 is an important concern for both policy makers and regulators They are highly aware that bank instability may trigger severe economic downturns and impose large negative externalities on economies (Hasman, López & Samartín, 2011) Moreover, the importance of banks as participants in the payment system and as intermediaries in the economy, coupled with the vulnerability

& Tarazi, 2011; Berger, Klapper, & Turk-Ariss, 2009; Boyd & Runkle, 1993; Iannotta, Nocera, & Sironi, 2007)

of their own financial structures, has led governments to accord them special attention (Fama, 1985)

Government has a major role in the banking sector from the stability perspective Healthy competition and the high quality infrastructure require strong prudential supervision as does mitigating the adverse impact of a banking crisis (Greene, 2012) Alternatively, the government could impose regulatory restrictions or regulate financial products to limit bank risk-taking activities and so avoid the potential externalities associated with fragility Government also intervenes directly, such as taking ownership of banks, to pursue stability objectives (Hainz & Hakenes, 2012) This gives the government more control over their safety and soundness, but can become quite expensive for taxpayers (Körner & Schnabel, 2011) Suchinterventions could also create expectations about future actions and lead to higher bank risk-taking and a misallocation of resources As a result, banks might become more fragile and the system even more vulnerable to panics In addition, inappropriately timed government intervention might result in more severe consequences

Although of crucial importance, the effectiveness of government intervention in the banking sector remains inconclusive Some authors favour greater intervention whereas others do not Those favouring intervention argue that it is appropriate to correct market failures provided such failures exist (Dow, 1996; Pigou, 1962; Stigler,

1971).3 Furthermore, banking systems with government intervention have provided stability with fewer failed banks (Bhattacharya, Boot, & Thakor, 1998; De Graeve, Kick, & Koetter, 2008) In contrast, those against intervention claim that market failures are not large and governments may in fact act more in their own interests when addressing them (Benston & Kaufman, 1996; Frye & Shleifer, 1997; Selgin & White, 1994; Shleifer & Vishny, 1994) The relaxation of government banking controls might also cause the over-expanded credit that has resulted in bank failures

in many countries (Benston, 2000; Carletti, 2008; Freixas, 2010)

Moreover, the main contrast between the mid-2007 GFC and those crises that preceded it, is the developed economies were affected much more strongly and more directly than developing economies (World Bank, 2013a) Less-developed economies may seem to offer more scope for government intervention to spearhead development Nevertheless, such economies often have a less effective institutional framework, which in turn increases the risk of inappropriate interventions (Demirgüç-Kunt & Detragiache, 1998; Greenwald & Stiglitz, 1986; Loayza & Ranciere, 2006)

The remainder of this chapter proceeds as follows Section 1.2 provides the background on the forms of government intervention Section 1.3 discusses the motivations behind the three research questions Section 1.4 summarizes the research objectives and hypotheses Section 1.5 details the sample, data sources and

3

Market failure occurs when a system fails to perform desirable activities or to cease undesirable ones (Bator, 1958) Alternatively, market failure is also defined as when the market fails to produce desired goods and services, to produce them in sufficient quality/quantity or produces goods with harmful spillovers (Ulbrich, 2011)

methodology Sections 1.6 and 1.7 cover the thesis’s contributions to the literature and their significance Finally, section 1.8 outlines the organization of the thesis

Government intervention is designed to correct and/or limit the repercussions of market failures (Ulbrich, 2011) and can usually be classified into two groups (Figure 1.1): first, preventive measures designed to promote appropriate behaviour patterns and so enhance stability; and second, remedial measures used to contain the consequences of an actual or threatened outbreak of instability (also known as crisis management policies) (Allen & Wood, 2006)

Preventive measures, which are classified based on their effects, include direct and indirect intervention While government ownership is known as direct intervention, indirect intervention includes regulatory restrictions, other prudential regulations and supervision Regulatory restrictions are restrictions on competition and restrictions

on the scope and scale of banks (Gual, 1999; Ongena, Popove & Udell, 2013), including controls on interest rates and fees, credit controls, entry restrictions or restrictions of bank activities in other sectors (insurance, securities or non-traditional banking) Other regulations and supervision include capital adequacy requirements, liquidity management and deposit insurance (Allen & Wood, 2006; Borio, Vale, & von Peter, 2010)

