Handbook of Safeguarding Global Financial Stability: Political, Social, Cultural, and Economic Theories and Models 9780123978752, 0123978750

Citation preview

HANDBOOK OF SAFEGUARDING GLOBAL FINANCIAL STABILITY: POLITICAL, SOCIAL, CULTURAL, AND ECONOMIC THEORIES AND MODELS

Intentionally left as blank

HANDBOOK OF SAFEGUARDING GLOBAL FINANCIAL STABILITY: POLITICAL, SOCIAL, CULTURAL, AND ECONOMIC THEORIES AND MODELS Editor-in-Chief

GERARD CAPRIO Jr. Williams College, Williamstown, MA, USA

Editors

PHILIPPE BACCHETTA University of Lausanne, Lausanne, Switzerland

JAMES R. BARTH Auburn University, Auburn, AL, USA and Milken Institute, Los Angeles, CA, USA

TAKEO HOSHI School of International Relations and Pacific Studies University of California, San Diego, CA, USA

PHILIP R. LANE Trinity College Dublin, Dublin, Ireland

DAVID G. MAYES University of Auckland, Auckland, New Zealand

ATIF R. MIAN Haas School of Business, University of California at Berkeley Berkeley, CA, USA

MICHAEL TAYLOR Adviser to the Governor, Central Bank of Bahrain, Manama, Bahrain

Elsevier 32 Jamestown Road, London NW1 7BY, UK 225 Wyman Street, Waltham, MA 02451, USA 525 B Street, Suite 1800, San Diego, CA 92101-4495, USA First edition 2013 Copyright # 2013 Elsevier Inc. All rights reserved No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights, Department in Oxford, UK: phone (þ44) (0) 1865 843830; fax (þ44) (0) 1865 853333; email: [email protected]. Alternatively, visit the Science and Technology Books website at www.elsevierdirect.com/rights for further information Notice No responsibility is assumed by the publisher for any injury and/or damage to persons, or property as a matter of products liability, negligence or otherwise, or from any use or, operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress ISBN: 978-0-123-97875-2 For information on all Academic Press publications visit our website at elsevierdirect.com Typeset by SPi Global www.spi-global.com Printed and bound in United States of America 12 13

14 15 16

10 9 8 7

6 5 4 3 2

Editorial: Gemma Mattingley Production: Karen East and Kirsty Halterman

1

Editor-in-Chief GERARD CAPRIO Jr. Williams College Williamstown, MA, USA

Associate Editors THORSTEN BECK CentER and European Banking Center Tilburg University, The Netherlands and CEPR London, UK

CHARLES W. CALOMIRIS Columbia University New York, NY, USA

TAKEO HOSHI School of International Relations and Pacific Studies University of California, San Diego, CA, USA

PETER J. MONTIEL Williams College Williamstown, MA, USA

GARRY J. SCHINASI Independent Advisor, Global Financial Stability Washington, DC, USA

Intentionally left as blank

Section Editors for this volume PHILIPPE BACCHETTA University of Lausanne Lausanne, Switzerland

JAMES R. BARTH Auburn University, Auburn, AL, USA and Milken Institute, Los Angeles, CA, USA

TAKEO HOSHI School of International Relations and Pacific Studies University of California San Diego, CA, USA

PHILIP R. LANE Trinity College Dublin Dublin, Ireland

DAVID G. MAYES University of Auckland Auckland, New Zealand

ATIF R. MIAN Haas School of Business University of California at Berkeley Berkeley, CA, USA

MICHAEL TAYLOR Adviser to the Governor Central Bank of Bahrain Manama, Bahrain

Section Editors for related volumes DOUGLAS W. ARNER University of Hong Kong Pokfulam, Hong Kong, SAR, China

CHARLES W. CALOMIRIS Columbia University New York, NY, USA

STIJN CLAESSENS International Monetary Fund, Washington DC, USA, University of Amsterdam, Amsterdam The Netherlands, and CEPR, London, UK

LARRY NEAL University of Illinois at Urbana-Champaign Urbana, IL, USA

SERGIO L. SCHMUKLER World Bank Washington, DC, USA

NICOLAS VERON Bruegel, Brussels, Belgium; and Peterson Institute for International Economics Washington, DC, USA

Contents Extensions to the Sectoral Model 30 Level of Standardization 30 Reliance on Imported Inputs 31 Structure of Firms’ Balance Sheets 31 Partisan Preferences on Exchange-Rate Policymaking Voters 33 Institutions and Exchange-Rate Policy 33 Democracy 34 Elections 34 Electoral System 34 Number of Veto Players 35 Central Bank Independence 35 Conclusion 36 See also 36

Editors-in-Chief v Section Editors vii Preface xvii Contributors xxi

I POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION 1. China and Financial Globalization M.D. CHINN AND H. ITO

Introduction 3 A Brief History of China’s Financial Opening 3 China’s Current Account and Saving Behavior in Cross-Country Context 10 Explanations for China’s High Saving 11 Financial Development and Corporate Finance in China 12 Household Behavior 13 Government Saving 14 Financial Globalization and China’s High Saving 14 Conclusion 15 See also 15 Acknowledgments 15

32

4. Financial Institutions, International and Politics L.L. MARTIN

Introduction 37 Intellectual Background 38 International Financial Institutions: How Much Autonomy? International Financial Institutions: Effects 42 Conclusion 44 See also 45

39

5. Political Economy of Foreign Aid, Bilateral E. WERKER

2. Emerging Markets Politics and Financial Institutions

Political Economy of Aid Disbursement 48 With a Little Help from My Friends 48 Empirically Speaking 49 Does it Matter that Aid Allocation is Political? 50 Political Economy of Aid Receipt 51 Capital to the Capitol 51 Spending Hard-Earned Aid 52 It’s the Economy, Stupid 52 Making Good 53 Taking the Politics Out of Aid Allocation 53 Saving Aid from Itself, or, Taking the Politics Out of Aid Receipt and Disbursement 54 Conclusion 55

A. ABIAD

Introduction 17 Analytical Framework 18 Testing the Abiad–Mody Results on a Wider Sample 19 Differing Influences, Across Types of Countries and Types of Reform 21 Advanced Versus Nonadvanced Economies 21 Domestic Financial Reform Versus Opening up the Capital Account 23 Conclusion 23 Appendix 25

3. The Political Economy of Exchange-Rate Policy

6. Interest Group Politics

D. STEINBERG AND S. WALTER

E.J. MALESKY

Introduction 27 Economic Explanations of Exchange-Rate Policy: Important but Insufficient 28 Preferences: The Demand for Exchange-Rate Policy 28 Sectors 29 Policymakers’ Beliefs and Ideas 30

Political Economy Models of Economic Integration Interest Groups 61 Domestic Capital (Business Owners) 61 Labor 61 Land Owners 62 Domestic Financial Intermediaries 62

ix

60

x

CONTENTS

Distributional Implications of Financial Globalization 62 The Political Economy of Financial Globalization in Authoritarian Regimes 64 MNCs as Actors 65 Conclusion 67 See also 67

7. International Conflicts

11. Microstructure of Currency Markets

Z. KELLY AND E. GARTZKE

M.D.D. EVANS

International Conflict 69 Financial Globalization Promotes Peace 69 Economic freedom 72 Skeptics of Financial Globalization 72 Ambiguous Conclusions for Peace/War 74 The ambiguous consequences of financial upheaval 75 The Effect of War on Financial Integration and Markets 76 Civil War and Domestic Conflict 77 Finance and Peace 77 Scholarship with Implications for the Study of War and Peace 78 Conclusion 78

8. The Political Economy of International Monetary Policy Coordination J.A. FRIEDEN AND J.L. BROZ

Introduction 81 The Potential Gains from International Coordination 82 The Problem: Exchange Rate Externalities 83 Exchange Rate Coordination: Motivation and Modalities 87 Conclusion 89 See also 89

II THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION 9. Theoretical Perspectives, Overview P. BACCHETTA AND P.R. LANE

Introduction 93 Net Capital Flows and the Current Account 93 Gross Capital Flows and the Structure of International Balance Sheets 94 Capital Flows and Crises 94 Exchange Rates as Asset Prices 95 Financial Globalization and the Policy Environment 95 Conclusions 95 See also 95

10. Capital Mobility and Exchange Rate Regimes M.W. KLEIN

Origins and Representation of the Policy Trilemma 98 Dynamics of Exchange Rate Regimes in the Modern Era Capital Mobility in the Modern Era 101

Evidence on the Policy Trilemma 102 Other Economic Effects of the Exchange Rate Regime 103 Exchange Rate Regimes and Inflation Performance 103 Fixed Exchange Rates and Trade 103 Conclusion 104 See also 104

99

Introduction 107 Currency Trading Models 108 Features of the FX Market 108 The Portfolio Shifts Model 108 Overview 108 Equilibrium 109 Empirical Evidence 111 From Micro to Macro 112 A Micro-Based Model 112 Overview 112 Equilibrium 113 Empirical Evidence 114 Micro Perspectives on Exchange Rate Puzzles The Disconnect Puzzle 117 The News Puzzle 118 Conclusion 118

116

12. Intertemporal Approach to the Current Account P.R. BERGIN

Introduction 121 Intertemporal Theories 122 A Simple Theory with Two Periods and No Uncertainty 122 A Basic Stochastic Case with an Infinite Horizon 122 More General Theoretical Cases 124 Empirical Relevance of the Theory and its Implications 125 Present-Value Tests 125 Implications for Current Account Experience 127 Conclusion 128 See also 128

13. Endogenous Portfolios in International Macro Models M.B. DEVEREUX AND A. SUTHERLAND

Introduction 131 A Simple Example Model 133 General Properties of Approximate Solutions 133 Mathematical Foundations 135 Applications 135 Home Bias 135 Valuation Effects and Current Account Imbalances 135 Global Imbalances and Emerging Market Portfolios 136 Monetary Policy and Financial Globalization 136 Conclusion 136 See also 137

xi

CONTENTS

14. Financial Contagion

19. Monetary Policy and Capital Mobility

R. KOLLMANN AND F. MALHERBE

M.M. SPIEGEL

Introduction 139 Bank Balance Sheet Adjustments as a Channel of Contagion: The International Financial Multiplier 140 Financial Contagion Through Interbank Linkages 141 Bank Runs and Self-Fulfilling International Crises 142 See also 142

15. Financial Development and Global Imbalances E.G. MENDOZA AND V. QUADRINI

Introduction 145 Financial Globalization and Financial Underdevelopment: Stylized Facts 145 Explaining Global Imbalances: Financial Globalization with Financial Underdevelopment 147 Mendoza et al. (2009) 148 Quantitative findings, extensions, and robustness 150 Cabellero et al. (2008) 150 Other Contributions with Market Incompleteness 151 Consequences of Global Imbalances 152 Welfare 152 Imbalances as a Source of Crisis 152 Imbalances as a Mechanism for Amplification and Contagion of Crisis 153 Global Imbalances with Cross-Country Heterogeneity in Growth 153 Conclusions 154

16. Foreign Currency Debt M. CHAMON

Introduction 157 Risks of Foreign Currency Debt 157 Reasons for Holding Foreign Currency Debt 158 Development of Local Bond Markets 159 Conclusion 160

17. International Trade and International Capital Flows K. JIN

20. Theory of Sovereign Debt and Default MARK L.J. WRIGHT

Introduction 187 Why Do Countries Repay Their Debts? 187 Sovereign Immunity, Legal Sanctions, and Direct Punishments 187 Restrictions on Financial Market Access 188 Domestic Costs of Default 189 Why Do Countries Borrow So Much? 189 A Benchmark Model 189 Evaluating and Extending the Benchmark Model 190 Policy and Welfare 191 Rollover Risk and Self-Fulfilling Debt Crises 191 Debt Dilution and the Maturity of Sovereign Debts 191 Collective Action Problems in Debt Restructuring 192 Conclusion 192 See also 193

21. Tax Systems and Capital Mobility

Introduction 163 Specialization and Capital Flows 164 The ‘Composition Effect’ 165 Implications 166 Trade and Capital Flows with Frictions 166 Conclusion 167

M.P. DEVEREUX AND C. FUEST

18. International Macro-Finance A. PAVLOVA AND R. RIGOBON

Introduction 169 The Workhorse Model 171 Log-Linear Preferences 172 Log-Linear Preferences with Demand Shocks Next Steps 175 See also 176

Introduction 177 Growth in International Capital Mobility and Monetary Policy 178 Measures of Capital Mobility 178 Global Imbalances 179 Convergence in Yields 180 Monetary Policy Responses 180 Arguments for Improved Monetary Policy Under Increased Capital Mobility 181 Increased International Asset Substitutability 181 Increased Wage and Price Flexibility 182 Reduced Public Pressure for Output Stabilization 182 Theoretical Arguments for Reduced Monetary Policy Quality 183 Reduced Policy Effectiveness 183 Increased Exposure to Global Shocks 183 Empirical Evidence 184 Conclusion 185 See also 186

Introduction: Implications of Globalization for Tax Systems 195 National Taxation and International Mobility 196 Residence-Based Taxation 197 Destination-Based Taxation 197 Source-Based Taxation 198 International Tax Coordination 200 Summary and Conclusions 200

22. Trade Costs and Home Bias 174

N. COEURDACIER

Introduction 201 The Equity and Consumption Home Biases: Facts and Figures

202

xii

CONTENTS

The Equity Home Bias 202 The Consumption Home Bias and the Size of Trade Costs 202 Equity and Consumption Biases: Are they Empirically Related? 202 Why Investors Would Hold Different Equity Portfolios? 203 Home Bias in Equities and the Hedging of Real Exchange Rate Risk 204 From Partial Equilibrium . . . 204 . . . To General Equilibrium 204 Are Equities Empirically a Good Hedge against Real Exchange Rate Fluctuations? 205 Home Bias in Equities and the Hedging of Nontradable Risk 205 The International Diversification Puzzle is Worse than we Think . . . 205 Or Better Than We Think . . . 205 Trade Costs and Portfolio Home Bias: Alternative Stories 206 The Role of Expropriation/Sovereign Risk 206 The Role of Information and Behavioral Biases 206 Conclusion 206 Acknowledgments 207

23. Explaining Deviations from Uncovered Interest Rate Parity

24. Valuation Effects, Capital Flows and International Adjustment C. TILLE

Introduction 231 The 2007–09 Global Crisis: A Synopsis 233 Which Countries Had a Systemic Banking Crisis in 2007–09? 234 Policy Responses in the 2007–09 Crises: What Is New? 235 How Costly Are the 2007–09 Systemic Banking Crises? 245 Concluding Remarks 248 Appendix 250 Acknowledgment 257

27. Advantages and Drawbacks of Bonus Payments in the Financial Sector

D. SCHOENMAKER

217

SAFEGUARDING GLOBAL FINANCIAL STABILITY 25. Safeguarding Global Financial Stability, Overview

227

L. LAEVEN AND F. VALENCIA

28. Central Banks Role in Financial Stability

III

Financial Stability 225 Establishing and Maintaining Financial Stability

26. Resolution of Banking Crises

Introduction 259 Principal–Agent Theory: Why Bonuses may be Beneficial 259 Why Ideal Contracts may well be Unavailable 261 Implications for Contracts in the Financial Sector 261 Assessing the Case for Profit-Related Pay 262 Are Financial Sector Bonuses Actually Deserved? 263 Arguments Against Restricting Financial Sector Bonuses 264 Some Technical Difficulties 265 Some Specific Practical Problems with Bonuses 265 Implementation 267 Concluding Comments 268

Introduction 209 Risk Premium with Representative Investors 210 Limited Participation 210 Deviations from Rational Expectations 211 Conclusion 212

J.R. BARTH, D.G. MAYES, AND M.W. TAYLOR

228

P. SINCLAIR

P. BACCHETTA

Introduction 213 Financial Globalization and Valuation Effects 214 Stock and Flows in External Accounts 214 The Drivers of Net Foreign Assets 214 Yield and Return Differentials 216 International Portfolio Choice and Adjustment in Theory Modeling Financial Integration 217 Portfolio Choice and Capital Flows 218 Valuation Gains and External Adjustment 218 Interpretation of the External Accounts 219 Concluding Remarks 220 See also 221

Crisis Management and Avoidance Global Approaches 229 Other Issues 229 See also 230

Introduction 271 The Broader Monetary and Financial Framework 272 Financial Stability: National or International? 274 Single Country Model of Bailout 275 Multicountry Model of Bailout 275 Financial Stability Framework 276 Assessment 276 Preventive and Remedial Action 278 Private sector solutions 278 Liquidity support measures 278 Public intervention tools 278 Winding down 280 Financial Stability Functions of Central Banks 280 Financial Stability Boards 281 Financial Stability Oversight Council 281 European Systemic Risk Board 282 But Financial Stability Tools are Needed 282 Information Challenge 283 Conclusion 283

xiii

CONTENTS

29. Organization, Supervision and Resolution of Crossborder Banking P. ANGKINAND AND C. WIHLBORG

Introduction 285 The Nordea Case 286 Subsidiary and Branch Organizations in Theory and Practice Organization of Supervision and Crisis Management: Can National Responsibility Be Effective 292 Conclusions: Need for Reform of the Architecture for Supervision and Crisis Management 294

287

30. Dynamic Provisioning to Reduce Procyclicality in Spain ´ NDEZ DE LIS AND A. GARCIA-HERRERO S. FERNA

Introduction 297 The Housing Boom and Bust in Spain 298 The Introduction of Dynamic Provisions in Spain 300 How Was the System Expected to Work? 301 How Did the System Work? 303 Measures Introduced During the Crisis 304 Comparison with Other Countries: Peru and Colombia 304 Conclusions 307 See also 307

31. Varieties of European Crises T.D. WILLETT AND C. WIHLBORG

Introduction 309 A Brief Overview of the Varieties of Financial Crises with Illustrations from Europe 310 Currency Crises 310 Banking Crises 310 Sovereign Debt Crises 311 Different Models of Currency Crises 311 The Crises of the European Monetary System 1992–93 and Nordic Banking Crises 313 Nordic Banking and Currency Crisis 315 Europe in the Global Financial Crisis 2007–09 316 Financial Crisis 2007–09 316 Effects on Central and Eastern Europe 317 Sovereign Debt Crisis 2010 318 The Inadequacy of the Official Responses to the Crisis 319 Concluding Remarks; Lessons from European Crises 320 See also 321

32. The Financial Sector Assessment Program S. MARCUS

Origins of the International Monetary Fund/World Bank Financial Sector Assessment Program 323 The Objectives of the FSAP 324 How is the Program Doing? 325 The Program’s First Decade: Milestones and Country Participation 325 Program Milestones 325 Country Participation in the FSAP 326

Areas of Assessment 327 Indicators of Financial Structure, Soundness and Development (Handbook Chapter 2) 327 Assessing Financial Stability (Handbook Chapter 3) 327 Assessing Financial Structure and Financial Development (Handbook Chapter 4) 327 Evaluating Financial Sector Supervision: Banking, Insurance, and Securities Markets (Handbook Chapter 5) 328 Assessing the Supervision of Other Financial Intermediaries (Handbook Chapter 6) 328 Rural and Microfinance Institutions: Regulatory and Supervisory Issues (Handbook Chapter 7) 328 Assessing the Legal Infrastructure for Financial Systems (Handbook Chapter 9) 329 Standards Assessments in the FSAP 329 The Conduct of an Assessment 330 FSAP Team Composition 331 Interactions with Country Counterparts 331 Stress Testing Approaches 331 Documents of the FSAP and Their Publication 332

33. Financial Sector Forum/Board K. LANGDON AND L. PROMISEL

Establishment of the Financial Stability Forum 333 Initial Work 334 Establishment of Three Working Groups 334 Highly Leveraged Institutions (HLIs) 334 Short-Term Capital Flows 335 Offshore Financial Centers 335 International Supervisory and Regulatory Standards 336 First Decade 336 Working Model 336 Expansion of Membership 337 Main Messages 338 Financial Crisis in 2007 and Aftermath 339 Relationship to the IMF 340 From Financial Stability Forum to Financial Stability Board 341 Expansion to G-20 341 Mandate and Structure of the Financial Stability Board 341 Recent Developments 343 Key Areas of FSF/FSB Interest 343 Early Warning Exercise 343 Evaluation of Role of FSF 347

34. Financial Stability and Inflation Targeting K.N. KUTTNER

Introduction 349 The Separation of Monetary and Financial Policy 349 Does Price Stability Promote Financial Stability? 350 Does Price Stability Guarantee Financial Stability? 350 Does IT Constrain the Response to Financial Crises? 352 New Directions Following the 2007–09 Crisis 353 See also 354

xiv

CONTENTS

35. Financial Supervision in the EU

Basel Capital Requirements as Essential but not Sufficient Regulatory Tool 406 From Basel I to Basel III: What Has Changed? 409 Basel I: The First Step Toward Banking Regulation 409 Basel II and Basel III: Does a Flexible and Sophisticated Approach to Banking Regulation Regulate Banks’ Incentives? 411 What are the Unresolved Flaws of Basel Regulations? 414 Basel Regulation, Incentives, and Role of Pillars 2 and 3 415

D. SCHOENMAKER

Introduction 355 Prudential Supervision 356 Conduct of Business 359 Protecting Retail Customers 360 Market Functioning 361 Supervisory Structures 362 New European Financial Supervisory Framework European Supervisory Authorities 365 European Systemic Risk Board 368 Assessment 368 Conclusions 368

364

40. International Monetary Fund G.G.H. GARCIA

36. Groups: G-5, G-7/8, G-10, G-20, and Others A. BAKER

Introduction 371 A Short History of ‘G’ Group Cooperation 371 Existing ‘G’ Group Scholarship 373 Why Do Governments Participate in ‘G’ Groups? 375 ‘G’ Group Functions 377 Conclusions: The G20 and Political Conflict 378 See also 380

37. Market Structures and Market Abuse K. ALEXANDER

Introduction 381 Regulatory Rationale of Market Abuse Laws 382 Market Developments: Technology and Regulation 382 United Kingdom’s Market Abuse Regulation and Markets Advanced Trading Techniques and Market Abuse 386 Reforming the EU MAD 388 Conclusion 389

384

38. Development and Evolution of International Financial Architecture F. CAPIE

Introduction 393 Metallic Standards 394 What Did Gold Provide? 394 Interwar 395 War and Redesign 396 Bretton Woods 397 Post-Bretton Woods 399 Conclusion 400 See also 400

39. On the Role of the Basel Committee, the Basel Rules, and Banks’ Incentives R. AYADI

Introduction 403 The Evolving Role of the Basel Committee on Banking Supervision 404

Introduction 419 The Fund in June 2011 419 The Legal Foundation 420 The Articles of Agreement 420 Sources of IMF Funds 421 Voting 422 Borrowing 423 Using the Fund’s Resources 424 Financing (loans, grants, and debt relief) 424 The financing process 426 Conditionality reforms 427 Loan subsidies 427 Financing the fund’s own operations 427 Other fund activities 427 Criticisms of the IMF 428 Financing in Crises 428 Moral hazard from bailing out private creditors 429 Bad medicine 429 Global Imbalances and Surveillance 430 The mutual assessment process 431 Freedom of Capital Movements 431 Governance, Legitimacy, and Accountability 431 Not Foreseeing and Forestalling Crises 432 Conclusions and Recommendations 433 See also 434 Acknowledgments 434

41. Innovations in Lender of Last Resort Policy in Europe ´ LEZ-PA ´ RAMO J.M. GONZA

Introduction 435 Underlying Conceptual Issues 436 LoLR in the Euro Area under ‘Normal’ Market Conditions 436 LoLR in the Euro Area during ‘Exceptional Times’ 437 Providing Liquidity to the Market as a Whole 438 Money markets and bank funding liquidity 438 Covered bond markets 438 Sovereign debt markets 439 Assessment 439 ELA to Individual Institutions 439 Concluding Remarks 441 See also 441

xv

CONTENTS

42. Micro and Macro Prudential Regulation A.D. PERSAUD

What Is Macroprudential Regulation? 443 Macroprudential Regulation and the Cycle 444 The Cycle and the Political Economy of Macroprudential Regulation 445 Countercyclical charges and buffers 445 Can the cycle be measured? 446 Valuation and Mark-to-Funding Accounting 446 Macroprudential Regulation Beyond the Cycle 446 Risk Capacity 447 Too Big to Fail 447 Complexity: Financial Instruments and Systemic Risk 448 Incentives and Remuneration 448 Host- and Home-Country Regulation 448 Conclusion 449

43. Role and Scope of Regulation and Supervision D.T. LLEWELLYN

Key Issues 451 Opening Perspective 451 Rationale of Regulation and Supervision 452 Costs of Instability 453 Objectives of Regulation: Strategic Versus Incremental Instruments in a Regulatory Regime 454 Eight Types of Measures 454 Regulatory Strategy: Two Strategic Approaches 455 Reducing the Probability of Failures 455 Structural Regulation 455 Glass–Steagall Approach 456 Narrow Banks 457 Equity Banks 457 Behavioral Regulation 458 Connectedness 458 Intervention 458 Minimizing the Cost of Bank Failures 459 Structural Measures 459 TBTF and the Size Issue 460 Systemically Important Banks 460 Taxation and Insurance 461 Taxation 461 Insurance 461 Resolution Arrangements 462 Living Wills 463 Summary of the Argument 463

453

44. Independence and Accountability of Regulatory Agencies M. TAYLOR

Introduction 465 Independence and Accountability in Theory 465 Lack of Independence Worsens Financial Crises 465 The Example of Central Bank Independence 466 Four Dimensions of Independence 467 Critics of Agency Independence 468 Ensuring Accountability 469

Independence and Accountability in Practice 469 Making Regulatory Independence and Accountability Operational 469 The Importance of the Political Framework 470 The Institutional Structure of Regulation 470 Empirical Work on Independence and Accountability

471

45. Institutional Structures of Regulation M. TAYLOR

Introduction 473 Institutional Structures of Supervision 473 Institutional Regulation 473 Functional Regulation 474 A Mixed System: The United Kingdom Pre-1997 474 Integrated Regulation 475 Twin Peaks 476 The Role of the Central Bank 477 Macroeconomic Policy and Banking Supervision 478 Macroprudential Supervision 479 Crisis Management Arrangements 479 Concentration of Power, Independence, and Accountability 479

46. Organizations of International Co-operation in Standard-Setting and Regulation D.W. ARNER

International Financial Regulation 482 International Financial Standards and Standard-Setting Organizations 482 Policy Direction 483 Coordination 483 Financial Stability Forum/Financial Stability Board 483 Bank for International Settlement 483 Key Standards for Sound Financial Systems 484 Process of Standard Setting 484 International Standard-Setting Organizations 485 IFIs and Other Formal International Organizations 485 International Monetary Fund 485 World Bank 485 Bank for International Settlements 485 Organization for Economic Cooperation and Development 485 International Financial Organizations 485 Regulatory Standard-Setting Organizations 485 Basel Committee 485 International Organization of Securities Commissions 485 International Association of Insurance Supervisors 485 Financial Action Taskforce 486 International Association of Deposit Insurers 486 International Organization of Pensions Supervisors 486 Islamic Financial Services Board 486 OTC Derivatives Regulators’ Forum 486 Central Bank Organizations: CPSS and Committee on the Global Financial System 486 Professional Groups: IASB and International Federation of Accountants 486 Market Associations: International Swaps and Derivatives Association, International Capital Markets Association, Loan Market Association 486

xvi

CONTENTS

Expert Groups 487 Legal Groups 487 Implementation and Monitoring Expert Groups 487 Legal Groups 487

487

47. Prevention of Systemic Crises D.G. MAYES

How the Financial System should be Structured 490 How the Financial System should be Regulated 491 Harmonization 491 Cooperation 493 How Prudence can be Encouraged 494 How Crisis Resolution Tools can Contribute to Prevention See also 496

495

48. Lines of Defense Against Systemic Crises: Resolution ¨ PKES E. HU

International Efforts to Promote Effective Resolution Regimes FSB Policy Measures to Reduce Moral Hazard Risks 500 A New International Standard for Resolution Regimes 500 Building Cross-Border Cooperation 500 Corporate Insolvency 501 National and Regional Initiatives 501 Objectives of Resolution 502 Corporate Failures 502

499

Financial Failures 502 Cross-Border Failures 502 A Special Resolution Regime for Banks and Other Financial Institutions 503 Resolution Regimes as Key Component of the Financial Safety Net 503 Features of a Special Resolution Regime: The ‘FSB Key Attributes’ 504 Scope 504 Resolution Authority 504 Resolution Powers 504 Set-off, Netting, Collateralization, and Segregation of Client Assets 505 Safeguards 505 Funding of Firms in Resolution 505 Legal Framework Conditions for Cross-Border Cooperation 505 Crisis Management Groups (CMGs) 506 Institution-Specific Cross-Border Cooperation Agreements 506 Resolvability Assessments 507 Recovery and Resolution Planning 508 Access to Information and Information Sharing 508 Conclusion 509 Remaining Challenges 509 Promoting Effective Implementation of the New International Resolution Standard 510

Index 511

Preface

Although finance has been a cross-border business for centuries, there are many senses in which the world is becoming more globalized financially. To be sure, early banks carried out transactions to settle imbalances at international trade fairs in the Middle Ages, but the vast swath of society at that time lived untouched by or unconcerned with the financial world outside their village and certainly outside their region. Their world abounded with risks, yet these risks were largely of the type from which their ability to achieve any kind of protection was limited. Indeed, risk was a term that if understood at all would have very different connotations than it does today. Probabilistic thinking was not yet known, and fate or ‘God’s will’ was the more operative expression. Finance, even in basic settings, performs the same functions throughout history as those identified by Levine (1997): mobilize savings, allocate resources, monitor investments, provide payments, and mitigate risk.1 However, both the demand for and the ability of financial systems to provide these services have expanded and evolved in countless directions. Just as money long ago ceased to entail the burden of transporting precious metals (notwithstanding the desire of Congressman Ron Paul, a US presidential candidate as of this writing, to do away with central banks and return to a world of money backed by gold), more recently even international payments are made by electronic transfer. Residents of the world can now travel to other countries carrying only a piece of plastic to make payments, and likely soon will be dispensing with plastic, using a chip built into their cell phones. Many workers in middle-income countries today, and most in higher income economies, have bank accounts or more likely mutual or pension funds with investments outside their home country, though some may not be well aware of their exposure. Just as pensions deal with a risk previously unrecognized – the period after the working stage of life used to be death, and later in time the exceptionally brief interlude of care was provided by families – many risks covered by financial instruments today are relatively recent in being perceived, let alone addressed by finance. 1

The many advances in financial services come with a cost, including the cost of crises. Certainly, crises have been important as long as modern banks have existed – from the failure of banks in northern Italy (including the Ricciardi, the Bardi, and the Peruzzi banks) to the ongoing Euro crisis and the impairment of bank balance sheets that of this writing still is officially minimized. The persistence of crises – which Kindleberger once dubbed ‘A Hardy Perennial’ – might seem puzzling. Why do societies not learn and protect themselves and/ or regulate the financial system better? The answer of course is that finance arises due to information asymmetries, without which there would not only be no crises but also no return for financial intermediaries. Notwithstanding these many constants, the shape of finance has changed markedly in recent decades since the era of extensive domestic and international controls in the aftermath of World War II, when much of the world lived in a period in which the returns on many assets were controlled, instruments not allowed, credit guidance was directly or subtly provided by government, and, in socialist economies, mandated almost entirely by the hand of the state. Now for the first time in history, the residents of virtually all countries can participate in the global financial system, though many, especially in the lowest income countries, remain locally based. But controls have been lifted, capital flows freely across many borders, and financial innovations occur at a rapid pace, even if not all of these innovations contribute to society’s welfare. And efforts to control the financial system understandably advance as the industry itself changes. It is into this situation in the development of global finance that the present effort comes. Three volumes – the Handbook of Key Global Financial Markets, Institutions, and Infrastructure; the Handbook of Safeguarding Global Financial Stability: Political, Social, Cultural, and Economic Theories and Models; and The Evidence and Impact of Financial Globalization – have been put together online and in print to advance our understanding of the origins, requirements, and consequences of financial globalization. The chapters herein share a common overarching goal:

Levine, R., 1997. Financial development and economic growth: views and agenda. Journal of Economic Literature 35 (2), 688–726.

xvii

xviii

PREFACE

to describe the many issues related to financial globalization. The first volume looks at the historical roots of financial globalization, as this is not the first time financial systems have trod this path, and likely will not be the last. Given the technology of today, it then turns to look at the ‘plumbing’ that underlies a healthy financial system, both at the national and global level, including that which supports the ever-changing panoply of new instruments. Financial infrastructure shapes financial systems as surely as any regulator. Thus, some developing countries are at present just seeing the birth of a private bond market, which often is the last part of the system to develop as a result of the demands that it places on the legal system and on information. The second volume examines the political economy of finance. Since its inception, finance has been intimately linked with government. Just as goldsmiths and other early bankers were discovering that not all depositors demanded their money back every day, so that some funds could be profitably deployed, kings were desperately seeking funding for soaring armament expenditures, and a relationship that continues to this day was born. Once the sovereign’s security was assured, other governmental functions needed finance, as did the many projects, useful and not, of the sovereign’s supporters. Bankers’ financial clout quickly translated to political power, and the difficulty of getting financial regulation that works, in addition to being linked to the difficulty of the job, has also been linked to the influence of the industry. Thus, the volume also turns to a discussion of what is meant by financial stability, of attempts to safeguard the stability of the global financial system, and of the many international bodies that are involved in the effort and how they might contribute to this goal. Finally, this volume also looks at how various theories of financial globalization evolved with the developments in the markets. Interestingly, debates on flexible versus fixed exchange rates during the 1960s completely missed the story of what happened once the Bretton Woods system ended.2 As tumultuous as the ‘real world’ has been with more integrated financial markets and flexible exchange rates, one senses that the theoretical literature is evolving significantly as well. Last, but certainly not least, is the final volume that looks at the expanding literature on empirical research regarding the forces behind and the impact of financial globalization, how it has affected policies, and the crises associated with globalized finance, which are transmitted through many channels. The scope of the issues covered in this volume alone testifies to the complexity of the phenomenon.

In this investigation of financial globalization, it is worthwhile to remember that progress has not been linear. It was less than 90 years ago that Keynes could write that The inhabitant of London could order by telephone, sipping his morning tea in bed, the various products of the whole earth, in such quantity as he might see fit, and reasonably expect their early delivery upon his doorstep; he could at the same moment and by the same means adventure his wealth in the natural resources and new enterprises of any quarter of the world, and share, without exertion or even trouble, in their prospective fruits and advantages; or he could decide to couple the security of his fortunes with the good faith of the townspeople of any substantial municipality in any continent that fancy or information might recommend. He could secure forthwith, if he wished it, cheap and comfortable means of transit to any country or climate without passport or other formality, could dispatch his servant to the neighboring office of a bank for such supply of the precious metals as might seem convenient, and could then proceed abroad to foreign quarters, without knowledge of their religion, language, or customs, bearing coined wealth upon his person, and would consider himself greatly aggrieved and much surprised at the least interference. But, most important of all, he regarded this state of affairs as normal, certain, and permanent, except in the direction of further improvement, and any deviation from it as aberrant, scandalous, and avoidable.3 Keynes’ last sentence ranks among the most memorable in financial history, both because of how accurately it described the past and how little it applied to the ensuing decades. World War I of course interrupted the state of affairs that he described, but as we now know was only the first shock to disrupt the system. Caution in forecasting financial globalization therefore seems wise. Following the crisis that began in 2007, with the calls for a Tobin-type tax, the possibility as of early 2012 that one or more members will exit from the Euro, and even the fears of a new Middle East war, few would venture predicting that an immediate further deepening for financial globalization is inevitable. Still, the technology that was so evident in the ‘Arab Spring,’ namely cheap and easy communications, makes it hard to see how the globalization genie can be put back in the bottle. Then again, that is why true ‘shocks’ deserve their appellation! Whatever the immediate outcome, this stocktaking is timely. This project was a labor of love for a great set of professionals who worked tirelessly on this effort: Thorsten Beck, Charles Calomiris, Takeo Hoshi, Peter Montiel,

2

A discussion with Bob Aliber, David Love, Peter Montiel, and Ted Truman was useful in this regard.

3

Keynes, J.M., 1920. The Economic Consequences of the Peace, Harcourt, Brace and Howe, New York, pp 11–12.

PREFACE

and Garry Schinasi as associate editors were instrumental in early decisions on the shape of the effort and on desired content, as well as of great value in finding section editors and authors. The section editors – Thorsten, Charlie, and Takeo taking on this additional burden, along with Douglas Arner, Philippe Bacchetta, James Barth, Stijn Claessens, Philip Lane, David Mayes, Atif Mian, Larry Neal, Sergio Schmukler, Michael Taylor, and Nicholas Veron – were instrumental in finding the best authors for the targeted chapters and along with those wearing the associate editor hat in reviewing the chapters. Of course, the effort would not exist without the labors of the individual authors, who worked to

xix

bring the reader this unparalleled effort. A huge debt of thanks is owed them, both by me as editor and on behalf of all those using this resource in the future. I certainly learned much and was happy to see that so many busy, first-rate professionals were willing to devote the time and effort to this project. Like globalized finance, immense intellectual efforts such as the present one involve much debt! Unlike some financial debt, however, this one is a debt that will keep on paying. Gerard Caprio, Jr. William Brough Professor of Economics, Williams College February, 2012

Intentionally left as blank

Contributors A. Abiad USA

International Monetary Fund, Washington, DC,

Z. Kelly

US Department of Defence, PA, USA

M.W. Klein Tufts University, Medford, MA, USA

K. Alexander University of Zurich, Zurich, Switzerland; University of Cambridge, Cambridge, UK

R. Kollmann ECARES, Universite´ Libre de Bruxelles, Brussels, Belgium; Universite´ Paris-Est, Paris, France; Centre for Economic Policy Research, London, UK

R. Ayadi Centre for European Policy Studies, Brussels, Belgium

K.N. Kuttner Williams College, Williamstown, MA, USA

P. Angkinand

L.

Milken Institute, Santa Monica, CA, USA

D.W. Arner University of Hong Kong, Hong Kong, SAR, China P. Bacchetta University of Lausanne, Lausanne, Switzerland; CEPR, London, UK

Laeven Research Department of the International Monetary Fund, CEPR, Westminster, England, UK

P.R. Lane CEPR, London, UK; Trinity College Dublin, Dublin, Ireland K. Langdon International Monetary Fund, Washington DC, USA

A. Baker Queen’s University Belfast, Belfast, Northern Ireland, UK J.R. Barth Auburn University, Auburn, AL, USA; Milken Institute, Santa Monica, CA, USA; Wharton Financial Institutions Center, Philadelphia, PA, USA

D.T. Llewellyn Loughborough University, Loughborough, Leics, UK; CASS Business School, London, UK; Vienna University of Economics and Business, Vienna, Austria E.J. Malesky Duke University, Durham, NC, USA

P.R. Bergin University of California at Davis, Davis, CA, USA

F. Malherbe

J.L. Broz

University of California, San Diego, CA, USA

F. Capie

City University, London, UK

S. Marcus Financial Sector Assessment Program, Wakefield, RI, USA

London Business School, London, UK

International Monetary Fund, Washington,

L.L. Martin University of Wisconsin-Madison, Madison, WI, USA

M.D. Chinn Robert M. La Follette School of Public Affairs, Department of Economics and NBER, Madison, WI, USA

D.G. Mayes University of Auckland, Auckland, New Zealand

M. Chamon DC, USA

N. Coeurdacier

E.G. Mendoza University of Maryland, College Park, MD, USA; National Bureau of Economic Research (NBER), Cambridge, MA, USA

Sciences Po and CEPR, Paris, France

M.B. Devereux University of British Columbia, Vancouver, BC, Canada

A. Pavlova

M.P. Devereux Oxford University Centre for Business Taxation, Oxford, UK

A.D. Persaud Gresham College, London, UK; London Business School, London, UK

M.D.D. Evans Georgetown University, Washington, DC, USA

L. Promisel

S. Ferna´ndez de Lis

BBVA, Madrid, Spain

J.A. Frieden Harvard University, Cambridge, MA, USA C. Fuest Oxford University Centre for Business Taxation, Oxford, UK A. Garcia-Herrero China G.G.H. Garcia E. Gartzke

BBVA, Hong Kong, Hong Kong, SAR,

E. Hu¨pkes

Economic Consultant, Washington DC, USA

V. Quadrini University of Southern California, Los Angeles, CA, USA; Centre for Economic Policy Research (CEPR), London, UK; National Bureau of Economic Research (NBER), Cambridge, MA, USA R. Rigobon Sloan School of Management, MIT and NBER, Cambridge, MA, USA D. Schoenmaker Duisenberg School of Finance, Amsterdam, The Netherlands

Alexandria, VA, USA

P. Sinclair

University of California at San Diego, CA, USA

J.M. Gonza´lez-Pa´ramo am Main, Germany

London Business School and CEPR, London, UK

University of Birmingham, Birmingham, UK

M.M. Spiegel Federal Reserve Bank of San Francisco, San Francisco, CA, USA

European Central Bank, Frankfurt

Swiss Federal Banking Commission, Switzerland

D. Steinberg University of Oregon, Eugene, OR, USA

H. Ito

Portland State University, Portland, OR, USA

A. Sutherland

K. Jin

London School of Economics, London, UK

M. Taylor

xxi

University of St Andrews, Fife, UK

Manama, Bahrain

xxii

CONTRIBUTORS

M.W. Taylor

Bank of International Settlements

C. Wihlborg Chapman University, Orange, CA, USA

C. Tille Graduate Institute of International and Development Studies and CEPR, Gene`ve, Switzerland

T.D. Willett Claremont Graduate University and Claremont McKenna College, Claremont, CA, USA

F. Valencia Research Department of the International Monetary Fund, Westminster, England, UK

Mark.L.J. Wright Federal Reserve Bank of Chicago, Chicago, IL, USA; University of California, Los Angeles, CA, USA; National Bureau of Economic Research, Cambridge, MA, USA

S. Walter E. Werker

University of Heidelberg, Heidelberg, Germany Harvard Business School, Boston, MA, USA

S E C T I O N

I

POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

Intentionally left as blank

C H A P T E R

1 China and Financial Globalization M.D. Chinn*, H. Ito† *Robert M. La Follette School of Public Affairs, Department of Economics and NBER, Madison, WI, USA † Portland State University, Portland, OR, USA O U T L I N E Introduction

3

A Brief History of China’s Financial Opening

3

China’s Current Account and Saving Behavior in Cross-Country Context Explanations for China’s High Saving Financial Development and Corporate Finance in China

Household Behavior Government Saving Financial Globalization and China’s High Saving Conclusion Acknowledgments Glossary Further Reading

10 11

15 15 15 15

12

INTRODUCTION

national saving in a cross-country context so as to identify how much portion of China’s current accounts and national saving are unexplainable with cross-country variations. Third, we provide descriptive explanations for China’s uniquely high saving rates. Last, we provide some concluding thoughts regarding China’s saving behavior and financial integration with the rest of the world.

China’s impact on the world economy has grown substantially over the past two decades. Attitudes toward the consequences of this development can, at best, be described as ambivalent. Some economists, notably the previous and current chairmen of the Federal Reserve, have argued that China is partially responsible for the crisis; its excess savings – that is, a current account surplus at 11% of gross domestic product (GDP) as of 2007 (Figure 1.1) – fed the profligacy of several industrialized countries, most notably the United States and the United Kingdom. These ‘global imbalances,’ they argue, gave rise to asset bubbles that eventually burst and led to the crisis. Now, the question is how China’s current account surplus balances will evolve, as financial globalization proceeds. How does China’s access to global financial markets interact with its underdeveloped financial markets? Would opening up the Chinese capital account lead to a much trajectory? In order to answer those questions, we review the development of external financial policies and cross-border capital flows of China. Second, we survey empirical findings of the determinants of current account balances and

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00006-4

13 14 14

A BRIEF HISTORY OF CHINA’S FINANCIAL OPENING Since 1978, the Chinese government has very gradually liberalized product markets. Liberalization policies usually start with a limited scope; the policy implementation is often targeted to carefully chosen geographical areas, and narrowly restricted to strictly defined subjects. Only when they yield convincing success does the government expand the scope of coverage and finally make it into a national policy. Financial liberalization has also followed the same pattern. It started in 1980 when the government created the Special Economic Zones (SEZs) in four southern coastal cities and provided foreign firms in the cities (that were allowed to exist only in the form of joint

3

# 2013 Elsevier Inc. All rights reserved.

4

1. CHINA AND FINANCIAL GLOBALIZATION

ventures with local firms) with exemptions from the central planning and other special treatments including exemptions from corporate income tax and other generous tax incentives. Since then, there have been three waves of financial liberalization policies. In 1984, the experiments of the SEZs were expanded to 14 coastal cities, which led to a 98% increase in inward foreign direct investment (FDI). In 1992, when Deng Xiaoping made it clear that the country will pursue market-oriented economy (or ‘socialism with Chinese characteristics’ in his words) during his famous ‘Southern Tour,’ the government implemented further liberalization policies, which led to a surge in inward FDI in 1992 and 1993. The last wave, which is still underway, came when China joined the World Trade Organization (WTO) in 2001. In doing so, China committed to liberalize its financial markets. In this wave, FDI flows continued to be a dominant form of capital flows for the country. Only in the mid-2000s, in response to demands by foreign governments, and also in an attempt to manage an overheating economy, did the government gradually begin liberalizing other types of cross-border capital flows, such as portfolio flows and banking lending. Nevertheless, as of the beginning of 2011, the progress has been quite limited. The last two waves can be observed in Figure 1.2(a), which depicts the evolution of capital inflows to and outflows from China. In 1993, the amount of capital inflows increases dramatically, followed by an increase in capital outflows by a similar magnitude in the late 1990s. Figure 1.2(b)–1.2(d) highlight the fact that the biggest component of the increase in capital inflows was

IR (inc. gold), RHS scale

associated with FDI flows, which have been the main form of capital inflows ever since financial opening in the early 1990s. Both inflows and outflows of ‘other’ type of investment, which is comprised mainly of bank lending, became active after 2005 while portfolio outflows (which includes both equity and debt securities). These developments reflect the authorities’ efforts to cool down the then overheating economy and lessen the appreciation pressure on the exchange rate. The global financial crisis of 2008–2009 caused a significant drop in the outflows of portfolio investment and bank lending, both of which had just experienced a significant expansion in the preceding year. The crisis has more negatively affected FDI and bank lending inflows than portfolio investment inflows. In contrast to the present situation, at the earlier stage of postliberalization development, the primary motive for inviting FDI inflows was to increase accessibility to the then scarce foreign exchange. As of 1980, China held only $10 billion, or 5% of its GDP, of international reserves, a stark contrast to $2.5 trillion, or 49% of GDP, as of the end of 2009. FDI is typically perceived to be the most stable source of external financing compared to the other types of flows. Furthermore, the main motive for the Chinese government to focus on encouraging inward FDI in earlier years was to import corporate governance and other know-how for management and, in later years, banking practices. The relative stability of FDI inflows was much appreciated when other Asian economies with liberalized markets for portfolio investment were more directly exposed to the Asian crisis of 1997–1998. In fact, many agree that China’s tight

Current and capital account (LHS scale)

500

2500

400

2000

300

1500

200

1000

100

500

0

$ Billions

$ Billions

Financial account (LHS scale)

0 1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

−100

−500

FIGURE 1.1 China’s current account, financial account, and international reserves holding. Source: CEIC, World Development Indicators (WDI).

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

5

A BRIEF HISTORY OF CHINA’S FINANCIAL OPENING

250

$ Billions

200

150

100

50

0 1980

(a)

1985

1990

1995

Total inflows

2000

2005

Total outflows

160

140

120

$ Billions

100

80

60

40

20

0 (b)

1980

1985

1990

Foreign direct investment inflow

1995

2000

2005

Foreign direct investment outflow

FIGURE 1.2 (a) China’s capital inflows and outflows. (b) China’s foreign direct investment inflows and outflows. (Continued)

controls over portfolio flows shielded the economy from contagious speculative attacks on other Asian currencies at the time of the crisis. This experience seems to have convinced Chinese policy makers that they need to be careful about removing restrictions on other forms of capital flows than FDI. As we just saw, Chinese authorities started relaxing restrictions first on capital inflows and later on outflows when the Chinese economy started overheating and receiving criticism as a big contributor to the global imbalances only in the mid-to-late 2000s.

Although it has made significant progress toward more open cross-border financial transactions, China still lags behind other major economies including developing ones. While it is extremely difficult to compare the extent of financial openness, or that of capital controls, across countries, there are roughly two ways of measuring it in a cross-country context. One way is to look into the extensity and intensity of regulatory controls on cross-border capital transactions. Such a de jure approach usually uses information from the IMF’s Annual Report on Exchange Arrangements and

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

6

1. CHINA AND FINANCIAL GLOBALIZATION

160 140 120 100

$ Billions

80 60 40 20 0 1980

1985

1990

1995

2000

2005

−20 −40 Portfolio investment inflow

(c)

Portfolio investment outflow

160 140 120

$ Billions

100 80 60 40 20 0 1980

1985

1990

1995

2000

2005

−20 (d)

FIGURE 1.2, CONT’D

Other investment inflow

Other investment outflow

(c) China’s portfolio investment inflows and outflows. (d) China’s ‘other’ investment inflows and outflows.

Source: CEIC, IMF.

Exchange Restrictions (AREAER). The other approach is to construct a de facto measure of financial openness. Here, there are several approaches. One is to examine interest differentials, and another is to examine quantities. Whether we use de jure or de facto measures of financial openness, it is clear that China is a laggard in terms of its openness to cross-border capital transactions. While many emerging market economies removed or loosened regulatory restrictions on capital flows in the 1990s as shown in Figure 1.3, in terms of de jure financial openness, China has not made progress since the early 1990s. It must be noted that de jure measures fail

to fully capture the complexity of real-world capital controls. One de facto measure involves a direct measure of gaps in interest rates. In principle, one would want to examine the two measures: (i) the domestic–foreign interest rates adjusted for expected exchange rate changes (or deviations from uncovered interest parity), and (ii) the domestic–foreign interest rates adjusted for the forward discount (or deviations from covered interest parity). As expected exchange rate changes are not directly observable, the first measure is hard to examine. A recent study finds that the deviations from uncovered interest parity between the United States and China,

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

7

A BRIEF HISTORY OF CHINA’S FINANCIAL OPENING

FIGURE 1.3 De jure financial openness – China, IDC, LDC, and EMG. Reproduced from Chinn, M.D., Ito, H., 2008. A new measure of financial openness. Journal of Comparative Policy Analysis 10(3), 309–322 and updates.

KAOPEN, 1.00 = most open

1

0.8

0.6

0.4

0.2

0 1995

1990

2000 Year

2005

KAOPEN, IDC average KAOPEN, EMG average

2010

KAOPEN, LDC average KAOPEN, China

when the rational expectations are being imposed, declined over the 1996–2001 period. The development of a nondeliverable forward (NDF) market for the Chinese yuan has provided an alternative measure of expected depreciation. Another study which uses this alternative measure, on the other hand, finds no evidence of declining interest differentials in a sample over the 1997–2005 period while allowing for a structural break in 2001, and concludes that capital controls continue to bind. Since onshore rates are higher than offshore, the controls essentially prevent capital from flowing out. Figure 1.4 shows the 1-month covered interest differential (using offshore NDF rates), calculated using

Chibor and Libor. The evidence is, if anything, stronger for binding capital controls, in the post-2005 period, with the exception of a few months right after the depegging of the yuan in July 2005. The late 2008 decrease in the differential is attributable to distortions in Libor associated with the global financial crisis. In Figure 1.5(a), we examine the implications of using a quantity-based measure, namely the components of the international investment position normalized by GDP – Lane and Milesi-Ferretti’s (2007) measure of de facto financial openness that is calculated as the sum of total stocks of external assets and liabilities as a ratio to GDP. It appears that China has been catching up with

0.25

FIGURE 1.4 ‘Covered’ 1-month interest

0.20

differential, annualized. Reproduced from Cheung, Y.-W, Qian, X., 2010. Capital flight: China’s experience. Review of Development Economics 14 (2), 227–247.

0.15 0.10

Annualized one-month NDF ‘covered’ interest differential

0.05 0.00 −0.05 −0.10 1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

8

1. CHINA AND FINANCIAL GLOBALIZATION

(Total assets + total liab.)/GDP

2

1.5

1

.5 1990

1995

2000 Year

LMF index, LDC average LMF index, China

2005

2010

LMF index, EMG average

(a)

(FDI assets + FDI liab.)/GDP

.5

.4

.3

.2

.1

0 1990

1995

2000 Year

FDI, LDC average LMF index FDI, China

2005

2010

FDI, EMG average

(b)

FIGURE 1.5 (a) De facto financial openness – overall. (b) De facto financial openness – FDI. (Continued)

other developing countries since the mid-2000s. Based on Figure 1.5(b)–1.5(d), most of the catch-up is mainly driven by a rapid growth in the stock of portfolio investment (which does not include debt securities in this measure). Interestingly, the markets for debt securities have not shown any progress in terms of increasing openness toward international transactions (Figure 1.5(d)). Although most researchers agree that encouraging mainly FDI inflows has helped the Chinese economy to achieve impressive economic growth, this approach

to financial globalization did not come without cost. First, its asymmetrical approach to financial liberalization toward inflows and outflows of capital has made the country prone to experience surpluses in both current and financial accounts, resulting in a massive buildup of international reserves. Second, FDI inflows have also reinforced the government’s efforts to focus on industrialization through strengthening the manufacturing, capital-intensive sectors. As a result, the economy has had the tendency to experience overcapacity, which contributed to expanding exports and

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

9

A BRIEF HISTORY OF CHINA’S FINANCIAL OPENING

(Port. assets + port. liab.)/GDP

0.2

0.15

0.1

0.05

0 1990

1995

2000 Year

Portfolio flows, LDC average LMF index – Portfolio, China

2005

2010

Portfolio flows, EMG average

(c)

(Debt assets + debt liab.)/GDP

1.5

1

.5

0 1990

1995

2000 Year

Debt flows, LDC average LMF index – Debt, China

2005

2010

Dept flows, EMG average

(d)

FIGURE 1.5, CONT’D (c) De facto financial openness – portfolio investment. (d) De facto financial openness – debt equity investment. Reproduced from Lane, P.R., Milesi-Ferretti, G.M., 2007. The external wealth of nations mark II: revised and extended estimates foreign assets and liabilities, 1970–2004, Journal of International Economics, 73 (2), 223–250 and updates.

exacerbating current account imbalances. Third, the excessive focus on industry has also resulted in excessive capital intensity, driving down the share of national income going to labor. Many researchers have pointed out that labor income has been declining in the last decade, pushing down disposable income. Hence, the distorted industrial structure has raised savings in both the corporate and household sectors. Thus, the unique development of financial liberalization in China has contributed to the rise of the global

imbalances. To more closely examine the impact of financial globalization, we look at how saving and investment behavior has been influenced by these policies. We first investigate saving and investment determination in a cross-country context in the next section to identify common denominators of the saving and investment behavior across the countries. Once we identify the China-specific portion of current account and national saving behavior, we then focus on the peculiarities of China’s saving behavior in the following section.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

10

1. CHINA AND FINANCIAL GLOBALIZATION

CHINA’S CURRENT ACCOUNT AND SAVING BEHAVIOR IN CROSS-COUNTRY CONTEXT Estimating a simple empirical model of current account balances and national saving can be an effective way of identifying the commonalities and peculiarities of China’s saving behavior. Here, we discuss results from an empirical exercise based on several recent empirical studies and conducted for 23 industrial and 86 developing countries over the period of 1970–2008 to estimate the determinants of the current account balances, national saving, and investment. In this exercise, current account balances, national saving, and investment (all expressed as a share of GDP) are individually regressed against the same set of explanatory variables, which are selected based on the literature. The vector of explanatory variables includes budget balances (as a share of GDP); private credit creation (PCGDP) as a measure of financial development; the Chinn–Ito measure of financial openness; a measure of legal/institutional development; net foreign assets as a ratio to GDP; relative income (to the United States); its quadratic term; relative dependency ratios on young and old population; terms of trade volatility; output growth rates; trade openness (¼ exports þ imports/GDP); dummies for oil-exporting countries; and time fixed effects. The ordinary least squares estimation with heteroskedasticity-consistent standard errors is applied to the panels of nonoverlapping 5-year averages of the deviations from their GDP-weighted world means of each of the variables. Most of the variables are found to behave consistently with what has been found in the literature. Among the variables of our interest, the estimation yielded a result consistent with the hypothesis that countries with more developed financial markets should have weaker currents accounts. The estimation also identified significant interactions between capital account openness, financial development, and legal development. More specifically, emerging market economies with better-developed financial markets and open capital accounts are found to have weaker current account balances, as if they are on the receiving end of inflows (or experience the least tendency for capital to flow out). Consistently with the saving glut hypothesis, further financial deepening coupled with higher levels of legal development would worsen current account balances. When the model is estimated for national saving and investment separately, it is found that government budget deficits affect primarily national saving. Given that the Ricardian hypothesis predicts the estimated coefficient of budget balances to be zero, any change in public saving would be offset by the exact same change but with the opposite sign in private saving – this finding

can be interpreted as evidence that there is some nonRicardian effect of deficit spending. It is also found that dependency ratios affect both savings and investment in the way consistent with the lifetime income hypothesis. As the saving glut proponents argue, further financial development would lessen the need for precautionary saving. If a country is equipped with better-developed legal systems, the negative impact of financial development on national saving can be even larger. Financial development has a more consistent impact on investment than saving (something that would not be obvious a priori). However, one must be careful about this sort of exercise especially if it is intended to examine the factors that led to the unique situation of the global imbalances on the eve of the crisis. Because the global crisis can be interpreted as a large-scale correction of the imbalances, some of the saving and investment behavior of countries, which contributed to the global imbalances, can only be interpreted as anomaly. If that is the case, there must be some portions of current account balances or national saving or investment that cannot be explained by crosscountry variations of the explanatory variables. In fact, these regression results suggest the possibility that current accounts may have behaved atypically in the 2006–2008 period, a period with global imbalances prior to the global crisis. Figure 1.6 shows the Kernel density estimates of the distribution of the prediction errors for the groups of industrialized countries and emerging market economies when the predictions are made for the current account balances for the 2006–2008 period using the data up to 2005. Interestingly, for both groups, the distribution of the prediction errors from the regression estimation has become significantly wider in the 2006–2008 period. For the group of industrialized countries, the prediction errors are more skewed to the left and more widely distributed in 2006–2008. While industrialized countries seem to have experienced a wide variation of the prediction errors also in the 1980s and the 1990s besides the last period, the wider variation in the global imbalances period stands out for the group of emerging market countries, suggesting a possibility of a regime shift in the current account balance series in this period. The estimation model performs poorly for China as well. Figure 1.7 displays the implied current account balances for China along with 95% confidence intervals of prediction that are calculated using the estimation results. The figure shows that China’s current account is well outside the confidence interval. The same kind of underperformance of the regression model is also observed for the national saving estimation, a result consistent with other studies. These estimation results can be also used to see if any factors, which are not included in the estimation model and which can be more prevalent in the global

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

11

EXPLANATIONS FOR CHINA’S HIGH SAVING

Emerging market countries

Industrialized countries 20

15

Density

Density

15 10

10

5 5 0

0 −0.1

−0.05

0

0.05

0.1

−0.2

Residuals

−0.1

0

0.1

0.2

Residuals 1970s 1990s 2006–08

1980s 2001–05

FIGURE 1.6 Kernel distributions of prediction errors. Reproduced from Chinn, M.D., Eichengreen, B., Ito, H., 2011. A forensic analysis of global imbalances. Mimeo.

FIGURE 1.7

Predictions of current accounts. Reproduced from Chinn, M.D., Eichengreen, B., Ito, H., 2011. A forensic analysis of global imbalances. Mimeo.

China .1 .08 .06 .04 .02 0 −.02 −.04 −.06 −.08

1971–1975 1976–80

1981–85

1986–90 1991–95 1996–2000 2001–05 Period

Current acoount % of GDP Lower limit

2006–08

Fitted values Upper limit

imbalances period than other period, can explain the unexplained portion of current account balances for the countries. We can test to see if the portion of the current account balances that cannot be explained by the benchmark model can be explained by some variables that account for monetary or fiscal policy stance as well as those which represent the conditions of financial markets and, most importantly, housing markets. While the boom in the financial markets as well as housing markets explain some of the unexplainable portions of the current account balances, it is found that there is still a large portion of current account balances left unexplained for the countries with overly imbalanced current accounts such as the United States, the United Kingdom, Greece, Iceland, and China.

These results indicate that these countries need to implement policies that are particularly tailored for their country-specific situations that affect the saving and investment decisions in order to guide themselves toward rebalancing. In the next section, we review some of the characteristics of China’s policies and socioeconomic conditions that may have contributed to its unique saving and investment imbalances.

EXPLANATIONS FOR CHINA’S HIGH SAVING China’s unique situation has led the country to experience two types of imbalances. The first is the

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

12

1. CHINA AND FINANCIAL GLOBALIZATION

well-known external imbalances. The second imbalance is the multifaceted pattern of China’s economic growth, which is reflected in several gaps. The first pertains to the wide income gap between industrial, high-growth coastal areas and agricultural, underdeveloped inland regions, which was essentially a result of the longtime emphasis on market-driven economic experimentation in the coastal cities. The second pertains to the gap between growth in the returns to capital versus labor. While the corporate sector profits, especially those of the manufacturing sector, have risen continuously throughout the 2000s, labor income has been declining in the same period. Both manufacturing-oriented industrialization and declining labor income have contributed to the third aspect of unbalanced growth, which is the rapid rise in savings, especially those of corporate and household sectors. Figure 1.8 shows that, while the level of national investment of China has been fairly high in recent years, that of national saving has been even higher, the difference between the two accounting for the magnitude of the current account surplus. Hence, understanding the impact of financial globalization on China requires an examination of the growth imbalances that have contributed to China’s unique saving behavior. For that purpose, we need to examine China’s domestic savings from the perspective of the flow of funds. Figure 1.9 displays the development of national savings in three sectors: household, corporate, and gov-

ernment. Since 2001, the level of aggregate national saving has been rising steadily through 2008. While household saving was the main contributor to the aggregate saving before 2000, both household and corporate savings have been the main contributors since then. During the last few years of the sample period, or the global imbalances years, household saving became the largest contributor again. However, it is also noteworthy that during the same period, government saving has been rising rapidly after having played a minor role for a long while. Below, we will only briefly review what kind of economic and socioeconomic factors as well as government policies have contributed to the different paths of development for each of the three sectors’ savings.

Financial Development and Corporate Finance in China As was in the case with other East Asian economies such as South Korea and Japan, China’s rapid industrialization has been achieved through tight state controls on the financial system, which allowed (initially scarce) capital to be allocated to ‘strategically’ important industries. In such financially repressed financial markets, the cost of capital would usually be artificially maintained low. The government, hoping to jump-start economic development with robust export growth, would FIGURE

1.8 China’s national saving and investment. Source: World Development Indicator.

55 50 45 40 35 30 25 20 15 10 1980

1985 China – saving

1990

1995

China – investment

2000 LDC – NS

2005 LDC – Invest.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

13

EXPLANATIONS FOR CHINA’S HIGH SAVING

FIGURE 1.9 Compositions of China’s national saving (as a percentage of GDP). Reproduced from Ma, G., Wang, Y., 2010. China’s high saving rate: myth and reality. BIS Working Paper No. 312 (June); China National Bureau of Statistics; Additional information from Menzie Chinn, and Hiro Ito, Financial Globalization and China (8/3/11).

60

50

40

30

20

10

0 1992

1994

1996 Total

1998 Corporate

2000

2002

Household

2004

2006

2008

Government

encourage cheap capital to be allocated to capitalintensive industries such as heavy and manufacturing industries that would produce tradable goods. While this sort of developmental strategy is typical among emerging market economies, what is unique about China’s case is that: (1) because of its communist past, the state-owned enterprises (SOEs) have played an important role in industrialization and export growth as well as in capital allocation process; (2) because of more direct government involvement in industrial policy and corporate finance (in contrast to more private–government collaborations in the case of Korea and Japan), the government policies have been much less responsive to market forces, resulting in overinvestment in certain industries; and that (3) the lack of responsiveness to market forces also helped the country to lack a scheme that would redistribute the benefits of capital-intensive industrialization to workers in the forms of distribution of dividends. Such a state-dominant financial system may have been effective in capital allocation, but has clearly been an obstacle to the marketization process in the financial sector, making financial development lag behind overall economic development. It is the gap between the impressive economic development and China’s financial underdevelopment that has contributed to a rapid raise in corporate saving. That is, even after many corporations, including both state and nonstate owned, improved profitability in the robust economy in the 2000s, the financial sector continued to be dominated by SOEs and failed to provide attractive financial instruments, to which corporate profits could have been invested. Also, until recently, the government did not create a scheme to force corporations to redistribute dividends to shareholders (i.e., the government in the case

of SOEs). Furthermore, in such an environment, where financial resources are not allocated based on market signals, internal earnings functioned as an important alternative financing source for firms. The inevitable consequence of all these conditions is a rise in corporate saving; due to the lack of financial development, corporate profits are neither effectively reinvested in financial instruments nor redistributed as dividends. For this sort of financial system, one could argue that one effective way to lowering China’s high saving is to implement policies to allow corporate profits to be effectively reinvested or redistributed as dividends. However, that outcome is likely to occur only in the long term.

Household Behavior The peculiarities of China’s economic and financial development have also affected households’ saving behavior. The government’s focus on capital-intensive, tradable industries led to overconcentration of labor force in the manufacturing sector. The situation with labor surplus is worse in the urban areas due to constant migration from the rural areas while the government’s tight controls of labor unions has also discouraged workers’ demand for higher wages. All these factors have contributed to a declining labor income share in the economy. Furthermore, net interest income declined by about a half between 1992 and 2007, so did net transfers from the government, it has been found, mainly because of the increased contributions to pension funds and other welfare obligations.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

14

1. CHINA AND FINANCIAL GLOBALIZATION

While the household income share dropped, the average propensity to save (as a share to GDP) went up by 10 percentage points in the 2000s, resulting in a shrinkage of private consumption and a rise in household saving, both as shares in GDP. These changes in the household saving in China can be attributed to both macroeconomic factors as well as institutional factors. The life cycle, permanent income hypothesis can be a good macroeconomic factor. Since 1980, the working-age share of the population rose from 60 to 74% in China, undoubtedly contributing to increasing the household saving rate. A combination of sluggish change in the consumption behavior and rapid output growth also contributed to a rise in the household saving rate, which is quite common among high-growth developing economies. Furthermore, the restructuring and streamlining efforts as part of the marketization of the corporate sector after the 1990s, along with the large-scale influx of migrants from the rural areas, have made the labor markets highly fluid and led to a drastic shrinkage of the once comprehensive ‘cradle-to-grave’ social safety net, or ‘iron rice bowl.’ Many argue that these trends have motivated Chinese households toward precautionary saving. Limited accessibility to mortgage financing despite increased private house ownership has also been argued to be a factor for the high household saving rate in China. According to a recent empirical study, 82.3% of urban ‘registered city residents’ (or city hukou holders) own houses. This figure has been growing rapidly nationwide. However, due to the lack of financial development as well as risk averseness of the government authorities and financial institutions, mortgage financing has been relatively limited, requiring a high down payment requirement and thus motivating Chinese people to save.

Government Saving As Figure 1.9 illustrates, government saving has been playing a minor role compared to the other two sectors. However, it has been rising rapidly in recently years and becoming a major contributor to the rise in China’s national saving. The rise in government saving is a reflection of a rapid rise in government income, which is also an outcome of rapid economic growth. As it has taken a while for the households to change their consumption behavior to catch up with the rapid economic growth, the same phenomenon has been in place for the government. Now the question is, why has the government consumption level been relatively stable and low, making its saving high, despite a rapid increase in its income? The first reason for the recent rise in government saving is the government’s emphasis on investment

for infrastructure building and other growth-enhancing economic policies. This type of initiatives through active investment is a legacy of the communist style policy implementation. The central government also appropriate a share of fiscal revenue to less well-funded local governments or provide capital transfers to related SOEs to execute national growth-oriented policies. Growthenhancing projects are viewed as important at all government levels because promotions of government officials are often predicated on the performance of the economies under their jurisdiction. Whether it is implemented at the central or local levels, this type of investment is not counted as government consumption, but counted as government saving. Second, the pension system reform implemented in 1997 as a preparation for anticipated aging population has contributed to a rise in government saving. As a result of an increase in pension contributions, the government’s holding of both financial and physical assets has increased in recent years, adding to government saving. Thus, a strong emphasis on growth-oriented investment and preparation for future demographical changes (i.e., aging population) are the main contributors to the recent rise in government saving. However, these types of increase in government saving or investment will also mean that government consumption will have to rise in the future. That means government saving is to fall in the relatively near future, though probably not at the pace the critics of China’s high saving in the rest of the world hope for.

Financial Globalization and China’s High Saving China’s path of development has incorporated a unique approach to financial globalization, associated with a high degree of distortion, manifesting in excessively high levels of savings in both the private and public sectors. In the absence of determined measures to correct the distortions, the extent of both external and internal imbalances may very well become greater. In principle, the development of financial markets could mitigate these distortions. In particular, introducing more market mechanisms could help unclog the flow of funds within the Chinese economy and reduce the accumulated savings in the country. Thus, developing domestic financial markets is a necessary ingredient of China’s further economic development. However, it is easier said than done. As we have observed, the misallocation of funds is also rooted in institutions and systems in place for decades, which ironically contributed to the government’s reluctance to developing financial markets. Hence, drastically changing these institutions and systems will be necessary. Just as China decided to join the WTO to use external

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

CONCLUSION

pressure as leverage to push the economy onto the next stage of market liberalization, the country may need a further push by opening its financial market. The estimation exercise in the last section can be used to conduct simple forecasting exercises to examine what will happen to China’s current account balances in the near future if it develops and/or liberalizes its financial markets. It is found that financial liberalization would be more effective than financial development in reducing China’s current account surplus. Further, while financial development alone may help shrink only marginally the size of its current account surplus, when it is coupled with financial liberalization, it can contribute to reducing current account surplus significantly. This result highlights the potential impact of removing financial sector distortions on external imbalances.

CONCLUSION Our results indicate that China remains in some ways very incompletely integrated with the rest of the global financial system, even as it has an increasingly influential role in the world’s economy. First, de jure measures indicate the presence of substantial capital controls, while de facto measures indicate extant restrictions on capital movements, even as cross-border holdings of Chinese assets increase. Second, China stands out in terms of its saving and investment, and hence current account, behavior. The Chinese current account balance, particularly over the last decade, is anomalous, despite taking into account measurable financial development and level of institutional development. The abnormally high private sector saving can be attributed to the idiosyncrasies of the Chinese financial system. Third, Chinese accumulation of US government debt can then be seen as the outcome of incomplete financial integration of the Chinese economy, rather than financial globalization per se, combined with a quasi-pegged exchange rate set at a level persistently weaker than that determined by private flows alone.

SEE ALSO Theoretical Perspectives on Financial Globalization: Intertemporal Approach to the Current Account; Financial Development and Global Imbalances; International Macro-Finance.

15

Acknowledgments Chinn and Ito acknowledge the financial support of faculty research funds of the University of Wisconsin and Portland State University, respectively. We thank Xingwang Qian for providing data.

Glossary Foreign direct investment (FDI) Cross-border flow involving purchase of equity in excess of a threshold amount, or actual physical investment in plant and equipment. Net foreign asset position Value of holdings of foreign assets minus value of foreign holdings of domestic assets. Nondeliverable forward (NDF) A cash-settled, short-term forward contract on a thinly traded or nonconvertible foreign currency. Portfolio capital flow Capital flow involving purchases of equities and bonds. Sterilization Open market operations undertaken by the central bank to offset the impact of foreign exchange reserve changes on the monetary base.

Further Reading Caballero, R., Farhi, E., Gourinchas, P.O., 2008. An equilibrium model of ‘global imbalances’ and low interest rates. American Economic Review 98 (1), 358–393. Cai, F., Wang, D., 2005. Demographic transition: implications for growth. In: Garnaut, R., Song, L. (Eds.), The China Boom and ItsDiscontents. Asia-Pacific Press, Canberra. Cheung, Y.-W., Chinn, M., Fujii, E., 2005. Perspectives on financial integration in the Chinese economies. International Journal of Finance and Economics 10 (2), 117–132. Cheung, Y.-W., Qian, X., 2010. Capital flight: China’s experience. Review of Development Economics 14 (2), 227–247. Chinn, M.D., Ito, H., 2007. Current account balances, financial development and institutions: assaying the world ‘savings glut’. Journal of International Money and Finance 26 (4), 546–569. Chinn, M.D., Ito, H., 2008. A new measure of financial openness. Journal of Comparative Policy Analysis 10 (3), 309–322. Chinn, M.D., Eichengreen, B., Ito, H., 2011. A forensic analysis of global imbalances. Mimeo. Huang, Y., Wang, X., Lin, N., 2010. Financial reform in China: progresses and challenges. Mimeo, China Macroeconomic Research Center, Peking University, China. Hung, J.H., 2009. China’s approach to capital flows since 1978. In: Cheung, Y.W., Wang, K. (Eds.), China and Asia: Economic and Financial Interactions, Routledge Studies in the Modern World Economy, Routledge, New York. Kuijs, L., 2006. How will China’s saving-investment balance evolve? World Banking Working Paper Series #3958. Lane, P.R., Milesi-Ferretti, G.M., 2007. The external wealth of nations mark II: revised and extended estimates of foreign assets and liabilities, 1970–2004. Journal of International Economics 73 (2), 223–250. Lin, G., Schramm, R.M., 2009. A decade of flow of funds in China (1995– 2006). In: Cheung, Y.W., Wang, K. (Eds.), China and Asia: Economic and Financial Interactions, Routledge Studies in the Modern World Economy, Routledge, New York. Ma, G., Wang, Y., 2010. China’s high saving rate: myth and reality. BIS Working Paper No. 312 (June). Tyers, R., Golley, J., 2010. China’s growth to 2030: the roles of demographic change and financial reform. Review of Development Economics 14, 592–610.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

Intentionally left as blank

C H A P T E R

2 Emerging Markets Politics and Financial Institutions A. Abiad International Monetary Fund, Washington (DC), USA O U T L I N E Introduction

17

Analytical Framework

18

Testing the Abiad–Mody Results on a Wider Sample

19

Differing Influences, Across Types of Countries and Types of Reform

Advanced Versus Nonadvanced Economies Domestic Financial Reform Versus Opening up the Capital Account

21

INTRODUCTION

23

Conclusion

23

Appendix References

25 25

differences exist between advanced and nonadvanced countries in the determinants of reform. Finally, it makes use of the subcomponents of the reform index to analyze differences in the political economy of reform of the domestic financial sector and in opening up the capital account to international inflows and outflows. As in AM, three potential determinants of reforms are analyzed. First, reforms may be triggered by discrete events or ‘shocks’ that change the balance of decisionmaking power. These include crises, the formation of a new government, changes in global interest rates, and leverage exercised by international financial institutions (Krueger, 1993). Second, ‘learning’ may foster reform by revealing information that causes reassessment of the costs and benefits of the policy regime. Learning can also help resolve the impasse on account of uncertainty regarding the identity of winners (Fernandez and Rodrik, 1991). Thus, the term ‘learning’ is shorthand for both discovery and the consequent realignments in relationships. Domestic learning may be supplemented by international ‘diffusion,’ as countries move to global or regional norms to compete for international capital (Simmons and Elkins, 2004). And, finally, reforms may be conditioned by the ideology of the ruling government

Over the past 30 years, there has been a strong push across many countries toward financial liberalization – both in the matter of opening up the domestic financial sector to market forces and removing restrictions on inflows and outflows of international capital (Figure 2.1). But these movements toward greater liberalization also differ considerably across countries and regions in their timing, speed, and magnitude. What accounts for the worldwide advance of financial liberalization and for the differences in the pace of liberalization across countries? To what extent have global and regional forces shaped the dynamics of domestic financial reform and international financial integration? Have these influences differed in advanced and nonadvanced economies? This chapter utilizes the empirical methodology for analyzing the political economy of reforms, developed by Abiad and Mody (2005, AM), to explore these questions. It updates the results of that paper using the database of financial reforms of Abiad et al. (2010), which expands the original AM dataset, both in country coverage, from 36 to 91 countries, and in temporal coverage from 1973–96 to 1973–2005. It then explores whether

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00008-8

21

17

# 2013 Elsevier Inc. All rights reserved.

18

2. EMERGING MARKETS POLITICS AND FINANCIAL INSTITUTIONS

0.90 0.80 0.70

Opening the Capital Account 1.00

Advanced Emerging Asia Latin America MENA Sub-Saharan Africa Transition

Index of External Financial Liberalization

Index of Domestic Financial Sector Reform

Domestic Financial Sector Reform 1.00

0.60 0.50 0.40 0.30 0.20 0.10 0.00 1973

Advanced Emerging Asia Latin America MENA Sub-Saharan Africa Transition

0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00

1978

1983

1988

1993

1998

2003

1973

1978

1983

1988

1993

1998

2003

Source: Abiad, Detragiache, and Tressel (2010).

FIGURE 2.1

Financial reforms worldwide, 1973–2005. Reproduced from Abiad, A., Detragiache, E., Tressel, T., 2010. A new database of financial reforms. IMF Staff Papers 57(2), 281–302.

(Alesina and Roubini, 1992; Cukierman and Tommasi, 1998) and such structural features as openness to trade (Rajan and Zingales, 2003). The paper makes several contributions. First, it confirms that the results found in AM hold in a much wider set of countries than those included in the original estimation. However, it also finds that the factors that drive financial liberalization differ across countries, with differences most pronounced between advanced and nonadvanced economies. For example, the phenomena of regional diffusion, crises affecting reform likelihood, and the role played by International Monetary Fund (IMF) programs and low US interest rates feature mainly in the reform dynamics in emerging and developing economies. Finally, it also identifies significant differences in the determinants of domestic and external financial liberalization. While initial reforms of the domestic financial sector make further reforms more likely, this dynamic is absent in the opening up of the capital account. The negative effect of banking crises and the positive effect of IMF programs likewise are manifested mainly in domestic rather than in external financial reforms. The primary factors that influence the likelihood of greater de jure financial globalization are the regional diffusion effect and the presence of cheap international capital – suggesting that competition for inflows is what spurs the opening up of the capital account. The paper is structured as follows: The section ‘Analytical Framework’ lays out the analytical framework used in the analysis and reviews the various theories of financial reform that are to be tested. The section ‘Testing the Abiad–Mody Results on a Wider Sample’ updates and summarizes the results of AM using the more comprehensive financial reform dataset. The section ‘Differing Influences, Across Types of Countries and Types of Reform’ tests for differences in the factors

affecting financial liberalization across country groups and also examines domestic and external financial liberalization separately. The last section arrives at conclusions.

ANALYTICAL FRAMEWORK This section briefly reviews the political economy perspective used by AM in explaining the timing, pace, and extent of financial sector reforms. The starting point of the analysis is an observed bias toward retaining the status quo, as established interest groups compromise to maintain the existing policy regime (Table 2.1). Policy change is modeled as a function of the difference between the desired level of financial liberalization, FL*it, and the current level of financial liberalization, FLi, t
1, so that
ð2:1Þ DFLit ¼ a FLit
FLi;t
1 þ eit The adjustment factor, a, is a measure of status quo bias: the lower a is, the greater the status quo bias is. Our benchmark model assumes that the desired level of financial liberalization is full liberalization, so that FL*it ¼ 1; we also relax this assumption in subsequent specifications by assuming that FLit* is some constant c 2 (0, 1) or that FL*it varies with the level of development; the main results are robust to these changes. The adjustment factor, a, is likely to be time-varying; we assume that the resistance to reform is a function of the current state of liberalization, so that a ¼ y1FLi, t
1. The presumption is that y1 > 0, so that status quo bias is highest when financial sectors are repressed and the bias declines as the sector is liberalized. Such a dynamic would occur, for example, in models of a multistage version of the Fernandez and Rodrik (1991) model where

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

19

TESTING THE ABIAD–MODY RESULTS ON A WIDER SAMPLE

TABLE 2.1 Distribution of Financial Reform Policies, Full Sample, and by Region Full sample Large reform

Advanced

Emerging Asia

Latin America

Sub-Saharan Africa

Transition

MENA

4.7

4.0

2.7

6.6

3.1

9.5

3.6

Reform

25.2

21.0

27.7

24.8

22.8

40.1

23.2

Status quo

65.2

73.2

63.7

59.2

70.1

45.2

69.6

Reversal

4.4

1.7

5.6

7.7

3.6

5.2

3.6

Large reversal

0.5

0.1

0.3

1.7

0.5

0.0

0.0

100.0

100.0

100.0

100.0

100.0

100.0

100.0

Total

Source: Abiad, A., Detragiache, E., Tressel, T., 2010. A new database of financial reforms. IMF Staff Papers 57(2), 281–302.

earlier reforms help identify winners and losers. It is also consistent with a strengthening of ‘outside’ groups’ positions relative to incumbents and with the need to build technical and managerial expertise in reform implementation. We can thus rewrite Eq. (2.1) as
ð2:2Þ DFLit ¼ y1 FLi;t
1 1
FLi;t
1 þ eit The impact of regional diffusion is also explored. If such an influence were important, countries within a region would be induced to catch up with the highest level of liberalization reached within the region (the regional ‘norm’), either due to a reduction in uncertainty regarding the benefits of reform or due to competition for external capital flows. The larger the gap between the maximum level of liberalization achieved in the region (REGFLi, t
1) and the level of a country’s state of liberalization (FLi, t
1), the higher would be the probability of further liberalization:

DFLit ¼ y1 FLi;t
1 1
FLi;t
1 þ y2 REGFLi;t
1
FLi;t
1 þ eit ð2:3Þ Finally, various shocks can dislodge the status quo, and ideology and structure can influence the speed of reforms. For empirical analyses, a set of variables reflecting these influences (SHOCKSit, IDEOLOGYit, and STRUCTUREit, respectively) is included. This implies the following specification:

DFLit ¼ y1 FLi;t
1 1
FLi;t
1 þ y2 REGFLi;t
1
FLi;t
1 þ SHOCKSit þ IDEOLOGYit þ STRUCTUREit þ eit ð2:4Þ In the category of ‘shocks,’ dummies for balanceof-payments crises (BOPit), banking crises (BANKit), recessions (RECESSIONit), and high inflation periods (HINFLit) are included. For the political variables, the honeymoon hypothesis is examined by including a dummy variable indicating the incumbent executive’s first year in office (FIRSTYEARit). The influence of international financial institutions on policy reform is proxied by an IMF program dummy (IMFit). And to

explore the influence of global factors, international interest rates (USINTit) are included. For the political orientation toward reform, dummy variables for left-wing and right-wing governments (LEFTit RIGHTit) are included; centrist governments are the omitted category. Finally, the structural variable included here is trade openness (OPENit). All of these variables enter the regression contemporaneously, except for banking and balance-of-payments crises dummy variables, which take the value 1 if a crisis occurred within the past 2 years, as these may have prolonged effects.

TESTING THE ABIAD–MODY RESULTS ON A WIDER SAMPLE We begin by updating the benchmark regressions of AM using the new dataset of financial reforms found in Abiad et al. (2010). This new database records financial policy changes along seven different dimensions: six dimensions measuring liberalization of the domestic financial sector (credit controls and reserve requirements, interest rate controls, entry barriers, state ownership, policies on securities markets, and banking regulations), and one dimension measuring the lifting of restrictions on the financial account (often referred to as the capital account) of the balance of payments. Importantly, it expands the coverage of the AM dataset substantially, almost tripling the number of countries covered (from 36 to 91 countries) and adding an additional 10 years’ worth of observations (from 1973–96 to 1973–2005). The expanded coverage is an important test of the generality of the findings in AM, as the new database of financial reforms covers a much wider and more diverse group of nonadvanced countries, including lowincome countries and transition economies. It also now covers the period from 1997 to 2005, when many emerging market crises occurred. Table 2.2 presents the updated versions of AM’s benchmark regressions (specifically, updated versions of the regressions in columns 4–6 of Table 7 and columns

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

20

2. EMERGING MARKETS POLITICS AND FINANCIAL INSTITUTIONS

TABLE 2.2

Ordered Logit Estimates: Overall Financial Reform Index, Full Sample

FLi, t
1  (1
FLi, t
1)

5.972

6.132

5.209

[7.77]***

[7.94]***

[7.11]***

FLi, t
1

(FLi, t
1)

2

5.830

7.906

[6.95]***

[6.55]***


4.840


8.370

[6.65]***

[6.90]***

FLi, t
1  Yi, t
1

0.124 [5.13]***

REG_FLi, t
1
FLi, t
1

2.788

2.500

2.930

3.988

4.748

[6.44]***

[5.54]***

[6.30]***

[5.62]***

[5.98]***

0.374

0.424

0.458

0.421

[2.37]**

[2.53]**

[2.73]***

[2.52]**


0.301


0.447


0.461


0.530

[1.83]*

[2.36]**

[2.42]**

[2.85]***


0.179


0.305


0.305


0.285

[1.29]

[1.99]**

[2.01]**

[1.83]*

0.043

0.082

0.131

0.056

[0.14]

[0.21]

[0.33]

[0.13]

0.105

0.094

0.107

[0.90]

[0.80]

[0.91]

0.561

0.530

0.578

[3.58]***

[3.41]***

[3.66]***


0.068


0.043


0.033

[3.32]***

[1.71]*

[1.27]

0.320

0.293

0.253

[2.15]**

[1.94]*

[1.63]

0.465

0.442

0.450

[2.74]***

[2.60]***

[2.69]***


0.004


0.005


0.006

[2.17]**

[3.09]***

[3.70]***

BOPit

BANKit

RECESSIONit

HINFLit

FIRSTYEARit

IMFit

USINTt

LEFTit

RIGHTit

OPENit

Log L


3429.93


3187.87


2506.12


2503.67


2490.90

Wald test of joint significance (p-value)

0.00

0.00

0.00

0.00

0.00

Observations

2580

2367

1955

1955

1955

Notes: The dependent variable is the change in the Financial Liberalization Index, DFLit. Robust t-statistics are in parentheses, adjusted for clustering by country. Country dummies are included but are not reported. *** denotes significance at the 1% level; ** at the 5% level; and * at the 10% level.

3 and 4 of Table 8). The results confirm all of the findings of AM, namely: • The coefficient on FLi, t
1(1
FLi, t
1) is positive and significant at the 1% level (columns 1–3), confirming the conjecture that status quo bias decreases as

financial liberalization increases, and verifying the inverse U-shaped relationship between policy change and the level of liberalization described in AM. This relationship is consistent with several channels through which initial reforms increase the incentives and pressures for further reforms, including a better

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

DIFFERING INFLUENCES, ACROSS TYPES OF COUNTRIES AND TYPES OF REFORM

•

•

•

•

•

assessment of the value and distribution of reforms and greater voice for ‘outsiders’ who, as incipient insiders, have more say in the policymaking process. We relax the assumption that FLit* ¼ 1 by allowing FLit* to be some constant c 2 (0, 1) (column 4), or allowing FLit* to vary with the level of development (column 5); the results continue to hold. The coefficient on (REGFLi, t
1
FLi, t
1) is also positive and significant at the 1% level, providing evidence that regional influences are an important factor in the spread of financial reform. These findings are consistent with Simmons and Elkins (2004), who also find strong evidence for regional ‘diffusion’ effects and conclude that countries within a region compete for the same international pool of risk capital. The crisis/adversity dummies are added to the regression in columns 2–5. The coefficient on the balance-of-payments crisis dummy variable is positive and statistically significant, suggesting that these crises are an impetus to reform. In contrast, financial liberalization is typically set back following banking crises, as indicated by the negative and significant coefficient on the banking crisis dummy. Thus, greater government control of the financial sector appears to be a common temporary response to banking crises, possibly to prevent a collapse of confidence in the banking sector. These new estimates based on the broader sample also suggest that reform is less likely during periods of negative gross domestic product (GDP) growth, as the coefficient on the recession dummy is negative (as they were in AM) but occasionally significant at either the 5 or 10% level. The coefficient on the IMF program dummy is positive and significant at the 1% level in all specifications, indicating movement toward reforms during IMF programs and providing evidence that global financial institutions have played a role in the global tendency toward greater liberalization. As seen further, the influence of the IMF program dummy results mainly from increased liberalization of the domestic financial system, not from an opening of the capital account. The coefficient on US interest rates is negative and occasionally significant, so that a rise (fall) in US interest rates is seen to slow down (speed up) the pace of financial sector liberalization. This is consistent with Bartolini and Drazen’s hypothesis that when international capital can be accessed cheaply, incentives for reform are strong and the likelihood of liberalizing increases. The coefficients on left-wing and right-wing dummy variables are both positive and mostly significant, suggesting a greater tendency to reform on both sides of the ideological spectrum relative to centrist parties (which are the omitted group). The coefficient

21

magnitudes are also slightly larger for right-wing parties than for left-wing parties, in contrast to what was found in AM. However, chi-square tests of the difference in magnitudes between left-wing and right-wing dummies are insignificant, so that there is no clear evidence that right-wing parties are more reform-oriented than left-wing parties, or vice versa. • Finally, the coefficients on trade openness are negative and significant, a result not found in AM. The coefficient is small, however, and as shown later, this result is not particularly robust. In sum, an update of the baseline regressions of AM using the newer and broader dataset of financial reforms finds that all the key results continue to hold. In the following section, interesting differences in the determinants of reform in advanced and nonadvanced economies, and in the factors that influence domestic and external financial liberalization are seen.

DIFFERING INFLUENCES, ACROSS TYPES OF COUNTRIES AND TYPES OF REFORM Advanced Versus Nonadvanced Economies Do the factors that influence financial reform differ across countries? The regressions in AM and in Table 2.2 assume that these influences work in the same manner and with the same strength across all countries. This issue can be explored further by dividing the sample into advanced countries and nonadvanced countries. The first column of Table 2.3 replicates the fullsample results of our benchmark regression, column 3 of Table 2.2. The next two columns show the estimation results using the same specification but limiting the sample to the 22 advanced economies (column 2) and the 69 nonadvanced economies (column 3) covered by the expanded financial reform dataset. Columns 4 and 5 split the nonadvanced economy subsample further into emerging markets and developing economies. There are several notable results. First, the coefficient on FLi, t
1(1
FLi, t
1) continues to be positive and significant at the 1% level for both the advanced and nonadvanced country subsamples. The result that financial reforms tend to gain momentum – either through learning effects or by strengthening outsiders relative to incumbents – after initial reforms are implemented is something that seems to be true in both types of economies. Regional diffusion, however, is not a significant factor in the advanced country subsample but is positive and highly significant for the nonadvanced economies. The manner in which shocks such as crises influence reforms also differs substantially across the two types of country groups. The fact that balance-of-payments crises

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

22 TABLE 2.3

2. EMERGING MARKETS POLITICS AND FINANCIAL INSTITUTIONS

Ordered Logit Estimates: Overall Financial Reform Index, Advanced Versus Nonadvanced Nonadvanced subsamples Full sample

Advanced

Nonadvanced

Emerging markets

Developing

5.209

7.595

8.565

10.022

6.373

[7.11]***

[4.33]***

[5.88]***

[4.87]***

[3.28]***

2.930

1.282

5.752

6.533

6.037

[6.30]***

[1.41]

[8.29]***

[5.99]***

[7.43]***

0.424

0.071

0.416

0.382

0.249

[2.53]**

[0.19]

[2.19]**

[1.11]

[1.09]


0.447

0.231


0.684


0.469


1.030

[2.36]**

[0.47]

[3.87]***

[1.89]*

[4.12]***


0.305

0.270


0.502


0.246


0.724

[1.99]**

[1.16]

[2.76]***

[0.81]

[3.26]***

0.082


3.921

0.117


0.263

0.303

[0.21]

[10.75]***

[0.28]

[0.32]

[0.86]

0.105

0.081

0.128

0.370


0.119

[0.90]

[0.39]

[0.79]

[2.13]**

[0.42]

0.561


0.585

0.586

0.320

0.849

[3.58]***

[0.91]

[3.80]***

[1.19]

[4.06]***


0.068


0.064


0.044


0.030


0.076

[3.32]***

[1.59]

[1.99]**

[1.02]

[2.14]**

0.320

0.370

0.185

0.368

0.101

[2.15]**

[1.02]

[0.82]

[0.77]

[0.39]

0.465

0.602

0.351

0.867


0.208

[2.74]***

[1.51]

[1.67]*

[3.28]***

[0.79]


0.004


0.002


0.001

0.002


0.004

[2.17]**

[0.61]

[0.32]

[0.39]

[0.63]

Log L


2506.12


710.31


1730.70


818.09


878.73

Wald test of joint significance (p-value)

0.00

0.00

0.00

0.00

0.00

Observations

1955

682

1273

627

646

FLi, t
1  (1
FLi, t
1)

REG_FLi, t
1
FLi, t
1

BOPit

BANKit

RECESSIONit

HINFLit

FIRSTYEARit

IMFit

USINTt

LEFTit

RIGHTit

OPENit

Notes: The dependent variable is the change in the Financial Liberalization Index, DFLit. Robust t-statistics are in parentheses, adjusted for clustering by country. Country dummies are included but are not reported. *** denotes significance at the 1% level; ** at the 5%: and * at the 10% level.

tend to spur financial reforms is a phenomenon distinct to nonadvanced economies. Similarly, the reversal of previous reforms that tends to occur during banking crises is also constrained to emerging and developing economies. The negative effect that recessions have on financial liberalization, documented in the previous section, is evident mainly in the developing economies, which could explain the absence of this result from AM, which only had a limited number of non-emerging market (EM) developing economies. The one crisis that seems to affect the likelihood of reform in advanced

countries (but not in emerging or developing economies) is periods of high inflation, when reforms tend to be reversed. Caution should be exercised in interpreting this last result, however, as there is only one high inflation episode in the advanced country subsample, which lasted 8 years: Israel in 1978–85, a period in which Israel imposed restrictions on both the domestic financial sector and capital account transactions. Among the other variables that influence reform, there is some support for the opportunistic politician theory – where politicians take advantage of the

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

23

CONCLUSION

‘honeymoon period’ during their first year in office to push through needed reforms – but only for the emerging market subsample. In fact, the coefficient on the first year dummy is negative (albeit insignificant) in the developing country subsample, consistent with the evidence in AM, as well as anecdotal evidence cited in Krueger (1993) and Haggard and Webb (1993) that both reforms and reversals become more common during the early stage of the electoral cycle. Perhaps unsurprisingly, the influence of international organizations such as the IMF on the pace of financial reform is most evident in the developing economies, where IMF programs are more common and where conditionality of IMF programs tends to be more extensive. Finally, the influence of US interest rates on financial reform is again a phenomenon limited mainly to nonadvanced economies.

Domestic Financial Reform Versus Opening up the Capital Account Figure 2.1 shows that while both reforms of the domestic financial sector and opening up of the financial account of the balance of payments to international capital increased over the past three decades through most regions of the world, there seems to be differences in the pace of these two aspects of liberalization. For example, while the opening up of the domestic financial sector in Emerging Asia proceeded steadily and gradually from the early 1970s to the mid-2000s, much of the opening up of the capital account in this region occurred in the period from 1984 to 1996; prior to and after that period, there was little change in the extent of external financial liberalization in the region. In many countries in Latin America, the period from 1980 to 1987 was marked by the imposition of controls on capital inflows and outflows; there were fewer moves toward greater repression of the domestic financial sector during this period, in contrast. Just as in Emerging East Asia, the late 1980s and the early 1990s was a period of rapid de jure and de facto financial integration in Latin America, as well as in many other regions. To examine whether different sets of factors influence domestic and external financial liberalization, one can disentangle the domestic and external subcomponents of the financial reform index and use them separately as independent variables. The results are reported in Table 2.4. The first apparent difference in the political economy of domestic and external financial reforms is that while initial reforms of the domestic financial system make further reforms more likely, no such dynamic is evident in the liberalization of the capital account. The two reforms are similar, however, in that both are affected by regional diffusion: the greater the distance between the regional

reform leader and the country, the more likely it is that a country will reform, regardless of whether the reforms are in the domestic or external sector. This suggests that regional diffusion captures not only competition for international capital – which would only affect the opening of the capital account – but also some demonstration effects of the benefits and costs of domestic financial sector reform. The effect of crises is more pronounced for domestic financial sector reform than for increased financial integration. While both types of liberalization become more likely following a currency crisis – a somewhat surprising result for external liberalization, given the common perception that capital controls are increased following such crises – banking crises only affect the pace of domestic financial reform. This latter result is to be expected, as the common response to banking crises is to increase government control over the sector, such as through nationalization of troubled banks, possibly to prevent a collapse in confidence. Similarly, domestic liberalization tends to slow or reverse during recessions, but downturns have no significant effect on external financial liberalization. IMF programs have their effect solely on domestic financial sector reforms – the coefficient on the IMF dummy is positive and significant when domestic financial liberalization is the dependent variable, but the same coefficient is not significant in the capital account liberalization regressions. These results are consistent with other analyses of capital account liberalization, such as Quinn and Toyoda (2007). In their regressions, they find that the IMF program variable coefficient, while positive, never approaches statistical significance. As they note in their paper, “[while] it is widely argued that the International Monetary Fund (IMF) is able, through terms of conditionality in negotiating a program, to impose its policy preferences . . . the Fund rarely to almost never imposed capital account liberalization on nations as part of program conditionality (p. 346).” The lure of cheap international capital, on the other hand, as proxied by world interest rates, is mainly associated with opening up of the capital account. The coefficient on the US interest rate variable is positive and highly significant in the capital account liberalization regressions, although it is also occasionally significant in the domestic financial reform regressions. This is to be expected, as competition for international capital is more likely to spur an opening of the capital account.

CONCLUSION This paper updates the empirical analysis of the political economy of financial reforms reported in Abiad and Mody (2005). It first updates the Abiad–Mody study

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

24 TABLE 2.4

2. EMERGING MARKETS POLITICS AND FINANCIAL INSTITUTIONS

Ordered Logit Estimates: Domestic Versus External Financial Liberalization

FLi, t
1  (1
FLi, t
1)

REG _ FLi, t
1
FLi, t
1

Overall financial reform

Domestic financial sector reform

External financial reform/opening up the capital account

5.209

5.099

1.347

[7.11]***

[7.24]***

[1.18]

5.830

5.668

0.644

[6.95]***

[7.15]***

[0.46]


4.840


4.602


1.326

[6.65]***

[6.29]***

[1.15]

2.930

3.988

3.305

4.384

3.744

3.138

[6.30]***

[5.62]***

[7.28]***

[5.46]***

[11.91]***

[3.87]***

0.424

0.458

0.306

0.340

0.460

0.446

[2.53]**

[2.73]**

[1.93]*

[2.16]**

[2.25]**

[2.14]**


0.447


0.461


0.551


0.564


0.183


0.176

[2.36]**

[2.42]**

[2.49]**

[2.55]**

[0.89]

[0.84]


0.305


0.305


0.292


0.288


0.232


0.239

[1.99]**

[2.01]**

[1.73]*

[1.70]*

[0.95]

[0.97]

0.082

0.131

0.067

0.080


0.388


0.404

[0.21]

[0.33]

[0.18]

[0.21]

[0.72]

[0.73]

0.105

0.094

0.079

0.066

0.202

0.209

[0.90]

[0.80]

[0.66]

[0.56]

[1.03]

[1.05]

0.561

0.530

0.492

0.457

0.362

0.362

[3.58]***

[3.41]***

[3.32]***

[2.97]***

[1.30]

[1.29]


0.068


0.043


0.070


0.041


0.113


0.119

[3.32]***

[1.71]*

[3.26]***

[1.43]

[3.59]***

[3.76]***

0.320

0.293

0.313

0.284

0.492

0.492

[2.15]**

[1.94]*

[1.87]*

[1.66]*

[1.59]

[1.58]

0.465

0.442

0.664

0.638

0.231

0.232

[2.74]***

[2.60]***

[3.81]***

[3.64]***

[0.83]

[0.83]


0.004


0.005


0.002


0.004


0.006


0.006

[2.17]**

[3.09]***

[1.03]

[1.69]*

[1.54]

[1.43]

Log L


2506.12


2503.67


2208.27


2205.83


740.83


740.40

Wald test of joint significance (p-value)

0.00

0.00

0.00

0.00

0.00

0.00

Observations

1955

1955

1955

682

1273

1273

BOPit

BANKit

RECESSIONit

HINFLit

FIRSTYEARit

IMFit

USINTt

LEFTit

RIGHTit

OPENit

Notes: The dependent variable is the change in the Financial Liberalization Index, DFLit, or one of its two subcomponents: an index of domestic financial reform or of external financial reform. Robust t-statistics are in parentheses, adjusted for clustering by country. Country dummies are included but are not reported. *** denotes significance at the 1% level; ** at the 5% level; and * at the 10% level.

using the new dataset of financial reforms compiled by Abiad et al. (2010), which almost triples the number of countries covered, including a much more diverse set of countries including low-income developing countries and transition economies, and lengthens the time period covered to include the late 1990s and early

2000s, when a number of important emerging market crises occurred. The update confirms all of the results found in AM. These results include the following: (i) liberalization becomes more likely once initial reforms are implemented, consistent with domestic learning about the benefits and

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

APPENDIX

costs and the identification of winners and losers from reform, or with the strengthening of ‘outsiders’ relative to incumbents; (ii) shocks such as crises tend to spur action, but different crises had different effects – balanceof-payments crises tend to spur financial reform, while banking crises tend to lead to a slowing or reversal of previous reforms; (iii) regional diffusion effects were important: a country was more likely to liberalize when its state of liberalization was further from the region’s leader; (iv) reforms became more likely in the presence of an IMF program or when US interest rates were low; and (v) there was little evidence that differences in ideology or political structure affected the pace of financial reform. The paper also investigates differences across countries in the factors that influence reform and finds that such differences exist. Specifically, while domestic learning/reform momentum was a common feature across country subsamples, the phenomena of regional diffusion, crises affecting reform likelihood, and the effects of IMF programs and low US interest rates were evident mainly in emerging and developing economies. Finally, a comparison of the determinants of domestic and external financial liberalization also identifies significant differences. While initial reforms of the domestic financial sector make further reforms more likely, no such dynamic is evident for opening up the capital account. Banking crises and IMF programs likewise have the most sway over domestic rather than external financial reforms. The primary factors that influence the likelihood of greater de jure financial globalization are the regional diffusion effect and the presence of cheap international capital – suggesting that competition for inflows is what spurs the opening up of the capital account.

APPENDIX The dependent variables used here are taken from the new database of financial reforms of Abiad et al. (2010) and are described in detail in that paper. Three dependent variables are used: the overall financial reform index, the domestic financial reform component, and the external financial liberalization component. In all cases, the dependent variable was normalized to be between 0 and 1. The independent variables are defined as in Abiad and Mody (2005). Among the crisis variables, the balance-ofpayments crisis variable (BOPit) and the banking crisis variable (BANKit) are based on the crises identified in Bordo et al. (2001), and extended using the currency and banking crisis dates in Laeven and Valencia. Because

25

both types of crisis can be protracted, the dummy variables BOPit and BANKit are set equal to 1 if a balanceof-payments or banking crisis, respectively, has occurred within the last 2 years. The recession dummy variable RECESSIONit is defined as a year where annual real GDP growth is negative, and the high inflation dummy HINFLit is defined as a year in which annual inflation exceeds 50%. Both are based on data from the IMF’s International Financial Statistics database. The political variables are defined as follows: The first year in office dummy FIRSTYEARit is based on the YRSOFFC variable in the World Bank’s Database of Political Institutions. The political orientation variables, LEFTit and RIGHTit, were taken from the same database. The IMF program dummy variable, IMFit, was constructed using the program dates from the History of Lending Arrangements reported by the IMF’s Finance Department and available through the IMF Web site (www.imf.org). The proxy for the world interest rate USINTit is the United States Treasury Bill rate, and trade openness OPENit is calculated as the sum of exports and imports divided by GDP, as reported in the IMF’s International Financial Statistics. Finally, to measure the level of economic development Yit, we use GDP per capita in PPP terms from the Penn World Tables of Heston et al.

References Abiad, A., Detragiache, E., Tressel, T., 2010. A new database of financial reforms. IMF Staff Papers 57 (2), 281–302. Abiad, A., Mody, A., 2005. Financial liberalization: what shakes it? What shapes it? American Economic Review 95 (1), 66–88. Alesina, A., Roubini, N., 1992. Political cycles in OECD economies. Review of Economic Studies 59 (4), 663–688. Bordo, M.D., Eichengreen, B., Klingebiel, D., Peria, M.S.M., 2001. Financial crises: lessons from the last 120 years. Economic Policy 16 (32), 51–82. Cukierman, A., Tommasi, M., 1998. Credibility of policymakers and of economic reforms. In: Sturzenegger, F., Tommasi, M. (Eds.), The Political Economy of Reform. MIT Press: Cambridge, MA, pp. 329–347. Fernandez, R., Rodrik, D., 1991. Resistance to reform: status quo bias in the presence of individual-specific uncertainty. American Economic Review 81 (5), 1146–1155. Haggard, S., Webb, S.B., 1993. What do we know about the political economy of economic policy reform? World Bank Research Observer 8 (2), 143–168. Krueger, A.O., 1993. Political Economy of Policy Reform in Developing Countries. MIT Press: Cambridge, MA. Quinn, D.P., Toyoda, A.M., 2007. Ideology and voter preferences as determinants of financial globalization. American Journal of Political Science 51, 344–363. Rajan, R.G., Zingales, L., 2003. The great reversals: the politics of financial development in the twentieth century. Journal of Financial Economics 69 (1), 5–50. Simmons, B.A., Elkins, Z., 2004. The globalization of liberalization: policy diffusion in the international political economy. American Political Science Review 98 (1), 171–189.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

Intentionally left as blank

C H A P T E R

3 The Political Economy of Exchange-Rate Policy D. Steinberg*, S. Walter† †

*University of Oregon, Eugene, OR, USA University of Heidelberg, Heidelberg, Germany O U T L I N E

Introduction

27

Economic Explanations of Exchange-Rate Policy: Important but Insufficient

28

Preferences: The Demand for Exchange-Rate Policy Sectors Policymakers’ Beliefs and Ideas Extensions to the Sectoral Model Level of Standardization Reliance on Imported Inputs Structure of Firms’ Balance Sheets

28 29 30 30 30 31 31

Partisan Preferences on Exchange-Rate Policymaking Voters

INTRODUCTION

Institutions and Exchange-Rate Policy Democracy Elections Electoral System Number of Veto Players Central Bank Independence

33 34 34 34 35 35

Conclusion Glossary Further Reading

36 36 36

Unstable exchange rates can make it hard for economic agents to plan for the future, stymieing investment. Excessively rigid exchange rates imply a loss of control over the domestic money supply, which can intensify business cycle fluctuations or increase unemployment and inflation. From a political-economy perspective, a number of questions follow. Why do some policymakers fix the value of their currencies while others are more tolerant of exchange-rate fluctuations? Why do some policymakers allow their exchange rates to become misaligned, and why are others more successful at avoiding exchange-rate misalignments? How do interests and institutions shape exchange-rate policies? This chapter critically reviews the recent political economy literature on these questions, and demonstrates that while the understanding of the political economy of exchange rates has improved dramatically in recent years, many open questions remain. After briefly reviewing some prominent economic models of exchange rates, the remainder of the chapter examines the effects of domestic politics on exchange rates. The discussion is divided into different sets of

A country’s exchange rate conveys the price of the country’s currency in another currency. When the exchange rate appreciates, the national currency becomes more expensive. Conversely, when the exchange rate depreciates, the value of the national currency declines. In internationally integrated economies, the exchange rate is the most important price in the economy. Through their exchange-rate policies, governments can determine the price of foreign currency as well as the stability of this price; by doing so, governments influence the size and the stability of their country’s international financial and trade flows. By implementing suitable exchangerate policies, policymakers can foster exports and international competitiveness, and can increase domestic macroeconomic stability. Unfortunately, however, policymakers often implement their exchange-rate policy in a way that inflicts considerable damage on their own economies. Misaligned exchange rates reduce economic growth, increase unemployment, and often result in financial crisis.