Figure 1.1: Government intervention in banking sector

Source: Author’s own design based on Allen and Wood (2006); Borio et al (2010); Klapper and Zaidi (2005); Ongena et al (2013)

Government intervention

Preventive measures

Direct intervention

Government ownership

Indirect intervention

Regulatory restrictions

Prudential regulations &

supervision

Remedial measures

Containment measures

Liquidity support;

Blanket guarantees

Deposit freeze;

Bank holidays

Resolution measures

Recapitalizatio n; Restructure

The remedial measures are applied when a banking system comes under stress and include both containment and resolution measures (see Borio et al., 2010; Honohan

& Klingebiel, 2003; Laeven & Valencia, 2008, 2010) The containment policies include emergency liquidity support to financial institutions, deposit freezes, bank holidays, blanket guarantees, and so forth.4 In contrast, resolution measures cover government-assisted bank recapitalizations through injecting funds, government-assisted sales to new owners, and government-owned asset management companies buying and resolving distressed loans (Laeven & Valencia, 2008)

The public interest theory on which this thesis is based relates to market failures and the role of government (Stiglitz, 1988) According to this, government should intervene to correct market failures in the banking sector because private markets have done particularly poor jobs in certain areas, which results in negative externalities such as adverse selection, incomplete market conditions and information asymmetries.5 However, government intervention in the banking sector may result in moral hazard, and so distort bank risk taking and reduce banking system stability (Boyd & De Nicolo, 2005)

4 Liquidity support or emergency lending mean central banks’ lending to insolvent banks on good collateral at low or penalty rates; a deposit freeze is a deposit timeout that lasts for weeks or months; and a banking holiday is a brief timeout taken to allow supervisors the opportunity to diagnose individual-bank insolvencies (See Kane and Klingebiel, 2004 for more detail)

5

Negative externalities arise when the insolvency or operations of one firm negatively affects other

As discussed in Section 1.2, government intervenes in the banking sector using different approaches, such as preventive measures to promote bank-appropriate behaviour patterns and hence foster bank stability or remedial measures to shorten the duration, or costs, of a systemic banking crisis once it occurs Their efforts are to avoid the severe negative impact imposed on economies due to bank insolvencies or banking crises Although of importance, the effectiveness of such intervention is still largely unexplored, or a controversial issue, which requires further investigation

The following sections present how government interventions are discussed in the extant literature, and the motivations underlying the three research questions posed

by this thesis

1.3.1 Government ownership and bank stability

Among different types of intervention discussed in Section 1.2, government may favour bank ownership (complete or partial), and so control their operations and the banking system (La Porta, Lopez-de-Silanes, & Shleifer, 2002) Indeed, government ownership is claimed to be an easier way to correct market failures (Clarke, Cull, & Shirley, 2005) and may also be less costly (Sappington & Stiglitz, 1987) So while many governments have privatized government-owned banks, others have resisted this trend A few have nationalized or even renationalized previously privatized

banks,6 believing that such ownership affords some important policy tools (Borisova, Brockman, Salas, & Zagorchev, 2012) This revives the debate over the role and impact of government ownership, especially from the perspective of bank stability

The bank ownership literature has mainly focused on the relation between types of ownership (foreign, domestic private and government) and bank performance or profitability (as summarized in Table 1.1) Generally, government-owned banks are found less efficient than privately-owned or foreign-owned banks in developing, though not developed, countries (Bonin, Hasan, & Wachtel, 2005b; Fries & Taci, 2005; Lin & Zhang, 2009; Micco et al., 2007) Another aspect considered in the literature is the relation between privatization and bank performance Some studies find evidence that privatization generally improves bank performance, especially in developing countries (Bonin, Hasan, & Wachtel, 2005a; Boubakri, Cosset, Fischer & Guedhami, 2005; Megginson, 2005; Otchere, 2005) This result is consistent with that of non-financial firms: after being privatized, firms become more profitable, improve their operating efficiency and increase their capital investment (Boubakri & Cosset, 1998; Megginson, Nash, & Randenborgh, 1994) This is further enhanced when the government fully relinquishes control (Clarke et al., 2005)