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00007-6

32 33

27

# 2013 Elsevier Inc. All rights reserved.

28

3. THE POLITICAL ECONOMY OF EXCHANGE-RATE POLICY

causal variables, beginning with the preferences of various actors – sectors, politicians, political parties, and voters – and then discussing several institutional factors – democracy, elections, electoral systems, veto points, and central bank independence (CBI) – that influence exchange-rate policy. (See Broz and Frieden (2001) for a discussion of international aspects of exchange-rate policy.) Theory and evidence confirm that preferences and institutions both shape exchange-rate policy. However, no single variable always matters, and many variables appear to have different effects in different circumstances. Our review of the existing literature reveals that different studies come to opposite conclusions about the effect of many of the most popular explanatory variables. For example, while many studies find that political factors such as large tradable sectors, right-wing governments, and nondemocratic regimes promote fixed exchange-rate systems, other studies find that these factors are associated with more flexible and volatile exchange-rate regimes. This leads one to conclude that most political factors have contingent effects on exchange-rate policy. Recent research has started to give greater attention to how various political factors interact to jointly determine exchangerate policies, and future research needs to continue exploring these complex causal relationships.

ECONOMIC EXPLANATIONS OF EXCHANGE-RATE POLICY: IMPORTANT BUT INSUFFICIENT Most political explanations of exchange-rate policy build and extend upon theories of exchange rates that were developed by economists. The open-economy trilemma – which has also, more provocatively, been labeled the ‘unholy trinity’ – has been the standard framework for understanding the economic effects of exchange-rate policy since the 1960s. The trilemma states that maintaining a stable exchange rate requires countries to give up either international capital mobility or domestic monetary policy autonomy. This implies that, when capital is mobile internationally, fixing the exchange rate means that interest rates cannot be manipulated in pursuit of domestic economic objectives. (For a detailed discussion of the open-economy trilemma, see the corresponding chapter in this Handbook.) Likewise, the ability to gear monetary policy toward domestic objectives comes at the cost of giving up exchange-rate stability. The theory of optimum currency areas (OCA), a related economic model, argues that the characteristics of the national economy determine which types of exchange-rate policies are optimal. OCA theory suggests that larger, less trade-dependent economies should find the costs of exchange-rate adjustments lower in terms of

aggregate economic efficiency, while valuing monetary policy autonomy more. In contrast, small open economies prioritize fixed exchange-rate regimes because externally oriented economies will fare better with exchange-rate stability than with control over domestic interest rates. These economic models illuminate the costs and benefits of different exchange-rate policies, and provide a necessary starting point for a political analysis of exchange rates. However, economic theories of exchange rates, by themselves, leave much unexplained. The aggregate economic efficiency effects stressed by traditional OCA analyses are often not the major factor influencing policy. Rather, policymakers are often concerned with many considerations beyond aggregate economic efficiency. Their own political fortune is one such concern. After all, policymakers’ own survival is often at stake as a result of exchange-rate-related events: empirical research has demonstrated that finance ministers and prime ministers are significantly more likely to lose office if they devalue the currency. Moreover, exchange-rate policy has strong redistributive effects. Most authors observe that political considerations are particularly relevant in exchange-rate policy because the trade-offs governments face are between macroeconomic outcomes, which different sociopolitical actors value differently. Exchange-rate policy decisions are therefore not purely a question of economic contingencies, but a question of political priorities as well. The rest of this chapter explains how the preferences of various social and political actors and domestic political institutions codetermine exchange-rate policy choices.

PREFERENCES: THE DEMAND FOR EXCHANGE-RATE POLICY Exchange-rate policy has strong distributional consequences. Not surprisingly, opinions on the ‘right’ kind of exchange-rate policy therefore tend to vary among different socioeconomic groups. Some groups favor fixed exchange rates, while others benefit from more flexible exchange-rate regimes. Some benefit from an appreciating currency, while others gain when the exchange rate depreciates. How can the opponents and proponents of certain exchange-rate policies be identified? Also, to what extent do policymakers heed these preferences in the political process? Recent evidence convincingly demonstrates that distributional considerations influence exchange-rate policy. At the same time, qualitative and quantitative evidence both indicate that the effect of preferences and the ability of interest groups, parties, or voters to influence exchange-rate policy in line with these preferences are contingent upon a host of factors.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

29

PREFERENCES: THE DEMAND FOR EXCHANGE-RATE POLICY

Sectors Jeffry Frieden’s seminal 1991 article, ‘Invested Interests,’ on the distributional effects of exchange-rate policy has formed the basis for much of this work. Frieden argues that different sectors of the economy are divided over two aspects of exchange-rate policy: the degree of exchange-rate flexibility/stability and the level of the exchange rate. The open-economy trilemma discussed above implies that fixed exchangerate regimes reduce uncertainty about the value of the currency, whereas flexible exchange rates are beneficial because they enhance monetary policy autonomy (under conditions of capital mobility). Frieden hypothesizes that industries that are involved in international trade and finance, such as exporters and international traders and investors, favor fixed exchange rates because they care more about exchange rate predictability than domestic macroeconomic conditions. They face opposition from import-competing industries and producers of nontradable goods and services (e.g., real estate), who prefer flexible exchange rates because such an exchange-rate policy maximizes domestic monetary policy autonomy. The exchange-rate level also redistributes income across industries. A strong (appreciated/overvalued) currency makes imported goods cheaper, thus increasing actors’ purchasing power – the total amount that can be purchased with a given income. A weaker (depreciated/undervalued) exchange rate, conversely, makes foreign goods more expensive, which means that domestically produced goods are cheaper and more competitive in both home and foreign markets. According to

Frieden, nontradable firms and international traders and investors favor a strong exchange rate because they purchase imports and assets from abroad. Export-oriented and import-competing industries – collectively referred to as tradable industries – prefer a weak exchange rate to enhance their international competitiveness. Several studies have tested Frieden’s predictions that sectors hold differing exchange-rate policy preferences, and that these preferences, in turn, influence which exchange-rate policies are selected. Statistical analyses using time-series-cross-sectional datasets have been the most common methodology for this purpose. The standard setup for these regression models is to use a sector’s share of income as a proxy for its political influence as an independent variable, and dichotomous or categorical measures of exchange-rate regimes as dependent variables. The influence of the manufacturing sector, a tradable industry that is export-oriented in some countries and import-competing in others, on exchange-rate regime choice varies across these studies to a shocking degree (Table 3.1). A larger manufacturing sector significantly reduces the use of fixed exchange rates in three studies, whereas two other studies come to the exact opposite conclusion: larger manufacturing sectors increase the probability of fixing the exchange rate. Yet another study fails to find any significant effect for this sector. These findings are not irreconcilable, however, because it is possible that manufacturers favor flexible exchange rates in some contexts, but have more favorable views of fixed exchange rates in other circumstances. One possibility, consistent with these findings, is that manufacturers only dislike fixed exchange rates in regions or countries with experiences of high inflation,

TABLE 3.1 The Effects of Interest Groups on the Choice of Fixed Exchange-Rate Regimes Study

Manufacturing

Tradables

Sample

Bernhard and Leblang (1999)

x

þ

Industrial democracies, 1974–1885

Blomberg et al. (2005)





Latin America, 1960–99

Broz (2002)

x

þ

All countries, 1973–95

Frieden et al. (2001)



þ

Latin America, 1960–94

Frieden et al. (2010)



þ

Central and East Europe, 1992–2004

Hall (2008)

0

0

Developing countries, 1977–98

Leblang (1999)

x

þ

Developing countries, 1974–1994

Schamis and Way (2003)

x

0

Latin America, 1970–99

Shambaugh (2004)

þ



Developing countries, 1973–2000

Singer (2010)

þ

þ

Developing countries, 1982–2006

Thies and Arce (2009)

x



Latin America, post-1973

Note: þ indicates variable significantly increases probability of fixed exchange-rate regime;  indicates variable significantly decreases probability of fixed exchangerate regime; x indicates variable was not included; 0 indicates variable is not statistically significant in any models.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

30

3. THE POLITICAL ECONOMY OF EXCHANGE-RATE POLICY

because in these contexts greater exchange rate fixity usually implies more overvalued exchange rates. Existing research has also failed to find a consistent relationship between the size of the tradable sector, typically measured as (exports þ imports)/gross domestic product (GDP), and the exchange-rate regime. Consistent with the idea that firms involved in international trade favor more stable exchange rates, six published studies have found that larger tradable sectors increase the use of fixed exchange rates. On the other hand, three articles have found a negative association between trade dependence and fixed exchange rates. There was no statistically significant relationship between these two variables in two other studies. Interestingly, even studies using similar samples and measures come to opposite conclusions. There are strong theoretical grounds for believing that tradable sectors have different exchange rate preferences from nontradable sectors. The existing quantitative evidence confirms that sectors matter. But the evidence also suggests that the nature of particular sectors’ preferences remains elusive. Tradable sectors seem to favor fixed exchange rates sometimes, while opposing them at other times.

translates into exchange-rate policy in a straightforward manner. However, several challenges remain for this perspective. First, ideational theories tend to overstate decision-makers’ ability to independently choose exchange-rate policy, and understate how the complex interplay of societal preferences and state institutions encourage leaders to choose particular exchange-rate policies. Second, the measurement of beliefs and ideas is inherently difficult. Even if these scholars are correct that ‘interests’ are indeterminate, the effect of ideas appears indeterminate and often contradictory. For example, while some scholars have argued that the shift from Keynesian ideas to monetarist/neoliberal ideas accounts for America’s shift to floating exchange rates, other researchers have argued that the same ideational change explains Europe’s increased success maintaining fixed exchange-rate arrangements. Thus, ideas alone do not appear to provide a sufficient explanation for exchange-rate policy. One fruitful avenue for future research should be greater consideration of why similar ideas can produce different outcomes in different contexts. Likewise, ideational theories make the important point that interest group pressures are not always important. But rather than dismiss interest group theories altogether, it would be more useful to consider when ideas or interest groups are more important.

Policymakers’ Beliefs and Ideas In light of these contradictory findings, it is tempting to conclude that material interests are less important than subjective beliefs and understandings. Previous scholarship provides three theoretical grounds for skepticism about the importance of interest groups. The first is that interest group mobilization on exchange-rate policies is limited because the effects of exchange-rate policy are complex and uncertain. Second, even when interest groups can identify their objective interests, exchange rates are public goods that are subject to immense collective action problems. Finally, even if interest groups are able to organize, monetary policymakers are often insulated from the political process, and can therefore resist pressures from lobbyists. Given these supposed difficulties of interest groups to articulate clear policy positions with regard to the exchange rate, some authors have emphasized the importance of the beliefs and ideas of policymakers themselves. These authors argue that prevalent mental models – such as the belief in Keynesian state intervention or in neoliberalism – guide policymakers’ attitudes with regard to exchange-rate policy. Several case-study analyses have demonstrated this impact of ideas on exchange-rate policy choices. Ideational research has enriched our understanding of exchange-rate politics and has been particularly strong in pointing out that no single material factor

Extensions to the Sectoral Model Rather than discard the sectoral theory, some scholars have used it as a foundation to build a more complicated interest group theory. Political economists have built upon and extended Frieden’s sectoral model in several directions to better account for some real-world complications that were not addressed in the original theory. Recent scholarship points to three additional economic factors that determine group’s preferences with respect to exchange-rate stability and valuation: the level of standardization of the product a sector produces; a sector’s reliance on imported inputs into production; and the structure of firms’ balance sheets, particularly the reliance on foreign finance. An additional avenue of research has been to examine trade-offs among multiple policy issues. These extensions have improved the explanatory power of interest group theories. Level of Standardization The first extension distinguishes between ‘simple tradables’ and ‘complex and specialized tradables.’ Some products, such as agricultural commodities and textiles, are relatively standardized and homogeneous, and are not differentiated on the basis of quality. Producers of simple tradables compete on the basis of their price, and, consequently, they are highly concerned with

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

PREFERENCES: THE DEMAND FOR EXCHANGE-RATE POLICY

the level of the exchange rate because a depreciation of the exchange rate can make their products cheaper – and hence more competitive – in domestic and international markets. However, fixed exchange rates provide limited benefits for exporters of standardized products, and they may even oppose fixed exchange rates in order to preserve the ability to depreciate and maintain a favorable value of the exchange rate. By contrast, goods that are specialized so that they can be differentiated on the basis of quality are less price sensitive. Producers of specialized goods gain limited advantage from an undervalued exchange rate, but are very sensitive to currency volatility, which tends to disrupt international trade. Specialized exporters are therefore expected to favor fixed exchange rates – even if this means a less favorable rate of exchange. In support of this proposition, research by Frieden has shown that European countries that exported specialized manufactured goods to Germany and the Benelux countries experienced greater exchange-rate stability and less exchange-rate depreciation vis-a`-vis the German mark. Similarly, an analysis of Germany concluded that German firms with more differentiated products were less concerned with keeping the exchange rate undervalued than were German firms that produced more homogeneous goods. Reliance on Imported Inputs Many tradable firms use foreign products as inputs, and, as a result, exchange-rate depreciation raises their production costs. This suggests that tradable firms may not always favor more depreciated exchange rates; those that heavily rely on imported inputs into production should care little about the level of the exchange rate or they might even prefer a stronger exchange rate. This argument has been supported by several case studies: one article showed that Mexican exporters supported a fixed but overvalued exchange rate during the early 1990s in large part because they were heavily reliant upon imported inputs into production; and a study of Canada revealed that Canadian firms’ heavy reliance on imported inputs lessened their opposition to exchange rate appreciation. Structure of Firms’ Balance Sheets Firms and individuals with mismatched balance sheets – debts denominated in foreign currency but assets denominated in domestic currency – will prefer more appreciated exchange rates because this lowers their debt burden. Such foreign currency denominated debt has become increasingly common as international finance has become liberalized and now constitutes by far the most common form of private credit in some countries (such as Latvia). Not surprisingly, firms with foreign-currency debts were more opposed to the devaluation of fixed exchange rates in countries as diverse as

31

Russia, Argentina, Korea, Thailand, Hong Kong, and Indonesia. An additional set of complications arises from the fact that the exchange-rate regime and exchange-rate level are chosen neither in isolation from one another nor in isolation from other policies. The exchange-rate regime and the currency’s level are related because fixed exchange rates are much more likely to become misaligned than flexible exchange rates, which can easily adjust as market conditions change. As a result, fixed exchange rates tend to be more overvalued than most flexible regimes. While Frieden’s initial theory was ambiguous about how interest groups react when facing a tradeoff across these two dimensions of policies, his subsequent work concludes that exporters are often torn between a concern for currency stability and a concern for a competitiveness-enhancing level of the exchange rate. Qualitative and survey research supports the argument that preferences about the exchange-rate regime are influenced by the level of the exchange rate. Survey data of firms across many developing countries reveal that manufacturing firms dislike unstable exchange rates, and that their opposition to flexible exchange rates is strongest following an appreciation of the real effective exchange rate. Similarly, research on China has found that Chinese manufacturers supported the fixed exchange-rate regime in large part because of its favorable and competitive level. Exchange-rate stability and valuation are also functions of other policies, especially domestic monetary and fiscal policies. The relationship between exchangerate, monetary, and fiscal policy once more goes back to the open-economy trilemma model, which holds that in a world of internationally mobile capital, monetary policy is only effective under flexible exchange rates (but not under fixed exchange rates), whereas fiscal policy is most effective under fixed exchange rates, and much less so under flexible exchange rates. As a result, industrialized countries tend to have more stable exchange rates when their fiscal policies are tight, and this can reinforce some interest groups’ exchange-rate preferences: import-competing and nontradable industries oppose both tight fiscal policies and fixed exchange rates, while capital-intensive internationally oriented firms benefit from both low levels of government spending and stable exchange rates. Similarly, whether the financial sector supports or opposes fixed exchange rates may depend on whether doing so will increase or decrease inflation. Moreover, when fixed exchange rates come under severe speculative pressure, governments must raise interest rates to defend their pegs. Under these circumstances, interest group preferences for maintaining or abandoning a fixed exchange-rate regime will strongly depend on how vulnerable these groups are to tight domestic macroeconomic policy. For example,

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

32

3. THE POLITICAL ECONOMY OF EXCHANGE-RATE POLICY

increasingly restrictive monetary and fiscal policies reduced support for overvalued pegs in various East Asian countries. Furthermore, exchange-rate policy preferences can depend on the country’s trade policy as well, because tariffs and subsidies can serve as substitutes for devaluations: a 10% devaluation has identical effects to a 10% increase in tariffs and export subsidies. Interest group pressures over exchange-rate policy may therefore be dulled by the use of these commercial policies. For example, protectionist trade policies weaken the manufacturing sector’s demands for more flexible/competitive exchange rates. Targeted commercial measures, such as export subsidies and tariff barriers, were integral to building support for fixed and overvalued exchange rates in places as diverse as nineteenth-century America, Colombia in the 1970s, and China during the late 1990s. More indirect linkages between exchange-rate policy and other aspects of foreign policy, such as economic integration, foreign aid, and security policy, have increased support for currency unions in Europe and West Africa. In sum, interest groups are often divided over the exchange rate, but the nature of these divisions is not dictated by any single structural variable. Certain characteristics of firms, such as their export orientation, significantly shape firms’ exchange-rate policy preferences, but whether a given sector supports or opposes a particular exchange-rate policy also depends on the combination of exchange rate, macroeconomic, and commercial policies. As a consequence, preferences over the exchange rate can be dynamic in nature, and interest groups that support a given exchange-rate policy one day may change their positions as other characteristics of the policy environment change. Predicting interest group preferences may therefore require consideration of the package of policies under consideration. Even if a simple interest group explanation of exchange rates has limited explanatory power, interest group theories can still shed much light on exchange-rate politics. Incorporating various characteristics of the firm and of the economic environment has improved our understanding of exchange-rate policymaking. To be sure, many anomalies remain unexplained. The solution is not to abandon interest group approaches, but to continue theorizing about the context-dependent nature of exchangerate preferences.

Partisan Preferences on Exchange-Rate Policymaking Political parties are one mechanism through which the preferences of social groups can be translated into actual policies. The most common argument about the role

of partisanship is that right-wing and conservative parties are more likely to adopt and sustain fixed, stable exchange rates because their constituents value financial stability and low levels of inflation, which fixed exchange rates can help to achieve. Constituents of left parties, such as the working class and domestically oriented firms, in contrast, value domestic monetary autonomy and more expansive fiscal policies, which are inconsistent with fixed exchange-rate regimes under conditions of international capital mobility. In support of this argument, econometric evidence has shown that right parties are associated with more stable exchange rates among Organization for Economic Co-operation and Development (OECD) countries in the post-Bretton Woods system. Similarly, qualitative analysis reveals that conservative parties in France and Italy were more supportive of European monetary integration in the 1980s than left parties, while the general continent-wide shift from left to right governments during the early 1980s accounts for the greater success of exchange-rate stability in that decade compared to the 1970s. An examination of the interwar period produced a similar finding; right parties were more likely to stay on the gold exchange standard than left parties. More indirect evidence also supports the view that left governments are more inclined to forego exchange-rate stability. Several authors show that left-leaning governments experience more frequent speculative attacks on their exchange rates than right-leaning governments, as financial markets seem to expect that these governments are less likely to resist this pressure and to devalue the exchange rate instead, and the crisis probability also significantly rises when there is a shift in the government’s partisan orientation to the left. Moreover, foreign exchange markets become more volatile when they expect left-leaning parties to gain power. Despite this evidence, some studies question the association between conservative parties and fixed exchange rates. Several empirical studies find no effect of partisanship on exchange-rate policy; for example, one quantitative study of 20 industrial democracies in the post-Bretton Woods period found no relationship between partisanship and exchange-rate regime choice. Similarly, neither the incumbent’s partisanship nor partisan change of government have been found not to affect the risk premium on the exchange rate in election periods. Others argue that left governments are in fact more likely to maintain fixed exchange rates. Studies looking at both Western Europe and at former Communist countries have found that exchange rates are significantly more stable under left-wing governments; these authors argue that left governments face greater incentives to use the exchange rate as a short-run stabilization tool and/or have more incentives to signal monetary credibility.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

INSTITUTIONS AND EXCHANGE-RATE POLICY

Similarly, pairs of countries with left governments have more stable bilateral exchange rates than other types of dyads, including pairs with right governments. In addition, several studies have found that left-leaning governments are more likely to defend their exchange rates against speculative attacks. This discussion shows that while most researchers agree that partisanship influences exchange rate policy, the nature of its influence is debated. Nonetheless, the different findings may be reconciled. For example, left governments might face a higher probability of experiencing speculative pressure on their currencies, but defend their currencies more strongly than right governments. Left governments’ ability to stabilize exchange rates may also depend on the partisan orientation of foreign governments. Partisan differences also depend on domestic political institutions; some argue that left parties will favor fixed exchange-rate regime when the central bank is independent as fixing in this case helps them circumvent tight monetary policies and the trade-offs associated with achieving exchangerate stability versus other policy goals. Once more, it appears to be the interaction of partisan preferences with institutions and other policies that shape their overall effect on policy outcomes. These issues warrant further research.

33

Public opinion research also finds that not all voters are of the same mind, and some people oppose fixed and appreciated exchange rates. Several articles find that individuals with higher income and skills have more favorable attitudes toward European monetary integration than others. Support for the Euro was greater in Sweden among business owners and white-collar workers than among blue-collar workers. Education is positively related to Euro support, but educated people care less about an appreciated exchange rate than others. Individuals’ sector of employment also matters. Consistent with the argument that nontradable industries favor flexible exchange rates, one study found that those employed in the sales sector, an archetypal nontradable industry, opposed the Euro in Denmark. Similar to firms, voters’ preferences should also depend on their reliance on imports and their balance sheets, but thus far little research has addressed these issues. Public opinion research about the euro has also found that nonmaterial factors, such as concerns about national identity and sovereignty, influence preferences on exchange-rate policy. Voters’ preferences for exchange-rate policy appear to be influenced by a combination of various individual- and national-level factors. However, research has focused almost exclusively on Europe, and more analysis of other regions is needed.

INSTITUTIONS AND EXCHANGE-RATE POLICY

Voters Democracies ultimately rest on the support of the populace. Voters can influence exchange-rate policy during elections or through referenda. Of course, when citizens go to the polls, exchange-rate policy is, at most, only one of many considerations informing their vote choice. However, exchange-rate policy has probably been decisive in some elections. For example, Argentine voters reelected the incumbent, Carlos Menem, in 1995 because they overwhelmingly supported Menem’s policy of keeping the exchange rate fixed and overvalued. Previous studies indicate that voters often have wellidentified preferences on exchange-rate policy. In general, the average voter appears to favor stronger over weaker exchange rates, and stable to unstable exchange rates. This fact is supported by cross-national studies, which generally find that governments tend to keep their exchange rates fixed and overvalued during pre-election periods (see section ‘Elections’), and studies of public opinion toward the European Monetary Union. Several articles find that individuals favor replacing their national currency with the Euro when the Euro is strong against the dollar, but they prefer to keep their national currency when it has appreciated in value vis-a`-vis the Euro.

Preferences, whether of societal, partisan, or ideational origin, do not directly translate into policy outcomes. Rather, they are mediated by political institutions. Institutions are important because they ‘aggregate preferences,’ meaning that they can determine whether decision makers are more sensitive to the preferences of a specific interest group, voters as a whole, or other actors. For this reason, the same set of preferences can result in very different policy outcomes depending on a country’s institutional structure. Institutions also matter because they can constrain policymakers from implementing the policy that they and their constituents favor. For example, independent central banks and political systems with several veto points can make it difficult for the national leader to implement his/her preferred exchange-rate policy. Political economists have therefore focused a lot of attention on the influence of institutions on economic policy. The conclusion from this research is that institutions, such as the political regime type, elections, and other domestic political structures, can significantly shape the choice of exchange-rate regimes, daily exchange-rate management, as well as the crisis proneness of the currency. However, these effects are rarely uniform across all countries and times. Rather, just like

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

34

3. THE POLITICAL ECONOMY OF EXCHANGE-RATE POLICY

preferences, the effects of individual institutions typically depend on the larger context in which they are embedded.

Democracy The distinction between democracies and autocracies is probably the most fundamental categorization of political systems. Unsurprisingly, considerable attention has been paid to the effects of political regime type on exchange-rate policy. A number of quantitative studies demonstrate that democratic countries are more likely to implement flexible exchange-rate regimes than autocratic countries. One study found that several distinct characteristics of democracies, such as electoral competition and the presence of multiple veto points, are each associated with flexible exchange-rate regimes. Historical analyses similarly attribute the collapse of the Gold Standard in the 1920s, at least in part, to the enfranchisement of large segments of the population. Two arguments have been put forth to explain this effect of democracy on exchange-rate flexibility. First, exchangerate flexibility allows policymakers to autonomously conduct monetary policy in order to improve domestic economic conditions, a policy option which offers high political rewards when politicians need to maintain voters’ approval to remain in office. A second argument maintains that monetary commitment transparency and the transparency of the political system are substitutes. According to this theory, because autocracies have less transparent political systems than democracies, autocracies have a stronger need for a transparent monetary commitment device such as a fixed exchange-rate regime. However, some evidence appears to contradict the finding that democratic governance encourages flexible exchange-rate regimes. For example, ‘fear of floating’, that is, countries’ propensity to officially announce a flexible exchange-rate regime while in fact intervening so as to prevent true floating, is more pronounced in democratic countries, while ‘fear of pegging’ is less prevalent. More directly, an analysis of exchange-rate regime choices in 21 Eastern European countries found that democracies were associated with a higher likelihood of fixing their exchange rates. Other research suggests that the effect of democracy upon foreign exchange markets is context dependent. For example, the effect of the political regime type on de facto exchange-rate policy has been found to be conditional on the declared, or de jure, exchange-rate regime; autocracies are more likely to defend their exchange rates against speculative pressure when they have officially fixed exchange rates, but democracies are more likely to defend their exchange rates under intermediate than fixed regimes. Democracies therefore appear to

decrease the use of fixed exchange rates in many circumstances, but not in others. As a group, democracies behave differently from autocracies, but within the group of democracies heterogeneous behavior can be observed as well. Therefore, now, attention is turned to the question of how variations in democratic institutions such as the electoral cycle, electoral system, and CBI explain this variance in exchange-rate policy choices.

Elections Elections – the key constitutive feature of democracy – also matter for exchange-rate policy. Extensive evidence has documented that exchange rates are influenced by the electoral cycle. For example, exchange-rate-based stabilization programs are typically implemented when elections are pending. Similarly, there is a lot of evidence that devaluations tend to be delayed until after an election in order to preserve voters’ purchasing power on the day of election. The strength of this electoral exchange-rate cycle is conditioned by several factors. For example, when speculative pressure is very severe in the preelectoral period, devaluations are rarely delayed before elections, and devaluations are particularly likely after the subset of elections that have resulted in a transfer of executive power. Furthermore, research examining industrial countries operating freely floating exchange rates points in the opposite direction; exchange rates become more volatile in the run-up to elections as traders become more uncertain about future economic policies. Whether elections increase or decrease, exchange-rate stability therefore depends on the economic and political pressures facing the politicians.

Electoral System The electoral system is one of the main sources of institutional diversity among democratic countries. The most common distinction is between proportional representation systems and plurality–majoritarian regimes. Under proportional representation, a party’s vote share determines their share of electoral seats – an arrangement that tends to produce multiparty coalition governments. By contrast, in plurality systems the party with the most votes typically controls policy alone. It is not surprising that the electoral system affects incentives for choosing certain types of exchange-rate policies, even though the findings have been somewhat contradictory. Some research finds that democratic politics with majoritarian electoral systems are more likely to choose fixed exchange rates and experience less exchange-rate volatility after cabinet dissolutions than those with proportional systems. In contrast, other authors find that industrial

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

INSTITUTIONS AND EXCHANGE-RATE POLICY

countries with majoritarian systems experience higher exchange-rate volatility than countries with proportional electoral systems, whereas another study concludes that electoral systems do not affect currency markets in emerging markets. The range of estimates of whether and how the electoral system affects exchange rates once more suggests the need to consider how this institutional feature interacts with other aspects of the political system. The existing literature suggests that the level of opposition influence is one such conditioning factor. For instance, one article observed that countries with both majoritarian systems and low opposition influence are least likely to fix, whereas proportional representation (PR) systems in which the opposition exerts a lot of influence are most inclined to adopt some type of fixed exchange rate. The electoral system thus influences some countries’ exchange-rate policy, but the direction and strength of influence depend on other political-economic factors.

Number of Veto Players Several authors have pointed out that the number of veto players – actors whose consent is required to change policy – also shapes exchange-rate policy choices. Some research shows that developing countries with fewer veto players have a tendency toward choosing pegged exchange-rate regimes. One explanation for this finding is that in countries with few veto players, and hence strong political accountability, policymakers value their ability to influence domestic conditions through an autonomous monetary policy, whereas fixed exchange rates serve as a focal point for policymaking that can reduce conflicts about macroeconomic policy decisions in countries with many veto players. Another explanation is that divided governments find it more difficult to implement painful but necessary internal adjustment policies and therefore are more likely to devalue when confronted with speculative pressure than unified governments. As a consequence, divided governments also face a higher risk of speculative attacks on their currencies. Others argue that veto players and exchange rate outcomes have a nonlinear relationship; countries with very few and very many veto points are prone to suffer currency crises, whereas countries with intermediate numbers of veto players are least crisis prone. Once more, most research indicates that the effect of veto players on exchange-rate policy is context dependent. For example, increasing the number of parties in the governing coalition reduces the probability of fixing the exchange rate for developing countries, but the opposite occurs when using a sample that includes both industrial and developing countries. Along these lines, another study found that countries with multiple

35

partisan veto players commonly fix their exchange rates in unitary systems, but not in federal systems. This suggests that the number of veto players impacts the desire and ability of policymakers to maintain fixed exchange rates, but does so differently for various types of countries. Moreover, the number of veto players also conditions the ability of governments to operate their exchange-rate regime. As discussed above, more veto players tend to increase the risk that governments fail at this task and experience a currency crisis, but this risk is highest when they have adopted an intermediate exchange-rate regime. Paying attention to the institutional and country-specific context in which veto players operate has thus resulted in a more detailed understanding of how veto players affect exchange-rate policymaking.

Central Bank Independence CBI is very closely linked to exchange-rate politics because exchange rate and monetary policy are themselves tightly linked, and because independent central bankers are more sheltered from popular opposition to tight monetary policy or uncompetitive exchange rates. CBI and fixed exchange rates both can decrease inflationary bias and are therefore two possible solutions to the same time-inconsistency problem of monetary policy. The two institutions are therefore often seen as substitutes, even though more recent research indicates that they can act as complements when both are not fully credible. This latter argument is consistent with research that shows that CBI increases exchange-rate stability. For example, research has shown that European countries with independent central banks had greater currency stability visa`-vis the German mark than those with lower levels of CBI and that countries with independent central banks tend to have more appreciated exchange rates. Some authors also find that CBI lowers the risk of experiencing a speculative attack, even though other authors find that CBI is not statistically significantly related to the risk of currency crisis. The relationship between CBI and the exchange-rate regime is conditioned by several other factors. One study finds that CBI encourages the adoption of fixed exchange rates, but only for leftist governments. Another suggests that independent central bankers advocate fixed exchange rates when they are weak domestically and need an external anchor to tie the governments’ hands, an argument that echoes the argument of the complementarity of CBI and fixed exchange-rate regimes. On the other hand, one article argues that CBI often makes it harder to stabilize the exchange rate because independent central bankers are reluctant to cut interest rates to do so, and shows that CBI reduces exchange-rate stability in OECD countries with de jure fixed exchange-rate regimes.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

36

3. THE POLITICAL ECONOMY OF EXCHANGE-RATE POLICY

According to one comparative analysis, in countries where the private sector holds a strong preference for weak exchange rates, CBI reduces the private sector’s ability to convince policymakers to undervalue the exchange rate, but CBI does not affect the level of the exchange rate when private sector preferences are weak. In sum, countries with independent central banks select different types of exchange-rate policies than those with politically dependent central banks, but the direction and size of this difference appear to depend on other political factors, such as government preferences.

CONCLUSION Like most other policy issues, the choice of exchangerate policy is a politically driven one. Considerations of aggregate welfare are not irrelevant, but they typically do not fully determine exchange-rate policy either. To make sense of exchange-rate policy choices therefore requires an understanding of political incentives. This chapter has shown that a variety of political factors – the preferences of sectors, parties, voters, and the nature of political institutions – shape exchange-rate policy. Scholarly understanding of how such preferences and institutions shape exchange-rate politics has dramatically improved in recent years. But many puzzles remain. Why, for example, do tradable industries favor fixed exchange rates in some conditions but not others? Or, why does democracy encourage the adoption of flexible exchange-rate regimes in most times and places while promoting fixed exchange rates in some situations? It would be wrong to conclude that interests and institutions do not matter. The task at hand is to develop better theories of why the same political variables have different effects on exchange-rate policy across different contexts. Therefore, it is argued that the effects of preferences and institutions on exchange-rate politics are conditional upon one another. This message may not be surprising, but it is often under-appreciated. There is not one major political variable that always has a consistently strong effect upon exchange-rate policy, and many variables have opposite effects in different circumstances.

SEE ALSO

Glossary Central bank independence The freedom of monetary policymakers from governmental interference. Central banks are considered independent when they are able to define their own policy objectives, and can implement policy without requiring the government’s approval. Democracy A type of political system in which government officials are selected through free and fair elections. Economic sector Economies are divided into various distinct sectors, or industries. Agriculture, manufacturing, and services are three major sectors in the economy. Exchange rate The price of foreign currency. An exchange rate is defined as the amount of domestic currency that is required to purchase one unit of foreign currency. Exchange-rate regime The system, or set of rules, used to determine the currency’s exchange rate. Exchange-rate regimes vary from fixed exchange-rate regimes, where the government keeps the currency’s foreign exchange value stable, to floating exchange-rate regimes, in which the currency’s external value fluctuates with market supply and demand.

Further Reading Bearce, D.H., 2007. Monetary Divergence: Domestic Policy Autonomy in the Post-Bretton Woods Era. University of Michigan Press, Ann Arbor, MI. Bernhard, W., Leblang, D., 2006. Democratic Processes and Financial Markets: Pricing Politics. Cambridge University Press, Cambridge. Bernhard, W., Broz, J.L., Clark, W.R., 2002. The political economy of monetary institutions. International Organization 56 (4), 693–723. Broz, J.L., Frieden, J.A., 2001. The political economy of international monetary relations. Annual Review of Political Science 4, 317–343. Cooper, R., 1971. Currency Devaluation in Developing Countries. Essays in International Finance 86. Eichengreen, B., 1996. Globalizing Capital. A History of the International Monetary System. Princeton University Press, Princeton, NJ. Frieden, J., 1991. Invested interests: the politics of national economic policies in a world of global finance. International Organization 45 (4), 425–451. Frieden, J.A., Stein, E., 2001. The Currency Game. Exchange Rate Politics in Latin America. Johns Hopkins University Press, Washington, DC. Henning, R.C., 1994. Currencies and Politics in the United States, Germany, and Japan. Institute for International Economics, Washington, DC. McNamara, K., 1998. The Currency of Ideas: Monetary Politics in the European Union. Cornell University Press, Ithaca. Simmons, B., 1994. Who Adjusts? Domestic Sources of Foreign Economic Policy During the Interwar Years. Princeton University Press, Princeton, NJ. Walter, S., 2008. A new approach for determining exchange-rate level preferences. International Organization 62 (3), 405–438.

Political Economy of Financial Globalization: Interest Group Politics; The Political Economy of International Monetary Policy Coordination.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

C H A P T E R

4 Financial Institutions, International and Politics L.L. Martin University of Wisconsin-Madison, Madison, WI, USA O U T L I N E Introduction

37

Intellectual Background

38

International Financial Institutions: How Much Autonomy?

39

International Financial Institutions: Effects

42

Conclusion Glossary Further Reading Relevant Websites

INTRODUCTION

problem is to encourage beneficial flows of capital while avoiding moral hazard problems that would result from unfettered access to external resources. As a result, these organizations constantly balance political and economic interests, and much research has treated the IFIs as principals of their state agents. A second major theme is the balance between rule-based interaction and the unconstrained exercise of economic and political power. Most scholarly work on IFIs concentrates on the Bretton Woods institutions (the World Bank and International Monetary Fund (IMF)). This is not to say that other IFIs are unimportant. For example, the Organization for Economic Cooperation and Development is a vital group of developed economies that collects and exchanges substantial economic information, regional development banks play an increasingly important role in development, and regulatory accords (such as the Basel Accord) have at times had profound effects. Nevertheless, concentrating on the major financial institutions has advantages. The scholarly work on these organizations is richer and deeper than that on other IFIs. In addition, the general analytical questions addressed in studies of these organizations should provide substantial insight into other types of IFIs. Providing some background on the study of institutions generally in international relations (IR) is a good beginning. This discussion shows how the study of institutions moved from being purely descriptive or

In this globalized era, many of the world’s international financial transactions are organized by international financial institutions (IFIs). The international political economic environment is highly institutionalized, with international financial organizations playing an important role in the worldwide distribution of wealth. As such, these organizations have become subject to intense public scrutiny, some supportive and some hostile. (There is a valid distinction between institutions and organizations, as other chapters in this handbook make clear. In the international relations literature, the term ‘institutions’ is used to refer more generally to sets of rules and norms. ‘Organizations’ embody these norms and are empowered to take actions. However, in this case, the distinction does not hold any great analytical consequences. Most institutions also have substantial organizational structure.) IFIs have also increasingly been the subject of rigorous scholarly study. These political institutions are studied particularly by political scientists, using the same intellectual framework applied generally to international organizations. This chapter considers the framework used to study IFIs and highlights major findings. The focus is on the following themes: first, understanding the causes and consequences of IFIs by specifying the fundamental strategic problems that they address. For most IFIs, the basic

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00001-5

44 45 45 45

37

# 2013 Elsevier Inc. All rights reserved.

38

4. FINANCIAL INSTITUTIONS, INTERNATIONAL AND POLITICS

normative to developing strong analytical foundations. The modern study of IFIs is firmly grounded in this more general IR tradition. The focus is then turned on the IFIs themselves, considering the issue of their relationship with their member states, and their impact on the economies of their borrowers. A summary of the current political study of IFIs and the most promising directions for future research forms the conclusion.

INTELLECTUAL BACKGROUND Our understanding of the functioning and effects of IFIs generally has its roots in the modern scholarly study of international institutions and in the early 1980s study of international organizations (IOs). Prior to the early 1980s, the study of IOs was quite policy-oriented and descriptive and lacked an overarching analytical framework. This lack of a theoretical foundation meant that, although individual studies generated strong insights, they did not cumulate to create a coherent picture of, or debate about the role of, IOs in the world economy. This situation changed in the early 1980s, when new work cast international institutions in a new light, suggesting a novel explanatory framework for studying them. The puzzle that motivated this research began with two observations: first, that international economic cooperation in the 1970s was stable in spite of substantial shifts in the distribution of international economic power, and second, that organizations such as the Bretton Woods institutions and the GATT were prominent features of the economic landscape. These two observations were connected to one another: the existence of institutions and IOs explained the persistence of economic cooperation. For states to be able to cooperate, they must overcome a range of collective-action problems. No external enforcement exists in the international economy, so any agreements must be self-enforcing. This means that states must find ways to avoid temptations to cheat, for example, by reneging on agreements to encourage trade by erecting protectionist barriers. Avoiding such temptations requires high-quality information about the actions and preferences of other states, and about the likely consequences of cheating on agreements. In addition, states must coordinate their actions, for example, agreeing on common technological and public health standards. IOs provide forums in which states can mitigate collective-action problems that threaten stable patterns of cooperation. IOs can perform monitoring functions and provide assurance that others are living up to the terms of their commitments. They are forums for negotiating to resolve coordination problems and to learn about the preferences and constraints facing other governments. IOs create structures for enforcement

and dispute resolution, although actual enforcement powers typically remain in the hands of member states. Through these functions, IOs become a valuable foundation for cooperation and for the global economy. IOs enable more resilient patterns of cooperation in the face of underlying shifts in economic power and interests. The initial work applying this ‘contractual’ view of institutions concentrated on international regimes, which, in turn, are defined as sets of principles, norms, rules, and decision-making procedures. One advantage of examining regimes, as compared to the earlier focus on individual IOs, is that this shift allows researchers to consider informal institutions as well as formalized bodies. While in more recent years much attention has shifted back to formal IOs, the understanding that informal bodies of norms sustain cooperation in the global economy underlies today’s work on individual organizations. While research on international regimes represented a major step forward in the analysis of international institutions, it was subject to criticism from a number of perspectives. It may have moved too far from the analysis of specific IOs, thus missing some important internal organizational dynamics. The concept of regimes was broadly defined, and regimes were difficult to observe independent of their effects. Consequently, much effort went into determining whether or not regimes actually existed in various issue areas. Further research explored whether changes in patterns of behavior reflected changes within regimes or of regimes. It is not clear that these descriptive debates added a great deal to our understanding of the causes and consequences of institutions in the international environment. Other major weaknesses of the literature included its state-centric focus and neglect of domestic politics. IOs may fail in their attempts to manage difficult problems in international relations. The inability of IOs to resolve serious conflict could reflect not just random mistakes but a systematic pattern of failure. IOs could even have perverse effects, exacerbating conflict rather than mitigating it. For these reasons, it may be unwise to rely too heavily on formal IOs to manage international relations. In addition, the regime literature may neglect the role of political leadership. Many of these criticisms have been echoed in recent years in the analysis of IFIs. One of the most telling critiques of the regime literature is that it was too focused on market failures. The failures involve instances where all could potentially benefit from mutual cooperation, but where collective-action problems such as high transaction costs prohibit states from reaching the ‘Pareto frontier.’ For example, if we consider cooperative communications efforts, states seem to have little trouble reaching the Pareto frontier. States found it relatively easy to identify the set of bargains that would benefit all participants. Distributional conflict trapped them; they found themselves having

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

INTERNATIONAL FINANCIAL INSTITUTIONS: HOW MUCH AUTONOMY?

to choose among bargains that benefited some while harming others. Thus, the most significant problem plaguing efforts at international cooperation was not providing a good contractual environment to overcome transaction-cost problems such as informational limitations, but a coordination problem in which states disagreed over which of the multiple Pareto-efficient equilibria they preferred. This insight has led to a revision of early work on regimes, which claimed that coordination problems would be relatively easy to solve. A new focus on how institutions might aid in resolving coordination problems has added depth to our understanding of IOs’ functions. The theory of international institutions became deeper and richer in the 1990s. Scholars brought the concept of multilateralism back into the study of institutions. Multilateralism may be defined, simply, as cooperation among three or more states. A more elaborate definition indicates a set of norms that prescribed certain patterns of behavior, such as nondiscrimination. Either definition redirects attention to variation among types of institutions, a highly productive move for the field. Another debate arose regarding the problem of compliance with the rules of IOs and with international agreements more generally. The managerial school, representing primarily the views of legal scholars, argued that states generally wanted to comply with international rules and that variation in compliance was therefore not a compelling puzzle. Political scientists responded by noting that the managerial argument was plagued by selection bias: if states almost always complied with the rules, it was likely because they would only accept rules that demanded minimal changes in their patterns of behavior. The appropriate question, therefore, was not so much compliance, but rather how different structures of rules would promote far-reaching changes in behavior that left states open to exploitation, or ‘deep cooperation.’ Interestingly, both the managerial and contractual schools agreed on the conclusion that variation in patterns of compliance was not a terribly important or interesting question, although they came to this conclusion by different paths. The managerial school argued that little variation in compliance could be observed because states are obliged to comply. The formal analysis of compliance argued that minimal observed variation in compliance simply reflected the fact that states are unlikely to make commitments on which they intend to renege. Nevertheless, empirical research on variation in compliance has continued, leading to some intriguing findings. For example, human rights treaties lead to greater compliance in situations where they lead to the mobilization of domestic groups who share the goals of the treaty, and lead to more government respect for human rights in areas including women’s rights and nonuse of torture.

39

Other theoretical developments focus on the form and design of IOs. One body of work asks why IOs are becoming more ‘legalized’: they more often incorporate legalistic features such as third-party dispute resolution mechanisms. Researchers have begun to explore the advantages and possible disadvantages of legalization in promoting international cooperation. Another body of work focuses on design principles for IOs. Starting from the assumption that IOs are designed to resolve collective-action problems, analysts have derived a number of hypotheses about the form of IOs. For example, if states design an IO to reduce the transaction costs of monitoring members’ behavior, we would expect the organization to have relatively centralized monitoring capacities. Using logic like this, dimensions of IOs such as their centralization and autonomy from member states can be explained. It is important to note that the typical IO constitutes only one point on a wide spectrum of forms of international organization, ranging from complete anarchy to hierarchical organization, the latter exemplified by empires. A related question is why states sometimes cooperate informally, while at other times they choose to create formal IOs. The answer to this puzzle likely lies in transaction costs and trade-offs between autonomy and the benefits of commitment. Overall, these developments in the study of international institutions provide a firm foundation for more specialized studies of IFIs. They suggest that one of the first questions to be asked when studying a particular organization is to ask about the problems it was designed to address. An understanding of such issues then leads to predictions about the form and functioning of the organization and about its effects on economic flows and conditions.

INTERNATIONAL FINANCIAL INSTITUTIONS: HOW MUCH AUTONOMY? IFIs such as the IMF and the World Bank play a major role in the world of international finance and money. We are just beginning to understand how the interaction of politics and economics works in such institutions. To understand what IFIs do, we need some insight into the fundamental strategic problems that they confront. These problems have led many analysts to use a principal-agent framework to study the IFIs, asking about the relative freedom of maneuver available to these organizations, given patterns of state interests. This problem is played out in an ongoing battle pitting rules that attempt to constrain state behavior against the continual exercise of state power. The empirical evidence supports both perspectives: that rules matter

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

40

4. FINANCIAL INSTITUTIONS, INTERNATIONAL AND POLITICS

and also that they have not succeeded in fully defeating power politics. The IMF and World Bank are known as the Bretton Woods institutions, created at the end of World War II at the Bretton Woods conference. Initially, the main purpose of the IMF was to oversee the functioning of a fixed exchange-rate regime. To make this regime work, the IMF was to organize short-term support for members that were facing balance-of-payments crises. Over time, the Bretton Woods exchange-rate regime fell apart. However, by then the IMF had proved itself valuable in providing relief for states facing financial crises, and it has continued to play the central role in these situations. The World Bank (also known as the International Bank for Reconstruction and Development) was initially intended to provide funding for development efforts, particularly for states too poor to reliably access private international capital markets. Thus, the World Bank funded longer-term development projects, such as infrastructure construction. Over time, the specific types of programs funded by the IMF and World Bank have tended to converge, but some distinction remains. In any financial transaction, institutions need to walk a fine line between encouraging the provision of funding that will be beneficial for both the borrower and the lender and encouraging moral hazard. Moral hazard is a serious concern in these transactions. Consider the typical case addressed by the IMF. A country has fallen into a financial crisis, either through poor policy or exogenous shocks. The government finds itself unable to make good on its commitments to make payments on its outstanding debt, and the value of its currency is collapsing. If the cause of the crisis will pass, the provision of temporary financing will benefit both the country that receives the financing and its lenders. The lenders would be likely to recover more of their assets once the crisis has passed. However, a government that knows that it will be bailed out of such crises is likely to behave more recklessly, adopting inappropriate policies and overborrowing. This is the moral hazard dilemma. The IMF has addressed the moral hazard problem by imposing conditions on its lending programs, attempting to force states to adopt more responsible fiscal policies. Initially some IMF members opposed the use of such conditionality. They argued that the organization’s role was to provide funding as needed, with the major creditors (especially the United States) insisting on the imposition of conditions. The number and types of conditions have expanded substantially over the years. Governments wishing to conclude a program with the IMF must typically commit to reduce public spending, increase collection of taxes, liberalize their international economic relations, and even improve other areas of governance. Of course, such

conditions are not popular among the governments that must accept them. Even if they are economically justified (a point some would dispute), there are occasions on which the major creditor states would prefer looser conditions for purely political reasons. For example, it is widely understood that the United States opposed the imposition of tough conditions on Russia in the early 1990s, wishing to assure Russia’s political stability. In addition, states that are home to private creditors with substantial exposure in the crisis country are likely to prefer looser conditions and flows of capital. This basic strategic problem – potential benefits from capital flows, but a moral hazard problem – has led many scholars to use a principal-agent framework to study the IMF and, less extensively, the World Bank. In this framework, the members of the IFIs, especially the major creditors, are treated as the principals that use the IFI to implement their preferred policies. IFIs, as agents, have their own interests, usually understood as technocratic economic interests. The question is then the extent to which the IFIs can pursue their own agenda versus responding to the specific demands of their principals. As such, the ongoing tug-of-war between rules and power describes the dynamics of the IFIs. How much freedom of maneuver are IMF staff and management likely to have? In part, the answer lies in the distribution of interests among the principals. In practice, the IMF’s Executive Directors (EDs), who represent either individual member states or groups of states, serve as the principals. When considering the stringency of the conditions that the IMF puts on its structural adjustment programs, we can conceive of the preferences of the EDs as ranging from a desire for lenient conditions to a preference for more stringent conditions. These preferences will be driven by a combination of factors: economic conditions in the country negotiating a program, the state of the world economy, and political relations with the negotiating country. At times, ED preferences will not diverge much. For example, if a borrower is suffering an intense crisis and presents a fundamental threat to the stability of the world economy, all EDs will want conditions designed to address the immediate situation. On the other hand, if a borrower is less crucial to the world economy, or has varying political relations with different principals, we observe that ED preferences diverge significantly. It is also possible for ED preferences to diverge from those of the IMF staff, who are more likely to be driven purely by technocratic, narrowly economic considerations. For example, preferences over the conditions attached to programs for Russia in the early 1990s diverged substantially. IMF staff, as well as some EDs, thought that stringent conditions were necessary to force the Russian government to enact essential reforms. On the other hand, the United States has a close and

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

INTERNATIONAL FINANCIAL INSTITUTIONS: HOW MUCH AUTONOMY?

complicated relationship with Russia, and the US government felt that stringent conditions would threaten its stability. When the preferences of EDs diverge from one another, spaces open up for IMF staff to pursue their own agenda. They can play EDs off against one another and thus gain autonomy. On the other hand, when ED preferences converge, there is little scope for staff autonomy. We thus observe that circumstances that cause ED preferences to converge – such as systemic crises – tend to decrease IMF staff flexibility, as they will be more tightly constrained by their principals. Informational considerations also affect the degree of autonomy that IFI staff have. A fundamental reason for principals to delegate authority to agents is that they need agents to acquire information and respond appropriately. The design of IMF programs, for example, requires extensive information about the particular economic circumstances in the borrowing country. The negotiating team working with the country will collect and analyze large quantities of economic data. If these agents were tightly constrained by their principals, their ability to gain access to such information would be compromised. This general situation is especially complex in the case of the IMF, as borrowing countries often have tense political relationships with some of the most powerful member states. As a result, they are often reluctant to fully reveal sensitive economic information, being reluctant to do so if it will quickly get into the hands of their political adversaries. For this reason, IMF staff have gained a striking degree of autonomy in the process of working with borrowers in the design of structural adjustment programs. In addition, EDs as a rule simply approve the programs designed by the staff, without submitting the program to a process of amendment or revision. In the earliest years of the IMF, such informational and political considerations played a major role in the evolution of IMF procedures. Initially, EDs were deeply involved in the process of negotiating and designing programs, even participating in field visits to borrowing countries. However, it quickly became evident that such political interference led borrowing governments to conceal information and hinder the negotiation of programs. As a result, EDs engaged in a massive delegation of authority to IMF staff, removing themselves from the process of negotiating programs. Despite dramatic economic and political changes in the world since then, delegation of authority with respect to negotiating programs and establishing conditionality has barely changed. One exception to the largely stable delegation of authority to IMF staff lies in the use of so-called preconditions. Preconditions are requirements that the staff impose on borrowing governments before they will submit a program for approval by the Executive Board. With the debt crises of the 1970s and 1980s, use of

41

preconditions grew and came to represent a large increase in the ability of the staff to act independently of ED oversight. EDs did react to this situation, by imposing limits on the use of preconditions. However, the staff do still rely on preconditions – either formal or informal – and they have been a mechanism by which the staff has increased its autonomy from powerful member states. Some analysts demonstrate that the IMF’s patterns of lending respond to the geopolitical interests of the United States, its dominant member. Further development of this line of analysis asks about the conditions under which the United States actually exerts substantial influence over IMF lending. The informal rules by which the IMF operates allow ‘exceptional access’ for powerful countries. Such powerful countries – especially the United States – can use this access to intervene and relax the conditions imposed on some borrowers. The United States only intervenes in this manner when a country has economic or political significance and when it is vulnerable enough to be willing to draw on its political relationship with the United States. In other circumstances, powerful states appear to delegate broad authority to the IMF, and it has refrained from maximizing the scope of conditionality. One difficulty in assessing political influence within the IMF lies in differentiating political favoritism from the role of the IMF in maintaining systemic stability. ‘Important’ countries could appear to receive favorable treatment either because member states press the IMF to do so, serving their narrow political interests, or because their economic difficulties threaten the system as a whole. One standard for the IMF is ‘technocratic impartiality,’ in which IMF programs are driven solely by the individual economic conditions facing individual borrowers. It is clear that IMF activities do not match this strict condition, but in what way do they deviate from it? A major determinant seems to be the capacity of large economies in distress to create regional or global crises. The possibility of contagion drives the IMF to intervene in some circumstances when it otherwise would not, leading to apparent preferential treatment in some cases. Thus, the threat of serious crises leads to IMF behavior that is not even-handed. When circumstances are less dire, we also observe some variability in IMF treatment, but this does not appear to reflect systematic biases, but rather reactions to local and regional conditions. The ‘public choice’ school has studied the IMF as a self-interested organization attempting to assert itself in the face of constant political demands from its powerful member states. On the other hand, states are often able to exert substantial influence over the IMF’s activities. Thus, while the IMF is an agent with some autonomy, it has a hard time escaping its political confines. Scholars have applied this analysis to examine the evolution of conditionality over time, asking about the

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

42

4. FINANCIAL INSTITUTIONS, INTERNATIONAL AND POLITICS

content and number of conditions imposed. The IMF can usefully be studied as a bureaucracy with its own internal rules and interests. In this sense, it has more autonomy than most scholars have recognized. An autonomous IMF should impose stringent conditionality on states that have a poor record of living up to past commitments, and authors have found evidence to support this argument. Overall, the evidence suggests that the IMF is an agent constrained by the political interests of its principals, but one that is able to exert autonomy under certain conditions. Other scholars, working within the same general principal-agent framework, focus on the delegation of authority to IFIs. They ask why states would choose to allow IFIs what appears to be a substantial degree of autonomy. Some analysts find that delegation has not undermined the interests of the most powerful member states, as delegation is itself a strategy for promoting their interests. For example, the World Bank’s environmental policies correlate highly with the environmental interests of the United States. In another example, the United States exerts a substantial impact on World Bank and IMF programs, for example, arguing for more lenient conditions for its political allies. On the other hand, the IMF, in its use of conditionality, often responds to private financial actors rather than state interests. Scholars have also approached the question of political influences in the IMF by considering the process of its funding. Periodically, the US Congress must pass legislation to increase the US contribution to the IMF. These pieces of legislation provide an opportunity to determine who supports the activities of the IMF and who sees less benefit from them. Banks that specialize in international lending should support the IMF, as it reduces their exposure to risk. Similarly, individuals who generally benefit from globalization (highly skilled workers and holders of capital in the United States) should favor enhancement of IMF activities. We can test and find strong support for these hypotheses by looking at congressional voting on IMF funding: Members of Congress who represent districts with many international banks and high-skilled workers tend to vote in favor of greater financial support by the IMF. Thus, the institutionalist perspective has given rise to insights into the design of the IFIs, particularly focusing on issues of delegation and influence. Some have begun to critique this view of the IFIs, arguing that it underestimates the autonomy of the staff of IFIs. The IFIs may in fact be able to pursue agendas that have little relationship to the interests of either major donors or borrowers. This can happen through the exercise of legitimately perceived authority, especially because it has the veneer of science. This line of analysis presents a potentially strong threat to the entire contractual framework; it conceives of a very different relationship between states and

institutions. For example, it suggests that we should spend much more time analyzing processes of socialization within institutions. Another perspective suggests that we need to draw on alternative theories of accountability to make sense of the role of IFIs. While this perspective does not directly challenge the contractual one, it does suggest that the contractual approach with its emphasis on principal-agent relationships is too narrow a prism through which to study IFIs. Issues of accountability have long been a concern both within the World Bank and among those who study it. This led, for example, to the creation of an Inspection Panel in 1993 to investigate complaints about the bank’s activities.

INTERNATIONAL FINANCIAL INSTITUTIONS: EFFECTS The other major set of questions with respect to the politics of IFIs, of course, concerns their impact on the economies of the states where they are active. There is an emerging literature on this topic, and space constraints prevent doing full justice to it here. However, it is safe to say that, in particular, the evidence on the IMF suggests that it has not been terribly effective in bringing high levels of growth to program participants. Countries that enter IMF programs, rather than relying on them temporarily and then resuming a ‘normal’ growth pattern, tend to remain under IMF aid for long periods of time. IMF programs may increase income inequality even while failing to promote aggregate growth. The reasons for this lack of systematic success – or even evidence of systematic failure – lie in the complex interplay between external political and economic forces and internal bureaucracy and ideology. Determining whether an IMF program improves the macroeconomic fortunes of a country is a difficult task. Analysts need to take into account many potentially confounding factors. The same conditions that lead a government to enter an IMF program could lead to poor macroeconomic performance. In order to identify the causal effect of an IMF program, therefore, we need to control for these factors. In addition, we need to explicitly model the decision to enter an IMF program, a process that involves a complex set of negotiations between the borrowing country and the IMF. Statistical analysis of the effect of IMF programs therefore needs to be taken with a grain of salt. Taking all of these considerations into account, it is safe to say that we can find no convincing evidence that IMF programs improve the long-term macroeconomic performance of countries that receive them. Perhaps the major effect of IMF activity cannot be found on the level of individual countries but may be focused on its capacity to limit the contagion of debt and financial

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

INTERNATIONAL FINANCIAL INSTITUTIONS: EFFECTS

crises. Or perhaps the IMF is acting not primarily in the interest of borrowing countries but of lending countries. At the extreme, some consider the IMF an enforcement mechanism for a lending cartel. Scholars have debated the causes of this apparent lack of efficacy. Some argue that it is precisely the autonomy of the IMF, which wishes to loan large amounts of money, that causes conditions not to be enforced and that undermines programs. However, the fundamental problem is probably the reverse: the Fund’s principals frequently intervene to promote leniency toward favored states. This persistent influence of political pressures means that the conditions the IMF so painstakingly negotiates are rarely imposed with any consistency or credibility. Thus, the problem with IMF programs is not that they are poorly designed or based on an inappropriate economic ideology. It is that even well-designed programs are not enforced because of political interference. Similarly, IMF loan decisions appear to reflect domestic political considerations in the United States. The degree to which IMF member states choose to push IMF programs in particular directions is closely linked to their global economic interests. For example, consider countries (such as the United States) that are home to large international banks. These banks find themselves exposed when they make loans to poor and emerging markets which then face debt and financial crises. Specifically, such banks risk default or, at a minimum, delayed or partial repayment. An IMF program provides direct assistance to the crisis country, catalyzes the flow of private investment, and imposes conditions that should enhance the capacity of the government to eventually repay its loans. For these reasons, internationally exposed banks should strongly desire the intervention of the IMF in debt crises. This is what analysts mean when they say that the IMF serves, at least in part, as a ‘creditors’ cartel.’ Large international banks are also powerful actors in domestic politics. Therefore, we should expect them to bring significant pressure to bear on their home-country governments to facilitate the flow of IMF resources to troubled debtor countries. In fact, this is precisely what we observe: IMF programs are more likely to support countries in which large international banks are deeply involved. In addition, the composition of capital flows has an impact. Banks, as the source of loans, are wellorganized politically and well-placed to pressure homecountry governments. Other creditors do not benefit from these advantages. In particular, bondholders tend to be more numerous, have smaller-scale individual investments, and are more dispersed. Bondholders therefore find it more difficult to organize and convince governments to represent their interests in IMF discussions. Thus, IMF bailouts tend to benefit the holders of loans rather than bonds.

43

Domestic politics are not the only political influence on the design of IFI programs. Individuals who work for IFIs such as the IMF, come from particular organizational and educational backgrounds, their backgrounds likely affecting the kinds of decisions they make. For example, the IMF has moved toward more neoliberal policy prescriptions, such as demanding liberalization of capital controls. It appears that part of the momentum in this direction came from greater hiring of individuals who trained in neoliberal economics departments such as the University of Chicago. IMF staff drove much of this movement, in the face of lack of direction from top management or from powerful member states. In other words, the ‘Wall Street – Treasury Complex’ may be less responsible for the broad direction of IMF policies than are internal IMF processes and the ideologies of its staff. Another approach to understanding the consequences of IMF programs is to look at the details of how governments under a program change their spending priorities. In the absence of an IMF program, democratic states tend to spend more on public services than do nondemocracies. However, this distinction disappears when countries enter an IMF program. The IMF requires that governments accessing its funds decrease their budget deficits. For this reason, the impact of domestic politics on government spending is mitigated by participation in an IMF program. Scholars have also devoted extensive attention to the effects of World Bank activities. Many of their findings echo those regarding the IMF, although the World Bank appears to be somewhat more autonomous from direct political influence than is the IMF. One common observation is that the World Bank suffers from a common pathology of international organizations: apparently systematic inconsistency between its words and deeds. For example, while the World Bank’s stated agenda is to alleviate poverty, often projects flow to middleincome and emerging markets rather than the most poverty-stricken areas. In addition, in spite of abundant rhetoric about good governance, the World Bank has historically turned a blind eye to systemic corruption in governments that draw on its resources. What might explain such ‘organized hypocrisy?’ The answer likely lies in the fact that this organization, like all IOs, needs to respond to a complex, conflicting mix of demands, many of which have been discussed in this chapter. Member states demand that the World Bank engage in economically viable projects while tending to their political interests. Nongovernmental actors demand that the World Bank takes issues such as the environment, governance, and human rights into account when allocating its resources. The internal bureaucracy of the bank and the ideas held by its staff influence their choice of programs to support. And, like any large

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

44

4. FINANCIAL INSTITUTIONS, INTERNATIONAL AND POLITICS

organization, the Bank needs to continually prove its worth, creating pressures to engage in large-scale, visible projects. Not surprisingly, the combination of these demands leads to patterns of project development that do not fit any neat conception of ‘consistency.’ One particularly interesting recent analysis considers how participation in a World Bank structural adjustment agreement (SAA) influences government human rights practices, especially government respect for physical integrity rights. If an SAA improves economic performance, we might expect it to also lead to less government repression, as the government would confront less domestic unrest. On the other hand, if the conditions of an SAA exacerbate domestic conflict, then it could increase levels of repression. Unfortunately, data on physical integrity rights suggest that structural adjustment leads to higher levels of government repression. One fascinating new line of analysis explores the relationship between participation in an IMF structural adjustment program and geopolitics, especially the politics of the United Nations Security Council. The Security Council is made up of five permanent members (the United States, Britain, France, Russia, and China) and a rotating set of ten other members. The Security Council takes action on issues that threaten global security, and in these circumstances the Permanent Five desire the votes of the other ten members. In an intriguing manner, having one of the rotating members of the Security Council under the auspices of an IMF program provides the Permanent Five a mechanism through which to leverage their votes. Building on the discussion above, we know that the powerful states that are members of the Permanent Five also have influential roles within the IMF. In particular, they can exert influence over the degree of stringency of IMF conditions, and the rate at which IMF funds are dispersed to states under IMF programs. Using the threat and promise of more lenient treatment within the IMF – which directly contributes to the fortunes of borrowing governments – the Permanent Five could use IMF programs as a means of securing more votes in the Security Council. If this conjecture is correct, we should expect to see that temporary members of the Security Council are more likely than nonmembers to have an IMF program in place: this program is one method of exerting leverage in the Security Council. Careful statistical analysis has found that this is in fact the case. Controlling for other factors, states that are rotating members of the Security Council are more likely to have an IMF structural adjustment program in place. Determining the effect of World Bank activities is not quite as complex an endeavor as for the IMF, as the vast majority of World Bank operations involve discrete development projects. The World Bank provides funds for building dams, roads, and other infrastructure

projects. On one level, the effects of these projects are easy to see, via the physical presence of these large investments. However, on another level, we can ask about the larger effects of such projects. Do they stimulate private investment and thus encourage more rapid economic development? What externalities might be attached to them, in terms of environmental or human rights consequences? These larger questions are subject to the same difficulties of causal inference as the effect of IMF programs. The World Bank has arguably been more responsive to outside criticism about its activities and their effects than the IMF has been. As mentioned above, when the environmental effects of World Bank projects were criticized, it began implementing more serious environmental reviews and evaluations of its activities. It also established an independent review panel, although many question the seriousness with which it takes the panel’s findings. There is less evidence of direct political influence in the World Bank’s activities than in the case of the IMF. Perhaps the fact that the World Bank focuses primarily on development, rather than the containment of financial and debt crises, has allowed it a measure of autonomy and independence which the IMF has not been able to acquire. Overall, we find that the struggle between political influence and rule-based behavior defines the impact of the IFIs on the world economy and domestic politics. Powerful member states do undoubtedly intervene to influence IMF activities. But there are also periods of ‘normal politics’ during which IMF staff are allowed to implement their technocratic preferences without much interference. What might look like political interference is instead often a convergence of preferences between IMF staff and powerful principals. This can occur when countries that have the capacity to generate rippling crisis effects receive IMF assistance. Isolating the effects of IMF programs is difficult, as these programs likely are not designed as much to benefit individual countries as to enhance global stability. However, as the crucial role of the IMF in the 2008 financial crisis illustrates, even an institution that operates with significant limitations can perform an invaluable role in containing financial instability.

CONCLUSION The new global economy is highly institutionalized. Understanding this phenomenon has led to the development of a vibrant field of political science centered on the study of international institutions and IOs. This field continues to hold a primarily contractual view that sees institutions as solutions to collective-action problems. Thus, the study of IFIs begins by identifying the

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

CONCLUSION

underlying strategic problems that IFIs address. Their fundamental challenge is to provide flows of needed capital while avoiding moral hazard problems. This tension sets up the IFIs as agents of their state principals who frequently have conflicting interests. Thus, the contractual approach with its emphasis on principals and agents has been a powerful tool for studying IFIs. New perspectives are beginning to emerge, as noted in this essay, with a focus on socialization, legitimacy, and accountability. However, they are not yet developed to the degree that they present a fundamental challenge to the contractual approach. The study of the IFIs consistently shows that their dynamics, design, and effects reflect an ongoing struggle between the exercise of power and the rule of law. While some authors find more evidence for the weight of one side in this battle than the other, careful empirical research reveals that neither side triumphs. The IFIs will continue to have a major influence on the global creation and distribution of wealth. Those studying them need to push further to understand the sources of their specific design features and move toward more conditional, precise statements of their effects. However, the analytical framework so far developed has proved insightful and appears to provide a strong foundation for this research agenda.

SEE ALSO Political Economy of Financial Globalization: Emerging Markets Politics and Financial Institutions; International Conflicts; The Political Economy of International Monetary Policy Coordination; Theoretical Perspectives on Financial Globalization: Capital Mobility and Exchange Rate Regimes; Safeguarding Global Financial Stability: International Monetary Fund.

International institutions Sets of rules and norms that are intended to regulate state behavior. International organizations Formal bodies that administer international institutions. Moral hazard problem A situation in which an agent’s actions are not verifiable, and steps taken to shield the agent from risk lead to inappropriate behavior from the perspective of the principal.

Further Reading Abouharb, M., Cingranelli, D., 2006. The human rights effects of World Bank structural adjustment, 1981–2000. International Studies Quarterly 50, 233–262. Broz, J.L., Hawes, M.B., 2006. Congressional politics of financing the International Monetary Fund. International Organization 60, 367–399. Copelovitch, M.S., 2010. The International Monetary Fund in the Global Economy: Banks, Bonds, and Bailouts. Cambridge University Press, New York. Easterly, W., 2001. The Elusive Quest for Growth: Economists’ Adventures and Misadventures in the Tropics. MIT Press, Cambridge, MA. Gould, E.R., 2003. Money talks: supplementary financiers and International Monetary Fund conditionality. International Organization 57, 551–586. Keohane, R.O., 1984. After Hegemony: Cooperation and Discord in the World Political Economy. Princeton University Press, Princeton, NJ. Krasner, S.D., 1983. International Regimes. Cornell University Press, Ithaca, NY. Krasner, S.D., 1991. Global communications and national power: life on the pareto frontier. World Politics 43, 336–356. Martin, L.L., Simmons, B., 1998. Theories and empirical studies of international institutions. International Organization 52, 729–757. Nielson, D.L., Tierney, M.J., 2003. Delegation to international organizations: agency theory and World Bank environmental reform. International Organization 57, 241–276. Stone, R.W., 2008. The Scope of IMF conditionality. International Organization 62, 589–620. Vreeland, J.R., 2003. The International Monetary Fund and Economic Development. Cambridge University Press, Cambridge. Vreeland, J.R., 2007. The International Monetary Fund: Politics of Conditional Lending. Routledge, Oxon. Weaver, C., 2008. Hypocrisy Trap: The World Bank and the Poverty of Reform. Princeton University Press, Princeton, NJ. Woods, N., 2007. The Globalizers: The IMF, the World Bank, and Their Borrowers. Cornell University Press, Ithaca, NY.

Relevant Websites

Glossary Autonomy The ability of an agent to take action without explicit approval from a principal. Delegation The transfer of authority from a principal to an agent for the purpose of carrying out a particular task.

45

http://www.adb.org. http://www.ebrd.com. http://www.iadb.org. http://www.imf.org. http://www.worldbank.org.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

Intentionally left as blank

C H A P T E R

5 Political Economy of Foreign Aid, Bilateral E. Werker Harvard Business School, Boston, MA, USA O U T L I N E Political Economy of Aid Disbursement With a Little Help from My Friends Empirically Speaking Does it Matter that Aid Allocation is Political?

48 48 49 50

Making Good 53 Taking the Politics Out of Aid Allocation 53 Saving Aid from Itself, or, Taking the Politics Out of Aid Receipt and Disbursement 54

Political Economy of Aid Receipt Capital to the Capitol Spending Hard-Earned Aid It’s the Economy, Stupid

51 51 52 52

Conclusion References

Kyrgyzstan is a landlocked, mountainous nation of 5 million people bordering three other landlocked former Soviet states and the remote Chinese province of Xinjiang. With the fall of the Soviet Union, Kyrgyzstan’s production fell significantly, sending the economy back a generation to animal herding and cotton growing. Since the attacks of September 11, 2001, in the United States, however, there has been a new foreign-exchange earner, stemming from Russo-American competition over the use of an airfield – and the very allegiance of Kyrgyzstan itself. In preparation for its campaign in Afghanistan, the United States secured access to air fields in Kyrgyzstan and neighboring Uzbekistan with a mixture of foreign aid and infrastructure upgrading. However, in 2005, the United States was kicked out by the Uzbeks for criticizing a government-led massacre (Walsh, 2005), and the following year, a new Kyrgyz government argued that the base contracts had disproportionately benefited cronies of the old regime. They demanded a 100-fold increase in ‘rent’ from the base, to $200 million (Cooley, 2009). The Americans were able to maintain Kyrgyz air access for the Afghanistan campaign through a $150 million aid package, including $18 million in rent. This uneasy balance remained for a few years, with US support buying access to the air base. Meanwhile, with global commodity prices rising, a cash-rich Russia began to reassert its

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00009-X

55 56

influence in what had historically been its backyard. While on a visit to Moscow in early 2009, the Kyrgyz president announced that the Americans had 180 days to vacate the base. Russia had offered Kyrgyzstan a $300 million loan for economic development, a $150 million grant for budget stabilization (it was the financial crisis), and forgiveness of most of the $180 million in debt that the Kyrgyz state owed Russia (Nichol, 2009). After extensive negotiations, the United States managed to keep air access to Kyrgyzstan, but not before raising the rent on the base and offering additional support for economic development, counternarcotics, and counterterrorist programs. Foreign aid has always been political. This fact has come as no surprise to scholars and practitioners of statecraft. Describing one of the functions of aid, Hans Morgenthau, one of the founders of realist international relations theory, noted: “the transfer of money and services from one government to another performs here the function of a price paid for political services rendered or to be rendered” (1962, p. 302). Indeed, since the late 1940s, every US administration considered foreign aid to be important in achieving foreign policy goals (Ruttan, 1996). The ‘political services’ Morgenthau referred to are not usually as blatant as the case of the Kyrgyz air base. While episodes of discrete political quid pro quos abound in aid,

47

# 2013 Elsevier Inc. All rights reserved.

48

5. POLITICAL ECONOMY OF FOREIGN AID, BILATERAL

more frequently, the donor nation is attempting to realize a broader ambition. Donor nations may give aid in order to further their economic interests, from the assistance of their companies’ commercial ventures to a larger pursuit of trade and market access. Donors may seek to achieve far-reaching political goals like support on international initiatives undertaken by the donor nation. Indeed, donors may even offer assistance to realize the ‘values’ of their electorates to be good global citizens. Aid is, for the most part, a normal good, and the list of aid donors has grown just as the number of nations with global interests and an effective source of revenue has increased. Even as aid disbursement is driven by political goals in the donor countries, these goals frequently coincide with the needs of populations in recipient countries. For a variety of reasons, most donors now view the reduction of poverty and suffering in far-flung countries as consistent with their national interest. Whether from an ethos of responsibility or a calculated decision to reduce terrorist activity, multiple political justifications currently align most donor countries around a common development and humanitarian agenda. But that has not freed aid from political interference. Moreover, the second that foreign aid crosses the border, the receipt and disbursement of funds becomes intertwined with domestic politics in the recipient country. As part of its attempt to explore the political economy of bilateral foreign aid, this chapter examines the politics of aid allocation from the perspective of the donor country, and then the politics of aid receipt and implementation from the perspective of the recipient country. When helpful, it draws from studies of multilateral aid. Following those discussions, the chapter explores solutions, employed by the development community, to the distortions brought about by the political economy of bilateral aid – distortions that steer aid away from achieving economic development in the recipient country. As it turns out, none of these solutions can shield foreign aid from the heavy hand of politics.

POLITICAL ECONOMY OF AID DISBURSEMENT Nearly all donor nations give aid through at least two channels: a bilateral aid agency that answers directly to the country’s government and multilateral agencies such

as the World Bank. Countries typically have less control over multilateral aid allocation, though multilateral agencies are able to share the fixed costs of operating a development agency with sector specialists and various disbursement locations. (In many cases, national governments have quite a bit of control over their donations through multilateral agencies, which often allow them the ability to earmark aid dollars for specific projects in specific countries. Yet, the coexistence of bilateral agencies and multilateral ones for nearly every donor suggests that bilateral agencies allow a level of control over the aid not available when working through development banks.) Looking around OECD capitals, one would question whether Morgenthau had been completely mistaken. According to Britain’s Department for International Development (DFID), “The UK government believes it is in all our interests to help poor people build a better life for themselves.”1 The Danish Ministry of Foreign affairs states that “Poverty reduction remains the fundamental challenge for Danish development cooperation,”2 while the Norwegian Agency for Development Cooperation states that its aid should be administered in such a way that it “contributes effectively to poverty reduction.”3 US foreign assistance has the purpose of “expanding democracy and free markets while improving the lives of the citizens of the developing world.”4 Many bilateral agencies recognize that they are pursuing foreign policy interests at the same time but never that they are buying anything resembling Morgenthau’s ‘political services.’ Yet when Alesina and Dollar (2000) set out to determine whether aid flows corresponded to politics or need, they found that political variables explained “a large, but not exclusive extent” of cross-country differences (p. 34). In particular, measures of former colonial status, voting in the United Nations, and being Israel or Egypt were far better predictors of foreign aid receipt than income per capita, trade policy, or democratization. How to explain the apparent disconnect between Alesina and Dollar’s results and the development rhetoric?

With a Little Help from My Friends A look back to one of the most highly regarded episodes of aid, the Marshall Plan, offers some clues. The Marshall Plan involved around $13 billion in aid

1

‘Who we are and what we do,’ DFID website: http://www.dfid.gov.uk/About-DFID/Quick-guide-to-DFID/Who-we-are-and-what-wedo/, accessed 04/12/10.

2

‘Danish Development Policy,’ Ministry of Foreign Affairs website: http://www.um.dk/en/menu/DevelopmentPolicy/ DanishDevelopmentPolicy/, accessed 04/12/10. 3

‘About Norad,’ NORAD website: http://www.norad.no/en/AboutþNorad/AboutþNorad.125317.cms?show¼all, accessed 04/12/10.

4

‘This is USAID,’ USAID website: http://www.usaid.gov/about_usaid/, accessed 04/12/10.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

POLITICAL ECONOMY OF AID DISBURSEMENT

(around 5% of the US gross domestic product at the time) in mostly foreign-exchange loans for Europeans to purchase critical imports, often from the United States. But, the aid was not merely humanitarian or a narrow stimulus to US producers. The Americans, in providing aid to Europe, were rebuilding the very legs of the international economic system. As George Marshall said at Harvard University when introducing the policy: [T]he consequences to the economy of the United States should be apparent to all. It is logical that the United States should do whatever it is able to do to assist in the return of normal economic health in the world, without which there can be no political stability and no assured peace . . . Its purpose should be the revival of a working economy in the world so as to permit the emergence of political and social conditions in which free institutions can exist. (Marshall, 1947)

American industry needed a market. Moreover, economic interdependence would help reduce the risk of conflict that had hung over the interwar period like a storm cloud. The recipients of Marshall Plan aid would not have corresponded perfectly to the ‘political’ variables in Alesina and Dollar’s analysis. After all, while the Western Europeans may have had similar views on global political debates, this was by no means secure, as the precarious balances in Greece, Turkey, and Italy attested. The United States had no former colonies in Europe; in fact, they gave aid to their former colonizer. Yet, to a student of history (e.g., Kunz, 1997), this aid was as political as any that had been used to sway a vote in the United Nations General Assembly. And its goal was the economic development of the recipient countries. A similar project occurred during the Cold War, except instead of the shattered pieces of the global economy being reassembled and redirected, foreign aid was used as part of a new but slow-moving political competition. Aid was employed by both the Soviets and the Americans (as well as their allies) to affect the balance of power and the depth of allegiance to one governing system or another. They did not simply spend money to change governments or their preferences, although that certainly happened (Faye and Niehaus, 2010). Aid was also spent to improve the economic capacities of one’s allies. In Vietnam, for example, the Soviets built massive hydroelectric facilities, while in South Korea, the Americans funded schools and land improvement. In this game of chess or dominoes, it was not just the number of pieces one had on the board, but their strength as well. Today, the closest thing to the Marshall Plan is also inherently political to the donor country and developmental in its aims. The largely Anglo-American global ‘war on terror’ led to a reorientation of development

49

with security, completing the vision, observed by Howell, of “an emerging view of the South as a source of international crime, terrorism and conflict that contributed to global instability” (Howell, 2006, p. 123). The development policy resulting from such a world view includes a heavy emphasis on reducing the incidence of so-called failed states, as well as eliminating the economic situations that benefit terrorist recruitment. (These policies have continued to be enthusiastically pursued in spite of careful microeconomic evidence that finds that terrorists are neither poor nor poorly educated (Krueger, 2007).) The take-home message, of course, is the same: a vast amount of aid dollars are spent in order to improve the perceived security of the donor countries’ populations.

Empirically Speaking The vast majority of recent empirical papers on the determinants of aid have sought to make the case that much of aid allocation is politically motivated. What motivates one donor, however, may not motivate another. In one analysis, Schraeder et al. (1998) find that, on balance, Japanese aid is motivated by economic and trade interests, Swedish aid supports progressive, socialistminded regimes, while France’s aid is almost exclusively targeted toward francophone countries. Neumayer (2003) finds that OPEC donors, including Saudi Arabia, Kuwait, and the United Arab Emirates, favor other Arab and non-Arab Muslim recipients. This not insubstantial literature has forcibly established that, across the global sample of recipient countries, so-called economic or humanitarian variables are insufficient to explain the direction and quantity of aid donations. Instead, variables that correlate with connections to, or interest of, the donor country seem to explain a great quantity of aid flows. However, what most of this literature fails to argue convincingly is that the political motivations behind aid imply that developmental motivations are absent, or that politically motivated aid is any less successful in reducing poverty. As the Marshall Plan, the war on terror, and even the Cold War demonstrate, just because aid may be designed to benefit the donor country, it does not easily follow that it is any less developmentally motivated. (In fact, one could make the case that donors would be more effective when their self-interest is served by the progress of the recipient country.) Giving more to one’s allies than one’s enemies, or more to one’s trading partners than to countries with no common economic interests, seems benign, if not thoughtful. Yet, in Kyrgyzstan, the US aid seemed to be designed to secure nothing more than Air Force access, and after the Russians bumped up the price of that access, the Americans were forced to

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

50

5. POLITICAL ECONOMY OF FOREIGN AID, BILATERAL

award lucrative procurement contracts to cronies of Kyrgyz President Bakiyev (Koring, 2010). When Bakiyev was overthrown in a violent coup, the outcome could hardly be described as developmental. Accounts of aid being used for strictly political ends of the donor country – with no corresponding developmental aims for the recipient country – like in Kyrgyzstan have been reported in isolation. Economists seeking to determine whether these anecdotes are indicative of broader trends have run into the challenge noted earlier: there are rare cases in which political variables determining aid flows have no corresponding story in which the donor wants the recipient to succeed. One exception is a paper by Kuziemko and Werker (2006) that examines aid to nonpermanent members of the UN Security Council. The Security Council has arguably been the world’s most important committee, with the power to authorize multilateral sanctions and military action. As 10 of the 15 seats on the Security Council are held for a nonrenewable 2-year term, when a nation enters the Council, it experiences a temporary increase in its political value to donor nations. Membership led to an increase in US and UN foreign aid – particularly in years in which the Council discussed matters of high importance – that disappeared once the country was no longer serving. Like the recent aid to the Kyrgyz government, this practice appears to have little to do with promoting development. What fraction of foreign aid is politically motivated, without any intention to promote development, is still an open question. In all likelihood, that number today may be quite small.

Does it Matter that Aid Allocation is Political? Beyond the case that the political economy of aid may occasionally result in ‘assistance’ flowing to the donor, there is a suspicion that politically motivated aid is somehow less effective than aid given through pure altruism. Alesina and Dollar (2000), for example, conclude from their analysis identifying political aid flows that the allocation of aid “provides evidence as to why it is not more effective at promoting growth and poverty reduction.” Others come to the same conclusion. As it turns out, it is not easy to test whether politically motivated aid does not work as well: most settings in which the donor has a political motivation are also indicative of an omitted-variable bias when testing for the effectiveness of the aid. For example, Cold War allies of the United States may have been more likely to face a military or socialist threat, which would certainly affect the likelihood an aid project would succeed or fail. Countries with similar voting records in the United Nations might be the sorts of places where one’s aid technocrats had the most success in helping out. In other words, it is hard to find

natural variation in the amount of politically motivated aid that is not correlated with its underlying potential effectiveness. In one example, Werker et al. (2009) examine the economic impact of politically motivated foreign aid donated by Arab oil producers. Donors such as Saudi Arabia and Kuwait have been among the World’s most generous, giving over 1.5% of their GDP (Neumayer, 2003). Although some of this money did go to famines and other disasters around the world, the vast majority went to other Muslim nations. No doubt the motivation for this was largely political: the Gulf countries were trying to quell unrest due to the huge inequality among their coreligionists (between the oil haves and havenots), as well as to “assure them[selves] a clear position of dominance within the Muslim world” (Kepel, 2002, pp. 69–70). What makes the Arab aid amenable to further analysis is its extraordinary timing. Donations from the Gulf oil producers essentially tracked movements in the price of oil. Their foreign aid programs only began in earnest with the oil crisis of 1973, picked up steam through the oil crisis of 1979, and fell off abruptly as oil markets flooded in the early 1980s. This allows for a differencein-differences analysis, comparing (nonoil-producing) Muslim aid recipients with non-Muslim countries that are similar in other respects. The analysis found no significant effects on economic growth, yet large increases in imports – especially noncapital goods, consumption, and even investment. How does this compare to nonpolitical aid? Estimates of the growth effects of aid that is not politically motivated are actually quite hard to identify, due to the methodological challenges of finding a suitable instrument (Roodman, 2007). While the Gulf aid was politically motivated, it very likely had a developmental component as well. The determination of the impact of foreign aid flows that are politically motivated yet with no plausible development component has proved even harder. The first challenge, as noted earlier, is that it is hard to systematically isolate this type of aid. The second challenge is that outcome measures like economic growth are very noisy, not to mention those caused by a number of factors other than aid itself. One recent attempt by Dreher et al. (2010) attempts to measure the effectiveness of aid going to nonpermanent members of the UN Security Council. On account of the noisiness of economic growth as an outcome, the paper focuses on effectiveness ratings attached to World Bank projects by their own evaluation team. Although World Bank aid is not bilateral, the findings provide some context to scholars of bilateral aid. Surprisingly, the authors found little reduction in project effectiveness ratings when they were awarded to Security Council members – unless those countries were economically mismanaged at the time of award.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

POLITICAL ECONOMY OF AID RECEIPT

On the whole, the economics research on the political economy of bilateral foreign aid has shown quite conclusively that aid is allocated according to political interests. This corresponds to accounts of aid disbursement by diplomats as well as by journalists and scholars in other fields. Where the research has been less conclusive is in making the case that politically motivated aid has no developmental component, and that such aid is any less effective than aid given for purely altruistic or humanitarian reasons. Part of this lack of conclusiveness in the research on the political economy of foreign aid is no doubt driven by the difficulty of a foreign actor improving the economic situation of another country, and part by the more mundane empirical challenges of measuring aid effectiveness.

POLITICAL ECONOMY OF AID RECEIPT Once aid has been disbursed to a recipient country, it is by no means free of political economy challenges. When foreign aid enters the recipient country, it enters a political system far more complex than the foreignaffairs networks in the donor countries that spawned it. In this domestic context, foreign aid can increase the conflict over scarce governmental resources, it can shift the competitive balance across sectors of the economy, and it can change the very relationships the government has with its opposition and its citizens.

Capital to the Capitol One of the most basic effects of aid is to increase the size of the governmental ‘pie,’ and if there are multiple groups dividing the pie, aid can lead to increased fighting over it (Grossman, 1992). This is an example of Tornell and Lane (1999) ‘voracity effect,’ in which a positive shock perversely reduces growth in an economy with weak institutions due to a more than proportionate increase in fiscal redistribution. Svensson (2000) articulates this in a paper on foreign aid and rent-seeking where he constructs a repeated game with stochastic shocks, where the increase in rent-seeking behavior results from a failure of coordination across the interest groups. This is not simply a theoretical construct. At its most extreme – when aid is one of the few sources of hard currency in the country – there can be very deliberate fighting over foreign assistance. During the barren war in Somalia in the 1990s, when the government and the economy were all but destroyed, shipments of

51

humanitarian aid were quite frequently the objects over which militias fought (Peterson, 2000). When the situation is more subtle, aid can still lead to more conflict at the margin: increasing the return to a coup d’e´tat, for example, or luring more actors into the political space. While aid can increase the size of the pie, thus leading to a generic increase in conflict, it can also affect the political sphere in a more predictable manner. Most aid will tend to reinforce existing power structures since it is disbursed through, or in consultation with, governmental actors. Examining decades of aid in pregenocide Rwanda, Uvin (1998) describes a systematic but largely unintentional favoring of the Rwandan Tutsi. He argues that the development aid in pregenocide Rwanda contributed to and exacerbated the structural differences between the Hutu and the Tutsi, reinforcing the conditions that eventually led to violence. Even today, ‘responsible’ donors are supposed to help the government of the day. Governments put together a poverty reduction strategy paper (PRSP) that, according to the World Bank, “sets out a country’s macroeconomic, structural, and social policies and programs to promote growth and reduce poverty, as well as associated external financing needs.”5 Donors are expected to fund projects that are part of the government’s PRSP. (Of course, this may simply set off another round of strategic behavior between the government and its citizens, in which voters in aid-dependent developing countries evaluate their leaders not only on observed public spending, but also on their manipulation of the donors.) Aid has been criticized of keeping bad regimes in power (Moyo, 2009). However, not all political actors benefit equally from foreign aid. During the Cold War, aid was used very deliberately to affect internal politics – promoting allies at the expense of potential adversaries, whether in Italy or in Congo (Devlin, 2007; Miller, 1983). Analyzing aid patterns around elections in the recipient countries, Faye and Niehaus (2010) found that administrations that were politically aligned with a donor received more aid during close elections. Similarly, the United States gave more aid to nongovernmental entities during election years in which the administration was not aligned. Even today, aid officials quite deliberately try to work with the “reformers within the government so that their preferred projects will have the best chance of working.” Nelson (2009) finds that International Monetary Fund (IMF) loans are larger and more forgiving when they are granted to countries with ‘neoliberal’ policymakers who are defined as those who have earned at least a Master’s degree at a top US economics department, or who

5

‘What are PRSPs?’ World Bank website: http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/EXTPOVERTY/EXTPRS/0,, contentMDK:22283891pagePK:210058piPK:210062theSitePK:384201,00.html, accessed 04/14/10.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

52

5. POLITICAL ECONOMY OF FOREIGN AID, BILATERAL

have experience working at the Bank or at the Fund. In Sierra Leone’s conflict resolution and postconflict development, the more successful interventions have involved building deliberate coalitions between donors and political actors in the recipient country who are the natural allies of reform (Thomson, 2007).

Spending Hard-Earned Aid Recipients, for their part, use the aid strategically. Governments regularly attempt to direct infrastructure projects toward the regions with the highest political return and to use donor-funded budgetary support in order to pay government workers, many of whom are part of the government’s patronage network. A number of the aid projects that have been proposed by recipients of the US government’s Millennium Challenge Account funds, which asks the recipient country for a ‘wish list’ of major projects, include very targeted regional requests.6 In China, Zhang (2004) found that Chinese political and bureaucratic interests did a far better job of explaining World Bank project allocation at the provincial level than did humanitarian or economic variables. For example, coastal provinces would have more likely got Bank projects in the 1980s when government priorities were aimed at developing the East; however, when the government shifted in the late 1990s to focus state-led development efforts on the interior provinces, the coastal provinces received less in Bank financing. (Governments also block aid strategically. The Ethiopian government in the 1980s and the Sudanese government, in much of its regional conflicts, regularly blocked basic relief aid to areas with opposition support (Kaplan, 2003; Perlez, 1990).) The net result of the strategic political behavior on the part of the recipient government is to increase their power and control. Yet not all ‘unearned’ resources are good for the government and its relationship with its constituencies. Much of the literature on the so-called resource curse follows the traditions in history and institutional economics that sees the evolution of the state as the result of bargaining between revenue-maximizing leaders and their citizens (e.g., North and Weingast, 1989; Tilly, 1992). In a nutshell, rentier state literature (e.g., Mahdavy, 1970) discusses the ramifications for mostly oil-producing countries that derive their income from the sale of a commodity as opposed to through a tax base. The political bargains that these states make with their citizens center on the distribution of largesse in exchange for staying in power, as opposed to ceding influence over policy choice in exchange for taxes. Moore (1998) applies rentier state arguments to foreign aid. He argues that the greater the dependence of 6

state income on unearned income, the less likely is that state/society relations are to be “characterized by accountability, responsiveness, and democracy” (p. 85). In other words, when governments do not need to collect taxes from their citizens because they are being financed by foreign aid, likewise they may not be pressured to respond to the needs of their citizens to the same degree. Through this channel, foreign aid reduces the responsiveness of governments to their citizens and leads to governments choosing policies that their citizens would not necessarily choose – or at least those citizens not lucky enough to be in government patronage networks. Like CEOs who may not govern fully in the interest of the firm’s shareholders, leaders of aid-rich nations may seek to capture government resources for their private benefit. The list of corrupt governments that receive substantial amounts of aid is certainly long, from Karzai’s Afghanistan (New York Times, 2010) to Mobuto’s Zaire (Transparency International, 2004). Alesina and Weder (2002) examine whether less corrupt governments are rewarded with increased aid. They find that more corrupt countries receive, if anything, larger amounts of aid. But we could ascribe this to political economy channels in aid allocation, as donors used aid to further their foreign policy goals. Whether aid corrupts recipient country governments is a different matter. Here, the evidence is inconclusive, no doubt complicated by the challenges of arguing causality. Svensson (2000) and Alesina and Weder (2002) argue that the data show a possible increase in corruption from foreign aid, while Tavares (2003) finds that aid reduces corruption. This points to the sensitivity of the results to different methods of managing the inherent endogeneity in the data.

It’s the Economy, Stupid Foreign aid can also affect the political economy of the recipient country through economic channels. One of the primary channels echoes the natural resource problem, just as it does via political channels. Just as natural resources can lead to ‘Dutch disease’ through currency appreciation, so too can the capital inflows associated with the unearned income of foreign aid lead to a rise in the real exchange rate. In Dutch disease, the demand for local goods and services associated with a boom in the natural resource sector (or aid projects, in this case) raises the real exchange rate, which harms the competitiveness of the manufacturing sector, since it competes with the aid sector for inputs. There is no shortage of descriptions of aid booms that drive up factor prices in the economy.

Project descriptions are posted clearly on www.mcc.gov.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

MAKING GOOD

In postwar Mozambique, for example, the price movements were evident to a contemporary observer: Donors buy up the best people – most engineers will not work for the government for $100 a month when they can earn 20 times that much working for a donor. Mozambique now has about 3000 foreign aid workers employed by the United Nations, World Bank, bilateral donors and non-government organizations. Often they simply fill gaps caused by other donors having hired Mozambican technicians at high salaries . . . The streets of the capital, Maputo, are full of new luxury four-wheel-drive vehicles. There is a building boom of expensive houses. (Hanlon, 1996)

But when countries have a vast pool of labor and land, can aid significantly affect prices that the productive sectors actually face? Rajan and Subramanian (2011) pool data on foreign aid and growth in manufacturing and find evidence suggesting that Dutch disease is prevalent in aid more generally. Their analysis finds that when countries have received more in foreign aid, the ‘exportable’ sectors of the economy experience slower growth. This reduced performance appears to be driven by an increase in the real exchange rate. How is this seemingly abstruse economic effect political? For one, if manufacturers once commanded extensive political power, the gradual erosion of their competitiveness could reduce their control over political outcomes. Second, it could generate new power brokers, as construction magnates and property owners (not to mention politicians themselves) become the new titans. A critical analysis of the Mozambique boom, for example, found that the prime beneficiaries were “large Mozambican trading companies, a new Mozambican aid and comprador group, white South Africans and foreign companies” (Hanlon, 1996). These changes could have nonrandom effects on longer term economic and political development, given a possibly vital role of the tradable sector (Rodrik, 2008). Foreign aid, then, can enter and change the political equilibrium of the recipient country when it constitutes a large resource flow. Unlike in aid disbursement, the political economy effects of aid receipt are invasive, and the challenges affect the recipient population at all levels. That said, donors can also be a force for positive political change, when they target their aid to reformers within the country, and when they provide positive incentives that outweigh the negative effects of aid on the political climate.

MAKING GOOD It is not particularly insightful to say that aid is political, since political forces are behind why aid was developed and why it continues to survive. However, the

53

political economy of aid allocation and receipt does interfere with its optimal distribution from a developmental standpoint. Aid policymakers, seeking to maximize the developmental impact of foreign assistance, have devised a number of ways to attempt to subvert the political forces at work. Morgenthau, on the other side of the coin, believed that the developmental aims of aid interfered with its political functions. Of the aid that is truly political, Morgenthau (1962) observed that: Bribery disguised as foreign aid for economic development makes of giver and recipient actors in a play which in the end they may no longer be able to distinguish from reality. In consequence, both may come to expect results in terms of economic development which in the nature of things may not be forthcoming. (p. 303)

That critique is worth keeping in mind as we examine the efforts to depoliticize aid allocation and spending.

Taking the Politics Out of Aid Allocation Politicians have managed, in many cases, to fix politically driven inflationary monetary policy by making independent central banks with clear missions. Looking at the field of foreign aid, one would see institutions such as the World Bank or the regional development banks (African, Asian, Inter-American, and European) and believe that they might be a corollary to independent central banks. After all, the World Bank is not beholden to any one government; its staff enjoys diplomatic privileges and the members are chosen based on their commitment to development; and the mission of the World Bank is to ‘fight poverty’ and ‘help people help themselves.’7 What about this isn’t independent? For starters, its management structure is not politically neutral. The Bank is overseen by an executive board composed of representatives from member-state governments, both donors and recipients. It is run by an American president. (The Europeans get to run the IMF.) This direct supervision provides incentives for governments to meddle with the lending decisions of the Bank. Kaja and Werker (2010) investigate whether the executive board, which is made up of only 24 members, receives special privileges. As it turns out, developing countries on the board can expect to receive approximately double the loans from the International Bank for Reconstruction and Development, the largest of the World Bank’s divisions. So, one downside of trying to insulate donor politics from aid allocation decisions is that the recipients themselves can try to capture the agenda.

7

‘About Us.’ World Bank website: http://web.worldbank.org/WBSITE/EXTERNAL/EXTABOUTUS/0,,contentMDK: 20040565menuPK:1696892pagePK:51123644piPK:329829theSitePK:29708,00.html, accessed 04/14/10.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

54

5. POLITICAL ECONOMY OF FOREIGN AID, BILATERAL

But donors, in fact, do a very poor job of stepping back from aid allocation decisions made at international financial institutions. Lending at the World Bank’s lowincome-targeted International Development Association is correlated with US political interests as measured by UN voting on important issues (Andersen et al., 2006). The big decisions at the IMF are controlled by the G7 (Fratianni and Pattison, 2005) and the United States in particular (Woods, 2003), and loans are larger and more frequent to US allies (Barro and Lee, 2005). Indeed, both IMF and World Bank loans increase to countries when they become nonpermanent members of the UN Security Council, suggesting a complex system of global horsetrading between votes at the Security Council by developing countries and aid decisions by donors through the Bank and the Fund (Dreher et al., 2009a,b). In some situations, donor governments even prefer donating highly political aid through these institutions as it allows them to bypass their legislatures. On the whole, then, multilateral aid institutions do not do a good job of insulating aid allocation decisions from politics. To be fair, this was never their intention in the first place. The structure of the development banks and the IMF was designed to ensure that their biggest funders had an extensive oversight over day-to-day operations. Donor governments have had somewhat more success in designing institutions within the government and such institutions are mandated to put developmental needs on the front burner. Britain established a cabinet-level position and endowed the new DFID with broad responsibilities, not to mention a seat at the table voicing ‘development concerns’ in larger discussions on foreign policy. DFID has since come to be recognized as the premier bilateral development agency by development commentators around the world (e.g., Richard and Rupp, 2009). In the United States, the Millennium Challenge Corporation (MCC) was created in 2004 alongside the more established Agency for International Development. According to its charter legislation, the MCC gives aid to poor countries committed to “just and democratic governance . . . economic freedom . . . [and] investments in the people” based on “objective and quantifiable indicators.”8 In practice, this has meant selecting countries by income category according to 17 published indicators, such as the advocacy nongovernmental organization (NGO) Freedom House’s indicator on political rights. Income-eligible countries receive a score, and (in a slight oversimplification) those above the line are theoretically eligible, while those below the line are not. This system has been subverted on at least one occasion, granting a $295 million aid package to Georgia – even as the

8

country was below the bar on corruption and several other indicators (Phillips, 2006). Compared to the politics of other aid flows, however, such a transgression appears quite minor, and the fact that it brought so much negative attention is a testament to the relative success of the MCC. Yet another way to insulate foreign aid allocation decisions from international and domestic politics goes relatively unnoticed. The tax laws of donor countries such as the United States allow for tax-deductible donations to NGOs, even as these entities may spend their charity abroad. The foregone tax here is essentially a contribution of the donor government to the activities of the NGO abroad, those activities having been decided outside of government channels. (Governments fund NGOs abroad as well, often hiring them to implement foreign aid activities in the national interest, but NGOs are free to spend their privately raised donations where they like.) Economists have examined whether NGO aid disbursement does any better at responding to humanitarian or developmental needs, rather than the political forces of the country in which they are based. Available accounts, using a similar methodology as Alesina and Dollar (2000), find mixed results on whether NGO aid disbursement is more correlated with humanitarian and developmental variables than is official foreign aid (e.g., Nancy and Yontcheva, 2006; Nunnenkamp et al., 2008). Moreover, utilizing a decentralized approach to development assistance, while having some advantages, introduces a new set of problems (Werker and Ahmed, 2008). Immunizing foreign aid from donor-driven political economy distortions is by no means an easy task, nor has it been successfully accomplished through any particular institutional innovation.

Saving Aid from Itself, or, Taking the Politics Out of Aid Receipt and Disbursement At the level of the recipient, there have been a number of institutional innovations to minimize the negative political economy from aid disbursement. Recall that some of the negative political economy effects were conflict over resources (including corruption), incumbency advantage (including geographic distribution of funds), insulation from citizenry through not having to raise revenue domestically, and the reduced competitiveness of the tradable sector. Individually, these innovations may circumvent particular negative effects of aid on the recipient country’s political economy. Collectively, however, their success is far less obvious.

‘Millennium Challenge Act of 2003,’ http://www.mcc.gov/mcc/bm.doc/mca_legislation.pdf, accessed 04/15/10.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

55

CONCLUSION

To minimize the chance that aid will be siphoned off by corrupt officials, donors often set up so-called project implementation units (PIUs). These are essentially structures parallel to the recipient government with rigorous procurement, financial management, and reporting controls. By providing tighter seals around aid projects, PIUs can reduce the potential for aid to be captured by the domestic elite for their own benefits, and they are widely used for this purpose (Asian Development Bank, 2005). To reduce the incumbency advantage of the acting government, and to prevent the geographical targeting of aid for political reasons, aid agencies have sought to ‘decentralize’ the provision of aid (e.g., USAID, 2000). Decentralization involves a combination of promoting subnational governance structures, local elections, and local management of natural resources. For the aid practitioner, it may mean offering technical assistance to the state and local government or planning and funding development projects in collaboration with the local government. It may also involve lobbying for more local control over the rents from natural resource extraction. In this way, aid can respond more closely to the needs of the citizens than the political strategies of their national governments. To increase the state–citizen obligations that may be reduced through foreign aid substituting for domestic taxation, aid agencies generally require extensive public consultation. To make sure that aid projects represent the desires of affected populations, aid agencies and the government entities they are working with hold a number of ‘consultative processes’ along the way. These normally entail holding meetings in affected communities, running workshops with representatives of civil society, and participating in media campaigns. Through these activities, input can be gathered from those ‘stakeholders’ who would be affected by the project and ideally incorporated into the project planning. Reducing the likelihood of Dutch disease from aid inflows is a challenging matter (IMF, 2005). If aid involves a real transfer to the recipient country, it must involve an increase in net imports to the country. Those imports can, simplistically, either be spent directly on goods or services, or indirectly through increased market activity and incomes, which stimulates demand for imports from the population. When import increases occur through an increase in demand, most likely at the same time (or, indeed, as the driving factor) there will have been an increased demand for nontradables – resulting in a price increase. Such is the cause of Dutch disease. So, if aid is spent directly on imports (i.e., it is given in kind), then it will have a smaller effect on the real exchange rate. Of course, depending on what is imported, such aid could still reduce the demand for domestically produced manufactured goods.

Perhaps the most sweeping intervention by donors to control any undesired political economy effects of foreign aid is to impose any number of conditionalities on the aid. These may take the form of granting aid, loans, or debt relief to a recipient country only after it satisfies a number of conditions. Those conditions might be broadly reaching, such as the free-market ‘Washington consensus’ policy recommendations favored by the structural adjustment programs, which reached their zenith in the 1980s. Or they might be more subtle, for example, requiring that the operations of a port be privatized before a loan is granted to refurbish the port. While structural adjustment programs are out of fashion, in favor of a ‘country ownership’ approach (World Bank, 2005), the presence of some conditionalities on foreign aid is still nearly ubiquitous. (Enforcing conditionalities, and other interventions to fix the negative political effects of aid receipt, may also be affected by the politics of aid disbursement; Kilby (2009) finds that the conditionalities associated with World Bank aid to US allies are more laxly enforced.) Thus, there are a host of strong, theoretically grounded policies that can be implemented to reduce individual political economy distortions inside the recipient country. Aid can be granted, in large quantities, and disbursed without being stolen, in projects that help the ‘regular’ people, even far from the capital. Yet, consider them as a package: a host of independent-fromgovernment PIUs handling the money; an emphasis on subnational government; countless streams of donordriven consultative processes; goods and services imported in kind; and requirements on how funds are spent. This portfolio of interventions to reduce the distortions of aid ends up with the potential to sideline the national government while creating a self-perpetuating, parallel, and big-spending government-by-donors. Of course, many countries that fit this description have also benefited from consistent growth, a stable macroeconomy, and reduced poverty, especially since the mid1990s. And the most competent leaders of less-developed countries are able to navigate the ills and the fixes of aid politics better than the foreign assistance bureaucrats themselves, using the funds as well as the associated restrictions to further their particular agenda. Whether the cure becomes worse than the disease is an open question, and much of the public debate on aid itself covers this question without being explicitly conscious of it.

CONCLUSION In analyzing the political economy of bilateral foreign aid, this chapter has explored the distortions present in aid allocation and spending, as well as the available fixes. As a profession, we economists know best the effects of

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

56

5. POLITICAL ECONOMY OF FOREIGN AID, BILATERAL

political forces in the donor country on the disbursement of aid. We have established that politics drive aid allocation decisions and by how much. We have examined the potential fixes to bilateral politics driving aid decisions and found that some of these fixes – most principally, the international financial institutions – come up short in insulating the aid process from donor-country politics. That said, we should not be so quick to jump to the conclusion that aid-for-development is doomed. We have been less successful in showing that politically driven aid entails worse outcomes for the recipients. Most politically driven aid still has a major developmental quotient, particularly today when donors see a prosperous and democratic Third World in their national interest. Moreover, there are a number of fora in which aid given for developmental reasons is in the national interest, particularly as it reflects the values of the taxpayers in the donor country. With national interests and values aligned with aid-for-development, donors are better able to insulate their own aid from short-term political pressures. It is in the increasing incidence of these situations that will be the most successful counter to the Morgenthau critique, of aid-for-development being a mere play, staffed with B-team actors. When it comes to the political economy of aid disbursement and spending in the recipient country, this chapter has argued that the potential negative effect of politics is much greater. Whereas the politics in aid allocation are consistent with the national interest of the donor at least, the politics in aid receipt are easily corruptive to the political equilibrium of the recipient country. Aid can disturb the balance, tipping the politics toward more conflict, more incumbency advantage, more strategic patronage networks, and away from manufactures. Economists understand these dynamics, but not as well as (and with much less attention than) the simpler and more benign politics of aid allocation. For better or for worse, there are a number of fixes available to mitigate the political economy distortions of aid receipt and implementation. Unlike with aid allocation, these fixes are used extensively – because the developmentally focused actors, the aid bureaucrats (and not the politicians), now control the purse strings. At their worst, these fixes can render impotent the national government in recipient countries, and its role in creating a healthy polity – replacing the usual governance relations with a heavy-handed development industry governed by analysis and not by politics. At their best, the donor fixes may better identify the needs of the recipient populations and better devise and monitor programs in their interest, meanwhile training a responsible generation of policy makers in developing countries who will have the ethos and experience to become responsible and competent leaders.