6 The prior empirical findings with regard to government ownership, record that such bank ownership was once common in both developed and developing countries (see La Porta et al., 2002; Megginson, 2005; Micco, Panizza, & Yañez, 2007; Olivero, Li, & Jeon, 2011) After 1995, privatization and openness to foreign participation in many countries resulted in the proportion of government-owned banks dropping dramatically (Yildirim & Philippatos, 2007) More recently, this decline was slightly reversed due to partial nationalization programs in some countries, such as the United Kingdom (Royal Bank of Scotland and Lloyds TSB), Ireland (Allied Irish Bank),

Table 1.1: Summary of the relevant literature to identify the research gaps addressed in this thesis

Panel A: Preventive measures

Angkinand and Wihlborg (2010), Cornett, Guo, Khaksari, & Tehranian (2010), Forssbæck (2011), Iannotta, Nocera and Sironi (2007, 2013)

Altunbas, Evans, and Molyneux (2001), Beck, Crivelli and Summerhill (2005), Berger, Hasan and Zhou (2009), Ghosh (2010), Karas, Schoors and Weill (2010)

Choi and Hasan (2011), Cornett et al (2010), Körner and Schnabel (2011), Mamatzakis, Staikouras, and Koutsomanoli-Filippaki (2008), Megginson (2005), Shen and Lin (2012), Williams and Nguyen (2005)

Level of

government

ownership

Not previously examined

Not previously examined – Addressed by RQ1 a

Beck, Cull, and Jerome (2005), Lin and Zhang (2009), Omran (2007), Sun, Tong, &

Agoraki, Delis, and Pasiouras (2011), Cubillas and González (2014), González (2005), Laeven and Levine (2009), Liu, Molyneux, and Nguyen (2011)

Wihlborg (2010), Barth, Caprio, and Levine (2008), Caprio and Klingebiel (1997), González-Rodríguez (2007), Levine (2001), Sturm and Williams (2004), Williams and Nguyen (2005) Interest rate

controls

Not previously examined

Not previously examined – Addressed by RQ2 b

Hendershott, and Lee (1998), Marco and Robles-Fernández (2008), Lee and Hsieh (2014)

García-Credit controls Not previously

examined

Not previously examined – Addressed by RQ2 c

and Phani (2008), Giannetti (2007), Levine and Zervos (1998), Quinn (1997) Foreign bank

Table 1.1: Summary of the relevant literature to identify the research gaps addressed in this thesis (continued)

Entry restrictions Not previously

examined

Not previously examined – Addressed by RQ2 e

Khiabani and Hamidisahneh (2012) Barth, Caprio, and Levine (2006), Huang

(2008), Naceur and Omran (2011), Peltzman (1965)

Activity

restrictionsf

Not previously examined

Agoraki et al (2011), Barth, Caprio, and Levine (2004), Beck, De Jonghe, and Schepens (2013), Laeven and Levine (2009) Morrison (2011)

Molyneux, and Pasiouras (2011), Fonseca and González (2010), Pasiouras, Tanna, and Zopounidis (2009), Turk-Ariss (2009)

Prudential

regulations and

supervision

Not previously examined

Agoraki et al (2011), Barth

et al (2004), Buch and DeLong (2008), Delis et al

(2011), Hadad, Agusman, Monroe, Gasbarro, and Zumwalt (2011), Laeven and Levine (2009), Morrison (2011)

Hadad et al (2011) Delis et al (2011), Fonseca and González

(2010), Pasiouras et al (2009)

Panel B:

Remedial

measures

Boyd et al (2005), Claessens, Ayhan Kose

et al (2010), Hoggarth et al (2002), Hoggarth et al (2005), Honohan and Klingebiel (2003), Honohan and Laeven (2005) Laeven and Valencia (2010)