References Alesina, A., Dollar, D., 2000. Who gives aid to whom and why? Journal of Economic Growth 5 (1), 33–63. Alesina, A., Weder, B., 2002. Do corrupt governments receive less foreign aid? American Economic Review 92 (4), 1126–1137. Andersen, T.B., Hansen, H., Markussen, T., 2006. US politics and World Bank IDA-lending. Journal of Development Studies 42 (5), 772–794. Asian Development Bank, 2005. The Role of Project Implementation Units. ADB, Operations Evaluation Department, Manila. Barro, R.J., Lee, J.-.W., 2005. IMF programs: who is chosen and what are the effects? Journal of Monetary Economics 52, 1245–1269. Cooley, A., 2009. How the U.S. lost its Kyrgyzstan air base. The New York Times. February 10. Devlin, L., 2007. Chief of Station, Congo: Fighting the Cold War in a Hot Zone. Public Affairs, New York. Dreher, A., Klassen, S., Vreeland, J., Werker, E., 2010. The costs of favoritism: is politically-driven aid less effective? Courant Research Centre Discussion Papers No. 26, March. Dreher, A., Sturm, J.-.E., Vreeland, J.R., 2009a. Development aid and international politics: does membership on the UN Security Council influence World Bank decisions? Journal of Development Economics 88 (1), 1–18. Dreher, A., Sturm, J.E., Vreeland, J.R., 2009b. Global horse trading: IMF loans for votes in the United Nations Security Council. European Economic Review http://dx.doi.org/10.1016/j.euroecorev.2009.03.002. Faye, M., Niehaus, P., 2010. Political Aid Cycles. Mimeo. UCSD. Fratianni, M., Pattison, J., 2005. Who is running the IMF: critical shareholders or the staff? In: de Gijsel, P., Schenk, H. (Eds.), Multidisciplinary Economics: The Birth of a New Economics Faculty in the Netherlands. Springer, Berlin, pp. 279–292. Grossman, H.I., 1992. Foreign aid and insurrection. Defence Economics 3 (4), 275–288. Hanlon, J., 1996. Strangling Mozambique: International Monetary Fund ‘stabilization’ in the world’s poorest country. Multinational Monitor 17 (7) Archived edition: http://www.multinationalmonitor. org/hyper/mm0796.00.htm. Howell, J., 2006. The global war on terror, development, and civil society. Journal of International Development 18 (1), 121–135. IMF, 2005. The Macroeconomics of Managing Increased Aid Inflows: Experiences of Low-Income Countries and Policy Implications. IMF, Policy Development and Review Department, Washington, DC. Kaja, A., Werker, E., 2010. Corporate governance at the World Bank and the dilemma of global governance. World Bank Economic Review 24, 171–198. Kaplan, R., 2003. Surrender or Starve: Travels in Ethiopia, Sudan, Somalia, and Eritrea. Vintage, New York. Kepel, G., 2002. Jihad: The Trail of Political Islam. Harvard University Press, Cambridge, MA (translated by Anthony F. Roberts). Kilby, C., 2009. The political economy of conditionality: an empirical analysis of World Bank loan disbursements. Journal of Development Economics 89 (1), 51–61. Koring, P., 2010. Kyrgyzstani government ousted in violent uprising. The Globe and Mail. April 7. Krueger, A., 2007. What Makes a Terrorist: Economics and the Root of Terrorism. Princeton University Press, Princeton, NJ. Kunz, D., 1997. The Marshall Plan reconsidered. Foreign Affairs May/ June, 162–170. Kuziemko, I., Werker, E., 2006. How much is a seat on the U.N. Security Council worth? Foreign aid and bribery at the United Nations. Journal of Political Economy 114 (5), 905–930. Mahdavy, H., 1970. The patterns and problems of economic development in rentier states: the case of Iran. In: Cook, M.A. (Ed.), Studies in the Economic History of the Middle East from the Rise of Islam to the Present Day. Oxford University Press, Oxford.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

CONCLUSION

Marshall, G., 1947. The ‘Marshall Plan’ speech at Harvard University. 5 June. Text at http://www.oecd.org/document/10/0,3343, en_2649_201185_1876938_1_1_1_1,00.html. Miller, J., 1983. Taking off the gloves: the United States and the Italian election of 1948. Diplomatic History 7 (1), 35–56. Moore, M., 1998. Death without taxes: democracy, state capacity, and aid dependence in the fourth world. In: Robinson, M., White, G. (Eds.), The Democratic Developmental State: Politics and Institutional Design. Oxford University Press, Oxford. Morgenthau, H., 1962. A political theory of foreign aid. American Political Science Review 56 (2), 301–309. Moyo, D., 2009. Dead Aid: Why Aid Is Not Working and How There Is Another Way for Africa. Penguin, New York. Nancy, G., Yontcheva, B., 2006. Does NGO aid go to the poor? Empirical evidence from Europe. IMF Working Paper No. 06/39. Nelson, S., 2009. Creating Credibility: The International Monetary Fund and the Neoliberal Revolution in the Developing World. PhD dissertation. Department of Government, Cornell University. Neumayer, E., 2003. What factors determine the allocation of aid by Arab countries and multilateral agencies? Journal of Development Studies 39 (4), 134–147. New York Times, 2010. President Karzai lashes out. Editorial, April 3, p. A16. Nichol, J., 2009. Kyrgyzstan: recent developments and U.S. interests. Congressional Research Service. August 9. North, D.C., Weingast, B.R., 1989. Constitutions and commitment: the evolution of institutions governing public choice in seventeenthcentury England. Journal of Economic History 49 (4), 803–832. Nunnenkamp, P., Weingarth, J., and Weisser, J., 2008. Is NGO aid not so different after all? Comparing the allocation of Swiss aid by private and official donors. Kiehl Working Papers No. 1405, March. Perlez, J., 1990. U.S. bars new aid in a Sudan famine. New York Times. October 5. Peterson, S., 2000. Me Against My Brother. Routledge, New York. Phillips, M., 2006. Controlling carrots and sticks: White House tries to rein in foreign-aid agency it recently created. Wall Street Journal A4. Rajan, R., Subramanian, A., 2011. Aid, Dutch disease, and manufacturing growth. Journal of Development Economics 94 (1), 106–118. Richard, A., Rupp, G., 2009. The ‘DFID model’: lessons for the U.S. The Wave: Currents in Transatlantic Policy. Center for Transatlantic Relations, Johns Hopkins University – SAIS, Washington, DC. Rodrik, D., 2008. The real exchange rate and economic growth. Brookings Papers on Economic Activity 2.

57

Roodman, D., 2007. macro aid effectiveness research: a guide for the perplexed. Center for Global Development. Working Paper No. 134. Ruttan, V.W., 1996. United States Development Assistance Policy: The Domestic Politics of Foreign Economic Aid. The Johns Hopkins University Press, Baltimore. Schrader, P., Hook, S., Taylor, B., 1998. Clarifying the foreign aid puzzle: a comparison of American, Japanese, French, and Swedish aid flows. World Politics 50 (2), 294–323. Svensson, J., 2000. Foreign aid and rent-seeking. Journal of International Economics 51 (2), 437–461. Tavares, J., 2003. Does foreign aid corrupt? Economics Letters 79, 99–106. Thomson, B., 2007. Sierra Leone: Reform or Relapse? Conflict and Government Reform. Chatham House, London. Tilly, C., 1992. Coercion, Capital and European States, AD 990–1992. Blackwell, Cambridge, MA. Tornell, A., Lane, P.R., 1999. The voracity effect. American Economic Review 89 (1), 22–46. Transparency International, 2004. Global Corruption Report 2004. Transparency International, Berlin.www.transparency.org. USAID, 2000. Decentralization and Democratic Local Governance Programming Handbook. US Agency for International Development, Center for Democracy and Governance, Washington, DC. Uvin, P., 1998. Aiding Violence: The Development Enterprise in Rwanda. Kumarian Press, West Hartford, CT. Walsh, N., 2005. Uzbekistan kicks US out of military base. The Guardian. August 1. Werker, E., Ahmed, F., 2008. What do non-governmental organizations do? Journal of Economic Perspectives. 22 (2), 73–92. Werker, E., Ahmed, F., Cohen, C., 2009. How is foreign aid spent? Evidence from a natural experiment. American Economic Journal: Macroeconomics 1 (2), 225–244. Woods, N., 2003. The United States and the International Financial Institutions: power and influence within the World Bank and the IMF. In: Foot, R., McFarlane, S.N., Mastanduno, M. (Eds.), US Hegemony and International Organizations. Oxford University Press, Oxford, pp. 92–114. World Bank, 2005. In: Koeberle, S., Silarszky, P., Verheyen, G. (Eds.), Conditionality Revisited: Concepts, Experiences, and Lessons. World Bank, Washington, DC. Zhang, G., 2004. The determinants of foreign aid allocation across China: the case of World Bank loans. Asian Survey 44 (5), 691–710.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

Intentionally left as blank

C H A P T E R

6 Interest Group Politics E.J. Malesky Duke University, Durham, NC, USA O U T L I N E Political Economy Models of Economic Integration

60

Interest Groups Domestic Capital (Business Owners) Labor Land Owners Domestic Financial Intermediaries

61 61 61 62 62

Distributional Implications of Financial Globalization

62

The impact of financial globalization has been dramatic, connecting wider swathes of countries and individuals, generating employment, and raising incomes around the world. Nevertheless, the process of financial globalization is almost never uniform in its impact on the various stakeholders in an economy. Certain interest groups are more likely to benefit (lose) from changes in globalization of capital than others. The creation of winners and losers from global capital creates interest groups with conflicting interests. Such conflict of interest can explain the frequent reluctance of countries to adopt changes that are otherwise beneficial for the aggregate economy. These issues have generated an enormous political economy literature and delivered important insights into the motivations for countries to open their economies to global financial flows, and the postopening distributional impact that such flows can have. A welldeveloped theoretical literature has provided useful forecasting tools, as to which groups will be benefited or injured by various flows. Using this information, predictions can also be made about the policies that will be favored by particular actors. And if the composition of such actors in a society is known, the eventual policy outcomes can be predicted. Empirically, the straightforward

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00004-0

The Political Economy of Financial Globalization in Authoritarian Regimes

64

MNCs as Actors

65

Conclusion Glossary Further Reading

67 67 68

predictions of the formal models have not always been borne out, but this does not mean globalization is unimportant. Careful research has shown that the impact of financial globalization on interest groups is not direct; rather, it is mediated by endowments, level of development, political institutions, and the preferences of leaders. Nevertheless, there is much room for advancement in this important literature. Very little of the work directly explores the impact of globalization on actual politics, which certainly includes electoral results, but could easily be expanded to include lobbying, fundraising, and other forms of political behavior. Inferences can be made about these political actions from the literature, of course, but the primary outcomes explored in the literature have generally been about policy choice and the interests that underlie those choices. A second complication of the literature is that a comprehensive theory of economic globalization that encompasses the multiple economic flows taking place between countries is yet to be developed. Some authors prioritize international trade, portfolio equity flows, foreign direct investment (FDI), or debt flows in their analysis (the latter three are generally thought to comprise financial globalization), and analyze the distributional

59

# 2013 Elsevier Inc. All rights reserved.

60

6. INTEREST GROUP POLITICS

effects of these flows. Other authors study the preferences of domestic actors on the policy choices that underlie these flows, such as exchange rate policy, capital account liberalization, or FDI regulations. But the different forms of globalization have differential effects on domestic actors. In some cases, these effects can be reinforcing, so that, for instance, greater exposure to FDI can exacerbate the distributional effects caused by greater openness to trade. In other cases, the effects can be countervailing. For example, competition from multinational corporations (MNCs) may threaten domestic entrepreneurs, but greater portfolio investment (especially with the rise of emerging market funds and private equity investors) offers opportunities for domestic firms to raise capital and compete with their larger foreign rivals. To date, the literature has not advanced enough to make sense of these overlapping and intersecting trends, when drawing conclusions about political processes. Part of this confusion has emerged in trying to understand the role of foreign direct investors (MNCs) in the domestic policy-making process of recipient nations. Work on capital account liberalization and exchange rate policy explicitly considers the role of MNCs as actors, in coalition with mobile domestic capital and exporters, arguing for greater capital account openness. On the other hand, most recent political economy work on FDI pays little attention to the political role of MNCs after their investment, preferring to study the impact of political institutions on the initial attraction of investment. This chapter starts by discussing the theoretical models that underpin predictions about financial globalization. Next, it introduces the critical interest groups in society that have been highlighted by the literature. Third, the key findings of the literature are explained, paying special attention to how the impact of globalization is often not direct, but depends on political institutions and endowments. Most work on the political impact of financial globalization privileges democracies, so the fourth section highlights the fascinating new work on political outcomes in nondemocratic societies. Finally, the piece highlights the unique role played by MNCs as additional actors in the policy process and consequently as ‘agents of change.’

POLITICAL ECONOMY MODELS OF ECONOMIC INTEGRATION Three economic models underpin existing political economy theory regarding the distributional effects of economic globalization. The first two originate from trade theory, but have been adapted to accommodate international capital flows as well.

First, the factoral Heckscher–Ohlin–Samuelson (HOS) posits that financial globalization should benefit the abundant factor in a country (land, labor, or capital), as returns rise absolutely and disproportionately to the owners of the factors that are used intensively in the production of goods from that country. As a result, when a country opens to international capital flows, one might expect competition between these various groups over appropriate policy. Political economists argue that HOS implies different effects on inequality for developed and developing countries. As wealthy countries are comparatively advantaged in terms of capital and high-skilled labor, financial globalization should benefit elites (business owners and the well-educated) at the expense of low-skilled labor. In developing countries, where less-skilled labor is abundant, HOS would predict that amelioration of inequality is due to international capital flows. An alternative trade model, Ricardo-Viner, that has been adapted to explore the implications of financial globalization, argues that it is not factor endowments which drive debates over financial globalization, but sectoral differences. In this model, factors specific to an industry or service bear the full weight of price changes in their unique product lines. A simple example is land used in sugar production. If land within a particular country is only suitable for growing sugarcane, the value of that land will be associated with changes in the price of sugar and not with agricultural products generally. Labor and capital specifically associated with sugarcane production will fair similarly. In a Ricardo-Viner framework, one expects globalization to generate conflict across sectors, but may actually unite owners of factors within those sectors. A large number of studies of the distributional effects of globalization test both models and find support for both. This is surprising because the two traditional trade models partly rely on contradictory assumptions about the level of factor mobility, which has shown to be quite limited, especially in regard to labor flows. A third workhorse model of the distributional effects of globalization has been developed more specifically for arguments about financial openness, particularly the establishment of exchange rate policy and capital account liberalization. These scholars distinguish between holders of mobile capital and diversified assets (e.g., ‘money center banks,’ MNCs, and large domestic conglomerates), who can benefit from the increased investment opportunities afforded by opening to the global economy. Due to their diversity and potential to shift resources abroad, mobile capital also tends to be more insulated from the potential for volatility brought by increased international exposure. Holders of mobile capital are often joined by producers of tradable goods in pursuit of national policies, such as floating exchange

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

INTEREST GROUPS

rates and capital account liberalization, which reduce the risk of and costs of currency volatility. Holders of immobile capital and nontradable sectors are likely to pose economic openness, which exposes them to increased competition and exposure to international economic forces, without the ability to shift resources to shield themselves.

INTEREST GROUPS While these models differ in their implications, they do provide a nice starting point as to where to look for evidence of distributional conflict over financial globalization. All three models highlight similar actors; where they differ is how they choose to disaggregate these basic groups. These actors include the following categories.

Domestic Capital (Business Owners) Domestic capital consists of companies and businesses of all sizes that operate within the borders of the host state. These may be privately held, publicly traded, or even state-owned companies. In the literature on financial globalization, the primary distinction drawn between domestic producers is whether they are producers of tradable (or exportable) or nontradable goods (services, construction, transport, and distribution). A secondary distinction is whether a producer competes in a market protected by trade restrictions. In general, producers of tradables tend to favor economic reforms that promote financial globalization because it affords them new opportunities to sell their goods abroad. Particularly important for exporters is capital account liberalization, which through easier currency convertibility allows more opportunities to hedge against exchange rate volatility and reduces the transaction costs of moving goods abroad. Producers of nontradables, and tradable producers who are protected by trade barriers, are thought to be less enthusiastic about financial globalization, which increases the opportunities for foreign competition that threatens their monopoly rents.

Labor These are the individuals who comprise the workforce in the domestic economy. They can be employed by either domestic or foreign operations. In general, individual laborers prefer economic policies that do not threaten their own employment while seeking higher incomes. More nuanced arguments about labor take into account their ability to inform themselves and hedge against economic risk.

61

Political economists primarily differentiate labor by their skill set. Unskilled laborers have relatively low education and little technical expertise. They have difficulty distinguishing themselves from others, so their wages are generally a function of their marginal productivity (determined by existing production technology) and the number of other laborers competing for the same position. Skilled laborers have more years of education and specific technical or management expertise. These workers face more limited competition and can demand higher wages from businesses. Most often skill set is operationalized as years of education or whether a worker is college-educated. As a result, some scholars have begun to question whether the opinions of skilled laborers regarding globalization are really a function of their skill set, or instead a result of exposure to economic theory in college. As with producers, laborers can also be distinguished by the sector in which they work, according to the Ricardo-Viner framework. Once again, scholars have hypothesized that laborers in nontradables or protected markets are more skeptical of the benefits of globalization. Laborers in sectors that benefit from increased competition for their employment are likely to see higher wages. On the other hand, laborers in industries not valued by foreign investment could lose their jobs or employment. A more sophisticated distinction has to do to with the production technology brought to a country by foreign investors. In the simplest terms, foreign investors may employ either labor-intensive or labor-saving technologies. Labor-saving technologies are often automated and require fewer workers per unit of output. A new contribution to the literature argues that one must further analyze whether labor is operating in industries that are complements or substitutes for the production technologies of foreign investment. In places where foreign investment technology complements labor, it will be supportive, but in cases where foreign investment technology substitutes for labor, labor will be against the investment, while domestic businesses will benefit. Other authors accept this theoretical determination of labor preference, but note that abundant labor and clear preferences do not necessarily lead to laborfriendly policy outcomes. One must also pay attention to how labor organizes to solve collective action problems. Countries with strong labor unions and tightly organized collective bargaining arrangements are more able to influence political outcomes than unorganized labor that lacks information and has difficulty demonstrating strength to policy-makers. With respect to trade policy it has been shown that outcomes do not reflect the position of the median voter, but are more influenced by organized lobbies, often in combination with partisan allies.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

62

6. INTEREST GROUP POLITICS

Land Owners Land owners are usually thought of as owning large plots of lands in agricultural economies that employ hired labor, but may also include real estate investors and small landholders in economies which have undergone land reform. In the HOS model explaining globalization preferences, land is considered a factor. In a Ricardo-Viner framework, one must pay closer attention to the type of product produced on the land. Landownership is most important in a mobile versus immobile capital theory of distributional effects; however, as land is the quintessential form of immobile capital, investors whose primary holdings are land cannot hope to ameliorate risk by moving capital abroad. For instance, distributional analyses of Malaysian politics over time emphasize the role of Bumiputras, ethnic Malays who generally earn their living off the land and are among the strongest advocates against open economic policies.

Domestic Financial Intermediaries These are the local banks and nonbank financial institutions which direct capital to domestic businesses. Scholars distinguish between two types of domestic financial institutions. Local banks and actors dependent upon them prefer financial rules that keep savings circulating in local credit markets and therefore are resistant to many aspects of financial globalization. ‘Money Center Banks,’ defined as banks that tap savings from the whole country, and their predominantly corporate clients, however, favor financial globalization because it provides them with a larger pool of assets and enhances their ability to match saving with assets.

DISTRIBUTIONAL IMPLICATIONS OF FINANCIAL GLOBALIZATION Depending on the type of financial globalization under investigation, different coalitions of the above actors can arise to support or challenge state economic policies. Political economy scholars have rarely tried to develop a general model of support for globalization, rather they have looked to see the constellation of actors that forms around specific policy issues. In general, political economy literature has been less successful at establishing robust empirical relationships between exposure to financial flows and distributional outcomes than international trade literature. One debate has been whether exposure to international financial flows leads to greater inequality. For instance, increased financial globalization was expected to increase inequality in the developed world and decrease it in the developing world. Instead, the literature has found little

evidence of globalization increasing inequality within countries over time. Indeed, some scholars have found microevidence that the distributional effects of globalization were off the mark. In developing countries, it is high-skilled labor that appears most supportive of globalization, not low-skilled labor as an HOS approach would predict. In panel analyses of countries exposed to financial openness, the most robust finding is that inequality differs more across countries than over time, indicating increased exposure to the international economy that has little to do with the distributional effects. Where inequality has increased in recent years, skillbiased technological change seems to be the greater culprit than financial globalization. What has been more important has been how the distributional effects of globalization have been mediated by the strength of organized labor, political institutions, and partisan political dynamics. In the developing world specifically, it has been shown that initial distributions of land and education drive inequality growth after economic integration. A second debate has taken place over what is known as the compensation hypothesis, whereby winners from globalization may compensate the losers with greater social assistance programs. Sometimes this has been referred to as the ‘embedded liberalism’ of international economic policy, indicating that liberal economic policies are accompanied by a generous social welfare state reflecting a societal debate and subsequent social contract. If the compensation hypothesis holds up, this may help explain the lack of correlation between globalization and inequality, but it offers an interesting prediction in its own right. That is, increased financial globalization should be associated with larger welfare states. This has been shown to be the case in the small states of Western Europe, within the Organisation for Economic Co-operation and Development (OECD), and among developing countries. The supportive findings for the compensation hypothesis use the level of trade openness as their core measure of globalization; studies focusing on changes in financial globalization, however, show that openness is actually negatively correlated with changes in social welfare spending. Some contributions offer more nuanced analysis, demonstrating that one must consider the interaction between globalization and labor strength and organization. Where unions are strong and have an influential say in public policy, changes in global integration are actually positively associated with the size of the welfare state. On the other hand, some have argued that increased economic integration restricts the independent monetary authority of nation-states, thereby limiting the size of welfare states, hypothesizing that in highly globalized

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

DISTRIBUTIONAL IMPLICATIONS OF FINANCIAL GLOBALIZATION

countries, governments would be forced to cut taxes, and consequently spending, in order to maintain an attractive environment for foreign portfolio and equity investors. Two political economy theories help explain this constraint. First, foreign portfolio investors demand fiscal discipline. Second, opportunities to borrow in foreign capital markets dry up during economic downturns. A tremendous amount of work has explored this hypothesis empirically, finding little evidence of a negative correlation between capital openness and government spending. Indeed most work has actually identified a positive relationship between the two, further establishing the compensation argument. Evidence of the positive relationship has come under fire. First, critics argue that omitted variable bias is influencing the results, believing wealthier states are both more likely to have larger welfare states and to be open to international financial movements. Related to this point, a second stream of criticism argues that the compensation hypothesis that has been observed among OECD countries appears to be nonexistent among developing countries. In repeated studies, scholars have found limited empirical support for the claims that spending on health, education, and targeted antipoverty programs has grown more rapidly in integrated economies. Indeed, comprehensive analyses of social spending programs around the globe conclude that the compromise of embedded liberalism in the advanced industrial states that occurred after World War II has not been replicated in the social policies of Latin America, East Asia, and Eastern Europe. Since World War II, globalization has proceeded without enhanced compensation, and perhaps even with less compensation, in the global South. Attempting to explain the weakness of the compensation hypothesis, researchers have pointed out that a key intervening variable between financial globalization and the size of the welfare state is economic insecurity. Globalization must make citizens more concerned about their individual economic prospects, and they must be willing to express these views at the ballot box, before national governments would agree to compensation as mollification. Some analyses rely on economic indicators of volatility, showing that international price volatility is often less than domestic price volatility. Secondly, greater openness actually diversified the risk faced by individual actors, thus helping them hedge against economic security. Others rely on survey data, but come to very similar conclusions. Individual perceptions of risk, even among unskilled labor, tend to be reduced with increased economic integration. These finding are at odds with previous survey work that financial globalization (particularly FDI) increased economic insecurity by raising the elasticity of demand for labor in England, which raised the volatility of wages

63

and employment. A more recent analysis finds evidence for every stage of the compensation hypothesis (globalization to risk, risk to vote choice, and vote choice toward welfare state provision), based on Swiss panel and survey data. Why the contradictory results? One important reason may be where scholars are looking. Those finding support for insecurity induced by globalization study developed countries (particularly, the United Kingdom and Switzerland), while those finding no evidence of insecurity study developing country data. The different research arenas may explain some of the differences, but the drastically different conclusions also warrant deeper thinking about the theory. A new literature has started to move beyond survey data, and test voting behavior of individuals directly to see if it conforms to the compensation hypothesis. This work, however, has also provided extremely ambiguous results. Studies of the 1996 US presidential election and of the 1998 election to the Australian Federal House of Representatives find that job insecurity increases the likelihood that an individual will vote for parties that run on an antiglobalization platform. At the same time, however, other scholars find that economic internationalization actually leads voters to care less about the economic performance of leaders, assuming that they have little control over international forces. Moving away from the distributional effects of globalization, other scholars focus on policy choice. If the strength of various actors in a polity is known, perhaps the particular policies that political elites might choose can be forecast. This has been best explored in terms of capital account liberalization and foreign investment regulation. Here again, scholars have found limited evidence of a direct effect of particular sectoral interests on policy choices. The first cut at the political economy literature was simply to see whether exposure to interest groups at all was more likely to lead to greater capital account liberalization. Because authoritarian regimes tend to be more insulated from public pressure, one might think it would be easier to achieve such policies in the authoritarian setting. Others claim that authoritarian regimes are predominantly captured by elites in the domestic economy, and economic policy therefore tends to follow the interests of those groups. In democracies, however, there is more opportunity for those who might benefit from globalization to assert themselves on the policymaking process. Recent work has found a positive correlation between democracy and capital account openness. The relationship between democracy and capital account openness is interesting, but also confusing. Democracies contain a wide variation in economic structures, constituent preferences, and constellations of powerful actors. Which actors exactly, actually benefit from greater liberalization? Direct testing of the three

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

64

6. INTEREST GROUP POLITICS

workhorse models of economic preferences has failed to provide robust empirical support. The lack of direct empirical support for interest group preferences and financial openness policies outcomes stands in stark contrast to trade liberalization. As a result, most scholarship has tended to focus on the negotiations of elite actors. The divergence between analyses of trade and financial openness has led some to conclude that because capital account liberalization involves opaque and confusing changes, but has the potential to create severe economic risk, the political strategy is one of blame avoidance. As a result, polities where political authority is fragmented and multiple parties are involved in decision-making are most likely to pursue extensive liberalization. As with the inequality debates, the real advances in scholarship have not come from the direct tests of economic models on openness to global financial flows. Rather, the most robust findings are those that show how preference for capital account openness is mediated by partisanship, political institutions, and the beliefs of political leaders.

THE POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION IN AUTHORITARIAN REGIMES Ironically, the most exciting advances in understanding financial globalization and political interests have taken place in authoritarian regimes. The research agenda was initiated by the suggestion that financial openness increases the likelihood of currency crises in authoritarian settings. These crises may in turn lead to rebellion against dictators, pushing them toward democratization. The analysis was based heavily on anecdotal reading of the collapse of the Soeharto regime, but his hypothesis piqued the curiosity of other scholars. Scholars have now refined the theoretical predictions by explaining not only the fall of Soeharto, but the stability of Mahathir in Malaysia. This new research agenda relies unapologetically on the theory of mobile and immobile capital proposed earlier, but rather than looking for evidence in electoral relationships, the theory is employed to understand the underlying coalitions of the two dictators. According to this line of thinking, Soeharto was unable to develop a coherent response to the crises, as he was constantly being pulled between Chinese business groups with extensive holdings of mobile capital, and military-linked firms and indigenous entrepreneurs, whose assets were rooted in Indonesia. Chinese conglomerates pushed hard for capital openness, whereas military and local entrepreneurs wanted a closed capital account to bolster domestic spending. Mahathir’s regime ultimately survived because the

support base for the ruling coalition, Barisian Nasional, are the ethnic Malay masses, who favor ethnic-based affirmative action and distributional public spending, and ethnic Malay entrepreneurs who earn money from fixed capital assets. Both the groups favored closed capital accounts and expansionary fiscal policy, allowing Mahathir to identify the appropriate monetary and fiscal medicine necessary to hold his coalition together and retain power. Soeharto’s fall and Mahathir’s resilience can be explained by shifting alliances within the base of support that result from debates over economic policy. Understanding the financial preferences of the support coalitions gives additional leverage over alternative explanations for authoritarian collapse, and the fall of the Indonesian New Order specifically, because it allows nuanced predictions about both the timing and manner of the downfall. Other explanations, such as declining legitimacy for the Soeharto regime or institutional discussions, are unable to account for swings in Soeharto’s popularity or the precise moment of his resignation in May 1998. By contrast, scholars analyzing the regime through the lens of mobile versus immobile capital show how the underlying coalition of interests responded to each new policy, demonstrating how various initiatives severed or reinforced the coalition. The eventual downfall came when anti-Chinese riots broke out and the Indonesian military failed to defend leading Chinese entrepreneurs. Chinese-Indonesians fled, taking their capital and support for the regime with them. Analysis does not stop with the study of two regimes in Southeast Asia. Quantitative cross-national analysis finds that proxies for coalitional support – capital account openness and change in capital account openness over the course of the crisis – are strongly correlated with regime breakdown. The histories of several Latin American countries (Chile, Argentina, Mexico, and Uruguay) provide additional support. Other scholars have explored how financial globalization has undermined the authority of a particular type of regime, the single-party dominant system, where regular elections are held, but the same party always manages to win. Single-party dominant systems have existed in both democratic settings (such as Japan and Italy), where elections are accompanied by executive constraints and independent judiciaries, and authoritarian regimes, where elections were the sole source of democratic accountability and even these were heavily undermined by ex-ante electoral engineering, coercion, and ex-ante electoral fraud. In both cases, scholars have found that exposure to international financial flows weakens the electoral grip of the dominant party. In Japan, Italy, and Mexico the hegemonic party, after decades in power, was toppled in elections experiencing some openness to international economic forces.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

MNCs AS ACTORS

Looking at the electoral results, scholars have found that the sectoral and regional predictions about the distributional effects of elections were consistent with support for and against the regime. Scholars of Japan explore increasing trade openness. In Italy openness is operationalized as the effect of joining the European Monetary Union. And, in Mexico, scholars show that the most internationally integrated regions (in terms of FDI attraction) were most likely to defect to the opposition from the Institutional Revolutionary Party (PRI), as they were less reliant on PRI controlled transfers.

MNCs AS ACTORS One particular actor deserves special attention because the role of this actor in the policy-making process is not well understood and controversial. As mentioned above, scholars attempting to understand capital account liberalization specifically look to MNCs as members of a proreform coalition in favor of greater reform. According to these theories, the more reliant a country is on trade and the more important FDI is to it, the more open its capital account. While it is easy to understand why MNCs may favor more openness, viewing them as political actors is more complicated. First, foreign investors do not have a vote and cannot run for office, so some channels of political influence are not open to them. More problematic than their parameterized modes of political influence, however, has been the widespread deference of scholars of MNCs to the obsolescing bargain theory (OBT) by assuming that investors have limited influence on the investment environment ex post, because the sunk costs associated with start-up shift the bargaining position to the host government. For example, a petroleum firm that has already committed substantial resources to site inspection and research or a manufacturing firm that has already hired staff and imported production equipment is less likely to walk away than a firm that has yet to commit financial resources to a particular project. This is not to say that foreign investors lose all bargaining power; the argument is that through strategic interaction their power is ex ante. To attract foreign investment, developing and transition countries need to make the institutional changes that investors desire and offer a credible commitment that these institutions will remain in place; otherwise they will lose to their neighbors in the investment attraction race. All bargaining over the regulatory regime, policy changes, and institutional structures is essentially concluded before investors even break ground. OBT was reinforced by early business strategy studies that concluded MNCs had no opportunities to influence policies of host governments. Political factors and

65

economic reform policies, therefore, came to be seen as just another variable to be added to the long lists in analysts’ existing economic models of FDI determinants. In contrast to OBT, when an earlier generation of scholars looked into the interaction of FDI and developing states in the 1960s and 1970s, they were explicitly interested in how FDI impacted local economic policy. Though their work was highly influenced by research on natural resource extracting industries, they took seriously the notion that foreign investors could influence political institutions of the host country, primarily because of the bargaining advantage that MNCs possessed over very poor and isolated developing countries due to their economic strength and access to external resources. For a while these arguments provided fuel for the raging dependencia fires because they confirmed developing countries’ and dependentistas’ fears that industrialized countries manipulated local institutions to exploit them for raw materials. This view of investors as agents of change spent several decades in the wilderness outside of mainstream IPE scholarship due to the predominant theoretical paradigm that investors needed credible commitments from host states. Without such iron-clad contracts, any investment deal with a developing nation would obsolesce, as local leaders began to incrementally rescind commitments. The term ‘obsolescing bargain’ was coined to describe this phenomenon of deteriorating investment arrangements with a host country. Because of the substantial costs of ex-ante contract infringement on large issues like expropriation and smaller regulatory and tax issues, foreign investors were much less likely to invest in sizable amounts where the risk of obsolescence and policy uncertainty was at a maximum. As a result, MNCs relied on overt signals of a country’s ex-ante credibility, such as whether it was a democracy or had plausible checks on the powers of key decision-makers. Seeing investors as agents in the economic process has recently begun to make a comeback, however. Strategy analysts have long understood that no contract is ironclad and that the political risk faced by MNCs has to be actively managed. Consequently, the strategy field has devoted great attention to how MNCs can mitigate host–government bargaining advantages through strategic alliances with local partners, broad alliances with other investors, staged entry and incentive alignment with local actors, the use of home governments and international organizations to spearhead discussions and the tactical allocation of proprietary technology and international finance. The general consensus in the business strategy literature is that the MNCs have become quite adept at functioning politically within host countries and that the obsolescing bargain is of limited utility today as a theoretical framework. According to many investors cited by the strategy literature, political risk is

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

66

6. INTEREST GROUP POLITICS

about relationship building – articulating your views and cultivating ties with people who share your goals. The consensus from the business strategy scholars has now begun to infiltrate the political economy literature in the wake of survey evidence that foreign investors believe that they have influence over economic policy in host states, detailed case study work of the specific mechanisms used to achieve policy change, chronicling of the strategic lobbying alliances MNCs forge with state-owned enterprises and subnational governments, and quantitative analysis isolating the impact of exogenous changes of FDI stocks and domestic economic reform choices. How do investors influence the political process once they are there? One rich area for research has been the wide variety of successful lobbying efforts by MNCs, including Fiat, Daewoo, and General Motors, which took stakes in Eastern Europe in 1991, prior to any serious progress on economic reform efforts. Along the same lines, when Volkswagen decided to buy Czechoslovakia’s Skoda car plant in 1992, there was no law on privatization, so legal principles on privatization and FDI regulation were developed in consultation between the firm and government. The lesson from Eastern Europe has been that most companies that have committed resources to transition states understood before embarking on their investment that the legislative and legal systems were inadequate. They were more interested in capitalizing on existing opportunities and believed their specific concerns would be addressed through subsequent discussions with policy-makers. Four distinct mechanisms have been identified for how investors, lured to a country by structural and market conditions, serve as agents for economic policy reform. These mechanisms closely track the causal logics described by the literature on policy diffusion across national borders. First, investors are important providers of information to transition country politicians. No economic policy reform can ever be completed in one sweeping action; it is a process of continual revision and adjustment. Each piece of commercial legislation must be revised as it proves inadequate to deal with the progressively more complex business environment. Because of their experience in other environments, foreign investors are critical in identifying these shortcomings and communicating them to policy makers and bureaucrats. Laws often lag behind the actual transactions taking place, such as the use of derivatives in the Czech Republic, and it falls to foreign investors to explain these complex mechanisms and the need for new laws to address them. Empirical evidence exists that such a process took place with the spread of ISO 14001 Management Systems across developing countries. Along these lines, foreign investors can offer critical information to inexperienced legislators about similar

laws in other countries and impart basic lessons on the day-to-day functioning of a market economy. To ensure that the foreign enterprise will continue its contributions to the country over the long term, through reinvestment or technology upgrading, it is likely that national governments will adopt a cooperative stance and be responsive to a foreign firm’s opinion about any new regulation affecting its business conditions. Certainly, policy makers will not see investor-provided information as unvarnished. They will understand that MNCs have their own agendas, as do all advocates, and will weigh their policy suggestions against alternatives provided by other domestic actors. What distinguishes countries with substantial FDI from those without, however, is that policy-makers hear and consider the MNCs’s information at all. Second, if politicians do not willingly seek out the advice of investors in these areas, investors may be forced to lobby for the changes in legislation that will affect their operations. This lobbying process has been modeled formally, demonstrating that the entry of foreign investment into an economy eventually leads to greater international openness. Foreign lobbying efforts often are abetted by alliances with domestic investors or regional governments who may have better access to policy makers than does the foreign investor. Third, existing investors can use coercion. Moderate forms of coercion include offering access to companyspecific financial and technical endowments, local charitable giving, and preferential access to proprietary technology. Less moderate coercion involves threatening to leave if legal and institutional changes are not made. In addition to the immediate loss of revenue and employment, investors closing up shop send a powerful signal to future investors interested in the country. This third mechanism can be costly for firms, which is why scholars argue that firms wait for diminished uncertainty before investing. On the other hand, it is the high cost of walking away from a large project that makes the signal to other investors credible. Not only can the threat of exiting investors provide an incentive for new reforms, but their bargaining power also provides a check against backsliding of reform. Governments are less likely to backtrack on reform initiatives when they risk upsetting a powerful group of investors. Testing this theory, the literature has found that the presence of a large stock of foreign investment leads to significantly fewer reversals of privatization projects. Fourth, the economic might of foreign investors can help free captured local politicians from the grips of entrenched interests, such as state-owned enterprises or Russian oligarchs. When there is a large presence of foreign investors, their combined power on key issues offers an external check on entrenched interests. Business environment survey data has been used to prove

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

67

CONCLUSION

this hypothesis, finding that foreign firms are more likely to believe they can influence the local regulatory environment of their host country than domestic firms. Perceived influence of foreign investment enterprises (FIEs) has been found to be highest in Eastern Europe and Central Asia, where regulatory institutions are most malleable. Addressing endogeneity bias through an instrumental variables analysis, recent research has provided further evidence for the fourth mechanism by demonstrating the impact of exogenously determined FDI on economic reform policies in transition states. This research finds persuasive evidence that stocks of FDI are positively correlated with reform progress in subsequent years. In short, OBT does not preclude the ability of existing investors to lobby for and succeed in shaping reform choices and new regulations that are advantageous to them. Viewing foreign investors as political actors also helps provide one plausible pathway for the well-documented empirical pattern of regulatory policy diffusion across similarly situated countries: states develop very similar economic regulations despite vastly different endowments and starting points. The literature on diffusion has highlighted coercion, competition, learning, and emulation as the four causal mechanisms undergirding policy diffusion. MNCs play critical roles as actors within all four mechanisms. States compete over the revenue and employment opportunities they provide, and can be coerced when preexisting investments threaten to leave for more welcoming policy regimes. MNCs also drive learning and emulation, by providing immediate feedback to policy-makers on issues under consideration and relating their experiences with policies in other countries.

CONCLUSION This chapter has demonstrated that there is a great deal of important work in the political economy of financial globalization. There are excellent theoretical models that provide clear predications about the distributional effects of increased openness. There is also a dense empirical literature that has been energized by spirited debate about core findings. While the core theoretical predictions of the theoretical models have not been upheld, there have been extremely important findings. The most important of them show quite clearly that institutions, endowments, ideology, and party systems help shape the direct effects of financial globalization. These results are critical, because they demonstrate that states are not solely at the whim of global forces. The institutions they construct help them take advantage of the

benefits of globalization, while ameliorating the political disruptive forces. Two new research agendas have been set in motion by the body of current scholarship. First, scholars are working to refine understanding by showing how different global flows interact and reinforce one another, so that the larger impact of the river of globalization is understood, and not only its independent tributaries. Second, scholars have begun to pay more attention to the political activities (elections, fundraising, activism, lobbying) that result from these distributional effects. A better understanding of these political mechanisms will enhance the ability to forecast how countries and individuals will respond to global forces and the policies that will be derived from their interaction.

SEE ALSO Political Economy of Financial Globalization: China and Financial Globalization; Emerging Markets Politics and Financial Institutions.

Glossary Heckscher–Ohlin–Samuelson (HOS) model The factoral model of trade that posits that globalization should benefit the abundant factor in a country (land, labor, or capital), as returns rise absolutely and disproportionately to the owners of the factors that are used intensively in the production of goods from that country. Multinational corporations (MNCs) A corporation that has its management headquarters in one country, known as the home country, and operates in another or several other countries, known as host countries. Obsolescing bargain theory (OBT) Raymond Vernon’s argument that investors have limited influence on the investment environment ex post, because the sunk costs associated with start-up shift the bargaining position to the host government. This is not to say that foreign investors lose all bargaining power; the argument is that through strategic interaction their power is ex ante. To attract foreign investment, developing and transition countries need to make the institutional changes investors desire and offer a credible commitment that these institutions will remain in place; otherwise they will lose to their neighbors in the investment attraction race. Ricardo-Viner model The sectoral theory of trade which argues that factors specific to an industry or service bear the full weight of price changes in their unique product lines. In a Ricardo-Viner framework, one expects globalization to generate conflict across sectors, but it may actually unite owners of factors within those sectors. The compensation hypothesis Conjecture that winners from globalization may compensate the losers with greater social assistance programs. Sometimes this has been referred to as the ‘embedded liberalism’ of international economic policy, indicating liberal economic policies are accompanied by a generous social welfare state reflecting a societal debate and subsequent social contract. In short, the trade-off implies that increased financial globalization should be associated with larger welfare states.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

68

6. INTEREST GROUP POLITICS

Further Reading Brooks, S.M., Kurtz, M., 2007. Capital trade and the political economy of reform. American Journal of Political Science 51, 703–720. Brune, N., Garret, G., 2005. The globalization Rorschach test: international economic integration, inequality, and the role of government. Annual Review of Political Science 8, 399–423. Frieden, J., Rogowski, R., 1996. The impact of the international economy on national policies: an analytic overview. In: Keohane, R., Milner, H. (Eds.), Internationalization and Domestic Politics. Cambridge University Press, Cambridge, UK. Garret, G., 1998. Partisan Politics in the Global Economy. Cambridge University Press, Cambridge, UK. Goodman, J., Pauly, L.W., 1993. The obsolescence of capital controls? Economic management in age of global markets. World Politics 46, 50–82. Haggard, S., Kaufman, R., 2008. Development, Democracy, and Welfare States: Latin America, East Asia, and Eastern Europe. Princeton University Press, Princeton. Hellwig, T., Samuels, D., 2007. Voting in open economies: the electoral consequences of globalization. Comparative Political Studies 40, 283–306. Hiscox, M.J., 2002. Commerce, coalitions, and factor mobility: evidence from congressional votes on trade legislation. American Political Science Review 96 (3), 593–608. Kayser, M.A., 2007. How domestic is domestic politics? Globalization and elections. Annual Review of Political Science 10, 341–362.

Malesky, E., 2009. Agents of economic transition: an instrumental variables analysis of foreign investment and economic reform. Quarterly Journal of Political Science 4 (1), 59–85. Pepinsky, T., 2009. Economic Crisis and the Breakdown of Authoritarian Regimes: Indonesia and Malaysia in Comparative Perspective. Cambridge University Press, New York. Pinto, P., Pinto, S., 2008. The politics of investment. Partisanship and the sectoral allocation of foreign direct investment. Economics and Politics 20 (2), 216–254. Quinn, D.P., Toyoda, M., 2007. Ideology and voter preferences as determinants of financial globalization. American Journal of Political Science 51, 344–363. Rodrik, D., 1997. Has Globalization Gone Too Far? Institute of International Economics, Washington, DC. Scheve, K., Slaughter, M., 2004. Economic insecurity and the globalization of production. American Journal of Political Science 48 (4), 662–674. Simmons, B., Dobbins, F., Garret, G., 2006. The Global Diffusion of Markets and Democracy. Cambridge University Press, Cambridge. Vernon, R., 1980. The obsolescing bargain: a key factor in political risk. In: Winchester, M.B. (Ed.), The International Essays for Business Decision Makers. Center for International Business, Houston. Walter, S., 2010. Globalization and the welfare state: testing the microfoundations of the compensation hypothesis. International Studies Quarterly 54, 403–426.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

C H A P T E R

7 International Conflicts Z. Kelly*, E. Gartzke† †

*US Department of Defense, PA, USA University of California at San Diego, CA, USA O U T L I N E

International Conflict Financial Globalization Promotes Peace Economic freedom Skeptics of Financial Globalization Ambiguous Conclusions for Peace/War The ambiguous consequences of financial upheaval The Effect of War on Financial Integration and Markets

69 69 72 72 74

Civil War and Domestic Conflict Finance and Peace

77 77

Scholarship with Implications for the Study of War and Peace

78

75

Conclusion Further Reading

78 79

76

INTERNATIONAL CONFLICT

derives from a similar understanding of international institutions and incentives to the trade-focused literature. Much of the current debate revolves around the underlying mechanisms behind globalization and peace. Does interconnectedness alter incentives for militarized conflict? If so, should our understanding of the process be drawn from expected utility or rationalism? The functional form of the relationship between globalization and peace invariably depends on logically prior assumptions about the origins of conflict. The first strand derives from expected utility theory, which predicts that wars will occur when the expected benefits of conquest outweigh the anticipated costs. This literature argues that the globalization of production and the free flow of capital in the late twentieth century significantly altered the ex-ante expectations of ex-post benefits. When the wealth of nations no longer resides in land or industrial capacity, returns on territorial conflict decrease. One cannot reasonably expect to capture the value of Wall Street by occupying the street itself. Others argue that financial globalization need not reduce the benefits (nor increase the costs) states can expect from conflict, so long as financial globalization provides fruitful alternative venues for conflict resolution. This branch of literature generally rejects expected

Financial Globalization Promotes Peace While empirical studies of capital flows are relatively new in political science, the importance of financial globalization has a long intellectual history. Montesquieu proposed that capital mobility could constrain leaders. Kant warned that a flexible system of international credit only facilitated conflict. David Ricardo and J.S. Mill warned of the domestic political economy impacts of raising debts during wartime as an alternative to taxes. David Hume decried Britain’s public debt in 1752, warning that ‘unchecked public credit will destroy the nation.’ Secretary of State Bryan argued (unsuccessfully) with President Wilson that prohibiting loans to the belligerent European nations during World War I would bring the conflict to a swift end. Despite this history, capital mobility and financial globalization garnered significantly less attention from political scientists than trade, until recently. The term financial globalization is used to refer to the integration of both regional and global capital markets alongside increased capital mobility in the form of lowered barriers to capital flows. This literature is distinct from that which prioritizes trade integration as an element of peace. Empirical work in this vein, however, Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00005-2

69

# 2013 Elsevier Inc. All rights reserved.

70

7. INTERNATIONAL CONFLICTS

utility as a theoretical model of conflict and instead draws inspiration from bargaining theories. When economic ties allow states to credibly demonstrate resolve, conflict can be avoided. Mutual membership in bilateral, regional, and global commercial institutions may facilitate security. Such institutions provide a forum for resolving disputes. A preliminary task for the commercial peace literature is specifying the mechanism through which privately accrued benefits of economic activity, such as stock market trading, currency speculation, or Foreign Direct Investment (FDI), can be translated into opportunity costs for leaders who may be willing to go to war. The state’s main role in financial transactions is regulatory. Its main benefit comes from increased revenue, mainly through taxes. The state benefits the most when society’s output is high, thus governments have an incentive to promote international commerce and avoid actions that would be disruptive to cross-border commercial exchange. International commercial institutions work in three ways to help states achieve their revenue-maximizing objective. One such mechanism is leaders’ increased expectation about future economic activity. Insofar as leaders anticipate growth, that expectation raises the opportunity cost of war. A second mechanism involves the coordination effect of commercial institutions themselves. Institutional membership may reveal private information about military capabilities and resolve, especially where security arrangements are built into the institutional structure. A third benefit to commercial institutions is that they bring representatives of states together with the goal of enhancing commercial coordination. These processes have spillover effects, making cooperation in other spheres more likely. To test these propositions, scholars turn toward both qualitative and quantitative studies. Early qualitative surveys of institutions suggest that a more nuanced measurement of institutional features is required in quantitative research beyond binary indications of membership. To measure financial interdependence, researchers first employed measures such as currency pegs, common currency areas, and capital openness. Statistically, all of these variables reduce the probability of conflict between states. Subsequent empirical studies expanded into dyadic analyses. In the absence of truly dyadic data, dyadic analyses in this field generally rely on the less constrained state or weak link assumption. That is, the state within a dyad that has a lower score on a given index is more likely to be the initiator of a militarized dispute. Conversely, the state that is more open to capital flows, trade, etc. is less likely to initiate a dispute. Dyads with lower overall joint scores are more likely to experience conflict than dyads with higher scores. Further research refines the quantification of capital markets. Working with data collected by the World

Bank and International Monetary Fund (IMF), others operationalize integration on multiple levels, first, by combining measures of capital restrictions into a larger index of capital openness, second, by collating data on FDI into a measure of exposure to foreign production capital, and third, by measuring exposure to financial capital markets. The dyadic model demonstrates that the likelihood of militarized disputes decreases significantly with increasing capital market integration. Like the institutional approach, this analysis argues that financial integration provides an alternative arena for states to demonstrate resolve and reveal private information. Constraints on leaders are not explicitly represented in this model, but the importance of capital markets instead causes leaders to signal their resolve in different, but still meaningful, arenas. The importance of private information comes from the influential bargaining model of conflict. This model posits that rational actors would avoid costly military conflict if they knew the outcome ex-ante. If outcomes are known, rational actors can reach a bargain that will be mutually preferable to the cost of war. Conflicts can occur between two rational actors when both have private information about their military capabilities and resolve and incentives to misrepresent that information to others. States have strategic incentives to withhold information about the costs of fighting, their military capabilities, and their resolve in order to achieve a more favorable settlement. This incentive undermines efforts to find peaceful resolutions to interstate disputes. Other empirical research approaches the issue of financial interdependence and peace from a slightly different perspective, arguing that portfolio investment does not represent a significant or costly tie to break; instead focusing on FDI as a proxy for costly interdependence. Using a dyadic research design, this research demonstrates that symmetrical FDI is more effective at reducing militarized conflict than trade ties. A key distinction here is between vulnerability and sensitivity interdependence. Sensitivity applies to relationships wherein the price of goods and services are altered, whereas vulnerability applies when breaking ties imposes real costs. Like others, this analysis relies on an assumption that the costs imposed on private investors will be reflected onto government and considered in foreign policy decision making. Capital was fairly globalized in the nineteenth century, prior to World War I. It was not until the postWorld War II period that market integration began again in earnest. While the end result of this process was peaceful, the process itself has the potential to stimulate conflict insofar as it creates winners and losers and exposes domestic economies to new competitive pressures. Nevertheless, the security externalities are potentially greater than the economic benefits. One such pacific benefit may be realized through investment from already

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

INTERNATIONAL CONFLICT

wealthy and developed countries to smaller, poorer economies. These rewards, however, are likely to be conditional on the degree to which they promote prosperity and democracy rather than exerting an independent effect on conflict. Ultimately, the proposition that capital market integration is more effective at promoting peace than is trade may be premature. Scholars returned to commercial institutions to test the proposition that mutual membership in multilateral institutions facilitates peace. Commercial institutions may have three effects. The first is taken from the commercial liberalism literature. If the opportunity costs of militarized conflict are low, leaders may prefer conflict to peaceful dispute resolution. Commercial institutions may increase opportunity costs of conflict when they facilitate interstate trade. Second, such institutions could serve as forums for revealing states’ private information about military capabilities and resolve. Some commercial institutions have evolved to include explicit commitments to share security information, although this is rare and largely includes only information that is already public. Third, commercial institutions can build relations between high-level diplomats, helping them overcome commitment problems, enhancing trust, and deepening communications. In the absence of trust, regular institutional meetings also provide more opportunities for states to find solutions to low-level disputes before they escalate into militarized conflict. Empirically, there is support only for the third proposition. Neither deeper economic integration nor security arrangements, when built into commercial institutions, have an independent statistical effect on promoting peaceful relations. This finding implies that quantitative research using only the density of commercial institutional arrangements obscures the most important variable: forums for high-level leadership. Such forums help resolve commitment problems. These problems occur when one or more parties to an agreement cannot trust one another to renege, that is, in the presence of incentives to defect. Trust, therefore, means having positive expectations about others’ future behavior. Defined this way, trust makes peaceful negotiations more likely, although it does not obviate the possibility of future changes in capabilities. Trust is operationalized as the number of contact forums contained in joint commercial institutions. Overall, studies that concentrate on the internal aspects of commercial arrangements are rare. Recent scholarship rarely combines institutions with capital integration. One empirical focus is on bilateral capital ties. Here, the analysis revolves around resolving the security dilemma. The dilemma originates with the idea that one state’s efforts to enhance its own security, for example by building arms, decrease the security of others. Capitalism can alleviate the security dilemma by aligning states’ goals such that they have little cause to fight.