Note:

a One exception is Iannotta et al (2013) who also employ the ownership percentage of the bank’s equity capital held by government to control for the differences between government-owned and privately-owned banks Their study, however, focuses only on the bank’s credit ratings instead of bank stability as discussed earlier Their sample was also limited to 16 European (developed) countries during the period from 2000–2009

b ,c Hellmann, Murdock, and Stiglitz (2000) is a theoretical paper examining the deregulation of interest rate control and credit controls in relation to bank behaviour The most relevant studies are Lee and Hsieh (2014) and Angkinand et al (2010) Lee and Hsieh, however, provide evidence of the impact of interest rate controls and credit controls on the relation between bank competition and stability while Angkinand et al focus on the impact of liberalization of such controls on the likelihood of systemic banking crises In addition, their interest rate and credit controls data, provided by Abiad, Detragiache, and Tressel (2010), is only up to 2005 Using 2005 regulation information to examine the impact on bank stability in 2009 might be ambiguous as such controls might have changed over the period 2005–

2009

d Foreign bank entry restrictions are defined as restrictions imposed on the entry of new domestic banks or of other potential competitors, for example foreign banks or nonbank financial intermediaries (Abiad et al., 2010; Gwartney, Hall, and Lawson, 2013) This is only considered a part of the restrictions on new bank entry

e While Barth et al (2004) also examine new bank entry restrictions, their study investigates the association between entry restrictions and the likelihood of a banking crisis Keeley (1990) and Marcus (1984) limit their study to the United States and focus only on branching restrictions

f

Activity restrictions reflect the ability of banks to be involved in securities, insurance and real estate activities

g ,h,i Although Mayes (2011) and Peek and Rosengren (1997) address early government intervention, they focus only on the importance of early government intervention to prevent bank problems or bank insolvencies, instead of limiting the consequences of banking crises (i.e., duration and costs) Hoggarth et al (2005) and Honohan and Klingebiel (2003) also discuss the impact of early government intervention in systemic banking crises They, however, use dummy variables for crisis containment policies to indicate early intervention (as opposed to resolution policies) This approach could not show the different impact of the same type of intervention at a different time Thus, the literature remains unexplored in regard to the timing of government intervention in relation to crisis duration and crisis costs

A few other studies, such as those by Barry et al (2011), Fraser and Zardkoohi (1996), García-Marco and Robles-Fernández (2008), also examine the relation between bank ownership and risk These, however, focus on publicly7 versus privately-owned banks, management controls or shareholder controls, instead of the nature of the owners (government/foreign/domestic private)

Despite the extensive evidence of the effects of government ownership on bank performance or economic development,8 limited empirical work exists on the relation between government-owned banks and stability whereas the theoretical argument on this issue remains inconclusive Such debate offers different theories, both encouraging and discouraging government ownership, known as political, agency and social views On the one hand, political view suggests that when governments actually own banks, their political influence on resource allocation and corruption might result in riskier activities and threaten stability (Dinç, 2005; Sapienza, 2004) Corruption is considered a serious problem, which is defined as when borrowers use political pressure and bribery to obtain bank loans at less than market rates causing the government-operated allocation mechanisms to be potentially less efficient than private markets (Hainz and Hakenes, 2012) Politicians might use this centralized information and enforcement power for their own benefits, especially during election years (Khwaja and Mian, 2005) Furthermore, an agency view, consistent with the political view, emphasizes that banks with government ownership might experience

more severe agency costs than their privately-owned counterparts (Banerjee, 1997)

or may simply not be run effectively (Andrianova et al., 2012)

On the other hand, according to the social view, government-owned banks, which benefit from government support, might achieve better risk sharing as they also have the power to redistribute and collect information These banks may more easily overcome information, contracting and coordinating problems and so enhance bank stability (Megginson et al., 1994) Moreover, they could become more effective with

a better regulatory governance environment imposing controls on politicians’ behaviour and limiting the agency costs (Laeven & Levine, 2009)

The few empirical studies that do focus on government ownership and bank stability only compare stability between government-owned banks and their counterparts (foreign or privately-owned) using a government ownership dummy variable set at a certain threshold (e.g., 50% or 20%) For example, Angkinand and Wihlborg (2010), Beck et al (2009), Berger et al (2005) and Tabak et al (2011), all define government-owned banks as when a government owns more than 50% of bank equity In contrast, the Cornett et al (2010) and Iannotta et al (2007, 2013) threshold

is 20% Such approaches could not determine whether a higher or lower level of government ownership, as well as changes of ownership over time, is related to bank stability.9 Moreover, a lower equity threshold, such as 10% or less, may nevertheless

be adequate to experience the advantages or disadvantages of government ownership, especially in publicly-held banks It is argued that banks can be considered as

9 Level of government ownership can be captured by the proportion of government ownership of bank equity (Saunders, Strock, & Travlos, 1990)

controlled if 10% or more of their voting stock is held by a single party such as the government (Vernon, 1970) Furthermore, these previous studies comparing stability between government-owned and privately-owned banks are also limited to a single country, such as Germany (Beck et al., 2009), Argentina (Berger et al., 2005) and Brazil (Tabak et al., 2011), or to a certain point of time, i.e., in 1999, 2002 and 2005 (Angkinand & Wihlborg, 2010) Such focus on a single country could not provide insights into the stability of government-owned banks in different institutional environments and countries as does a cross-country study

The findings of these diverse studies remain mixed and inconclusive owned banks are found to be riskier than privately-owned banks with more loan losses (Iannotta et al., 2007) and higher operating risks (Iannotta et al., 2013) or higher non-performing loan ratios (Berger et al., 2005) Furthermore, government-owned bank lending in Italy seems to favour large firms and depressed areas (Sapienza, 2004) and so threatens bank stability through this riskier lending In contrast, others argue that government-owned banks are more stable than privately-owned banks due to their lower volatility in profits in Germany (Beck et al., 2009) or have lower default risk than private banks in some European countries10 (Iannotta et al., 2013)

Government-In summary, to the best of the author’s knowledge, research of the relation between levels of government ownership in banks and bank stability has not yet been

10

Their sample included 15 European countries: Austria, Belgium, Finland, France, Germany, Greece,

published, especially across both developed and developing countries This gap, as summarized in Table 1.1, and the inconclusive debates on government ownership of banks in the literature, motivate this thesis’s first research question (RQ1) to be phrased as follows:

RQ1: Is bank stability related to level of government ownership?

1.3.2 Regulatory restrictions and bank stability

In addition to direct ownership, government might also intervene indirectly in banks via regulatory restrictions to pursue their objectives (Figure 1.1 in Section 1.2) These restrictions were originally intended to limit moral hazard problems and reduce excessive risk taking (Bhattacharya, 1982) There are, however, contradictory views about such restrictions Likewise the impact of various types of restrictions, such as interest rate controls, credit controls and entry restrictions, on bank stability has not been fully examined (as summarized in Table 1.1)

With regard to different viewpoints, one suggests that restrictions prevent banks from reaping necessary diversification and scale benefits which hamper bank stability (Agarwal & Ramaswami, 1992; Avgouleas, 2009; Beck et al., 2013; Matutes & Vives, 2000) These restrictions could also exacerbate the negative impact of competition on bank stability (Agoraki et al., 2011) In contrast, relaxing restrictions

in the banking sector has often been followed by instability (Caprio & Klingebiel,

1996; Kaminsky & Reinhart, 1999).11 Furthermore, restrictions on say, interest rates

or entry restrictions, are seen as fostering stability (Hellmann et al., 2000) This is because banks have better profit opportunities, capital cushions and therefore fewer incentives to take aggressive risks, with positive consequences for financial stability

Despite theoretical debates on bank restrictions and banking sector stability, there is little empirical evidence available Some studies, such as Caprio and Klingebiel (1997), Hellmann et al (2000) and Keeley (1990), identify the relaxation of branching restrictions being related to the higher probability of systemic banking crises In contrast, Barth et al (2004) provide evidence that tighter regulatory restrictions on bank activities and foreign bank entry are associated with a higher probability of a banking crisis Previous studies also concentrate on the impact of deposit guarantees, or capital requirements on bank risk taking behaviour (see Angkinand & Wihlborg, 2010; Antão & Lacerda, 2011; Fu et al., 2014; Laeven & Levine, 2009; Morrison & White, 2005; Pennacchi, 2006; Repullo, 2004) but largely overlook the association between other key regulatory restrictions (entry, interest rates and credit) and bank stability