71

Functionally, there are three mechanisms that reduce the benefits of victory or the cost of defeat. First, economic development reduces the value of using force to seize territory or assets. As economies shift from landintensive to capital-intensive production, the value of territorial conquest declines. However, economic development and concurrent technological advances increase states’ ability to project military force. Thus, there may be a distance effect rather than a blanket pacific effect. Second, capitalist development can align state interests along a track of mutual gain from development, although this does not mitigate the possibility of security competition. Third, globalization of capital creates forums for revealing information. Insofar as investors are risk averse, belligerence, even overtures toward conflict behavior, can reverberate through markets. More integrated states face higher potential costs than less integrated competitors. Using a dyadic research design, there is evidence that financial openness has a stronger pacific effect than either democracy or trade, contrary to the numerous studies that have found positive links between trade interdependence and peace. Other scholars focus empirically on internal politics. Drawing on selectorate theory, they find that higher degrees of fiscal independence provide leeway for governments to pursue more belligerent foreign policies. Conversely, increasing levels of privately held property, a hallmark of capitalist development, should create the conditions for interstate peace. The relationship between citizens and governments is key. Selectorate theory states that leaders must depend on two groups, the selectorate and the winning coalition. The selectorate is a subset of a state’s population, those who have a voice in crafting public policy. The members of the winning coalition, a subset of the selectorate, are those sufficient to appoint a leader to office. Selectorate theory has been extended by the observation that the pool of available resources shapes leaders’ calculations with regard to providing public and private goods. Publicly owned assets allow governments to lower the marginal tax rate, meaning that private benefits to the winning coalition can increase without increasing social costs (e.g., taxes). Financial autonomy influences conflict behavior by significantly lowering the ex-post costs of failure in war. Even if they lose, leaders can distribute sufficient private benefits to supporters (the winning coalition) to remain in office. Measuring public property by using the IMF’s Government Finance Statistics and the World Bank’s World Development Indicators provides statistical support for the proposition that states with higher amounts of nontax revenue are more likely to participate in militarized disputes than those that rely more heavily on taxes. Such states can build support for aggressive policies by either diverting revenue to buy off opponents or providing private benefits to supporters. This finding is an important

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

72

7. INTERNATIONAL CONFLICTS

step in linking domestic finance and international outcomes, but one caveat is in order. Because the dependent variable in this study is a count of militarized dispute participation, the results do not tell us whether states with high levels of fiscal autonomy are more likely to initiate conflict or whether they are more likely to experience more disputes because they are frequently targeted by other states. It may be the case that governments with high fiscal autonomy are attractive targets because the same mechanism that weakens the citizen– government connection leads to decreased or less efficient military spending. Extending that argument, subsequent research looks at different ways in which domestic institutions can promote peace or facilitate conflict. This approach privileges internal features of states, namely, their mix of private property protection and competitive markets, as an explanation for international behavior. The overall theme, that individual actors pursuing their own economic selfinterest can reduce the probability of interstate war, has two dimensions. First, high levels of public property facilitate conflict vis-a`-vis higher levels of privatization. Second, states that more actively restrict trade and capital markets are more belligerent than those that do not. These effects are illustrated with statistical analyses. Higher levels of public property, for example, domestic assets owned by the government, increase the likelihood that a dyad will experience violent conflict. This analysis relies on a common assertion in dyadic studies: the weak-link assumption that the less constrained state exerts more influence on a dyad’s conflict potential. In this case, higher public property acts as a lower behavioral constraint. It is not entirely clear from this analysis whether the converse is true, that private property promotes peace. The public variable captures governments’ nontax revenue. It may be that the indicator implies greater autonomy from publics rather than the comparative role of the private sector in the economy. Subsequent statistical tests demonstrate the impact of protectionist policies on militarized conflict. Economic freedom A related subset of this literature focuses on economic freedom. The concept is most frequently applied to economic growth, but there is overlap with financial globalization. The subfield merits more attention from students of war and peace.1 One such study links the empirical findings of the dyadic democratic peace with the norms of markets and contract enforcement that are common in advanced industrial democracies. Democratic social values follow from economic values. Further research asserts that the

economic development conditions the pacific, dyadic benefits of democracy. Only above a threshold of about $1400 GDP per capita does democracy reduce the likelihood of militarized conflict. An assessment of Schumpeterian creative destruction in the context of globalization notes that a system of open trade is generally thought to privilege some over others, undermining an international hierarchy based partly on wealth and economic size. However, while some predict great power conflict as the result of changes in the status quo, others point to the potential for peace as a positive externality of globalization. Insofar as the empirical observations of (1) per capita income on democracy promotion and joint democracy on reducing conflict, (2) trade on promoting economic growth, and (3) trade dyadically reducing conflict are accurate, then an international system that promotes economic freedom can also encourage peace. With respect to civil conflict, studies of economic freedoms across the last three decades of the twentieth century find that countries where governments take a strong hand in the economy are more likely to experience internal violence. There are several propositions for the relationship between freedom and peace. Free markets, not trade flows or financial integration per se, have a substantial impact on international peace. Like much of the globalization literature, this approach draws on historical thinkers such as Kant, Mill, Angell, and Cobden but measures the concept of commerce more broadly, focusing on free markets rather than on the volume of commercial interactions. The effect in question is monadic; countries with higher levels of economic freedom have a lower probability of participating in a militarized dispute. These results imply that countries that are already economically free have a chance for sustained peace but they say little about the prospects of development – and thus peace – for others. Developing countries that lack domestic economic freedoms face a different set of conflict incentives. It is premature to conclude that economic freedom, growth, institutions, or democracy alone can promote a more peaceful world; these factors are likely complements to one another.

Skeptics of Financial Globalization Not all observers agree that financial integration will create the conditions for lasting peace. Free, open markets have created inequalities and spurred social backlashes in the past, especially in the late nineteenth century, generally considered the first era of modern globalization. Free market policies led to antimarket political movements that called for (and implemented) top-

1

Economic freedom is commonly conceptualized in terms of levels. To say that some are more free than others implies greater access to markets, fewer restrictions on buying and selling goods, judicial protection of economic rights, etc.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

INTERNATIONAL CONFLICT

down regulation aimed at reducing domestic exposure to international market forces. Some see open-market reforms as especially destabilizing in less developed countries where comparatively weak regimes cannot cope with market forces. The primary concern here is economic security rather than political security, but the implications cross domestic and international levels of analysis. Globalization can overturn existing hierarchical economic structures both within and between countries but it can also reinforce such relationships. For example, capital investment and FDI have primarily been between already capitalrich nations, rather than into capital-scarce areas such as Africa. Meanwhile, economic power and influence are slowly shifting toward China and East Asia, away from Europe, America, and Japan. Whether this will alter the balance of power, from US hegemony to multipolarity, in a traditional sense remains unclear. Ultimately, there is a possibility of retrenchment and reregulation of global markets as well as tension between democracy and market fundamentalism. Where democracy gives representation and voice to sectors that are hurt by globalization, the potential for withdrawal increases. Two potentially fruitful avenues for research are mentioned in passing. First, weak regimes, exposed to external economic pressures, may seek to bolster their position with military adventurism. Second, the breakdown of US hegemony, combined with increasing resource scarcity, could reignite imperial competition for natural resources by the remaining great powers. This last prediction may be pessimistic but others have predicted similar collapses and competition in the twenty-first century. Neorealists have leveled several criticisms against the optimistic conclusions of the financial peace scholarship. Acknowledging that integration of trade and capital flows increased during the late twentieth century, they suggest several cautionary notes. First, integration has taken place, and seems to be accelerating, primarily among the already-most-developed states. Others, for example in Africa, Southeast Asia, and Latin America, largely lagged behind this trend. Second, capital and trade integration in the 1990s scarcely equaled the levels achieved just before World War I (while the migration of labor today is actually more restrictive than it was 100 years ago). Much of the economic growth since World War II has simply been a process of repairing the enormous devastation caused by two world wars and depression-era protectionist economic policies. Capital markets are one of the few sectors that are truly global – interest rates have fallen across the developed world and barriers to capital movements are at historic lows. Despite these gains, the tale is a cautionary one. Economic cooperation does not necessarily mitigate the circumstances that create military and political competition. Even given the apparent pressure for

73

greater policy coordination, significant differences remain between regimes and domestic economies. Economies around the world persist in being more local than global, while developed nations conduct most of their trade between themselves, rather than more broadly among a great many states. What remains to be seen is whether global trends in inequality, coincident with increasing integration of markets, will create grievances that lead to conflict. In defense of the structural realism paradigm, scholars argue that globalization has not eclipsed the importance of anarchy and the politics of self-help in the international arena. The crucial point is that the structure of the system itself remains substantially unaltered. While states may interact in new ways through economic-centered institutions or by giving increased autonomy to capital markets, these changes do not reorder the fundamental nature of the system. From this perspective, interdependence has anemic effects on the prospects for peace. Economic interdependence may even increase nationalistic sentiment, while domestically, mutual economic interests are insufficient to mitigate civil conflict. Rather than shifting power away from states and toward autonomous actors, globalization reflects the existing balance of power between states. The trend toward economic integration must create winners and losers; what remains to be seen is whether the losers will be content with a small share of the benefits. Similarly, international institutions mirror the goals and power structures of the governments that create them. While concern focuses on security organizations, the point is equally applicable to international and regional economic institutions or agreements. Using the example of Bretton Woods, when the system no longer benefited the United States, it was abandoned virtually overnight. It is worth noting that these propositions do not explicitly address or refute the empirical findings offered by financial globalization optimists. Rather, they reflect skepticism about their mechanisms and predictions. Institutional membership and capital market integration may have done a great deal to condition peace between nations during the latter half of the twentieth century. However, they have not always done so and there is little basis for predicting that they will continue to do so indefinitely in the future. Admittedly, little research has gone into the circumstances under which these mechanisms break down. The conclusion that institutions are subject to the whims of their creators offers us little in the way of predictive capacity. Entirely distinct from the neorealist paradigm is an economic and institutional critique. This argument is directed primarily at the IMF and the policies it adopted during the 1990s, but it has much broader implications for financial globalization writ large and emphasizes the deep socioeconomic divisions created (or enhanced)

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

74

7. INTERNATIONAL CONFLICTS

by processes of global market integration. In this view, these processes have failed to deliver on their many promises – poverty reduction, widespread economic growth, or financial stability. In fact, much of the world’s population has suffered rather than benefited from opening markets in terms of environmental degradation, unemployment, and even violent conflict. Two of the primary institutions of globalization, the IMF and the World Bank, have not fulfilled their mandates while economic crises have grown more frequent. With respect to domestic politics, and perhaps international relations, globalization has done little to create conditions of lasting peace. Instead, relaxing restrictions on capital flows has, in many cases, destabilized domestic markets and led to widespread resentment, class or sectoral tensions, and even civil conflict. Much of this critique revolves around the management of globalization through international financial institutions as opposed to their potential capacity for harm or good. From this perspective, perhaps best exemplified by Joseph Stiglitz, the danger of financial globalization rests with ideological adherence to the efficient markets hypothesis. In the absence of institutional oversight, increasingly large and rapid capital flows can destabilize economies and facilitate serious crises such as that experienced in East Asia in 1997. Such crises have the potential to dramatically realign domestic politics, perhaps in unexpected ways, by creating social strife, exacerbating income inequalities, and depressing economic growth for long periods of time. The issue of when and how such economic upheavals precipitate militarized conflict deserves further attention from political science. By reshaping domestic preference patterns and political coalitions, financial crises can potentially affect noneconomic relations. Notably absent ex-post from the cases presented by Stiglitz (2002) are interstate conflicts that can be reliably traced to financial crises or economic collapse. This perspective has further implications for the institutional perspectives elucidated by others. The politicization of international institutions can weaken their effectiveness or produce suboptimal policies. For example, the IMF promoted capital market liberalization in developing economies without corresponding evidence that such openness would promote growth. This promoted a rapid expansion of currency speculation and hot money flows that developing countries were illprepared to accommodate in the absence of robust domestic banking institutions. Further research should take into account the efficacy of IFIs in addition to their presence.

Ambiguous Conclusions for Peace/War Not all research on financial globalization and security leads to clear-cut conclusions on the future of war

and peace. Much of the extant research on the connection between finance and state capacity offers ambiguous conclusions. Early works on this subject draw on Modelski’s long cycle theory to address the role of debt in the context of hegemonic competition. One such argument holds that credit and debt management illuminate the pattern of long cycles. Modelski’s theory predicts that periods of hegemonic stability are followed by global wars, after which a new leading power emerges. This trend proceeds in approximately 100-year cycles. Historically, the losers of hegemonic competitions are those who are unable to compete financially, while the winners are states that are able to generate and manage high levels of public debt. Unfortunately, winners find themselves forced to retrench in the wake of lengthy great power competition, setting the stage for their own decline and the rise of a new challenger for dominance of the global system. For example, access to low interest loans was crucial, though not sufficient, to push the Dutch and British into positions of international prominence during the seventeenth and eighteenth/nineteenth centuries, respectively. Ultimately, both states were weakened and foundered under the pressure of their debt burdens, as others, who were able to afford the mantle, took up the task of world leadership. Other research in the same vein suggests an alternative specification of the relationship between war and public finance. This work asserts that discontinuous patterns of state development can be explained by the role of large (global) wars. The growth and expansion of the state is conditioned by wartime financial decisions. In other words, state spending expands during the largest international conflicts without subsequent drawdown. Subsequent research revisits this point with attention to the institutional foundations of financial power within the context of the Cold War. Scholars extend the argument that hegemonic competition is conditioned by borrowing by arguing that the United States prevailed in the Cold War due, in part, to the borrowing advantage afforded by democratic institutions. A limited, representative government leads to easier, cheaper access to credit because leaders in such states can make more credible commitments to repay their debts. Democracy allows citizens to impose penalties on leaders who bankrupt the state. Default imposes broad economic costs that can be substantial enough to translate into real political costs. During the Cold War, the United States was able to leverage the importance of its currency and bonds because default would have been disastrous for virtually the entire population. In contrast, the erstwhile Soviet Union had almost no market for sovereign debt and often coerced citizens into purchasing bonds at very low interest rates.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

INTERNATIONAL CONFLICT

One alternative is to shift the focus of globalization from governments to multinational corporations (MNCs). As investment and the production of goods become increasingly cross-national, advanced capitalist countries are likely to be more peaceful. The same is not true, however, for developing countries. This research touches on a familiar theme identified earlier, namely that inequalities created or exacerbated by a globalized world increase conflict in some areas. The presence or absence of trade itself is secondary to the role that MNCs play in a diversified economy. Since World War II, MNCs have reorganized themselves to take advantage of the free movement of capital across national borders, although their geographic dispersion has been highly restricted to the already advanced, developed economies. This change altered not only the capabilities and incentives of states but their nature as well. Geographic dispersion of production reduces gains from conquest. However, this benefit accrues primarily to those nations that are integrated into the globalized production system. Regarding capabilities, self-sufficiency in weapons production is more difficult today than in the past. In terms of the nature of states, MNC dispersion can promote additional regional integration on other levels, particularly security. The changes described here would not have been possible on such a massive scale had they not coincided with increasingly open capital flows (especially in Europe since the 1970s). Furthermore, a major reason why the globalization of production alters the calculus of conflict is precisely that so many modern assets are financial in nature. The wealth of modern firms has a very low degree of asset specificity, making it difficult to conquer in a traditional sense. One large-scale study of reputation asks why any creditor would lend money abroad and governments repay their debts in the absence of international enforcement mechanisms. Over time, borrowers can improve their access to credit and help their credibility. Countries that repeatedly default on payments find themselves cut off from international credit markets. This reputational theory rejects the notion that fear of physical intervention, that is, gunboat diplomacy, keeps debtors in line. Rather, investors form expectations based on past behavior. Stalwarts are expected to repay at all times, while others are fair-weather and repay only during good times. While the model asserts that debt in and of itself does not cause conflict, it would be premature to assume that debt relationships cannot create linkages to belligerence. For example, trade sanctions and threats of force, although rarely enacted and perhaps ineffective, commonly underlie such agreements. However, one might conclude from the author’s evidence that debts promote peace. Lenders and borrowers find common ground – the former seek repayment and the latter wish to preserve their reputation and future ability

75

to borrow in times of need. Further research into secondary, or even tertiary, linkages between sovereign debt and conflict is required to sort out these questions. Others take the study of finance and war in a different direction by highlighting the role of crucial domestic actors in the decision to go to war, namely, bankers and other financiers. Here, the main assumption is that bankers (more broadly speaking, financial actors) have a deep-seated preference for international stability. Violent conflict represents a condition of extreme uncertainty and instability; ergo financiers should prefer that states avoid war. There are several destabilizing consequences of conflict, including inflation, currency depreciation, and negative returns to debt holders from wartime debts. This is explicitly not a theory of war onset, although it implies that the more important international finance is in a state’s decision-making calculus, the greater the likelihood that financial aversion can help deter war. Instead, this theory explains the key role of a certain set of actors by focusing on their own material preferences and the ways in which they attempt to influence state behavior based on those interests. This argument is tested with several case studies. For example, prior to the Spanish–American War, the financial community opposed the war, despite its popularity, for fear of debasement, inflation, and suspension of the gold standard. The generalizability of the approach in terms of the study of war and peace is limited by two factors: (1) its emphasis on powerful states with large, entrenched financial interests and (2) the observation that even at their most powerful role, financial actors have not been able to dissuade states from entering into conflict. This suggests limits to the value of financial globalization as a systemic variable. The ambiguous consequences of financial upheaval There is a long-standing belief that large-scale economic crises directly influence, if not cause, wars, the classic example being World War II. Prior to the war, waves of trade protectionism swept through the developed world. Reciprocal tariff barriers exacerbated political tensions between the great powers, culminating in war. One argument posits that low levels of interdependence in the 1930s meant that states had no alternatives for resolving domestic issues. Presumably, if trade opportunities had been available, they would have offered an alternative to territorial conquest by pacifying domestic economic concerns. Yet, Germany and Japan were two of the most integrated economies in the world on the eve of the war. While popular interpretations lay the blame for the Second World War at the foot of the Great Depression and subsequent economic stagnation, that is to say, a self-conscious turn away from economic integration by the major powers, this does not seem to be the case. The Second World War had many possible

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

76

7. INTERNATIONAL CONFLICTS

causes, only one of which was economic stagnation. Subsequent economic crises (in Latin America, in East Asia, in the former Soviet Bloc) are not generally credited with spurring military competition or other international violence. Nor does that appear to be a likely outcome of the 2009 financial crisis. Some discard economic explanations as unsystematic and not representative of the strategic nature of conflicts. Perhaps the least opportune time for one country to launch a war against another country is when the latter is undergoing an economic turmoil; this is contradictory to the popular interpretation of World War II. World War I is often regarded as the converse of World War II, a failure of interdependence to prevent interstate conflict. The high degree of interdependence, especially financial, challenges the notion that such ties prevent war. While Angell asserted in 1913 that modern economies rendered territorial expansion and war obsolete, not 1 year later Britain and Germany went to war in spite of their important trade relationship. This conception is problematic in light of World War II. If the First World War was the result of the failure of interdependence to prevent conflict, then the same logic cannot apply to World War II. Nor can the causal logic applied to the Second World War apply to the first. If one war results from a failure of interdependence and the other from the absence of interdependence, then we cannot generalize an underlying mechanism from those contradictory cases. The examples of the two World Wars highlight the difficulty in making probabilistic assertions about the likelihood of war based on interdependence. If economic disintegration and hardship contributed to World War II, then the popular interpretation of the First World War is incorrect, or perhaps neither case is more broadly generalizable. The same causal logic cannot apply to both. If one argues that the lack of financial prosperity and integration were root causes of World War II, then their presence must have made other conflicts, including World War I, less likely. Because war, especially a global war, is a very low-probability event, we lack sufficient evidence to generalize from war to a theory about the effects of financial integration (or lack thereof) on war. Research using one or a few cases as evidence must provide compelling reasons why contradictory evidence is wrong and why competing arguments cannot be correct. The two World Wars provide competing evidence for the outbreak of conflict, leading to ambiguity about cause and effect. Because such cases often contain apparently contradictory or theoretically inconsistent evidence, scholars in international relations increasingly turn to statistics and probabilistic inference. Consider that a theory of war induced from World War I might lead to the hypothesis that economic interdependence does not, nor

cannot, prevent war. Statistical research has made strides in demonstrating that, on the average, trade and financial interdependence go a long way toward reducing the incidence of conflict. While statistical inference is not without its own pitfalls, it can help researchers avoid problems of design bias, selection on the dependent variable, and inference from a small number of observations. Political science offers several large-N approaches to the questions posed by these competing examples. One substantive finding is that democracies are more likely than nondemocracies to launch diversionary wars during hard economic times. Additionally, recessions can increase the probability of internal conflict, an effect exacerbated by simultaneous external military conflict. Given the paucity of conflict events, and considerable fluctuation in growth rates, any causal effects of growth and instability are difficult to isolate. Some quantitative research concludes that economic growth has a small positive effect on the likelihood of interstate disputes, contrary to the common assertion that lower growth should be associated with conflict. On the whole, there is no clear pattern suggesting that either systemic crises or economic booms, in and of themselves, make violent conflict more or less likely.

The Effect of War on Financial Integration and Markets A separate variant of the finance and war literature reverses the relationship between globalization and conflict, asking ‘how do wars affect financial markets?’ Some answers to this question can be found outside political science. Evaluating the degree to which markets in 2003 predicted the US invasion of Iraq demonstrates that futures markets appeared optimistic, expecting oil prices to jump and then return rapidly to normal levels. Markets predicted small declines in equities as a consequence of the war in Iraq with more extreme effects among more globally integrated economies and oil importers. Historical inquiries offer some perspective. Using the Civil War, Greenback market shows correlations between large conflict and political events and price swings. Financial markets, however, do not view all historically salient events with equal weight. Scholars agree that increases in the expected future costs of the war will decrease the present value of holdings because it is unlikely that their expected future value will hold. There is a different result for the simultaneous Confederate Grayback market, suggesting that investors in the North and South had different views (or information) about the course of the war. Similarly, a study of French bond markets during World War II finds that the spread

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

CIVIL WAR AND DOMESTIC CONFLICT

behind Vichy bonds and 3% rentes reflected the progress of the war and expectations of the post-War government’s commitment to make good on Vichy debt. However, political events appear more salient in this case than military action. Expectations of conflict duration and intensity have a negative effect on stock markets and commodity prices. Financial markets can also reveal illegal arms trade if investors have private information about companies that produce weapons. Focusing on countries under arms embargoes, the effect is concentrated on companies in countries where the legal and reputational costs of engaging in illegal arms trade are low. A broad analysis of three wars between 1990 and 2000 finds that major stock markets were negatively affected by conflict. Rather than a uniform expectation that markets should respond negatively to interstate conflict, the effect of conflicts on markets depends on the severity of the conflict in question and expectations of market actors. Markets generally react negatively to war but where events reduce uncertainty over future costs, markets occasionally rally, often at the beginning of conflicts. Market fluctuations thus depend on both expectations and information about the likely intensity of militarized conflict.

CIVIL WAR AND DOMESTIC CONFLICT Finance and Peace Although many researchers touch on the domestic implications of financial globalization, there is significantly less research in this area than on international dimensions of finance and conflict. The extant literature on civil war generally treats economic globalization as an exogenous process and inequality as a control variable rather than self-consciously modeling the full effect of market integration on violence. The fields of comparative politics and development economics explore these questions about the relationship between globalization and development at length. International relations, however, has relatively little to say about when and whether globalization is the source of inequalities and thus conflict. Only three quantitative empirical studies examine this relationship, and the first is presented here. The others deal only with trade. Two alternatives offer quite different visions of globalization’s consequences for domestic conflict. The optimistic view, described, in part, earlier, holds that increasing investment and economic growth will have peace dividends. In contrast, pessimists worry that investment and trade create inequalities that increase the likelihood of political instability and violence.

77

Assessment of both of these propositions finds mixed evidence. Different varieties of globalization create different domestic conditions. On average, trade is more associated with peace, but higher levels of foreign investment lead to greater political instability. If some types of financial globalization exacerbate inequality, the question remains whether inequalities lead to violent conflict. FDI is often associated with the natural resource extractive industry. Natural resources themselves are associated with a higher risk of conflict by some, while others argue that natural resources have no effect on the likelihood of civil war. Conflict may be a function of domestic economies, volatility, and the type of FDI, rather than an absolute effect of foreign investment. Scholars identify three possible relationships. Globalization could (1) promote peace, (2) lead to conflict, and (3) have no connection with civil war. Economic globalization here includes a number of factors, trade, FDI, FPI (foreign portfolio investment), as well as Internet use after 1990. FDI and portfolio investment are both negatively associated with the presence of civil war but, crucially, not the likelihood of civil war onset. One interpretation is that the longer civil wars proceed, the greater the disruption to productive economic activity and the greater the pressure to end the war. To our knowledge, no such hazard model has been published. Nevertheless, their findings suggest that there is no direct effect between foreign investment and domestic conflict. If others are correct, it is likely that the distributional consequences of financial globalization are related more to political violence than to investment itself. Recent work has empirically tested competing claims about the consequences of economic globalization. On the one hand are claims that political stability is an externality of interdependence. On the other hand are skeptics, positing that the opening up of domestic economies to international markets will result in violent protests and perhaps civil war. One hypothesis is that financial flows should destabilize domestic politics, for example, through the consequences of speculative attacks in places where institutions are inadequate to regulate hot money. Another twist is the possibility that the winners of globalization can compensate the domestic losers, perhaps through increased social spending or foreign aid. Both the level of economic openness and FDI inflows reduce the probability of internal conflict but the process of liberalization slightly increases that likelihood. The more rapid the process of opening up, the stronger the destabilizing effect. Neither the supply of redistribution, measured in the form of government consumption and foreign aid, nor the demand, measured as inequality, has a significant effect on openness and liberalization.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

78

7. INTERNATIONAL CONFLICTS

SCHOLARSHIP WITH IMPLICATIONS FOR THE STUDY OF WAR AND PEACE A third subset of the literature deals with indirect or inferred links between financial globalization and its consequences for war and peace. Some authors in this camp emphasize state capacity, while others focus on the international dimensions of sovereign debt. Continued growth of American public debt has drawn increasing attention in recent years. Given the dollar’s role as a reserve currency and the US role as the sole superpower, to the extent that the American state is threatened by rising debts, there are implied consequences for international security and stability. A study of British war finance asks how a small island nation became the preeminent naval and economic power of Europe. Faced with competition from a powerful continental enemy, France, the British state had several advantages over its fiscal-military rivals. Namely, Britain had a weaker monarch, uniform tax policies, and a relatively more empowered parliament. The importance of parliament in tax policy-making lent strength and legitimacy to Britain’s high wartime taxes. The perceived fairness and oversight of tax policy lent credibility to British sovereign debt and allowed Britain to borrow from continental powers, such as Prussia, at favorable terms. It is worth noting that the conclusions are at odds with the more common assumption that investors favor repayment to regime type. Japan, for example, had little trouble securing credit abroad during the Russo-Japanese War, particularly after the fall of Port Arthur presaged the likelihood of Japanese victory. Other work reaches different conclusions with respect to the growth of American debt and power during the twentieth century, arguing that the substantial increase in US debt poses a danger to America’s ability to execute foreign policy and exercise power abroad. This loss for the United States jeopardizes international stability. In contrast to early debtors or cases compared elsewhere, the United States has several borrowing advantages. Today, fiscal deficits and borrowing are much smaller in scale than during the sixteenth and seventeenth centuries, but the increasing complexity of sovereign debt arrangements exposes modern states to potential crises. Borrowing can facilitate the use of force in the short term but, as others note, the exercise of power funded by debt is difficult to maintain in the long term. The implications for current American policy are clear – the United States will gradually lose the ability to project power in its accustomed manner. Revisiting these implications for state capacity and power leads to the conclusion that indebtedness threatens (American) foreign policy autonomy. Financial ties could create common interests, shifting state goals toward absolute, rather than relative, gains.

Lending could create interest coalitions in creditor nations that encourage peaceful external relations. This optimism comes with historical caveats. Creditors and debtors sometimes have different interests, especially when repayment could mean costly austerity measures or adjustments within debtor nations. Foreign investment is occasionally viewed with suspicion, if not outright hostility. Lastly, foreign preferences for American debt are not unlimited. Exogenous shocks, particularly the 2008–09 financial crisis, raise the question whether the United States can continue to provide international public goods, namely, financial stability and international security. The discussion of US debt would be incomplete if special attention is not paid to the likelihood of security competition from China. If creditors can, and do, leverage their status, the ramifications for the United States would be enormous. Optimists believe that China will find it very difficult to coerce American foreign policy. Tools of financial statecraft that may be effective against weak states are unlikely to work against the powerful. China has tried to exploit her financial position several times but has experienced limited foreign policy gains. The Chinese were unable to coerce the American government to abandon its asset protection program in 2008 and have not been able to force the dollar’s reserve currency status onto the international negotiating table. Although great powers may be resistant to such pressures, weak states should find themselves far more susceptible to financial leverage.

CONCLUSION The empirical study of finance and conflict garners increasing attention by political scientists. Early inquiries do well by extending the findings and hypothesis of existing research programs such as the democratic/ liberal peace, the democratic advantage, and international institution building. To date, no consensus exists with respect to the role or substantive impact of financial institutions and actors in averting or contributing to conflict. Some argue that financial integration and capital openness reduce states’ willingness to engage in costly military competitions but others warn against optimism, cautioning that economic gains in one issue-area do not a priori reduce incentives for political and military competition. The focus on the domestic consequences of international politics has led some to conclude that there are differential effects of integration. Some states are enriched but others experience few gains from an open financial system. Such differences may create domestic turmoil in the future, possibly with spillover effects. Compared to other political science research paradigms, the study of financial globalization is relatively new. At the time of writing, these issues can come dramatically to the forefront of politics

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

CONCLUSION

and popular media. There is a great deal of room for political scientists to explore questions of debt, financial interconnectedness, crisis, and conflict.

Further Reading Barbieri, K., Reuveny, R., 2008. Economic globalization and civil war. The Journal of Politics 67, 1228–1247. Bearce, D.H., Omori, S., 2005. How do commercial institutions promote peace? Journal of Peace Research 42, 659–678. Bussmann, M., Schneider, G., 2007. When globalization discontent turns violent: foreign economic liberalization and internal war. International Studies Quarterly 51, 79–97. Drezner, D.W., 2009. Bad debts: assessing China’s financial influence in great power politics. International Security 34, 7–45. Gartzke, E., 2007. The capitalist peace. American Journal of Political Science 51, 166–191. Gartzke, E., Li, Q., 2003. War, peace, and the invisible hand: positive political externalities of economic globalization. International Studies Quarterly 47, 561–586. Gissinger, R., Gleditsch, N.P., 1999. Globalization and conflict: welfare, distribution, and political unrest. Journal of World-Systems Research 5, 327–365.

79

Gray, J., 1998. False Dawn: The Delusions of Global Capitalism. The New Press, New York, NY. Leigh, A., Wolfers, J., Zitzewitz, E., 2003. What do financial markets think of war in Iraq? Stanford GSB Research Paper No. 1785. Available at SSRN: http://ssrn.com/. McDonald, P.J., 2009. The Invisible Hand of Peace: Capitalism, the War Machine, and International Relations Theory. Cambridge University Press, New York, NY. Mousseau, M., 2000. Market prosperity, democratic consolidation, and democratic peace. Journal of Conflict Resolution 44, 472–507. Rasler, K., Thompson, W.R., 1983. Global wars, public debts, and the long cycle. World Politics 35, 489–516. Rosecrance, R., Thompson, P., 2003. Trade, foreign investment, and security. Annual Review of Political Science 6, 397–398. Schultz, K.A., Weingast, B.R., 2003. The democratic advantage: institutional foundations of financial power in international competition. International Organization 57, 3–42. Stiglitz, J.E., 2002. Globalization and its discontents. WW Norton, New York, NY. Waltz, K.N., 2000. Structural realism after the cold war. International Security 25, 5–41.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

Intentionally left as blank

C H A P T E R

8 The Political Economy of International Monetary Policy Coordination J.A. Frieden*, J.L. Broz† †

*Harvard University, Cambridge, MA, USA University of California, San Diego, CA, USA O U T L I N E

Introduction

81

The Potential Gains from International Coordination

82

The Problem: Exchange Rate Externalities

83

INTRODUCTION

87

Conclusion Glossary Further Reading

89 89 90

participants. We also need a clear picture of the implications of monetary coordination among the major monetary authorities for countries elsewhere, both in times of crisis and in more normal times. The view that international monetary policy coordination is undesirable or infeasible, or both, has left us with too little understanding of what it might indeed entail. And yet, especially in the context of recent problems, all these issues deserve much more focused analytical attention. We begin with a brief discussion of the status of the literature on international monetary coordination. We then consider whether the lack of international monetary coordination among major powers is a source of uncertainty and instability in the world economy, and whether the world would be better off with a more encompassing governance structure to manage international monetary relations. The rationale for greater coordination hinges on the extent to which national exchange rate policies impose negative externalities on other countries. This provides a justification for cooperative action to internalize – and reduce – such externalities, and for an institutionalized mechanism to encourage such coordination. We then assess both the feasibility of such institutionalized coordination and some specifics of how it might work.

Exchange rates have been the focus of a great deal of international attention over the past two decades. Spectacular currency crises in emerging markets have thrown whole regions into economic and political turmoil. Major countries have accused one another of purposely manipulating their currencies, leading to acrimonious disputes that implicate broader economic ties. Questions have been raised about the future role of the dollar and the euro as international currencies. Yet despite these problems, the consensus academic view is that the economic rationale for international monetary policy coordination is not strong. Especially in the context of the global ‘subprime’ crisis that began in 2007, the problems and prospects of international monetary coordination warrant more scholarly and policy attention. It would be particularly useful to have a clearer understanding of the implications of the actions of the principal financial centers, as opposed to small open economies whose global impact is limited. It would also be helpful to explore the conditions under which the major powers could arrive at a collaborative solution to problems of monetary coordination, one that would not impose unacceptably asymmetric costs on

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00003-9

Exchange Rate Coordination: Motivation and Modalities

81

# 2013 Elsevier Inc. All rights reserved.

82

8. THE POLITICAL ECONOMY OF INTERNATIONAL MONETARY POLICY COORDINATION

THE POTENTIAL GAINS FROM INTERNATIONAL COORDINATION The rationale for international policy coordination stems from the potential external effects that independent domestic policy actions can have on other countries. For example, when one country loosens monetary policy to stimulate domestic demand after a global shock, it will result in a depreciation of the exchange rate, which can adversely affect the prospects of other countries by reducing demand for foreign goods. This is an example of a negative externality, commonly referred to as a ‘beggar-thy-neighbor’ policy because the stimulus comes at the expense of other countries. When other countries engage in the same policy, the policies cancel each other out, resulting in excessive inflation but no gain in output. The potential role for international coordination is to limit such counterproductive policy externalities. The issue of external effects has become more prominent as the world economy has become more integrated in recent years. This is because the size of the externalities depends critically on the extent of economic integration between economies. If goods markets are not well integrated (i.e., exports are but a small fraction of a country’s total output), a currency depreciation will have a negligible impact on domestic production, and the international implications of the policy will not matter very much. Likewise, if financial markets are not well integrated (i.e., capital does not flow readily in or out of a country in response to interest rate differentials), then monetary policy will not have a large effect on exchange rates or on foreign demand for domestic goods. Given the pace of globalization in recent years, it is all the more surprising that the consensus academic position is that the gains from international coordination on monetary policy are negligible at best, and possibly even counterproductive. This position – that an inward-looking policy is approximately optimal and there is not much to gain from policy coordination – has persisted for nearly 40 years, even though national economies have become more closely integrated over that span of time. In the 1970s and 1980s, strategic models were introduced to provide a theoretical rationale for policy coordination. Formal game theoretic models showed that policy bargains could be found that left some countries better off without others being worse off, in terms of their own policy objectives. The key insight was that coordinating policies to internalize the externalities could lead to higher welfare for all. While such models justified international coordination from a theoretical point of view,

empirical studies from the 1980s revealed that the benefits of coordination were quantitatively small, compared to measurement error in the data and other metrics. This was somewhat intuitive since international integration was actually quite low at the time. The gains from monetary coordination across the major economic regions were not expected to be large because the United States, Europe, and Japan were still relatively closed economies in the early 1980s. However, goods and financial markets have become more integrated since that time but the consensus position has changed very little. In fact, more recent work suggests that the benefits of coordination may actually decrease with globalization. This is the surprising conclusion of twenty-first century work that incorporates the microfoundations of the new open economy macroeconomics. These models include optimizing individuals, nominal rigidities, and a representative agent’s utility function serving as the welfare metric for optimal policy. The role of government is to counteract distortions, such as wage stickiness and the possible failure of international consumption risk sharing, which prevent individuals from acting in their own interests. With perfectly integrated global markets, international monetary cooperation is redundant in this context, as inward-looking monetary rules can replicate the Pareto-efficient (flexible wage) allocation and global integration can help insure against country-specific shocks. For example, if an individual country suffers a fall in production, the scarcity of domestic goods relative to foreign goods induces a rise in price of domestic goods. This change in relative prices can compensate domestic agents for the smaller quantity of domestic goods they have to consume and export, as they will be able to import more foreign goods for the smaller quantity of domestic exports. The result is that they enjoy a level of consumption and utility comparable to the foreign country. In short, integrated goods markets can do the job of pooling risk internationally and thereby leave policymakers free to focus on eliminating the sticky wage distortion. Counteracting the wage distortion, in turn, need not involve any international policy coordination because, unlike the beggar-thy-neighbor policy of manipulating exchange rates, policymakers mimic the flexible wage outcome by manipulating the domestic labor market. The implication of this research is that there is little need for institutionalized global cooperative mechanisms to coordinate national monetary policies. Rather, the best solution lies in having each individual country keep its own house in order and maintain a stable domestic price level. In fact, improvements of monetary policy institutions at the domestic level, conjoined with

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

THE PROBLEM: EXCHANGE RATE EXTERNALITIES

the further integration of world markets, may render international cooperative schemes superfluous.1 A limitation of this work is that the gains from cooperation are analyzed entirely within the macroeconomic realm. For tractability, monetary policies do not spill over into other domains, like trade policy, in these models. In the next section, we discuss how monetary policy may generate international spillovers outside of the macroeconomic realm, for political economy reasons.

THE PROBLEM: EXCHANGE RATE EXTERNALITIES Any argument for explicit collaboration among national governments rests on the presumption that purely national policymaking can produce collectively suboptimal results. If national governments acting in their own self-interest adopt policies that are best for them and for others, then there is no scope for coordination to improve welfare. With respect to exchange rate policy, many analysts believe that the best outcomes will be obtained if individual governments adopt responsible macroeconomic policies, in their own interest. While at some level this is true, almost by definition – who could oppose responsible policies? – it is incomplete, and may lead to incorrect conclusions. In particular, the view neglects the implications of a theoretically grounded analysis of the political economy of macroeconomic policy. An insistence that all that is necessary for optimal global policy outcomes is sensible national policy misses two important points, both relevant to international coordination. First, whatever theory may say about aggregate welfare, governments typically face political pressures to satisfy particular groups in the population with policies that deviate from whatever the textbook welfare-maximizing ideal may be. These politically driven policies can have a negative impact on other countries and coordination among nations can help mitigate the effects of such particularistic policies. Second, even government policies that are optimal from a national standpoint (whether in purely economic or political-economy terms) may impose externalities on other countries, whether due to purposive strategic behavior such as free riding or due to the fact that the external effects are not internalized.

83

This logic is widely accepted in many arenas, such as trade policy. Each country is, under most circumstances, best served by unilateral liberalization.2 Yet governments face powerful political incentives to erect or maintain protective barriers, despite the fact that this imposes efficiency costs on the national economy and on the nation’s trading partners. In this context, scholars have argued that a wide variety of international trade institutions can help governments cooperate to improve the welfare of all concerned. For example, the World Trade Organization (WTO) institutionalizes reciprocity, thereby empowering domestic interests (especially exporters) who can oppose protectionist pressures; and its dispute settlement mechanism helps resolve complaints about national policies while reducing the risk of unilateral retaliation. There seems little doubt that the institutions of international trade perform the transactions cost-reducing role scholars expect from an international governance structure, and do so in ways that increase the likelihood of a cooperative trade policy outcome and a joint increase in welfare. In other words, there are at least two objections to the view that international coordination is unwarranted because all that is necessary is that governments adopt nationally appropriate policies. First, we can reliably anticipate that government policy will often be driven by motives other than aggregate social welfare – such as getting reelected, satisfying powerful domestic interest groups, or achieving noneconomic goals. In this context, the uncoordinated pursuit of what is ‘nationally appropriate’ in a political economy sense may lead to outcomes that could be improved upon by interstate coordination. Second, there are many instances in which national governments may not fully internalize the negative (and positive) externalities created by their own policies. These two factors are indeed the justification for virtually all international organizations. They can be combined in the following argument: the uncoordinated pursuit of national policies, determined in politicaleconomy equilibrium, can create negative externalities for other nations, and thus can lead to a collectively inferior outcome. In applying this reasoning to international monetary relations, the first step is to assess the extent to which this actually might pertain to national exchange rate policies. After all, if a government manipulates its currency, pulling it away from a defensible or realistic exchange

1

The strong conclusion about the lack of gains from international coordination has stimulated a great deal of new research on the sources of coordination gains. For example, the gains from coordination may be related to the degree of exchange rate pass-through. Policy coordination may also produce welfare gains if the international financial markets are incomplete, if policymakers have imperfect information, and if domestic shocks are not perfectly correlated across countries.

2

The exception is in the case of a country that can use its market power to improve its terms of trade by an optimal tariff. While there may be evidence for the relevance of optimal tariff considerations for some aspects of trade policy, it seems clear that the structure and level of protection in most nations responds to political economy pressures as well.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

84

8. THE POLITICAL ECONOMY OF INTERNATIONAL MONETARY POLICY COORDINATION

rate, the principal effects will be felt at home. And it might be argued that to the extent that the movement of one country’s currency imposes costs on another country, the latter can simply reverse the charges by counteracting the currency movement. Nonetheless, in currency affairs, as elsewhere, there are varieties of external effects that go beyond the impact on the national economy and national economic actors, and for which a unilateral response is either not possible or not desirable. Some might see currency volatility as a strong argument for international coordination. The strongest case for this is that it dampens incentives for private actors to engage in welfare-improving cross-border (or crosscurrency) trade, investments, and payments. Certainly there is plenty of evidence that currency stability encourages international economic exchange; this in turn implies that the world would be better off if volatility were reduced. There are at least a couple of problems with this. First, while exchange rate volatility may be a serious problem for the international economy, it is not really the result of uninternalized externalities. Volatility is often the result of national policy, to be sure; but its negative effects are almost entirely domestic – national traders, investors, borrowers, and others are deprived of market opportunities. To the extent that the decision to allow a national currency to move around is made by the government, it presumably takes into account the negative impact of this volatility on domestic economic actors. The second argument against concerted international action to reduce currency volatility is that this volatility may to some extent represent a desirable flexibility in exchange rates. For example, a currency movement to correct an unsustainable misalignment is not a bad thing. The problem typically discussed is ‘excess volatility,’ but this is much more likely to be the result of the operation of modern currency markets than the result of government policy. Market actors are probably at the root of much of today’s ‘excess volatility,’ and it is not clear that cooperative government intervention to restrict foreign exchange markets would be a good thing. So the argument for international coordination to reduce exchange rate volatility is relatively weak. Exchange rate misalignments, on the other hand, can be the source of substantial problems for other nations and for international economic relations more generally. A government may purposely keep its currency relatively weak, in the expectation that a depreciated currency will stimulate exports (such policies cannot prevail forever, but there is strong evidence that the rate at which exchange rates converge toward purchasing power parity (PPP) can be quite slow – certainly slow enough to allow such misalignments to have substantial effects on the real economy). Of course, a depreciated

exchange rate has a negative effect on national purchasing power, but this is solely a domestic matter in which the government has decided to trade off the welfare of exporters and import-competers, on the one hand, for that of consumers. However, a depreciated currency puts competitive pressure on the country’s trading partners, and can stimulate protectionist sentiments abroad. The result may be to trigger commercial discord between countries, and even to endanger broader trade agreements. Recent conflicts between the United States and China, indeed, illustrate the potential for a currency misalignment to provoke trade tensions. For years, there have been indications that the weak renminbi was inflaming Congressional protectionism, and these sentiments were only moderated when the dollar began to depreciate against other currencies. In an earlier era, the dramatic appreciation of the US dollar in the early 1980s led to major protectionist legislation in the US Congress, and an unprecedented spike in complaints to the International Trade Commission: antidumping cases, for example, tripled from an annual average of 18 between 1979 and 1981, to an annual average of 56 between 1982 and 1984. These exchange rate-provoked conflicts have placed a significant strain on the international trading system, both in bilateral relations between the countries in question, and more generally inasmuch as they have called into question the commitment of major countries to the multilateral resolution of trade disputes. It was not only in the early 1980s that exchange rates spilled into the trade arena: from the late 1970s to the present, protectionist activity in the United States has been positively related to the level of the real exchange rate. Figure 8.1 plots the association between the real effective exchange rate (REER) of the US dollar and antidumping cases investigated by the United States International Trade Commission (USITC) – the quasijudicial Federal agency that conducts antidumping and countervailing duty investigations. These data clearly indicate that the number of antidumping cases investigated by the USITC increases with the appreciation of the US dollar. The one outlier – 1992 – is the exception that tests the rule. On 8 July 1992, the steel industry filed 47 separate antidumping petitions on various countries for four types of steel products. If we reduce these 47 cases to four – because this flurry of steel-related cases were not really separate – the fit of this simple model improves to R2 ¼ 0.32 from R2 ¼ 0.21. This relationship is meaningful in a substantive sense as well. Simulating the effect of increasing the REER of the US dollar by one standard deviation above its mean – a roughly 10% real appreciation – increases the number of antidumping cases filed at the USTIC by 10.7 cases per year (the 95% confidence interval ranges from 4.09 to 17.63 cases per year). Given that

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

85

THE PROBLEM: EXCHANGE RATE EXTERNALITIES

1985

90 Antidumping filings per year

FIGURE 8.1 US antidumping cases, 1979– 2009. Notes: The base year for the REER index is 2005. The REER data are from the Bank for International Settlements, available at http:// www.bis.org/statistics/eer/index.htm. Antidumping data are from Bown, C.P., 2010. Global Antidumping Database. Available at http://econ. worldbank.org/ttbd/gad/.

1992

100

2001

80 70

1998

60 50 40

1986

1991

1994

1999 2000

1993 1990

2003 1998

2004 1989 2007 1980 1996 2009 2008 1987 1981 1997 1979 1995 2005 2006

30 20 10

1982

1984

1983 2002

y = 1.129x – 74.44 R2 = 0.21

0 85

90

95

105 110 115 100 REER of the $US (based on CPI)

120

only 39.5 cases are filed per year on average, this is a very large effect. Another set of examples of how national currency policy can affect other economic ties comes from the interaction of regional currency relations and regional trade agreements. In January 1999, the Brazilian government allowed the real to float, which led to a very substantial depreciation of the currency. This came on the heels of a dramatic expansion of Brazilian–Argentine trade in the context of Mercosur, a trade agreement strongly favored by both governments. But the devaluation, and the overvalued peso associated with Argentina’s currency board, provoked a flood of imports into Argentina: in the first 8 months of 1999, Argentine imports of Brazilian textiles and footwear rose by 38% and 66%, respectively. This in turn provoked protests from Argentine manufacturers, who forced the Argentine government to impose barriers on Brazilian iron, textiles, and paper. The Brazilians retaliated, complained to the WTO, and even threatened to dissolve Mercosur. In Argentina, just as in the United States, the level of the exchange rate correlates positively with protectionist activity over time. Figure 8.2 plots the number of antidumping cases investigated by Argentina’s Comisio´n Nacional de Comercio Exterior (National Foreign Trade Commission) against the REER of the peso between 1995 and 2009. The positive relationship illustrates that when the peso appreciates against the currencies of major trading partners like Brazil, political pressure for trade protection intensifies. The most antidumping cases were initiated in 2000 and 2001, following the Brazilian devaluation and the spike in the REER of the peso. Moreover, the exchange rate appears to have a large substantive impact on antidumping investigations in Argentina. Simulating the effect of increasing the REER of the peso by one standard deviation (55.56) above its mean (147.93) results in 4.98 additional

125

130

antidumping cases investigated by Argentina’s trade commission (the 95% confidence interval for this estimate is 0.75–9.06 cases). Since the commission only investigates 16 cases per year, on average, this is a powerful (and statistically significant) result. The Mercosur crisis was reminiscent of an earlier episode in the European Monetary System (EMS). The 1992–93 currency crisis in the EMS led to large devaluations of some EMS currencies. As a result, producers in countries whose currencies had been stable – in particular France and Germany – came under competitive pressures. This in turn led to domestic complaints about imports from the countries whose currencies had depreciated, which threatened the core commitments of the European Union (EU), especially in the wake of the completion of the single European market. This posed a stark choice for EU members: whether to retreat to more flexible exchange rates, which might jeopardize the single market, or to move forward to full monetary union, which would eliminate the problem of exchange rate variability altogether. Returning to more freely fluctuating exchange rates threatened to fuel a popular backlash against the single market, since currency depreciation seemed to confer an arbitrary competitive advantage on certain national producers. In this instance, the alleged undervaluation of some European currencies endangered other economic and noneconomic agreements into which EU member states had entered. Just as relatively depreciated currencies can create negative externalities for governments’ trade relations, so can relative appreciated currencies cause problems. This is most clearly the case when a government’s attempt to sustain an appreciated exchange rate that is widely regarded as misaligned leads to an attack on the currency that results in a contagious currency crisis. In fact, most medium-to-large real appreciations are reversed by nominal devaluations. While there are

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

86

8. THE POLITICAL ECONOMY OF INTERNATIONAL MONETARY POLICY COORDINATION

35

FIGURE 8.2 Argentina’s antidumping cases, 1995–2009. Notes: The base year for the REER index is 2005. The REER data are from the Bank for International Settlements, available at http://www.bis.org/statistics/eer/ index.htm. Antidumping data are from Bown, C.P., 2010. Global Antidumping Database. Available at http://econ.worldbank.org/ttbd/gad/.

2000

Antidumping filings per year

30 25

2001

1995 1996

2009

1999 2008

20 15

2004 2002 2006 2007 2005

10

1997

1998

5 y = 0.087x + 3.200 R2 = 0.30

2003

0 70

90

110

150 170 190 130 REER of the Peso (based on CPI)

210

continuing debates over the precise sources and nature of devaluation crises such as those in Latin America in 1994–95 or in East Asia in 1997–98, there is a strong case to be made that in both instances, misaligned national currencies created the conditions for speculative attacks on both the initially misaligned currencies and the currencies of their regional neighbors. While regional spillovers may justify regional coordination of exchange rate policies, the case for international coordination rests with currency crises that spread across regions to affect nations that are very distant geographically and economically from the originating country. Examples of this type of contagion include the spillover of the crisis in Hong Kong to Mexico and Chile in 1997, and the transmission of the Russian crisis of mid-August 1998 into Mexico and other Latin American countries. In these cases, and others, there were few trade and financial linkages between the country experiencing the initial crisis and the distant countries that subsequently were attacked; there were also few macroeconomic similarities that would suggest spillovers. These were cases of ‘pure contagion’ in the sense that the spillover effects were unanticipated, or larger than expected, on the basis of the observable macroeconomic, trade, and financial interdependencies. None of this is meant to imply that the actual policies adopted in these instances were inappropriate. Indeed, in most cases they were probably the best the governments in question could do. The United States did not actively attempt to strengthen the dollar after 1981; Brazil should have allowed the real to depreciate in 1999; the EMS members in 1992 should have devalued, and so on. But that is not the relevant question. The relevant question is: could these outcomes have been improved by international coordination? Could the United States and its principal

230

250

partners have cooperated more organically before 1985 (when growing concern led to the Plaza Accord)? Could Brazil and Argentina have worked out a collaborative arrangement to allow both their misaligned currencies to depreciate? Could Germany and its EMS partners have developed a cooperative response to German unification? If so, could such coordination have moderated the currency misalignments, and mitigated their effects on trade and other policies? The answer is a qualified yes. Strongly misaligned currencies create problems not just for their home countries but for their economic partners as well, and in some instances for regional or global economic relations more generally. Intergovernmental cooperation could in principle help avoid some of the problems that arise as a result. For while all these cases are different, they have certain features in common. In all instances, policies pursued by national governments concerned about domestic conditions created serious difficulties for other countries. In all instances, the difficulties gave rise to frictions that spilled over into – and typically were most significant for – the nonmonetary realm, often trade policy. And this is not surprising. Exchange rate movements are effective substitutes for trade policy – a 10% devaluation is equivalent to a 10% tariff and a 10% export subsidy – but in most cases there is no international or bilateral check to such movements while there is to increasing trade barriers or export subsidies. A government that is constrained by international agreements not to slap on protective tariffs or proffer export subsidies can instead depreciate its currency and achieve much the same effect. Inasmuch as the former policies in the trade realm create negative externalities and thus provide scope for welfare-improving international coordination, the latter must as well.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

EXCHANGE RATE COORDINATION: MOTIVATION AND MODALITIES

The precise nature of the externalities created may vary over time and across place. They would almost certainly include: 1. Artificially weak currencies to gain competitive advantage, which risk imposing costs on trading partners and threatening commercial cooperation. 2. Active policies to depreciate currencies in hard times for ‘competitive’ reasons, which risk provoking a spiral of ‘competitive devaluations.’ 3. Artificially strong currencies whose collapse threatens crisis contagion. The next step is to consider how international coordination, and perhaps some form of international governance structure, might help alleviate the problems associated with this behavior.

EXCHANGE RATE COORDINATION: MOTIVATION AND MODALITIES Whatever theory may say about the desirability of international coordination, and international institutional innovation, they are only feasible if governments are motivated to move in that direction. There is some evidence that the time may be ripe for innovative measures to institutionalize international currency coordination. The first indication of at least a latent interest in greater coordination is that policymakers mention it on a regular basis. While actions speak louder than words, words are not irrelevant, and virtually every major summit meeting on economic issues has expressed a desire for coordination on currency issues. For example, the 11 April 2008 Statement of G-7 Finance Ministers and Central Bank Governors reads in part: “[T]here have been at times sharp fluctuations in major currencies, and we are concerned about their possible implications for economic and financial stability. We continue to monitor exchange markets closely, and cooperate as appropriate” (available at http://www.ustreas.gov/press/ releases/hp919.htm). So there is at least a verbal recognition of the desirability of coordination. This recognition includes arguments in favor of more active international monetary coordination from some of the world’s leading academic experts on the subject. William Cline and Morris Goldstein have been explicit about their view that exchange rate coordination is highly desirable. Barry Eichengreen has made similar arguments, both in general and with respect to East Asia. It is not inconsequential that these highly respected analysts have been outspoken in their views. There has also, in fact, been some official action in addition to all the many words. A few times in the past 25 years, major governments have felt strongly enough about the global implications of national currency

87

policies to try to work out collaborative agreements. The best known are the Plaza and Louvre Accords of 1985 and 1987, respectively, meant to address problems associated with the appreciation of the US dollar in the early and middle 1980s, and its aftermath. Most recently, the United States and China have engaged in ‘quiet diplomacy’ over China’s currency policy, which Fred Bergsten has likened to an off-budget export and job subsidy that has evaded all international sanctions to date. Although skeptical that the renminbi is significantly undervalued, Ronald McKinnon also supports greater G2 (US–China) coordination, on the grounds that exchange rate coordination between the world’s two largest economies could help limit the devaluation tendencies of other nations around the world. Perhaps due to the imbroglio over the value of the renminbi, the member states of the International Monetary Fund (IMF) have recently expanded the IMF’s mandate quite substantially to include regular surveillance and reporting on exchange rate issues. Of course, the IMF’s original activities were heavily oriented toward the exchange rate regime, but with the end of the Bretton Woods system, this fell into disuse. In the aftermath of currency crises and substantial misalignments, however, the Fund has been under substantial pressure to reinvigorate this dimension of its activity. The IMF’s Independent Evaluation Office produced a series of reports that were highly critical of the Fund’s actions in currency crises, culminating in a critical broad review of the IMF’s exchange rate policy advice published in May 2007. In June 2007, the Executive Board adopted a new ‘Decision on Bilateral Surveillance Over Members’ Policies,’ which was intended to strengthen the role of the IMF in overseeing national currency policies (the decision is available at http://www.imf.org/external/np/sec/ pn/2007/pn0769.htm#decision). Neither of these initiatives, among G-7 members or at the IMF, has been particularly productive, although the IMF endeavor is too new to judge fairly. Nonetheless, their existence is evidence of at least a superficial commitment to do something about the misalignment and volatility of currencies. So there is some evidence of a desire on the part of major governments to work out more fully established mechanisms to cooperate in the determination of national currency policies. There is, nonetheless, room for skepticism about the international political feasibility of such plans. Perhaps the most powerful argument that nothing substantial is likely to happen is a simple one: none of the major actors has an incentive to change the situation. The United States and the euro zone, in particular, can simply impose their exchange rate policy preferences on others. This gives them little reason to want any movement away from the status quo. While there is undoubtedly a lot of truth to this, it is also the case the American

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

88

8. THE POLITICAL ECONOMY OF INTERNATIONAL MONETARY POLICY COORDINATION

and European policymakers do sometimes find themselves facing politically difficult interactions with other governments (as with China); and that American and European policymakers sometimes find themselves at loggerheads with each other. This implies that even for the two great currency powers, there may be room for improvement in the context of a more cooperative international monetary order. The problem is more general, for it may be that those monetary authorities with the least incentive to act are those whose actions would be most helpful. The governments of the world’s principal financial centers (including the European Central Bank) have both the greatest potential to impose costs on others and the greatest ability to ignore those costs. The central position of the major currencies in international trade and payments means that the policies of the home governments of these currencies have an impact far beyond their borders. At the same time, precisely because the major currencies are issued by governments (or supranational authorities such as the European Central Bank) with responsibility for large and relatively closed economic areas, they may have little incentive to take into account the effects of their actions beyond their borders. Policymakers in Washington, Tokyo, or Frankfurt might simply regard the negative externalities they create as less important than the domestic political pressures to which they have to respond. This reality is compounded by the fact that whatever arrangement might be adopted would almost certainly privilege one currency over another, and the major financial powers are unlikely to be of one mind about which currency should be central. An even more encompassing observation, in the tradition of modern Political Economy, might be that it is difficult to imagine the incentives for support-maximizing politicians to undertake policies that impose concentrated and visible costs on constituents in order to reap diffuse and hardto-observe benefits. This may in fact help explain why there has been so little international monetary coordination. Nonetheless, especially in times of crisis, policymakers may find it in their own interest to seek out innovative measures – including those involving international coordination – in order to reduce the impact of severely negative economic shocks on politically relevant constituencies. Whether conditions are of this type remains to be seen. Setting aside these reservations for a moment, what might mechanisms to encourage coordination in exchange rate policies involve? The first question to answer is what their goals should be. Here it seems clear that the principal purpose envisioned by most, and which the above discussion implies, is to avoid or reverse substantial currency misalignments. Inasmuch as undervalued currencies exacerbate intergovernmental tensions in trade and other policies, and inasmuch as

overvalued currencies run the risk of subjecting other countries to contagious crises, international coordination should aim to reduce the incidence of these misalignments in exchange rates. While the goal of reducing currency volatility may also be relevant, there would seem to be somewhat less scope for that, given the ability of free-wheeling currency markets to affect short-term movements in exchange rates. How, then, might international initiatives lead governments to alter their policies in a more cooperative direction? Certainly there is little or no scope for coercive measures in this, as in most other international economic relations. International institutions are strictly limited in their ability to enforce compliance with their directives on countries that join voluntarily. They can, however, help national governments converge on punishment strategies for countries that do not conform to accepted principles, and they can also provide information that can affect the behavior of private (and public) actors. In this context, perhaps the most useful purpose for an international institution charged with monitoring exchange rates would be to make explicit, and public, a determination that a particular national currency was inappropriate, and perhaps unsustainably, misaligned. While some might argue that public identification of the misalignment of a particular currency by an international organization might impose some sort of psychic shaming costs on a government concerned about its reputation, this is unlikely. A government has pragmatic reasons to be concerned about this sort of misalignment ‘badge of shame.’ Such a finding might be seen by currency markets as an indication of the likely future path of the country’s exchange rate, prompting market actors to try to anticipate this movement. In this way, an institutional ‘finding’ of overvaluation could prompt a sell-off of the currency and force the currency down. This is somewhat less likely in the case of an undervaluation, for governments find it easier to sterilize a run toward their currency than to defend against a run on it (this is reminiscent of the adjustment asymmetry between surplus and deficit countries). Nonetheless, a declaration of misalignment from a respected, circumspect, international organization would be almost certain to provoke a market response, which would tend to push the exchange rate in the ‘right’ direction. Put differently, such a public announcement might serve to help resolve uncertainties about a government’s true commitment to an exchange rate target zone, where the target was some generally accepted notion of a realistic real exchange rate. In a sense, this sort of machinery would be analogous to the consensual establishment of target zones for all currencies. The difference is that the target would be one established by some independent authority, and presumably the bands would be quite wide – a currency

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

89

CONCLUSION

typically needs to be at least 20% above or below some notional equilibrium exchange rate to be considered seriously misaligned. The IMF’s Consultative Group on Exchange Rate Issues has long worked on a variety of methods to assess exchange rates, with an eye toward identifying deviations from the rate necessary to obtain macroeconomic or external-account balance, or to come close to an equilibrium real exchange rate. In this context, public and explicit recognition by an authoritative and respected international institution that a currency was misaligned would go a step farther than anything currently existing. A step even farther past this would be a norm, or a rule, that permitted countries faced with the impact of a recognized misalignment to act to protect themselves. This would be analogous to the right to impose trade sanctions granted by the WTO to winners in WTO disputes. In the case of a depreciated currency, the most likely corrective measure would be temporary trade barriers, although one would hope that the threat of such barriers would make their imposition unnecessary. Indeed, the fact that temporary protection would be permitted in the event of a recognized undervaluation could make the threat more credible. In the case of an appreciated currency, corrective measures might include capital or exchange controls targeted explicitly at the country, or anticipated commitments to coordinated intervention to reduce the threat of contagion if and when the currency were attacked. Which international institution could be charged with this task of managing coordination in exchange rate policies? The IMF makes the most sense, given its historical commitment to monitoring exchange rate relations and its more recently redoubled effort in this regard. One problem is that the Fund has moved so far in the direction of a development-oriented agency that it might be difficult for Fund findings to have an impact on developed governments. The Bank for International Settlements (BIS) has maintained much of its credibility with advanced industrial countries. Perhaps, as in the sovereign debt realm, some collaborative effort by the IMF and the BIS might be most appropriate. Whatever institution or mechanism was to embark on this difficult endeavor would need several things to maximize the likelihood of success: 1. A serious commitment from the monetary authorities of the major financial centers. 2. An internationally respected professional staff, at least of the caliber of the IMF and BIS research teams. 3. The political will, on the part of the participating states and the professional staff, to name names as necessary. 4. Consultation with the WTO, to both monitor tradepolicy spillovers of currency policies and ensure that

exchange rate-based trade-policy measures are consistent with WTO principles. Whether any, let alone all, of these things will be forthcoming in the near future is highly debatable. But at least it should be debated. And the recent global financial turmoil makes the stakes that much higher. The incentives for countries to act unilaterally rise substantially in times of crisis, and the seriousness of the negative externalities unilateral action can impose also rises substantially. It is easy to imagine circumstances in which uncoordinated macroeconomic policies could seriously deepen a financial crisis. While the difficulties of coordination may be great, its desirability is almost certainly growing.

CONCLUSION This essay makes the case for systematic intergovernmental cooperation on international monetary affairs. It argues that there are clear and present externalities associated with national exchange rate policies. Whatever one may believe about purely economic externalities in this realm, they are clearly present in a broader political-economy sense. Substantial currency misalignments create especially obvious negative externalities. An artificially depreciated currency creates powerful pressures for trade protection in trading partners, and can threaten the very structure of international commercial relations. An artificially appreciated currency risks ‘infecting’ other countries in the event of a contagious currency crisis. We think that the case for the existence of these ‘spillovers’ is clear, that the scope for non-cooperative strategic behavior in the currency realm is also clear, and that the intellectual case for international coordination is similarly clear. Scholars can help illuminate the options available to national governments, especially in times of crisis. Whether the intellectual clarity of the case is accompanied by incentives for policymakers to act is another matter.

SEE ALSO Political Economy of Financial Globalization: China and Financial Globalization; Financial Institutions, International and Politics.

Glossary Competitive devaluation Competitive devaluation is an example of a ‘beggar-thy-neighbor’ policy in international relations, in which

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

90

8. THE POLITICAL ECONOMY OF INTERNATIONAL MONETARY POLICY COORDINATION

governments compete to depreciate their currencies so as to shift effective demand away from imports onto domestically produced goods. The policy is counterproductive when other countries retaliate, as in the 1930s, when devaluation by one country was countered with corresponding devaluations by trading partners. In a generalized ‘currency war,’ global trade declines sharply, hurting all economies. Currency misalignment Currency misalignment indicates a sustained difference between the prevailing real exchange rate and the longrun equilibrium real exchange rate. Although there are a variety of methods to estimate equilibrium real exchange rates, the oldest is based on PPP: the idea that, in the long run, exchange rates should equalize prices across countries (the Big Mac index from The Economist is a version of this). European Monetary System (EMS) An exchange-rate system introduced by the European Community in March 1979, to reduce exchange rate variability and achieve monetary stability in Europe. Over a 10-year period, the EMS did much to reduce currency variability in Europe. However, in the early 1990s, the EMS was strained by the differing economic policies and conditions of its members, especially the newly reunified Germany. Most of its members eventually adopted the euro. Mercosur Mercosur (Spanish: Mercado Comu´n del Sur, the Common Market of the South) is a regional trade agreement among Argentina, Brazil, Paraguay, and Uruguay. Several other South American countries are associate members. Founded in 1991, Mercosur aimed to facilitate the free movement of goods, services, and factors of production by eliminating customs duties and nontariff barriers between members, and establishing a common external tariff on imports from nonmembers. While political opposition from affected industries has prevented Mercosur from eliminating all internal barriers to trade, the agreement has increased intraregional trade and deepened economic integration among members. Plaza and Louvre Accords The Plaza Accord of 22 September 1985 and the Louvre Accord of 22 February 1987 were agreements among the major industrial countries to depreciate the US dollar by engaging in coordinated intervention in currency markets. Between 1980 and 1985, the dollar appreciated by over 50% against the Yen, Deutsche Mark, and British Pound as a result of a sharp increase of the US federal deficit and high real interest rates in the United States. The strong dollar led to rising protectionist sentiment in the US

Congress, which spurred the Reagan administration to begin the negotiations that led to the Plaza and Louvre Accords.

Further Reading Bergsten, C.F., 2006. The U.S. trade deficit and China. Testimony Before the Hearing on US–China Economic Relations Revisited, Committee on Finance, United States Senate, March 29. Available at http://www.iie.com. Carranza, M., 2003. Can Mercosur survive? Domestic and international constraints on Mercosur. Latin American Politics and Society 45, 67–103. Chung, D.K., Eichengreen, B. (Eds.), 2007. Toward an East Asian Exchange Rate Regime. Brookings Institution, Washington, DC. Cline, W., 2005. The case for a new Plaza Agreement. Policy Brief in International Economics 5-4. Institute for International Economics, Washington, DC. Edwards, S., 1989. Real Exchange Rates, Devaluation and Adjustment: Exchange Rate Policy in Developing Countries. MIT Press, Cambridge, MA. Eichengreen, B., 2004. What macroeconomic measures are needed for free trade to flourish in the western hemisphere? Latin American Politics and Society 46, 1–27. Frieden, J., 2009. Avoiding the worst: international economic cooperation and domestic politics. VoxEU (February 2). Available at http:// www.voxeu.org. Goldstein, M., 2006. Exchange rates, fair play, and the ‘Grand Bargain’. Financial Times, April 21. McKibbin, W.J., 1997. Empirical evidence on international economic policy coordination. In: Fratianni, M., Salvatore, D., von Hagen, J. Handbook of Comparative Economic Policies: Macroeconomic Policy in Open Economies, vol. 5. Greenwood Press, Westport, CT, pp. 148–176. McKinnon, R.I., 2009. Solidifying a new G2. The International Economy 23, 27–29. Obstfeld, M., Rogoff, K., 2002. Global implications of self-oriented national monetary rules. Quarterly Journal of Economics 117, 503–536. Oudiz, G., Sachs, J., 1984. Macroeconomic policy coordination among the industrial economies. Brookings Papers on Economic Activity 1, 1–75. Research Department, International Monetary Fund, 2006. Methodology for CGER Exchange Rate Assessments. International Monetary Fund, Washington, DC.

I. POLITICAL ECONOMY OF FINANCIAL GLOBALIZATION

S E C T I O N

I I

THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

Intentionally left as blank

C H A P T E R

9 Theoretical Perspectives, Overview P. Bacchetta*†, P.R. Lane†{ *University of Lausanne, Lausanne, Switzerland † CEPR, London, UK { Trinity College Dublin, Dublin, Ireland O U T L I N E Introduction

93

Exchange Rates as Asset Prices

95

Net Capital Flows and the Current Account

93

Financial Globalization and the Policy Environment

95

Conclusions References

95 96

Gross Capital Flows and the Structure of International Balance Sheets

94

Capital Flows and Crises

94

INTRODUCTION

classic references mentioned in the various chapters or should simply consult well-known textbooks.

From a macroeconomic perspective, the recent process of financial globalization has raised several important issues. First, net capital flows increased substantially, in ways that have not been fully consistent with existing theories of the late twentieth century. In particular, the phenomenon of global imbalances has generated a strong debate. Second, the increase in gross capital flows has been even more spectacular, with numerous implications for asset prices, exchange rates, and macroeconomic variables (Lane and Milesi-Ferretti, 2007). Third, the financial crisis has highlighted new aspects of financial interdependence. Fourth, financial globalization has altered the policy environment in several areas. All these issues have naturally influenced recent theoretical research. While there is still much to be understood, many interesting developments have emerged in recent years. The chapters in this section of the encyclopedia give an overview of these developments. The list of topics is not exhaustive, but it still gives a wide coverage of the field. It has been deliberately chosen to focus on more recent ideas. It is noticeable that the chapters have been written mainly by researchers who have contributed directly to the recent developments. For older existing theories, interested readers should consult the more

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00048-9

NET CAPITAL FLOWS AND THE CURRENT ACCOUNT In the past decades, the basic theoretical framework used to analyze net capital flows, or the current account, has been the intertemporal approach. By focusing on the dynamics of saving and investment, this approach has been very insightful. Many of the numerous extensions are reviewed in the reference textbook of Obstfeld and Rogoff (1996) and elsewhere. Taking a more recent perspective, Paul Bergin gives a nice overview of this literature, from both a theoretical and an empirical viewpoint. He shows that this approach has empirical success in various contexts. However, several aspects of the data are difficult to explain with traditional models. One empirical fact that standard saving and investment decisions are unable to explain is the so-called Lucas paradox, which is the empirical pattern that capital tends to flow from capital-rich to capital-poor countries instead of the reverse. Related to this puzzle is the phenomenon of global imbalances, where emerging markets lend to developed countries.

93

# 2013 Elsevier Inc. All rights reserved.

94

9. THEORETICAL PERSPECTIVES, OVERVIEW

These features have generated a growing literature, extending the basic models in various directions. One direction is to introduce trade flows with sectoral heterogeneity. For example, the basic Heckscher–Ohlin twosector two-country model gives little role to capital flows as trade is a substitute for capital flows. However, in richer dynamic environments, sectoral differences lead to interesting capital flow patterns. Some of these issues are reviewed in the chapter of Keyu Jin. She shows, for example, that these features can explain the direction of capital flows from poor to rich countries or the complementarity between trade and capital flows. There has also been a growing literature on global imbalances. Although many factors influence global imbalances, a lower level of financial development has been identified as a key driver of the net capital outflow from emerging markets. In particular, a less-developed financial system has two major implications. First, the supply of domestic financial assets may be limited, which naturally leads to a demand for foreign assets. Second, the degree of individual-level risk is higher as instruments for risk sharing are limited, which implies a demand for precautionary saving that can spill over to international capital markets. These interesting developments are reviewed in the chapter written by Enrique Mendoza and Vincenzo Quadrini.

GROSS CAPITAL FLOWS AND THE STRUCTURE OF INTERNATIONAL BALANCE SHEETS For analytical convenience, earlier models of international capital flows have assumed either a single internationally traded bond or a complete set of markets (allowing full cross-border risk sharing). However, the explosion of gross capital flows combined with the obvious market incompleteness cannot be explained by these models. A natural way to introduce gross capital flows is to introduce (incomplete) portfolio diversification in dynamic models. This dimension, however, is technically challenging. A major reason is that portfolio decisions are influenced by the various assets’ risk characteristics. However, the traditional stochastic dynamic models are solved by using first-order linearizations, which are not able to measure risk. Recent research has made progress in tackling this challenge by taking two distinct paths. One direction is to add specific assumptions to the models, so that they can be solved without approximation methods. One approach in this direction comes from the finance literature and is reviewed by Anna Pavlova and Roberto Rigobon. In their review, they present the basic structure of the model and its implications for international asset pricing. They indicate the interesting recent extensions that

allow application of this approach to international macroeconomics. The second approach in introducing portfolio decisions is to consider approximation methods of higher order. These methods have recently been developed and are applied in many contexts. The chapter of Michael B. Devereux and Alan Sutherland reviews the basic idea behind these methods and several recent applications. Once the relevance of international portfolio positions is recognized, there are several further issues to consider. First, it is well known that investors tend to prefer assets from their own countries, so that there is a home bias in portfolio holdings. There can be various reasons behind this home bias. The chapter by Nicolas Coeurdacier reviews the main explanations proposed in the extensive literature on this issue. Another interesting question linked to portfolio decisions is the valuation effect. Such an effect has typically been ignored in theoretical models, but it has been prominent in the discussions of the dynamics of external balance sheets. While the implications of valuation effects are still open to debate, Ce´dric Tille reviews the issue and the related literature. Among the determinants of valuation effects are exchange rate movements. This is the main determinant when capital flows are in the form of debt contracts and are expressed in foreign currency. The experience has shown that exchange rate movements may have strong negative implications when foreign-currency debt contracts are on the liability side. Despite being a source of instability, many emerging-market countries are not able to borrow in their own currency, a phenomenon sometimes called the original sin. Marcos Chamon reviews the literature on local bond markets and the explanations for foreign-currency debt.

CAPITAL FLOWS AND CRISES The recent financial crisis is sparking growing interest in and is shedding new light on the behavior of international financial markets. It is too early to take stock of this emerging literature in the encyclopedia. However, there are two chapters that are directly related to the financial crisis. First, Mark Wright reviews the theories of sovereign debt and of debt crises. The theoretical literature was initially inspired by sovereign debt crises in emerging-market countries. While the initial theories were more conceptual, they have been refined over the years, with an attempt to match stylized facts. The ongoing sovereign debt crisis makes this literature even more relevant and the review even more useful. One of the challenges of the recent financial crisis is to explain the transmission of a shock that started in the US financial sector and spread internationally. There is active research and debate about the transmission

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

95

CONCLUSIONS

channels, but the consensus is that shocks were first transmitted through financial variables rather than real variables. A basic mechanism, coined by Krugman (2008) as the international financial multiplier, gives a central role to balance sheet adjustments of financial intermediaries. Robert Kollman and Fre´de´ric Malherbe review the main ideas and the first developments in this literature.

EXCHANGE RATES AS ASSET PRICES Financial globalization increases capital mobility, but it also increases the number of actors and the functioning of financial markets. This is particularly the case for the foreign exchange (FX) market. These changes have been accompanied by a growing literature focusing on heterogeneity and microstructure in the FX market. Better access to data has also allowed extensive empirical analysis. Part of this literature, analyzing the macroeconomic implications of microstructure models, is reviewed in the chapter written by Martin Evans. Some puzzling features in the FX market, such as excess return predictability or profitability from carry trade strategies, are related to the observed deviations from uncovered interest rate parity or the forward premium puzzle. From a theoretical perspective, it is fundamental to understand the sources of these deviations as this would give us a better understanding of the functioning of the FX market in particular and of financial markets in general. Consequently, there has been a lively literature developed in recent years proposing different approaches. The chapter of Philippe Bacchetta reviews the major contributions in this literature.

FINANCIAL GLOBALIZATION AND THE POLICY ENVIRONMENT The international policy trilemma states that international capital mobility requires each country to choose between an independent monetary policy and a stable exchange rate. However, even under an independent monetary policy, financial globalization affects the optimal conduct of monetary policy and the transmission mechanisms by which monetary policy influences the real economy. In his chapter, Mark Spiegel emphasizes that theory is ambiguous as to whether the conduct of monetary policy is made easier or more difficult by international capital mobility. This is consistent with the mixed empirical evidence as to whether monetary policy has been affected by financial globalization. However, an important task for future research will be to assess how the global financial crisis has prompted the deployment of new types of nonstandard monetary policies.

Michael Klein examines whether the conceptual framework provided by the international policy trilemma is indeed a useful guide to explaining variation in monetary policy and macroeconomic outcomes across exchange rate regimes. Indeed, as predicted by the trilemma hypothesis, an increase in international capital mobility has been associated with a sharper trade-off between exchange rate stability and an independent capability to set interest rates. While trade linkages and other microeconomic factors also influence the choice of exchange rate regime, the current European crisis underlines the strict implications for the monetary regime of giving up exchange rate flexibility. Finally, a major driver of international capital flows is differences in tax systems across countries (Lane and Milesi-Ferretti, 2008). Devereux and Fuest describe the implications of capital mobility for the design of tax systems, with a particular focus on corporate taxes and capital taxes. Although the inherent spillovers from competition in the taxation of mobile assets mean that the potential gains for international policy coordination are substantial, there has been relatively little cooperation across national authorities so far. Still, this is an area in which there may be greater scope for coordination in the future, especially within Europe and in relation to offshore centers.

CONCLUSIONS A fair assessment is that the theoretical literature on financial globalization is in a transition phase. While benchmark models of polar cases have been around for decades, it is only in recent years that researchers have started to make progress in building theoretical models that can better approximate the complexities inherent in a world of incomplete financial integration. This effort requires the synthesis of ideas and techniques from the macroeconomics and finance fields and is relevant for a broad range of modeling purposes, from the highly abstract to applied policy analysis. It is hoped that the chapters contained in this section provide a useful guide to this exciting area of research.

SEE ALSO Theoretical Perspectives on Financial Globalization: Capital Mobility and Exchange Rate Regimes; Microstructure of Currency Markets; Intertemporal Approach to the Current Account; Endogenous Portfolios in International Macro Models; Financial Contagion; Financial Development and Global Imbalances; Foreign Currency Debt; International Trade and International Capital Flows; International Macro-Finance; Monetary Policy

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

96

9. THEORETICAL PERSPECTIVES, OVERVIEW

and Capital Mobility; Theory of Sovereign Debt and Default; Tax Systems and Capital Mobility; Trade Costs and Home Bias; Explaining Deviations from Uncovered Interest Rate Parity; Valuation Effects, Capital Flows and International Adjustment.

Lane, P.R., Milesi-Ferretti, G.M., 2007. The external wealth of Nations Mark II: revised and extended estimates of foreign assets and liabilities. Journal of International Economics 73 (2), 223–250. Lane, P.R., Milesi-Ferretti, G.M., 2008. The drivers of financial globalization. American Economic Review 98 (2), 327–332. Obstfeld, M., Rogoff, K., 1996. Foundations of International Macroeconomics. MIT Press, Cambridge, MA.

References Krugman, P., 2008. The international financial multiplier, unpublished manuscript, Princeton University.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

C H A P T E R

10 Capital Mobility and Exchange Rate Regimes M.W. Klein Tufts University, Medford, MA, USA O U T L I N E Origins and Representation of the Policy Trilemma Dynamics of Exchange Rate Regimes in the Modern Era

98 99

Capital Mobility in the Modern Era

101

Evidence on the Policy Trilemma

102

Other Economic Effects of the Exchange Rate Regime Exchange Rate Regimes and Inflation Performance Fixed Exchange Rates and Trade

103 103 103

Conclusion Glossary Reference

104 104 105

This chapter on the link between capital mobility and the exchange rate regime offers more of an opportunity than that afforded while standing on one foot, and provides an overview of the study required to understand this topic in the modern era, that is, since the end of the dollar-based Bretton Woods exchange rate regime in 1973.2 The section ‘Origins and Representation of the Policy Trilemma’ presents a basic intellectual history of this term, and explains the reasoning behind the trilemma. The section ‘Dynamics of Exchange Rate Regimes in the Modern Era’ moves from theory to empirics and characterizes the dynamics of exchange rate regimes in the modern era. This is a period of increasing liberalization of capital markets, as discussed in the section ‘Capital Mobility in the Modern Era.’ Thus, the policy trilemma would dictate that, with capital mobility, countries could choose either to relinquish tight command of monetary policy in order to maintain

Forced to state all of the insights of international macroeconomics while standing on one leg, one could do worse than raise a foot off the ground and say something like “Governments face the policy trilemma – the rest is commentary.” Perhaps an economist with good balance could manage a brief explanation of the three corners of the policy trilemma triangle (as shown in Figure 10.1, these represent exchange rate management, monetary policy autonomy, and free capital mobility), and how a government must choose two of these three policies, represented by a single side of this triangle. And an economist with the ability to maintain a particularly stable personal equilibrium could go on about the trade-offs between these options. But this would likely be the full extent of an explanation that one could manage before wobbling, putting both feet firmly on the ground, and, similar to Hillel, conclude by telling the inquisitor to go and study.1

1

The first century Rabbi Hillel was asked to distil the teachings of the Torah while standing on one leg and responded “That which is hateful to you, do not do to your fellow. That is the whole Torah; the rest is commentary. Now go and study.”

2

The designation of the period since the end of the Bretton Woods exchange rate system as the modern era comes from the author’s work with Jay Shambaugh in Exchange Rate Regimes in the Modern Era (2010, MIT Press). I would like to express my debt to Shambaugh for years of discussion of this topic, and to Maurice Obstfeld and Alan Taylor for conversations on the origins of the ideas of the trilemma and its modern presentation.

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00015-5

97

# 2013 Elsevier Inc. All rights reserved.

98

10. CAPITAL MOBILITY AND EXCHANGE RATE REGIMES

currencies of the various members of the system in terms of the international standard, and to preserve at the same time an adequate local autonomy for each member over its domestic rate of interest and its volume of foreign lending. (p. 272)

Free capital mobility

A

Exchange rate management

This dilemma becomes a trilemma when capital mobility is a third possible choice. The term ‘trilemma’ was coined by Maurice Obstfeld and Alan Taylor in their 1998 article ‘The Great Depression as a watershed: international capital mobility over the long run.’ The first use of this term appears in their statement:

B

C

Monetary autonomy

FIGURE 10.1 The policy trilemma.

greater control over the exchange rate (corresponding to side A in Figure 10.1) or to have monetary policy autonomy but with the loss of an ability to manage the exchange rate (corresponding to side B in Figure 10.1). The section ‘Evidence on the Policy Trilemma’ shows that these two choices are the most common ones today, as most countries allow capital mobility, or at least find it difficult to episodically stifle the volume of capital flows. Some economic consequences of the choice of a pegged or a floating exchange rate regime, beyond that implied by the policy trilemma, are the subject of the section ‘Other Economic Effects of the Exchange Rate Regime.’ The section ‘Conclusion’ offers concluding comments.

ORIGINS AND REPRESENTATION OF THE POLICY TRILEMMA ‘The Dilemma of an International System’ is the title of the first section of Chapter 36 of John Maynard Keynes’ A Treatise on Money (this chapter itself is titled ‘Problems of International Management – III. The Problem of National Autonomy’). Keynes explains that with international capital mobility and “an international standard” (i.e., a worldwide fixed exchange rate), “. . . the rate of interest would have to be the same throughout the world.” Any deviation from this interest rate would lead to gold flows that would force a country to lose its gold holdings (if it kept its interest rate lower than the world level) or amass all of the world’s gold (if it kept its interest rate higher). The consequence would be that “. . . the degree of its power of independent action would have no relation to its local needs” (p. 271). More specifically, Keynes (1930) writes This, then, is the dilemma of an international monetary system – to preserve the advantages of the stability of the local

Secular movements in the scope for international lending and borrowing may be understood, we shall argue, in terms of a fundamental macroeconomic policy trilemma that all national policymakers face: the chosen macroeconomic policy regime can include at most two elements of the ‘inconsistent trilogy’ of (i) full freedom of cross-border capital movements, (ii) a fixed exchange rate, and (iii) an independent monetary policy oriented toward domestic objectives. (p. 354)

The first presentation of the now familiar trilemma triangle diagram (shown earlier) was in the fifth edition of the textbook International Economics: Theory and Policy by Paul Krugman and Obstfeld (2000, p. 647) although, in that presentation, one corner of the triangle represented a currency board, a special case of a fixed exchange rate. The consequences of the trilemma are sharpened when it is understood in the context of explicit international macroeconomic model. Some version of uncovered interest parity (UIP) will be included in that model. The simplest form of UIP requires that expected returns of bonds that differ only in their currency denomination are equal, that is,

 ðPt þ Itþ1 Þ=Pt ¼ Eetþ1 Pt
þ Itþ1 =Et Pt

e   ¼ Etþ1 =Et  Pt þ Itþ1 =Pt where Pt is the price of a one-period domestic currency bond purchased in period t , Itþ1 is the interest paid on * are the principal that bond in period t þ 1, Pt* and Itþ1 and interest, respectively, on a comparable one-period foreign-currency denominated bond, which has the same maturity, riskiness, and liquidity as the domestic bond, Et is the exchange rate (units of domestic currency needed to purchase one unit of foreign currency) e is the expectation, at time t, of the exat time t, and Etþ1 change rate at time t þ 1. Define the interest rate on the domestic bond as Rt ¼ (Pt þ Itþ1)/P and that on the * )/Pt*), and noting that foreign bond as R*t ¼ ((Pt*þItþ1 e /Et) represents 1 þ %DEet, tþ1 where %DEt,e tþ1 is the (Etþ1 expected rate of depreciation of the domestic currency between time t and time t þ 1. The above equation can be rewritten as  
1 þ Rt ¼ 1 þ %DEet;tþ1  1 þ Rt

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

DYNAMICS OF EXCHANGE RATE REGIMES IN THE MODERN ERA

which is approximately equal to Rt ¼

Rt

þ

%DEet;tþ1

since (R*t  %DEt,e tþ1 ) is an order of magnitude smaller than either R*t or %DEt,e tþ1.3 The UIP equation can be directly related to the trilemma diagram. The domestic interest rate represents monetary autonomy, the expected rate of depreciation represents exchange rate management, and capital mobility is represented by the equals sign because of the arbitrage that occurs when investors can freely purchase or sell foreign currency-denominated assets. This equation also points out some more subtle points. Domestic monetary autonomy may be retained in an asymmetric fixed exchange rate system if the exchange rate peg is maintained by foreign monetary authorities following the policies of domestic monetary authorities (i.e., the peg is maintained by adjustments in R*t rather than Rt). Also, to the extent that there is a belief that fixed exchange rates may not be maintained (i.e., %DEt,e tþ1 6¼ 0 even though the currency is pegged), the domestic interest rate and the foreign interest rate will diverge.4 Finally, the constraints of the policy trilemma are eased if interest parity includes a risk premium that can be systematically affected by sterilized intervention that alters the relative stocks of bonds of the two countries without changing either’s money supply.5

DYNAMICS OF EXCHANGE RATE REGIMES IN THE MODERN ERA A person whose understanding of international economic affairs was based solely on a close reading of an average textbook might think that the countries of the world neatly divided into those that peg their currency and those that allow their currency to float.

99

Alternatively, a person whose view of exchange rate regimes was shaped by news events could be excused for thinking that most, if not all, efforts to fix exchange rates are unsustainable and end in tears. An analysis of the actual behavior of exchange rate regimes, however, shows that the modern era is characterized by distributions of fixed-rate episodes and floating-rate episodes that look surprisingly similar to each other. The experience of countries in general, and even of some single countries in particular, are represented by long-lived pegged exchange rates, sustained continuous periods during which exchange rates float and are determined by market forces, and flips from pegs to floats and back again. Figures 10.2 and 10.3 present histograms of the duration of fixed-rate spells and floating-rate spells, respectively, for 125 countries during the period 1973–2004. An exchange rate spell is an episode during which a country has one or more years with either a fixed exchange rate or a flexible exchange rate.6 There are 398 peg spells and 395 float spells in this data set. The average length of a peg spell is 4.7 years, the average length of a float spell is 5.2 years, and the median length of both is 2 years.7 The skewness in the distribution implied by the difference between the mean and the median lengths of spells is evident in the two figures. These show only a few instances where countries pegged or floated over the entire 32-year period. But this should not be taken to mean that all spells are fleeting. Fiftyeight percentage of the peg spells and 64% of float spells lasted for five or more years. These high percentages reflect the fact that long-lived spells are not the province of only a small set of countries. Of the 125 countries in the sample, 78 had at least one peg spell that lasted for five or more years, and 88 had at least one float spell that lasted for five or more years.8 The large number of very short spells represented in Figures 10.2 and 10.3 imply that there are many instances

3

Alternatively, a risk premium can be introduced as an additive term in this equation if one assumes that investors are concerned with the expected variance of their portfolio, and not just the expected return.

4

Note that Keynes implicitly assumes fully credible fixed rates in his analysis quoted above.

5

In practice, there is little support that sterilized intervention is, in fact, effective. Similarly, the trilemma is also eased if fiscal policy can be used to manage the economy as readily and nimbly as monetary policy. The practical relevance of this challenge to the trilemma, however, is limited.

6

These figures are from Klein and Shambaugh (2010). As discussed in this book, these spells are based on actual bilateral exchange rates between a country and its base country. A peg spell represents a series of years during which this bilateral exchange rate varies by 0; ð11:4cÞ

110

11. MICROSTRUCTURE OF CURRENCY MARKETS

where At1 denotes the investors’ aggregate holding of II FX at the end of day t  1, and OD, t denotes the common information of dealers and the broker at the start of round II. 2. The trades initiated by dealer d in round II are II ¼ az Zd;t þ aA At1 ; Td;t

ð11:5Þ

and the customer orders received by dealer d in round I are

where

PD

ZId;t ¼ ðb=DÞYt þ ed;t ;

b < 0;

ð11:6Þ

d ¼ 1 ed,t ¼ 0.

There are several noteworthy features of this equilibrium. First, consider its implications for the behavior of the spot exchange rate at the daily frequency:  II 1 III  SIII t  St1 ¼ Vt þ lX ðXt  E½Xt Ot Þ r

ð11:7Þ

Daily changes in the spot exchange rate are driven by shocks to dividends and unexpected interdealer order flow. The former reflect the effects of public news while the latter conveys information that was initially dispersed across investors and was then aggregated via trading in the FX market. Dividend shocks play a familiar role in the determination of the spot rate. Realizations of Vt are public information, and affect the forecasts of future dividends by all dealers and investors in exactly the same way. Consequently, it should come as no surprise that Vt shocks are immediately reflected in the equilibrium spot rate. That said, it is important to remember that quotes are chosen optimally in this model, so Vt shocks affect only the spot rate because dealers have an incentive to adjust their quotes once the value of Vt is known (see Eq. (11.4a)). The role played by aggregate interdealer order flow in Eq. (11.7) is more complex. Notice that it is unexpected interdealer order flow that affects the exchange rate in Eq. (11.7). The reason is that dealers adjust their quotes between rounds II and III to account for the new information contained in the aggregate interdealer order flow (see Eq. (11.4a)). The customer orders received by each dealer reflect the difference between the desired and actual FX positions of individual investors. As such, they convey information to dealers about both the current income and the overnight positions of a subset of investors. This information is effectively shared between dealers via interdealer trading in round II. As a result, Xt conveys information about aggregate income and the prior overnight FX position of all investors. Since the latter is already known to dealers, Xt  E½Xt jOIt  is proportional to the new information concerning income that is incorporated into the round III quote, StIII. Of course, the mere fact that unexpected interdealer order flow conveys new information about aggregate

income to dealers does not explain why Xt  E½Xt jOIt  appears in Eq. (11.7). For that we need to understand why dealers find it optimal to incorporate the new income information they learn into their common round III quote. In short, why is information on aggregate income, Yt, relevant for the pricing of FX? The answer is quite simple. As in the actual market, dealers do not want to hold FX overnight – the risk of holding FX can be shared more efficiently by investors than by individual dealers. Consequently, each dealer’s aim in round III is to quote a price that will induce investors to purchase all the FX currently held by dealers. In other words, the round III quote is chosen so that the excess overnight return expected by investors is such that they want to hold the entire existing stock of the FX. Obviously, this would not be possible unless all dealers can calculate what the existing stock of FX is. However, since investors’ income is the only source of FX, the existing stock can be computed from the history of aggregate income. Thus, information on Yt is price relevant because it reveals to dealers what aggregate overnight FX position investors must be induced to hold. In sum, the interdealer order flow conveys information about the shift in the FX portfolios of investors, necessary to achieve efficient risk sharing. This is the origin of the term portfolio shifts. Another important feature of the model concerns timing. As noted above, public news concerning current and future dividends is immediately and fully incorporated into the spot exchange rate. By contrast, it takes time for the information concerning income to be reflected in the dealer quotes. The reason for the delay is important. Information about income is originally transmitted to dealers via the customer orders they receive in round I. Thus, each dealer has some information about Yt at the start of round II, but the information is imprecise. At this point, each dealer could choose to use his or her private information on Yt in setting the quote, but this is not optimal in the model’s trading environment. Instead, the dealers’ best strategy is to quote the same price as in round I (which is the same across all dealers and only a function of common round I information), because to do otherwise would expose the dealer to arbitrage trading losses. As a result, the equilibrium spot rate remains unchanged between rounds I and II even though dealers have information about aggregate income. The spot rate only incorporates this information when it becomes common knowledge among dealers. This process of information aggregation takes place via interdealer trading in round II. The best strategy of each dealer is to use their private information concerning income in determining the trade they wish to initiate with other dealers. It is for this reason that interdealer order flow provides information on aggregate income that becomes common knowledge to dealers by the start of round III.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

111

CURRENCY TRADING MODELS

Empirical Evidence

TABLE 11.1

Empirical support for the PS model first came from regressions of the daily depreciation rate on interdealer order flow and the change in the nominal interest differential: Dst ¼ b1 Xt þ b2 Dðrt  ^rt Þ þ Bt ;

DM/USD

Xt

Regressors Dðrt  ^rt Þ

I

2.14** (0.29)

0.51 (0.26)

II

2.15** (0.29)

ð11:8Þ

where Dst  stst  1 is change in the log spot rate between the end of days t  1 and t, rt is the nominal dollar interest rate and ^rt is the nominal nondollar interest rate. These regressions are motivated by Eq. (11.7) with Dst III replacing StIII  St 1, and actual interdealer order flow, Xt, replacing Xt  E½Xt jOIIt . The first of these substitutions makes the empirical specification comparable to standard macro models but has no significant effect on the estimation results; the second is motivated by the assumption that the expected interdealer order flow is zero. The nominal interest differential is included to account for the arrival of public news. Table 11.1 shows regression results for the DM/USD and JPY/USD spot rates between 1 May and 31 August 1996. The data on interdealer order flow comes from the Reuters Dealing 2000–1 system. This was the dominant platform for interdealer trading at the time, accounting for approximately 90% of all ‘direct’ dealerto-dealer trades. The table reports the coefficient estimates and standard errors in parenthesis for five versions of Eq. (11.8). Three features of the estimation results are particularly noteworthy: 1. The coefficient on order flow, Xt, is correctly signed and significant, with t-statistics above 5 in versions I, II and IV of the equation for both the DM/USD and JPY/USD. The positive sign indicates that net dollar purchases – a positive Xt – lead to a higher FX price for the dollar. For perspective, the estimated value of 2.1 for the order flow coefficient in the DM/USD equation translates to $1 billion of net purchases, increasing the DM price of a dollar by 0.8 pfennigs. 2. Almost all the explanatory power in the regressions is due to order flow. In specifications III and V where order flow is omitted, the R2 statistics are less than 1% in both the DM/USD and JPY/USD equations. Moreover, the explanatory power of order flow is extraordinarily high. The R2 statistics of 64% and 45% for the DM and JPY equations that include order flow are an order of magnitude higher than those found in other exchange rate models. 3. The coefficient on the change in the interest differential is correctly signed but is significant only in the JPY/USD equation. The positive sign arises in the sticky-price monetary model, for example, because an increase in the US interest rate requires an immediate dollar appreciation (i.e. an increase in JPY/USD) to

Estimates of the Portfolio Shifts Model

III IV

2.15** (0.29)

2.86** (0.36)

II

2.61** (0.36)

III IV

0.64

0.62 (0.77)

I

R2

0.63

V JPY/USD

rt1  ^rt1

0.01 0.02* (0.01)

0.64

0.02 (0.02)

0.00

2.47** (0.92)

0.46 0.40

0.57 (1.20) 2.78** (0.38)

V

0.00 0.02* (0.01)

0.42

0.01 (0.01)

0.00

Note: The table reports slope coefficients and standard errors. Statistical significance at the 5% and 1% level is denoted by * and **. Source: Evans, M.D.D., Lyons, R.K., 2002. Forecasting exchange rate fundamentals with order flow. Working Paper. Georgetown University.

make room for UIP-induced expected dollar depreciation. Figure 11.2 provides further perspective on the results in Table 11.1. Here the solid plots show the path of the spot exchange rate over the 4-month sample. The dashed plot shows cumulative order flow over the same period measured against the right hand axis. Cumulative order flow is Psimply the sum of daily order flow, that is, Xt1 ¼ it ¼ 1 Xi. If all variations in spot rates reflect quote revisions driven only by order flow, the daily change in the spot rate should be proportional to daily order flow; that is, St – St–1 ¼ lXXt. Under these circumstances, St ¼ lXXt1 so the plots should coincide. Figure 11.2 shows that this is not quite the case. Nevertheless, it is hard not to be impressed by the close correspondence between exchange rates and order flow over this sample period. Figure 11.2 highlights two further features of the relation between order flow and currency returns. First, although Eq. (11.8) is estimated at the daily frequency, the estimation results have implications for the behavior of spot rates over much longer periods. In particular, since there is almost no serial correlation in daily depreciation rates, the k-day change in the log spot rate is wellapproximated by

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

112

11. MICROSTRUCTURE OF CURRENCY MARKETS 600

112

1.54

3000

110 200

2500

0 - 200

1.48

- 400

X

1.5

108

- 600

1.46

YEN/$

DM/$

1.52

2000

106

1500 1000

104

500

- 800 1.44

- 1000 - 1200

1.42 1 8 15 22 29 36 43 50 57 64 71 78

DM/USD

FIGURE 11.2 Exchange rates (solid) and cumulative order flow (dashed). Source: Evans, M.D. D., Lyons R.K., 2002. Order flow and exchange rate dynamics. Journal of Political Economy 110(1), 170–180.

3500

400

X

1.56

102 100

0 - 500 1 8 15 22 29 36 43 50 57 64 71 78

JPY/USD

st  stk ffi b1 Xtk þ b2 Dk ðrt  ^rÞ þ Bt;k ; P where Xtk ¼ i¼0k  1Xti denotes order flow during the past k days, and b2 Dk ðrt  ^rt Þ þ zt;k identifies the impact of public news arriving during the same period. Thus, the rate of depreciation over k days will be well-approximated by the cumulative effects of order flow, Xt,k, and the arrival of public news. Figure 11.2 shows that there are many instances where st  stk ffi b1Xtk for significant horizons k. The second feature concerns the dynamics of order flow. The plots of cumulative order flow in Figure 11.2 display no significant mean reversion because there is no detectable serial correlation in daily order flows. This feature of the data is important because the empirical specification in Eq. (11.8) assumes that realizations of daily order flow represent news to dealers. Thus daily order flow must be serially uncorrelated if the results in Table 11.1 are to be consistent with the predictions of the PS model. The estimates in Table 11.1 are representative of those obtained in a large number of studies. The strong empirical link between interdealer order flows and daily changes in the spot exchange rate extends across different currencies, different time periods, and over horizons ranging from a few minutes to several weeks. The results also extend to customer order flow. Recall that in the PS model interdealer order flow conveys price-relevant information to dealers which initially entered the market in dispersed form via the customer orders received by individual dealers. Consequently, we should expect to find that customer order flows also have explanatory power in accounting for exchange rate changes. This is indeed the case. Customer flows disaggregated by customer type have more explanatory power for exchange rate changes than the aggregate flows received by individual banks. At the daily horizon, disaggregated flows account for less of the variation in exchange rate changes than aggregate interdealer order flows, but the explanatory power of customer and dealer flows are comparable over horizons of one to several weeks.

FROM MICRO TO MACRO Microstructure exchange rate models, such as the PS model, are silent on the potential links between the macro economy and order flows because they focus on the details of currency trading between different market participants rather than the macroeconomic factors that ultimately drive the individual trades. To address this deficiency, recent research focuses on the role that order flow plays in conveying macro information to the FX market. This research builds on two central ideas: First, only some of the macro information relevant for the current spot rate is publicly known at any point in time. Other information is present in the economy, but it exists in a dispersed microeconomic form. The second idea relates to determination of the spot rate. As it is literally the price of foreign currency quoted by FX dealers, the spot rate can only embed information that is known to dealers. Consequently, the spot rate will only reflect dispersed information once it has been assimilated by dealers – a process that takes place via trading. An overview of a micro-based model that incorporates these ideas is provided along with a discussion of its empirical implications.

A Micro-Based Model The economy comprises two countries populated by a continuum of risk-averse agents, indexed by n 2 [0, 1], and D risk-averse dealers who act as marketmakers in the spot FX market. The home and foreign countries are referred to as the United States and Europe, so the log spot exchange rate, st, denotes the dollar price of euros. The only other actors in the model are the central banks (i.e., The Federal Reserve and the European Central Bank), who conduct monetary policy by setting short-term nominal interest rates, rt and ^rt . Overview The model focuses on the currency trades between agents and dealers at a weekly frequency. At the start

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

113

FROM MICRO TO MACRO

of week t, all dealers and agents observe a data release that provides information on the state of the economy several weeks earlier. Let zt denote a vector of variables that completely describe the state of the macroeconomy in week t. This vector contains variables that are directly observable, such as short-term interest rates, and others that become publicly known only with a lag. Each agent n also receives private information concerning his or her current microeconomic environment: ztn ¼ zt þ vtn, where vtn ¼ [vi,nt ] is aRvector of agent-specific shocks with the property that 10 vi, nt dn ¼ 0 for all i. Notice that the ztn vector represents dispersed private information about macro conditions. Next, all dealers quote a log USD/EUR spot price equal to rt  rt  dt s t ¼ ED t stþ1 þ ^

ð11:9Þ

where ED t denotes expectations conditioned on the common information available to all dealers at the start of week t. This information set includes the histories of interest rates and data releases, and the order flows from earlier currency trading. The FX risk premium, dt, is chosen by dealers to share risk efficiently across the FX market. Macroeconomic conditions affect dealers’ quotes in Eq. (11.9) via their effects on current monetary policy (i.e., ^rt  rt ), and via ED t stþ1 insofar as these expectations are affected by dealers’ forecasts for the future interest differential. These forecasts are assumed to incorporate a view on how central banks react to changes in the macroeconomy. In particular, rtþi  rtþi Þ ¼ ð1 þ gp ÞED ptþ1þi  Dptþ1þi Þ ED t ð^ t ðD^ þ g y ED ð^ y  ytþi Þ  ge ED t t etþi ð11:10Þ tþi for i > 0, where gp, gy, and ge are positive coefficients. Equation (11.10) implies that dealers expect the future differential between euro and dollar rates to be higher (i) the future difference between EU and US inflation, D^ ptþ1  Dptþ1 , is higher; (ii) the difference between the EU and US output gaps, ^ yt  yt , widens; or (iii) the pt  pt , depreciates. These real exchange-rate, et  st þ ^ expectations are consistent with the widely accepted views on how central banks react to changing macroeconomic conditions. Dealers stand ready to fill the foreign currency orders of agents at their quoted price of st. Once agent n 2 [0, 1] observes st, they place their currency order with a dealer to fill their desired demand for euros. The order flow for euros from agent n is therefore atn  atn 1 where atn denotes the week-t demand by agent n. This demand is given by ant ¼ as ðEnt Dstþ1 þ ^rt  rt Þ þ hnt

ð11:11Þ

where as > 0 and Ent denotes expectations conditioned on the information available to agent n after the spot rate is

quoted at the start of week t. The demand for euros depends on the log excess return expected by the agent and a hedging term, htn, which represents the influence of all other factors. Without loss of generality, we may assume that htn ¼ azztn, for some vector az, where the vector ztn describes the microeconomic environment of agent n. The aggregate demand for euros by agents is thus given by Z 1  n stþ1  st þ ^rt  rt Þ þ ht ð11:12Þ ant dn ¼ as ðE at ¼ t 0

R where ht ¼ 01htn dn is the aggregate hedging demand  n denotes the average of agents’ expectations: and E t R R n  stþ1 ¼ 1 En stþ1 dn. Notice that ht ¼ az 10ztn dn ¼ azzt, so E t 0 t the aggregate hedging demand depends on the state of the macroeconomy, zt. For the remainder of the week, dealers trade among themselves. The model abstracts from the details of interdealer trade. Instead, consistent with the PS model, it assumes that all dealers learn the aggregate order flow that resulted from the earlier week-t trades between agents and dealers. This order flow is determined by Xtþ1 ¼ at  at1

ð11:13Þ

Notice that order flow is driven by changes in the average of agents’ expectations concerning future excess returns and the change in the aggregate hedging demand. (The t þ 1 subscript makes clear that order flow is only observed by dealers at the end of week-t trading.) Equilibrium An equilibrium in this model comprises a sequence of spot rates and market-clearing order flows that support and are supported by the endogenous evolution of dealers’ common information and agents’ private information. Recall that dealers and agents observe current interest rates and the data releases containing information on past macroeconomic conditions. Dealers also obtain information from their observations on order flow that is endogenously determined by the trading decisions of agents. Similarly, each agent observes the endogenously determined spot rate together with their microeconomic environment. Solving for equilibrium spot rates and order flows is complex. But, so long as public information about the current state of macroeconomy is incomplete, equilibrium order flows provide dealers with information about the macroeconomy that they will embed in the spot rate quotes, information that was initially dispersed across agents as their observations on ztn. In equilibrium, the spot rate quoted by dealers satisfies Eq. (11.9) subject to the risk-sharing restriction that identifies the risk premium, and dealers’ expectations concerning future interest rates in Eq. (11.10):

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

114

11. MICROSTRUCTURE OF CURRENCY MARKETS

st ¼ ð^rt  rt Þ þ ED t

1 X

ri ftþi  ED t

i¼1

1 X

ri dtþi

ð11:14Þ

i¼0

with r  1/(1 þ ge) < 1, where p  Dptþ1  Dptþ1 Þ þ gy ð^ y t  yt Þ ft ¼ ð1 þ gp ÞðD^   tþ1 1r pt Þ þ r ðpt  ^ The three terms on the right of Eq. (11.14) identify different factors affecting the log spot rate dealers quote at the start of week t. First, the current stance of monetary policy affects dealers’ quotes via the interest differential, ^rt  rt , because it directly contributes to the payoff from holding euros until week t þ 1. Second, dealers are concerned with the future course of macro fundamentals, ft. This term embodies dealers’ expectations of how central banks will react to macroeconomic conditions when setting future interest rates. The third factor arises from risk sharing between dealers and agents as represented by the present and expected future values of the risk premium, dt. These values are determined by   1 e ð11:15Þ s  dt ¼ E D t tþ1 as ht  n stþ1 . Intuitively, dealers stabilize where setþ1 ¼ stþ1  E t their euro holdings by lowering the risk premium when they anticipate a rise in the aggregate hedging demand for euros because the implied fall in the excess return that agents expect will offset their desire to accumulate larger euro holdings. Dealers also reduce the risk premium to offset agents’ desire to accumulate larger euro holdings when they are viewed as being too optimistic (on average) about the future spot rate; that is, when D n ED t stþ1 < Et Et stþ1 . In equilibrium, the aggregate order flow that dealers observe at the end of week-t trading reflects the change in aggregate demand for foreign currency by agents between weeks t – 1 and t: Xtþ1 ¼ at – atP – 1. As dealers know the history of order flow, and at1 ¼ 1 i ¼ 0 Xti by market . Consequently, unexpected order flow clearing, at1 2 OD t from week-t trading can be written as    n D

 D  Xtþ1  ED t Xtþ1 ¼ az zt  Et zt þ as Et stþ1  Et stþ1 ð11:16Þ Thus order flow contributes new information about the current macroeconomic conditions, zt, via the aggregate D hedging demand, ht  ED t ht ¼ az ðzt  Et zt Þ, and about  n stþ1 . These the average of agents’ spot rate forecasts, E t forecasts will embed agents private expectations concerning future interest rates, and the macroeconomic conditions that affect monetary policy.