Entry restrictions are typically only used as instrumental variables in those equations testing the relation between bank competition and stability (Beck et al., 2013; Berger,

11 Changes in regulatory restrictions have occurred not only in developed countries such as the U.S in the 1980s and Europe in the 1990s (Vives, 1991) but also in developing countries in South East Asia during the 1990s (Williams & Nguyen, 2005), and transition countries in Europe from the late 1990s to the early 2000s (Brissimis, Delis, & Papanikolaou, 2008) and Latin America over 1990–

1997 (Olivero et al., 2011; Yildirim & Philippatos, 2007) These changes have been widely examined but no research so far has studied the relationship between them and bank stability, especially across developed and developing countries These groups are expected to show different

Klapper et al., 2009; Demirgüç-Kunt & Detragiache, 2002; Schaeck, Čihák, & Wolfe, 2009) These studies, however, do not show whether a relation exists between those restrictions and bank stability This limitation may reflect their use of the Barth, Caprio, and Levine (2001) and the Barth et al (2004, 2008) database on regulation and supervision around the world, which has only three years (1999, 2003 and 2007) of cross-sectional data Such data limits the ability to observe the changes

of regulation and supervision over a much longer period and more frequently In addition, many countries experienced regulatory changes before 2000 (Angkinand et al., 2010).12 Furthermore, this data addresses only those policies preventing banks from engaging in securities, insurance and real estate activities Using the same database, Fu et al (2014) and Uhde and Heimeshoff (2009) find that tougher restrictions on foreign bank entry are likely to enhance bank stability Nevertheless, these studies only capture restrictions as the ratio of entry applications denied to applications received from foreign banks and so do not cover all the entry regulation imposed Their sample is also limited to selected Asian Pacific and European countries

In summary, the theoretical debate and the gaps in literature in relation to different dimensions of regulatory restrictions such as interest rate controls, credit controls and

12

Activity restrictions only capture bank potential to engage in securities insurance and real estate activities Moreover, this dataset was first observed in 1999 which means that it fails to capture the changes in these restrictions and regulations in earlier years

entry restrictions 13 (as summarized in Table 1.1) and bank stability, motivates this thesis’s second research question (RQ2):

RQ2: Is bank stability related to regulatory restrictions on banks?

1.3.3 Government intervention in banking crises

As discussed previously in Section 1.2, most governments use both direct and indirect intervention (known as preventive measures) to promote appropriate behaviour patterns and so enhance bank stability Nevertheless, banking crises may still occur and so require further approaches using remedial measures (Figure 1.1 in Section 1.2) Such interventions are to limit the repercussions of systemic banking crises

Indeed, governments have played an active role in most banking crises (Hoggarth et al., 2002) They usually rescue banks, offering emergency liquidity and other forms

of bail-out programs (Dell’Ariccia, Detragiache, & Rajan, 2008) This is because such crises are extremely costly not only in their economic impact but also in the rescue packages of insolvent banks (Ashcraft, 2005; Boyd et al., 2005) Moreover, costs also arise from misallocated resources revealed by loan losses and budgetary outlays leading to higher taxation and less public sector services (Honohan & Klingebiel, 2003)

The banking crisis literature has mostly concentrated on the cost of crises on economies14 (Bordo, Eichengreen, Klingebiel, Martinez-Peria, & Rose, 2001;

13

Dell’Ariccia et al., 2008; Freixas, Parigi, & Rochet, 2000; Kroszner, Laeven, & Klingebiel, 2007; Serwa, 2010) or on borrowers and bank stakeholders15 (Chava & Purnanandam, 2011; Ongena, Smith, & Michalsen, 2003; Peek & Rosengren, 2000).Some claim that governments lack adequate tools to respond to these situations without imposing a huge burden on public funds in order to avoid even more severe damage to the wider economy (Gwilym, Kanas, & Molyneux, 2013)

Other strands of the literature focus on the impact of different resolution policies on the crisis costs,16 whereas the timing effect of such intervention ex post remains unexplored The effectiveness of government intervention to cope with systemic banking crises might matter not only in terms of the types, but also in their timing The same policies may impact differently depending on when they are enacted (Hoggarth et al., 2005) Hence, the speed of intervention is also important in determining the duration and costs of the crisis

14 A banking crisis can impose severe costs across the economy The cumulative output losses during

a crisis and the fiscal costs to rescue insolvent banks are large, about 15–20% and 50% of GDP respectively (Hoggarth et al., 2002; Honohan & Klingebiel, 2003)

15 Banking crises may directly affect the failed banks’ five main stakeholder groups The first group includes shareholders whose equity value will decline or even disappear The second group, depositors, faces the risk of losing all or part of their savings as well as bearing the cost of any portfolio reallocation The third group consists of other bank creditors who may not be repaid The fourth group is the borrowers, who depend on the bank for funding and might face difficulties replacing them Finally, taxpayers bear the indirect costs of their governments crisis resolution policies (Hoggarth et al., 2002)

16

For an extensive literature review on how crisis resolution policies have been used and their tradeoffs, see Claessens, Dell’Ariccia, Igan, and Laeven (2010), Honohan and Laeven (2005), Dell’Ariccia et al (2008) and Laeven and Valencia (2012b) Some researchers argue that the current policy responses to systemic banking crises are adequate to handle major macroeconomic events (Bartram, Brown, & Hund, 2007; Giannetti & Simonov, 2013; Laeven & Valencia, 2013a) Others, however, claim that government actions and interventions before and during the financial crisis are ineffective or unsuccessful and may actually cause it, make it worse and prolong it (Kane

& Klingebiel, 2004; Klingebiel & Laeven, 2001; Schneider & Tornell, 2004; Taylor, 2009)

Theoretically, more appropriate types and timing of crisis intervention should reduce the costs and duration of banking crises (Hoggarth et al., 2002; Honohan & Klingebiel, 2003) It is conjectured that the authorities are likely to act before a complete loss of confidence occurs (Honohan & Klingebiel, 2003) Moreover, the large and visible costs of a banking crisis require policy makers to respond quickly with appropriate policies to cope with those difficulties Thus, delayed reaction might

be reflected in the crises lasting longer or becoming more severe (Hoggarth et al., 2002)

Dziobek and Pazarbasioglu (1997) and Honohan and Klingebiel (2003) are the most relevant empirical studies that test whether the timing of any intervention is important Dziobek and Pazarbasioglu argue that delays limit the success of resolving banking problems Their result, however, does not show the impact of government intervention on the banking crisis It focuses instead on the relation of the timing of banking restructuring programs to bank problems (non-performing loan ratios and profitability indicators) and only in countries which experience restructuring programs, not banking crises In contrast, Honohan and Klingebiel suggest that delayed intervention is better with lower expected crisis output losses Nevertheless, their findings are based on a comparison between resolution measures

in the second stage and containment measures in the first stage of a banking crisis, instead of the intervention timing

In summary, to the best of the author’s knowledge, while the association between government intervention and the costs of banking crises examined in previous studies remains inconclusive, the impact of the timing of those policies on the banking crisis

duration and costs is unexplored This motivates the following third research question (RQ3):

RQ3: Are the crisis duration and crisis costs related to the timing of government intervention?

The three main objectives of this thesis are to examine the association between: (1) the level of government ownership and bank stability, (2) the regulatory restrictions and bank stability, and (3) the timing of government intervention in systemic banking crises in relation to the crisis duration and crisis costs

As regards government ownership and bank stability (RQ1), the specific research objectives are to determine whether: (1) the level of government ownership is related

to bank stability; (2) the relation between the level of government ownership and bank stability is impacted by political influence, its control of corruption and regulatory governance quality To address these objectives, the following four hypotheses (developed later in Chapter 2) are tested

H1.1: Level of government ownership is negatively associated with bank stability

H1.2: Political influence negatively impacts on the association between government ownership and bank stability

H1.3: Control of corruption positively impacts on the association between level of government ownership and bank stability

H1.4: Regulatory governance quality positively impacts on the association between level of government ownership and bank stability