Empirical Evidence We can use Eqns. (11.14)–(11.16) to examine the models’ empirical implications. Consider, first, the

behavior or excess returns, ertþ1 ¼ Dstþ1 þ ^rt  rt . Equation (11.9) implies that, ertþ1 ¼ dt þ stþ1  ED t stþ1 , so substituting for the dealers’ forecast errors gives   rtþ1  rtþ1 Þ ertþ1 ¼ dt þ 1  ED t ð^ 1   X  D  þ Etþ1  ED ri1 ftþi  r1 setþi  ra1 s htþi1 t i¼2

ð11:17Þ Notice that the second and third terms depend on the flow of information reaching all dealers between the start of weeks t and t þ 1. This information flow comprises observations on current interest rates, data releases on past macro variables, and order flow from week-t trading, Xt þ 1. Equation (11.17) implies that any of these information sources will affect the excess returns insofar as they convey new information on current monetary policy, future fundamentals, and the risk premia. Thus the model attributes the strong correlation between foreign currency returns and order flows observed in the data to the important role order flow plays in this information transmission process. We can examine this implication of the model with a two-step regression procedure. In the first step we regress a macro variable wt on the P cumulation of current and past order flows, Xt1þ 1 ¼ i1¼ 0 Xt þ 1  i. Under effi1 ¼ Xtþ1  ED cient risk sharing Xtþ1 t Xtþ1 , so the regression coefficient on Xt1þ 1 should be statistically significant if unexpected order flow from week-t trading contains information about macro variable wt. In the second step we regress excess returns on the predicted values from the first stage. The estimated slope coefficient in this regression should be statistically significant when the information concerning wt conveyed by order flow is price relevant. Table 11.2 reproduces the results from estimating the second stage regression as reported in Evans (2010). The dependent variable is the excess return on the USD/EUR spot over a 4 week horizon. The dependent variables in panel A are the predicted values from regressions of gross domestic product (GDP) growth, consumer price index (CPI) inflation, and M1 growth in the United States and Germany on cumulated customer order flows for euros received by Citibank. In panel B, the dependent variables are the differences between the predicted values for the German and US macro variables. The regressions are estimated at the weekly frequency in data from January 1993 to June 1999. The results in Table 11.2 provide strong support for the idea that order flows convey price-relevant information about GDP, prices, and the money stock. The slope coefficients for German inflation, US GDP growth, inflation, and monetary growth are all significant at the 1% level. Insofar as spot rates reflect the difference between US and EU monetary policy, order flows should carry

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

115

FROM MICRO TO MACRO

TABLE 11.2

Excess Returns and Macro Information GDP

A

CPI

Germany

United States

0.14 (0.11)

0.68** (0.10)

Germany

M1 United States

0.11 (0.80)

B

United States

R2 0.10

0.50** (0.20)

0.14 (1.90)

Germany

0.51** (0.11)

1.14 (1.45)

0.48 (0.83)

0.11 0.12 (0.07)

0.70** (0.14)

0.19

1.65 (2.28)

0.69 (1.15)

0.29

0.43** (0.08)

0.15 0.17** (0.04)

0.90** (0.43)

0.04

0.52** (0.25)

0.26** (0.04)

0.11

0.19 (0.30)

0.23

Note: The table reports slope coefficients and standard errors from the second stage regression using the predicted values for the variables listed at the head of each column. Estimates are calculated at the weekly frequency. The standard errors correct for heteroskedasticity and an MA (3) error process. Statistical significance at the 5% and 1% level is denoted by * and **. Source: Evans, M.D.D., 2010. Order flows and the exchange rate disconnect puzzle. Journal of International Economics 80(1), 58–71.

more price-relevant information about the difference in macroeconomic conditions between countries. This seems to be the case. As panel B shows, the slope coefficients on the predicted values for GDP growth and inflation are highly significant. The model also holds implications for the links between the dynamics of spot rates, order flows and the future evolution of the macro economy. In particular, Eq. (11.14) implies that st ¼ st þ

1 X 1 ED ri ðDftþi  Ddtþi Þ 1  r t i¼1 tþi

ð11:18Þ

where st ¼ ð^rt  rt Þ þ ðr=1  rÞED t ft  ð1=1  rÞdt . This equation splits the factors affecting dealers’ spot rate quotes into two terms. Current conditions, such as interest rates and dealers’ estimates of week-t fundamentals affect dealers’ quotes via s*. t Expected changes in future conditions (i.e., inflation, the output gaps, and the risk premium), affect the spot rate via the second term. It is this term that provides the link between spot rates, order flows, and the future path of macro variables. Equation (11.18) implies that st  s*t will have forecasting power for any future macro variable, wt þ t, if dealers’ exD pectations, ED t wtþt , are correlated with Et ðDftþi  Ddtþi Þ. Intuitively, dealers will raise their spot rate quotes relative to st* when their forecasts of future changes in fundamentals increase; so if these forecasts are correlated with ED t wtþt , variations in st  st* will have forecast

power for wt þ t. Equation (11.18) also implies that order flow should generally have incremental forecasting power for future macro variables beyond that contained in st  s*. t To understand why, consider the following identity: D D D wtþt ¼ ED t wtþr þ ðEtþ1  Et Þwtþt þ ð1  Etþ1 Þwtþt ð11:19Þ

The first term on the right identifies dealers’ expectations concerning wtþt based on the information they use to quote spot rates at the start of week t. The second term identifies the revision in dealers’ forecasts between the start of weeks t and t þ 1. The incremental forecasting power of order flow comes from this term, which, by construction, is uncorrelated with ED t wtþt . In particular, any information conveyed by Xtþ1  ED t Xtþ1 concerning wtþt will lead dealers to revise their forecasts of wtþt. This implication of the model can be examined with a regression of wtþt on st and other variables known to dealers at the start of week t, and unexpected order flow from trades between dealers and agents during week t. The coefficients on the order flows should be statistically significant if they have incremental forecasting power for the macro variable in question. Furthermore, the portion of the predicted value for wtþt attributable to the order flows should make a significant contribution to the variance of wtþt if the information conveyed by the order flows is economically significant.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

116

11. MICROSTRUCTURE OF CURRENCY MARKETS

Figure 11.3 reproduces the plots from Evans (2011) that show the variance contributions of the order flows together with 95% confidence bands for the six macro variables considered in Table 11.2 for horizons t ¼ 1, . . ., 26 weeks. In five of the six cases, the contributions rise steadily with the horizon and are quite sizable beyond one quarter. The exception is US GDP growth, where the contribution remains around 15% from the quarterly horizon onward. These plots clearly show that order flows have considerable forecasting power for the future flows of information concerning GDP growth, inflation, and M1 growth at all but the shortest horizons.

Clearly, then, these order flows are carrying significant information on future macroeconomic conditions.

MICRO PERSPECTIVES ON EXCHANGE RATE PUZZLES Micro-based models provide a new perspective on some long-standing puzzles concerning the behavior of exchange rates over horizons ranging from a few minutes to several years. The micro-based perspective on the disconnect and news puzzles is discussed next.

0.50

0.50

0.45

0.45

0.40

0.40

0.35

0.35

0.30

0.30

0.25

0.25

0.20

0.20

0.15

0.15

0.10

0.10

0.05

0.05

0.00

0

2

4

6

8

10 12 14 16 18 20 22 24 26

0.00

0

2

4

6

A: US GDP growth

8

10 12 14 16 18 20 22 24 26

B: German GDP growth 0.50

0.60

0.45 0.50

0.40 0.35

0.40

0.30 0.25

0.30

0.20 0.20

0.15 0.10

0.10

0.05 0.00

0

2

4

6

8

10 12 14 16 18 20 22 24 26

0.00 0

2

4

6

0.50

8

10 12 14 16 18 20 22 24 26

D: German inflation

C: US inflation 0.60

0.45 0.50

0.40 0.35

0.40

0.30 0.25

0.30

0.20 0.20

0.15 0.10

0.10

0.05 0.00 0

2

4

6

8

10 12 14 16 18 20 22 24 26

E: US M1 growth

0.00

0

2

4

6

8

10 12 14 16 18 20 22 24 26

F: German M1 growth

FIGURE 11.3

Estimated contribution of order flows to the variance of future GDP growth, inflation, and M1 growth by forecasting horizons ^ is the standard error of the estimated contribution. measured in weeks. Dashed lines denote 95% confidence bands computed as  1.96^ s, where s Evans, M.D.D., 2011. Forecasting exchange rate fundamentals with order flow. Working Paper. Georgetown University.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

MICRO PERSPECTIVES ON EXCHANGE RATE PUZZLES

The Disconnect Puzzle The exchange-rate disconnect puzzle refers to the observation that most short- and medium-term variations in spot exchange rates appear to be unrelated to changes in current or expected future macro variables that could identify fundamentals. In other words, exchange-rate variations appear essentially disconnected from changing macroeconomic conditions over horizons up to several years. To gain a micro-based perspective on this puzzle, consider the week-by-week variations in the log spot rate implied by the micro-based model: Dstþ1 ¼ rt  ^rt þ dt þð1  ED rtþ1  rtþ1 Þ t Þð^ 1 D þ ðE  EEt Þðrftþ1  dtþ1 Þ 1  r tþ1 1 X 1 D ðED  E Þ ri1 ðDftþi  Ddtþi Þ þ t 1  r tþ1 i¼2 ð11:20Þ It is well established that interest differentials and other macro variables have very little forecasting power for short- and medium-term depreciation rates (see, e.g., Table 11.1). Thus, as a purely empirical matter, there appears to be little prospect of establishing the link between short-term variations in spot rates and macro variables known at the start of week t via interest rates and the risk premium terms rt, ^rt , and dt. Instead, we must focus our attention on the terms in the second and third lines of Eq. (11.20). The two terms in the second line identify the effects of new information concerning current macroeconomic conditions, that is, conditions at the start of week t þ 1. The first identifies the effect of unanticipated changes in short-term interest rates. In practice, central banks change interest rates relatively rarely; and when they do, they often communicate their intentions beforehand so as not to put undue stress on the financial system. As a result, unanticipated changes in short-term interest are not an important source of high-frequency spot rate dynamics over long time spans. This leaves the terms involving macro fundamentals, ft, and the risk premium, dt. It is in the identification of these terms that the macro and micro-based perspectives differ. Macro models typically assume that uncovered interest parity holds (or the risk premium is constant) so their focus is on the terms involving fundamentals. If the current state of the macroeconomy is known to all market participants (including dealers) as macro models assume, these terms become 1  X r  1  D ri1 Dftþi 1  ED Etþ1  ED t Dftþ1 þ t 1r 1r i¼2

ð11:21Þ

117

Under normal circumstances, the first term in this expression should be close to zero because macro fundamentals are unlikely to change significantly during a week. Thus, from a macro perspective, the lion’s share of the link between macro fundamentals and high-frequency spot rate dynamics must be attributable to the second term (i.e., the revisions in forecasts about future changes in fundamentals, Dftþi for i > 1). However, the time series properties of the macro variables used to construct fundamentals, ^f t , imply very little variation in the estimates of the present value of future changes in ^f t . As a result, estimates of the second term in Eq. (11.21) using ^f t account for roughly 5% of the variance of the depreciation rates at the monthly and quarterly horizons. By this metric, the high frequency movements in exchange rates appear largely disconnected from measured macro fundamentals. The micro-based perspective on the disconnect puzzle is rather different because all the terms on the right of Eq. (11.20) involving fundamentals and the risk premium could link short-term variations in spot rates to the macroeconomy. Consider the terms involving fundamentals. When the weekly changes in fundamentals are negligible (i.e., Dft þ 1 ffi 0), the terms involving fundamentals are well approximated by 1  X r  D 1  D D Etþ1  ED E þ  E ri1 Dftþi f t t t 1r 1  r tþ1 i¼2

ð11:22Þ From a micro-based perspective, dealers are not assumed to have contemporaneous information on the macro variables that comprise fundamentals, so the first term in Eq. (11.22) can differ from zero (because dealers are learning about past macroeconomic conditions that drive ft). Moreover, this term appears to be significant empirically. As Table 11.2 showed, order flow from week-t trading convey significant new information concerning current GDP, prices and money stocks, that is incorporated into spot rates. Thus, some of the high frequency behavior of exchange rates reflects the flow of new information reaching dealers concerning the slowly evolving state of the macroeconomy. The micro-based model also provides a new perspective on the second term in Eq. (11.22). Figure 11.2 shows that order flows have significant incremental forecasting power for future macro variables beyond that contained in current spot rates and other variables. Insofar as dealers use the information in these flows to revise their expectations, estimates of the second term in Eq. (11.22) based on macro data will understate the degree to which revisions in dealers’ expectations concerning future changes in fundamentals contribute to spot rate dynamics. In sum, the micro-based evidence complements rather than contradicts the empirical evidence derived

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

118

11. MICROSTRUCTURE OF CURRENCY MARKETS

from macro models on the link between spot rates and the macroeconomy. By concentrating on the flows of information available to dealers, micro-based models point to the role of order flow as a carrier of macroeconomic information. This information appears useful in revising forecasts of future changes in macro fundamentals, the channel emphasized by macro models. It also appears useful in revising dealers’ estimates of current macroeconomic conditions.

The News Puzzle One striking feature of the disconnect puzzle shows up in research on the effects of macro data releases. This research shows that spot rates react to data releases in a manner consistent with the predictions of standard macro models, but the releases account for less than 1% of the total variance in spot rates. In other words, most of the volatility in spot rates takes place during periods when there are no data releases. This represents a puzzle from the macro perspective. Macro exchange rate models predict that spot rates will respond immediately to the public release of macro data if it induces (i) a change in current interest rates and/or the risk premium and/or (ii) a revision in expectations concerning future fundamentals and/or the risk premia. The models do not predict that spot rates will respond to all data releases, or even to previously unexpected releases. Nor do they predict the size or direction of the spot rate’s response to particular releases. They do, however, rule out the possibility that the spot rate’s response is delayed. Consequently, only the variations in spot rates immediately following a data release are attributable to the effects of macro news. Micro-based models provide several new perspectives on how the release of macro data affects the behavior of spot rates. First, data releases will only affect spot rates insofar as they induce dealers to revise their spot rate quotes. Since data releases relate to past macroeconomic conditions, it is possible that most of the information they contain was already known to dealers from their past observations on order flows. A data release may contain new information about the macroeconomy relative to prior public information, but it may contain little incremental information to dealers. Consequently, the initial reaction of the spot rate to a data release will be smaller than if order flow were uninformative about current macroeconomic conditions. The immediate reaction of the spot rate to a data release may also be muted for other reasons. Although data releases are simultaneously observed by dealers and other agents, there may not be unanimous agreement on their implications for the value of foreign currency. For example, two firms may interpret the same

announcement on last quarter’s GDP as having different implications for future interest rates. When a data release contains imprecise but price-relevant information, dealers will immediately adjust their spot rate quotes to accommodate the new information on fundamentals and the risk of providing liquidity to the market. These effects can offset one another, so that the immediate impact of the data release on the spot rate is muted. Differing interpretations about the implications of a data release can also lead to changes in dealers’ spot rate quotes for some time after the release takes place. Some of these changes reflect the risk premia that were embedded in dealers’ original quotes. Others reflect the effects of order flow on dealers’ quotes as market participants reach a consensus on the price implications of the release. For example, suppose the data release represents good news for the dollar, but there is a diverse opinion on how far the dollar price of euros should fall. Under these circumstances, the initial fall in the USD/EUR spot rate may be viewed as too large by some market participants and too small by others. Those who view the fall as too small will place orders to sell the Euro, while those who view the fall as too large will place orders to buy it. In aggregate, the balance of these trades represents the order flow that dealers use to further revise their spot rate quotes. In particular, positive (negative) order flow for the euro signals that the initial USD/EUR spot rate was below (above) the balance of opinion among market participants concerning the implications of the data release. Empirical research supports this micro-based perspective on the effects of data releases. Studies of intraday data indicate that data releases affect spot rates directly and indirectly via induced order flows. Furthermore, order flows appear to carry more price-relevant information that is incorporated into spot rates following macro data releases than at other times. At the daily frequency, more than one-third of the total variance in spot rate changes can be related to the direct and indirect effects of macro data releases and other news sources. The indirect effects of news working via order flow contribute approximately 60% more to the variance of spot rate changes than do the direct effects. In short, micro-based research shows that data releases play a more important role in the determination of spot rates than was indicated by earlier research based on macro models.

CONCLUSION Although micro-based research has made some significant progress toward providing exchange rate models with empirically relevant microfoundations, much remains to be done. The models developed to date

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

CONCLUSION

are designed to study the behavior of major currency markets with lots of trading activity, such as the USD/ EUR and USD/JPY markets. They are less suited for studying the behavior of exchange rates between other currencies that trade with much less liquidity. Since the currencies of most counties fall into this category, adapting micro-based models to these markets is an important priority for future research. The models also need to expand their focus beyond the spot markets. Trade in forward contracts, swaps, and other exchangerate derivatives account for a large share of FX trading activity, but this portion of the FX market is yet to be studied from a micro-based perspective. This will surely change as data from electronic trading systems becomes more accessible to researchers.

Glossary Exchange rate fundamentals Macroeconomic variables that determine spot exchange rates.

119

Foreign exchange dealers Marketmakers in the foreign exchange market who quote prices at which they are willing to buy or sell foreign currency from/to others, and initiate currency trades with other dealers. Order flow The difference between the value of buyer- and sellerinitiated currency trades. Spot exchange rate The rate at which one currency can be immediately exchanged for another. Quoted as the domestic currency price of foreign currency.

Further Reading Evans, M.D.D., 2010. Order flows and the exchange rate disconnect puzzle. Journal of International Economics 80 (1), 58–71. Evans, M.D.D., 2011. Exchange Rate Dynamics. Princeton University Press, Princeton, NJ. Evans, M.D.D., Lyons, R.K., 2002. Order flow and exchange rate dynamics. Journal of Political Economy 110 (1), 170–180. Lyons, R.K., 2001. The Microstructure Approach to Exchange Rates. MIT Press, Cambridge, MA. Osler, C.L., 2008. Foreign exchange microstructure: a survey. Springer Encyclopedia of Complexity and System Science. Sarno, L., Taylor, M., 2002. The economics of exchange rates. Cambridge University Press, Cambridge.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

Intentionally left as blank

C H A P T E R

12 Intertemporal Approach to the Current Account P.R. Bergin University of California at Davis, Davis, CA, USA O U T L I N E Introduction

121

Intertemporal Theories A Simple Theory with Two Periods and No Uncertainty A Basic Stochastic Case with an Infinite Horizon More General Theoretical Cases

122 122 122 124

INTRODUCTION The current account summarizes transactions in goods, services, and factor income across national borders. Imbalances in the current account, in which import flows differ significantly from export flows, have become increasingly common and large. Figure 12.1 plots the current account values of several countries as a share of the levels of gross national income. This figure shows that the United States moved from current account levels fluctuating near zero in the 1970s to current account deficits that were an increasing share of national income in later years. The United States was not alone in this experience; countries such as Spain ran current account deficits that were even larger shares of their national incomes. Conversely, there are some countries, such as Germany and China, which ran large current account surpluses (positive values) repeatedly in later years. Large current account imbalances potentially may be unsustainable and prone to reversals, which could lead to instability in the overall macroeconomic condition of countries. Economic theory has proposed a useful way of understanding current account imbalances in terms of optimal consumption and saving decisions of private households within a country. Termed the intertemporal approach to the current account, this theory can address questions of where current account imbalances come

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00010-6

Empirical Relevance of the Theory and its Implications Present-Value Tests Implications for Current Account Experience

125 125 127

Conclusion Glossary Further Reading

128 128 128

from, and when they can be justified as a rational response to economic conditions. Running a current account imbalance implies costs. The balance of payments indicates that if a country has a current account deficit, this must be balanced by a surplus in the financial and/or capital accounts, such as sale of government bonds or other assets. Note that the sale of assets implies a loss in income from these assets in future periods, representing a long-run cost implied by running a current account deficit today. On the other hand, the ability to run a current account imbalance also offers potential benefits of consumption smoothing over temporary shocks. For example, consider the case of a small country devastated by an earthquake. If the country were restricted to a balanced current account, the only way the country could rebuild its capital and infrastructure would be to set aside more of the limited current production for investment expenditure and cut consumption. But if the country has access to a world financial market, it could borrow from abroad to finance this investment spending without such a large fall in consumption. This trade-off between the benefits and costs of a current account imbalance has been formalized in the intertemporal approach to the current account. This chapter summarizes the intertemporal theory, which characterizes when countries populated by rational, optimizing households will run current account imbalances.

121

# 2013 Elsevier Inc. All rights reserved.

122

12. INTERTEMPORAL APPROACH TO THE CURRENT ACCOUNT

0.08

Germany

0.04

China

0.00

United States –0.04

–0.08

Spain –0.12 1970

1975

1980

1985

1990

1995

2000

2005

FIGURE 12.1 Current accounts of selected countries. This figure plots the ratio of current account to gross national income for four countries. Source: International Financial Statistics.

A simple version of this theory is presented first, as it produces clear predictions about current account behavior. The chapter then discusses how the basic model can be generalized with various additional features of the world, and traces how the predictions of the theory change. The chapter concludes by discussing the empirical relevance of the theory, and its implications for understanding the recent experiences with rising current account imbalances.

INTERTEMPORAL THEORIES A Simple Theory with Two Periods and No Uncertainty Consider a simple case in which there are two periods, the present and the future, and for simplicity, there is no uncertainty about the future. Consider an open economy that trades goods and assets with the rest of the world. Suppose there is only one representative good in this world, used for consumption, government purchase, and investment. Suppose this good is not produced, but rather becomes available through an exogenous endowment process which can vary in amounts in the two periods. Suppose this country is able to issue or purchase bonds, which can be bought in the current period for one unit of the consumption good in the present period in exchange for an amount of the good in the future equal to 1 plus some constant extra fraction, which shall be referred to as the real interest rate. Assume this country is sufficiently small in relation to the rest of the world that it can borrow or lend as much as it wants without affecting the equilibrium in world real interest rate in the global financial market. Suppose this small open economy is populated by a representative consumer, who derives utility from consumption, and where the marginal utility in any given

period falls with higher levels of consumption. This assumption implies that the consumer will try to smooth consumption levels across periods. If consumption were high in the first period but low in the second, a reallocation of consumption from the first period to the second would raise utility in the second period more than it lowers utility in the first period. If it is assumed that the consumer prefers current consumption to future consumption by a discount factor that happens to equal the rate at which the world interest rate converts current saving into future consumption, then the consumer will prefer levels of consumption in each period that are exactly equal to each other. Assume also that there is a government which chooses an exogenous amount of spending. Lastly, assume for simplicity that the amount of investment spending is exogenous and has no effect on future output endowment. This implies that the key intertemporal decision in this simple case is the allocation of income between consumption and saving, rather than to investment decisions. Some of these restrictive assumptions are relaxed in the later cases studied in this chapter. Consider three scenarios. First, if future output is the same as the output in the current period, the consumer can achieve his goal of a smooth consumption level in both periods without requiring any borrowing or saving with the rest of the world. The current account will be zero. But suppose a second scenario in which the future output is higher than the current output. One option would be to run a zero current account balance and allow consumption to rise in the future with endowment. But the consumer knows that he can afford to consume more than his endowment in the present; he can issue debt today to finance extra consumption, and then use the higher level of endowment in the future to repay the debt. Because saving is negative in the present period, this implies a current account deficit. Far from being a bad thing, a deficit in this case is a rational response to an unequal distribution of endowment over time by a consumer who prefers to have a smooth profile of consumption. Finally, consider a scenario in which future endowment is lower than that in the present. In this case, a smooth profile of consumption implies a current account surplus, and the consumer saves for the future by purchasing bonds.

A Basic Stochastic Case with an Infinite Horizon Consider a somewhat more realistic version of this model of a small open economy, extended from two periods to an infinite horizon, and extended to include uncertainty about future endowments. In this case, the budget constraint of the household in the current period, denoted by t, can be written as follows: Bsþ1  Bs ¼ Ys þ rBs  Cs  I s Gs  CAs ;

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

s ¼ t; . . . ; 1 ð12:1Þ

INTERTEMPORAL THEORIES

where the endowment of goods (Y), investment spending (I), and government purchases (G) are exogenous and subject to random shocks. The variables C and CA stand for consumption and current account, respectively, and B represents holdings of the one-period noncontingent bonds defined above, with net return, r. If one adds up all of the budget constraints for each period and pins down the infinite future by assuming that households do not let their wealth to grow arbitrarily large, it implies an intertemporal budget constraint: st 1  X 1 ðCs þ Is þ Gs Þ ¼ ð1 þ rÞBt 1þr s¼t st 1  X 1 þ ðYs Þ 1þr s¼t ð12:2Þ The simple interpretation for this is that the present value of total expenditure in this economy must equal the present value of total income plus initial wealth. The household wishes to maximize the expected sum of utility from consuming goods over allPfuture periods, st U(Cs). discounted by the time preference, b : Et 1 s¼tb In this expression, Et represents the expectation for future variables based on information available in period, t. This problem implies that agents smooth marginal utility across periods, according to the following condition: 0

0

U ðCt Þ ¼ bð1 þ rÞEt ½U ðCtþ1 Þ

ð12:3Þ

Under the assumptions of a convenient quadratic form for the utility function U(Ct) ¼ Ct  (1/2)Ct2, and the assumption that households discount future periods at a rate related to the world real interest rate, b ¼ 1/(1 þ r), this condition simplifies to consumption smoothing in expectation: Ct ¼ Et[Ct þ 1]. This behavior is essentially the same as that described for the two-period model explained above: households like to maintain the same level of consumption across periods. The difference here is that households do not know for certain what level of consumption they can afford once the next period arrives as there may be unexpected shocks affecting the economy. Households will equate consumption in this period to what they expect consumption to be in the next period; that is, consumption is smoothed in expected terms. When the consumption behavior shown in Eq. (12.3) is combined with the budget constraints given in Eqns. (12.1) and (12.2), it implies the following behavior of the current account: CAt ¼ ðNOt Þ  ð1  bÞ

1 X

bst Et ½NOs 

ð12:4Þ

s¼t

where NOt is notation for net output, which summarizes the exogenous components of the economy, and is defined as Yt  It  Gt. Note that the second term on the right-hand side of the equation represents the

123

discounted sum of all future net output, representing a long-run average value of resources available for consumption. Its interpretation is that the current account is the difference between the current resources and the long-run average level of resources. For example, if output endowment in the present period is temporarily below its long-run average, households know that they can afford to borrow from the world financial market to finance a level of consumption today that is higher than the level of net output. This implies a current account deficit, which allows them to keep consumption smooth. The cost is that the country increases its debt to the rest of the world, which implies interest payments in all subsequent periods. Note that the equation would imply a very different outcome if the drop in output endowment instead were permanent and expected to persist in all future periods. In this case, both terms on the right-hand side of Eq. (12.3), the current net output and the long-run future average net output, fall together by the same amount. The country does not run a current account deficit, but instead cuts current consumption by the full amount of fall in the output. Because the endowment drop is permanent, the drop in consumption will apply to all future periods, implying that the objective of a smooth consumption profile is achieved without borrowing from abroad. The key lesson of this theory is that whether it is sensible for a country to run a current account deficit in response to a shock depends on the expectations of the consumers. If the shock is expected to be temporary, the current account imbalance is a useful mechanism for consumption smoothing; if the shock is permanent, there is no reason to run a current account imbalance. This lesson applies to any shock to net output, including government purchases. How should the current account respond to a rise in government spending? If the shock is temporary, it is a temporary fall in net output, and this induces a current account deficit such as that discussed above for a fall in endowment. This example offers insight into the theory of the Twin Deficits Hypothesis, which states that a government budget deficit tends to coincide with a current account deficit. Supposing the government budget deficit is defined as T  G, where T is a lump sum tax, and correspondingly augments the household budget constraint with taxes: Y  T ¼ C þ I. If one derives these equations under this notation again, it makes it clear that whether the twin deficits hypothesis holds depends on the persistence of government spending shocks. If the government budget deficit is due to a shock to government spending that households expect is temporary, they will run a current account deficit to boost current consumption. In this case the two deficits indeed will tend to coincide. But if the households expect the government spending to be permanent, they will lower consumption in response and

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

124

12. INTERTEMPORAL APPROACH TO THE CURRENT ACCOUNT

0.08

0.04

Government saving Current account

0.00

–0.04

–0.08

–0.12 1970

1975

1980

1985

1990

1995

2000

2005

FIGURE 12.2

US current account and government saving. This figure plots US current account and net receipts of the federal government, both as ratios to gross national income. Source: International Financial Statistics.

there will be no current account deficit. The twin deficits hypothesis will not hold in a case such as this. Figure 12.2 plots the US current account and government saving. Whereas the two seem to move together during some periods, such as the mid-1980s, recent experience shows that they can move independently of each other. In particular, the US current account deficit in 2000 coincided with a government budget surplus rather than deficit.

More General Theoretical Cases As discussed above, a prominent characteristic of current account data is a high degree of volatility. The simple intertemporal theory of the previous section has some difficulty generating sufficient volatility in its current account predictions. This failing comes from its simple version of consumption smoothing and the fact that most shocks to output appear to be of a highly persistent type. Recall that the basic theory implies that in the case of a permanent fall in output, consumption should fall by the same amount. If consumption falls with current output, there is approximately no impact on current account balance. This logic holds approximately for a shock that is less than permanent but still highly persistent. The failing of the theory in this particular dimension has motivated several theoretical extensions. First, consider a generalization of consumer preferences where utility today depends not only on consumption today but also on that in the previous period. Such preferences are referred to as non-time separable. For example, if the consumption level in the quadratic utility shown above is replaced by the change in consumption from the previous period, it implies that consumers try to smooth the change in

consumption across time periods rather than the level of consumption. Empirical evidence in the macroeconomics and finance literature supports this type of behavior, referred to as consumer habits. Such preferences alter the behavior of the current account. In the face of a highly persistent change in net output, consumers will move only gradually to a long-run equilibrium of higher consumption, so as to smooth changes in consumption from one period to the next. The small initial response in consumption implies that much of the initial rise in income is passed on to saving in the form of a current account surplus. So, the current account becomes more volatile than it was under the previous assumptions about preferences. A second source of extra current account volatility could be the presence of additional types of shocks. One possibility is shocks to the interest rate. To highlight the intertemporal response to interest rates, consider a different specification of utility, U(Ct) ¼ (r/(r  1))Ct1  (1/r), where the parameter r represents the intertemporal elasticity, as will become clear shortly. And consider the possibility that the world real interest rate, rt, is subject to shocks and can change over time. Equation (12.3) then would become Ct ¼ br(1 þ rt)rCt þ 1, indicating that a rise in the real interest rate can induce households to lower current consumption in relation to future consumption. In particular, a rise in the gross interest rate by 1% will lower the ratio of consumption today in relation to that in the future by r percent. Households are willing to accept a consumption profile that is tilted across periods rather than smooth, if the rewards for saving today in terms of extra consumption in the future are sufficiently large. Clearly, shocks to the world interest rate could induce greater fluctuations in the current account via their effects on consumption tilting. A second additional source of shocks could be variation in international relative prices such as the exchange rate. Extend the model now to consider endowments of two distinct types of goods, one of which can be traded internationally and the other which cannot be. Use T and N to indicate the traded and nontraded goods, respectively. Internationally traded bonds are denominated in terms of traded goods, with an interest rate likewise in traded goods units. Households consume both types of goods, where the overall consumption index in the utility function is specified as with an elasticity of substitution between goods equal to y. This means that when the relative price of nontraded goods in units of traded goods, called p rises by 1%, then the consumption of nontraded goods in relation to traded goods rises by y%. The analog to Eq. (12.3), written just for consumption of traded goods, CT,t, is " #  Pt yr r r CT;t ¼ b ð1 þ rt Þ Et CT;tþ1 ð12:5Þ Ptþ1

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

EMPIRICAL RELEVANCE OF THE THEORY AND ITS IMPLICATIONS

where Pt is the consumer price index in using units of traded goods as the numeraire, representing the cost of one unit of overall consumption index. Because traded goods are the numeraire, price index is an increasing function just of the relative price of nontraded goods in terms of traded goods, pt. This condition shows that changes in the price index between periods will cause a reallocation of consumption across periods, just like the real interest rate does. But note that the sign of this effect depends on whether the intertemporal elasticity, r, or intratemporal elasticity, y, is larger. Changes in the relative prices have two distinct effects, one intertemporal between periods and the other intratemporal across periods. Consider first the intertemporal effect. A price index falling from period t to t þ 1 means that domestic nontraded goods are becoming cheaper in relation to internationally traded goods. Because interest on debt must be paid in terms of traded goods, a debt accrued in period t will become more expensive in period t þ 1. This acts like a rise in the interest rate, and lowers current consumption in relation to future consumption by elasticity r. But there is also an intratemporal effect. If nontraded goods are currently expensive in period t, then this implies that traded goods are relatively cheap, making for a good time to substitute toward traded goods. A rise in traded goods consumption in relation to endowment implies a worsening current account. This intratemporal price effect raises current consumption by elasticity y. The two effects work in opposite direction, with the intertemporal effect dominating only if r > y. The question arises, would shocks to relative price have significant effects on current account volatility? Because exchange rate fluctuations can affect this relative price, and exchange rates are among the most volatile of macroeconomic variables, relative price changes potentially could raise current account volatility. Another promising source of current account volatility could be endogenous fluctuations in investment, because the current account equals saving less investment, and investment is the most volatile of national income expenditure categories. Consider an extension to the basic theory where production depends on capital and an exogenous technology term that is subject to shocks. Technology shocks can be either specific to a country or global in scope. The stock of capital is created by investment expenditure in the previous period. Consider the effects of a country-specific shock that permanently raises productivity. This implies a rise in the marginal product of capital in future periods, which motivates investment in new capital. Because current account equals saving less investment, the rise in investment expenditure would tend to lower the current account. But there is also a second effect at work. If the investment today raises the capital stock in the next period, this will also

125

raise the level of production in the next period even higher than the level that is today. As seen in earlier models, consumption smoothing under the expectation of a future rise in output implies that consumers will wish to borrow in order to finance extra consumption today. This drop in saving too implies a fall in the current account. The model implies that the current account will be falling both because of a rise in investment and a fall in saving. The basic intertemporal theory can also be extended from the case of a single small open economy to that of a two-country world. Whereas a small open economy takes the world real interest rate as exogenous, a twocountry world determines the interest rate endogenously by equating world saving to world investment. Consumption smoothing behavior in this case will imply a current account deficit when a country expects its share of world output to rise in the future in relation to the present. This result differs from the preceding models, in that it is the country’s share of world output, not the country’s level of output per se, that matters. If both countries in the world economy were expecting a rise in future output by an equal amount, both will try to smooth consumption by borrowing; given that there is no one from whom to borrow, the effort will simply drive up the world interest rate to the point that households would be willing to accept lower consumption today than in the future. Recent theoretical studies of current account behavior have used models with a host of extensions, originally designed to study how various shocks generate recessions. These dynamic stochastic general equilibrium models include production that uses labor as well as capital, where labor supply is endogenous. Also considered are adjustment dynamics for capital accumulation and frictions in the financial market in the form of a risk premium associated with larger foreign debts. Shocks to productivity, fiscal policy, and world interest rates follow calibrated stochastic processes. A lesson of such studies is that current account responses to shocks can play an important role in transmitting business cycles and macroeconomic policies across national borders.

EMPIRICAL RELEVANCE OF THE THEORY AND ITS IMPLICATIONS Present-Value Tests Because intertemporal current account theories are based on optimizing behavior by rational economic agents, they imply that even the large current account imbalances observed in recent decades could be defensible as rational responses to economic conditions, and do not require any intervention or correction by policy

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

126

12. INTERTEMPORAL APPROACH TO THE CURRENT ACCOUNT

makers. To evaluate the relevance of this claim, it is critical to review the empirical literature testing for whether the intertemporal theory is an accurate characterization of actual current account data. A brief summary is useful at the outset. There are several methods that have been used to test the intertemporal theories. Testing clearly is not a simple matter because the equations predicting the current account are present-value conditions involving expectations about periods far in the future. In summary, some versions of the theory are found to fit much better than others, and some countries have current account experiences that fit the theory better than others do. The greatest failing of the most basic intertemporal theory is that it understates the degree of volatility in actual current account data for some countries. One interpretation of this result is that actual current account imbalances are larger than can be explained by simpler optimizing models. This failure could reflect a rejection of any of the several parts of the intertemporal theory: optimizing consumer behavior, simple preferences assumed for consumers, rational expectations, or the specification of the international financial market. The understatement of the current account volatility comes as a surprise to some because it seems to contradict other empirical studies showing frictions in the international financial market and lack of perfect international capital mobility. Rather than a lack of capital mobility, the result here indicates if anything, an excessive amount of capital mobility. The most common method of testing the intertemporal current account theory is the present-value test. The theory’s current account P prediction in Eq. (12.4) can st Et[NOs  NOs  1], be rewritten as CAt ¼  1 s ¼ t þ 1b implying that a country would run a current account deficit in response to an expected rise in net output. To make this test operational, one must find a proxy for expectations of changes in net output. One could posit that households use lagged values of net output to form a forecast of the future. But households have more information at date t that could form a forecast. Because the theory says that current account balance reflects household forecasts, a researcher can include the current account value in the set of data used as proxy for household expectations of net output. The first step is to estimate a set of regressions (a vector autoregression), where changes in net output and the current account balance are regressed on lags of themselves and each other. Let C represent the matrix of coefficients from this set of regressions. A forecast for future change in net output can be expressed as a function of these parameters and the data are used in the regression. Substituting this forecast for the expected future net output in the current account equation shown above, and noting that it forms a geometric series, one can

derive a formula for the model’s current account prediction: ^ t ¼ ½1 CA

0bCðI  bCÞ1 ½DNOt

0

CAt 

ð12:6Þ

The simplest way to evaluate the model prediction is to ^ t , specified above and complot the predicted series, CA pare it graphically to the actual data, CAt. This is demonstrated in Figure 12.3, discussed below. A more precise statistical test makes use of the fact that the current account balance in period t is present in the formula for the current account prediction shown above, because it was included in the information set used to forecast net output. So, a formal way to test whether the model ^ t ¼ CAt , is to test statisprediction is correct, that is, CA tically if the expression [1 0]bC(I  bC)1 equals the vector [0 1] using a standard Wald test. This test can be easily expanded to allow for multiple lags of variables. The present-value test is easily generalized to test extensions of the basic intertemporal theory that include preferences with habits or time-varying interest rates and relative prices. In the latter case, the intertemporal budget constraint must be used as a long-linear approximation, producing a counterpart to Eq. (12.6), where the vector of the observable data includes not only net output and the current account but also the real interest rate and the relative price of nontraded goods. Figure 12.3 plots the current account data for Canada, CAt, along with the predictions for the current account, ^ t , arising from Eq. (12.6) and its extension for various CA versions of the present-value model when applied to these Canadian data. To accommodate the log 0.10 Data 0.08

Model with habits Model with relative prices

0.06 0.04 0.02 0.00

Basic model

–0.02 –0.04 –0.06 1970

1975

1980

1985

1990

1995

2000

2005

Year

FIGURE 12.3 Canadian current account: data and predictions of present-value models. This figure plots current account data for ^ t , implied by three versions Canada, as well as the predicted values, CA of the present-value model. The most basic model, as described in Eq. (12.6), assumes endowment of one world good and time-separable preferences. The other two predictions augment this simple model with time-varying relative prices and habits in preferences, respectively.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

EMPIRICAL RELEVANCE OF THE THEORY AND ITS IMPLICATIONS

linearization in the models, the data used for comparison are a transformation of the components of the current ac^ t and CAt count, ln NOt  ln Ct. The plots comparing CA permit a graphical test of how well the intertemporal theory captures movements in the actual data. The versions of the model represented are the simple current account model, along with the extensions of the theory that include habits and time-varying interest rates and relative prices. The values plotted are from the author’s calculations. Data for the ex-ante international real interest rate are computed from short-term nominal interest rates for the United States adjusted by a forecast of future inflation. Data on the relative price for nontraded goods are taken from a measure of the real exchange rate from the International Monetary Fund’s (IMF’s) International Financial Statistics. The habits parameter is calibrated at 0.8 based on outside empirical studies of habits. Note that all three model predictions do move over time in the same direction as the actual current account data; the prediction rises and falls with the actual current account. But the main failing of the simplest intertemporal model is that the predicted current account is insufficiently volatile; it rises by a smaller amount than the actual current account data. The standard deviation of the current account predicted by the simple model is only 39% as large as that for the actual data. This finding is typical among empirical studies of the intertemporal theory. Empirical tests tend to perform best for large countries such as the United States and worse for small countries. This result is surprising, because the assumption of the simple intertemporal model of an exogenous world interest rate is better suited to a small open economy than to larger economies. In the figures, the model augmented with timevarying interest rate and relative prices performs somewhat better in this regard. The standard deviation of the model prediction is 75% of that of the data. The model with habits in preferences performs quite well in capturing the direction and volatility of current account fluctuations, with the standard deviation of the model prediction at 90% of that of the data. However, it has been noted that the testable implication of the presentvalue model augmented with habits is observationally equivalent to a model augmented instead with persistent transitory consumption changes induced by world real interest rate shocks. In general, recent research points to world interest rates or habits as viable explanations for volatile current account behavior. An alternative statistical test of the theory focuses on its implication that a particular function of the current account should not be predictable based on past information. In particular, manipulation of condition (12.4) indicates DNOt  (1 þ r) that if one defines a statistic Rt  PCAt st {Et[DNOs]  CAt1, it should be equal to  1 s ¼ tb Et1[DNOs]}. Because the only difference between the

127

two terms on the right is the time subscript on the expectations operator, the R statistic should reflect only new information acquired between periods t  1 and t. This theoretical implication is tested by regression of the R statistic on lags of net output and current account data to see if these have forecasting power. Conclusions arising from this R test are usually similar to those from the presentvalue test given above. Other researchers have evaluated the theoretical model by calibrating and simulating it, and then comparing the simulated current account series to actual current account data. The results show that the model best replicates actual current account data when augmented with shocks to interest rates. A related approach estimates vector autoregressions of the variables in the model and imposes restrictions on the variables implied by the intertemporal theory. The results offer further support for the role of shocks to the interest rate as a primary mover of current account balances.

Implications for Current Account Experience A large amount of literature has applied the intertemporal theory to the particular cases of many countries, trying to interpret the origins of the widening current account imbalances observed in many of these cases. Consider the cases of European countries that have run large current account deficits, such as Spain and Portugal. One explanation proposed is that these countries may have expected high output growth in the future, as their admission to European Economic and Monetary Union could promote their convergence to European average income levels. The simplest of intertemporal current account models, rooted in consumption smoothing, predict that if a country expects future output growth, it may be reasonable for that country to borrow against future output to finance extra consumption in the present. Interestingly, because the present-value test expresses the current account as a function of expectations for future output, one can use the test to back out and examine the expectations of agents that are implicitly driving the current account. Empirical work estimates that the level of future output growth needed to justify the large current account deficit of Spain is beyond the level that can be regarded as reasonable on the basis of historical experience. This finding suggests that the current account deficits in some countries may be driven by excessive optimism about future output growth. The intertemporal theory likewise could be used to understand the large current account deficits run by the United States. Because the United States historically has experienced a higher average growth rate than have other developed countries, it could potentially be reasonable to borrow against future output growth to finance extra consumption in the present. Another

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

128

12. INTERTEMPORAL APPROACH TO THE CURRENT ACCOUNT

possible explanation for large current account deficits in the United States is a world interest rate at unusually low levels. The intertemporal theory augmented with interest rate shocks clearly indicates that low interest rates could discourage saving, leading to low current account balances. A potential cause, in turn, for low world interest rates could be shocks to the saving behavior of some emerging market countries such as China. The intertemporal theory further has been applied to offer an explanation for the rise in the current account balance in China. Recall that one version of the theory shows how changes in the relative price of nontraded goods can induce consumption tilting much like shocks to the real interest rate. If consumers in China predict that the prices of key nontraded goods such as housing and medical care will continue to rise in China depending on intertemporal elasticity, this could create an incentive to raise saving today and generate a current account surplus. Empirical research indicates that this incentive may explain more than half of the rise in the Chinese current account since 2001. The intertemporal model augmented with consumer habits might also be particularly useful in understanding the case of China. While Chinese income has experienced a dramatic rise that is likely permanent, habits in household preferences would prevent an immediate rise in consumption commensurate with income. It may take time for consumer habits in China to catch up with their permanent income. During this time of transition in consumer behavior, more of the rise in income is passed on to saving and hence a current account surplus happens.

CONCLUSION The intertemporal theory of the current account posits that a country’s current account balance with the rest of the world can be understood as a reflection of optimal saving decisions by residents in that country, as they smooth their consumption over time in response to random shocks to income. Because saving decisions are inherently an intertemporal decision, involving comparison of current and future periods, it is sensible to take an intertemporal approach for understanding current account. The basic theory can be augmented with extensions to consider how investment decisions affect the current account, as well as how alternative preferences and shocks to interest rates affect consumption smoothing. When the theory is augmented to consider these features, it is empirically relevant for numerous countries and offers understanding of the fact that

current account balances have become more volatile over time. Current research on this theory is considering the role of various types of shocks to international financial markets as a source of current account fluctuations.

SEE ALSO Theoretical Perspectives on Financial Globalization: Financial Development and Global Imbalances; International Trade and International Capital Flows; Valuation Effects, Capital Flows and International Adjustment.

Glossary Current account The sum of the balance of international trade (exports minus imports of goods and services), net factor income, and net transfer payments. Deterministic No randomness or uncertainty is involved. Elasticity The percent change in one variable as a ratio to the percent change in another variable. It indicates the responsiveness of a function to a parameter in that function. Financial account The net change in the ownership of internationally traded assets in the balance of payments accounts. Gross national disposable income Gross national income (see next entry) minus net transfer payments abroad (international gifts). Gross national income The total value produced within a country (gross domestic product), together with its net income received from other countries (such as interest and dividends). Present value The worth of a payment or stream of payments in future periods in terms of a current payment, discounted to reflect the time value of money. Regression A technique for estimating the statistical relationships among variables. Stochastic Subject to random shocks and fluctuations. Volatility Degree to which a variable fluctuates over time.

Further Reading Bergin, P.R., Sheffrin, S., 2000. Interest rates, exchange rates and present value models of the current account. The Economic Journal 110, 535–558. Bouakez, H., Kano, T., 2008. Terms of trade and current account fluctuations: the Harberger–Laursen–Metzler effect revisited. Journal of Macroeconomics 30, 260–281. Engel, C., Rogers, J.H., 2006. The U.S. current account deficit and the expected share of world output. Journal of Monetary Economics 53, 1063–1093. Feenstra, R.C., Taylor, A.M., 2008. International Economics. Worth Publishers, New York (Chapters 16 and 17). Ghosh, A.R., 1995. International capital mobility among the major industrialized countries: too little or too much? The Economic Journal 105, 107–128. Glick, R., Rogoff, K., 1995. Global versus country-specific productivity shocks and the current account. Journal of Monetary Economics 35, 159–192. Gruber, J.W., 2004. A present value test of habits and the current account. Journal of Monetary Economics 51, 1495–1507.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

CONCLUSION

Hoffmann, M., 2010. What drives China’s current account? HKIMR Working Paper No. 11/2010. Available at SSRN: http://ssrn.com/ abstract¼1628049. Iscan, T.B., 2003. Present value tests of the current account with durables consumption. Journal of International Money and Finance 21, 385–412. Kano, T., 2008. A structural VAR approach to the intertemporal model of the current account. Journal of International Money and Finance 27, 757–779. Kano, T., 2009. Habit formation and the present-value model of the current account: yet another suspect. Journal of International Economics 78, 72–85.

129

Nason, J.M., Rogers, J.H., 2006. The present-value model of the current account has been rejected: round up the usual suspects. Journal of International Economics 68, 159–187. Obstfeld, M., Rogoff, K., 1996. Foundations of International Macroeconomics. MIT Press, Cambridge, MA. Sachs, J., 1981. The current account and macroeconomic adjustment in the 1970s. Brookings Papers on Economic Activity 1, 201–268. Sheffrin, S., Woo, W.T., 1990. Present value tests of an intertemporal model of the current account. Journal of International Economics 29, 237–253.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

Intentionally left as blank

C H A P T E R

13 Endogenous Portfolios in International Macro Models M.B. Devereux*, A. Sutherland† *University of British Columbia, Vancouver, BC, Canada † University of St Andrews, Fife, UK O U T L I N E Introduction

131

A Simple Example Model

133

General Properties of Approximate Solutions

133

Mathematical Foundations

135

Applications Home Bias

135 135

INTRODUCTION Free international trade in equities and bonds has resulted in large and diversified international portfolios which, for many countries, imply gross holdings of asset and liabilities that are many times larger than net foreign asset (NFA) positions. The accompanying gross international capital flows, into and out of particular asset classes and across borders, are likewise many times larger than net capital flows between countries. Data reported by Lane and Milesi-Ferretti (2001, 2007) illustrate these facts very clearly. Table 13.1 shows external portfolio positions for a selection of countries for the years 1990 and 2007. These data show the dramatic rise in gross external positions over this period. For instance, gross asset and liability positions have grown by a factor of 2 or 3 for most countries so that gross positions can in some cases be as much as 300 or 400% of gross domestic product (GDP). This compares with NFA positions, which are usually (in absolute value) much less than 100% of GDP. The table also shows that total asset and liability positions are allocated across a range of different asset types, including equities, foreign direct investment (FDI), and debt (such as bonds and bank lending). Gross Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00018-0

Valuation Effects and Current Account Imbalances Global Imbalances and Emerging Market Portfolios Monetary Policy and Financial Globalization Conclusion Glossary References

135 136 136 136 137 137

positions in each of these types of assets are substantial, are much larger than the net positions, and have grown dramatically over the last two decades. Large cross-border portfolio holdings have important implications for macroeconomic interactions between countries. Gross asset and liability positions can generate large valuation effects following asset price and exchange rate fluctuations which can affect the international transmission of shocks and policy changes. The data in Table 13.1 can be used to illustrate the size of these valuation effects. For instance, the equity asset and liability positions of the United Kingdom are
50% of GDP in 2007. A fall in UK equity prices of 5% and a rise in non-UK equity price of 5% would result in an improvement of the UK NFA position of 5% of GDP (which is very substantial compared to the 2007 NFA position of  20% of GDP). As another example, consider the gross debt liability and asset positions of the United States which in 2007 were
90 and 50% of GDP, respectively. If the debt liability position is denominated in dollars and the debt asset positions is denominated in other currencies, a 10% depreciation of the dollar results in an improvement in the US NFA position (when expressed in dollars) of 5% of GDP

131

# 2013 Elsevier Inc. All rights reserved.

132

13. ENDOGENOUS PORTFOLIOS IN INTERNATIONAL MACRO MODELS

TABLE 13.1

External Asset and Liability Positions Equity assets

FDI assets

FDI liabilities

6

10

23

6

41

5

25

71

 46

28

37

32

37

26

81

3

89

155

 65

1990

5

5

15

19

17

54

3

40

78

 39

2007

39

23

36

35

31

53

3

109

111

2

1990

0

1

2

3

0

0

8

9

4

2

2007

1

13

3

21

20

11

45

70

45

22

1990

3

4

13

14

41

50

3

61

68

7

2007

33

39

94

62

158

171

2

287

272

12

1990

3

5

8

5

55

38

4

70

49

22

2007

30

25

38

30

151

138

1

219

193

26

1990

2

2

5

5

19

35

6

31

42

 11

2007

27

17

25

17

76

114

1

129

148

 21

1990

2

3

7

0

50

47

3

61

50

11

2007

13

28

12

3

74

40

22

121

71

50

1990

1

4

3

13

12

21

10

25

38

 12

2007

14

29

41

42

77

145

1

133

216

 85

1990

4

3

20

5

26

75

7

58

83

 25

2007

69

45

68

64

103

142

6

245

251

6

1990

19

11

23

23

124

139

4

170

173

3

2007

54

58

65

45

335

373

2

456

476

 20

1990

3

4

11

9

21

28

1

36

41

6

2007

38

23

25

18

49

89

0

112

130

 17

Country

Year

Australia

1990

4

2007 Canada

China

France

Germany

Italy

Japan

Spain

Sweden

UK

USA

Equity liabilities

Debt assets

Debt liabilities

Official reserves (minus gold)

Total assets

Total liabilities

NFA

Source: Lane, P., Milesi-Ferretti, G-M., 2007. The external wealth of nations mark II: revised and extended estimates of foreign assets and liabilities, 1970–2004. Journal of International Economics 73, 223–250. All asset and liability positions are expressed as a percentage of GDP.

(which is very substantial compared to the 2007 NFA position of  17% of GDP). While these large valuation effects have potentially important implications for transmission of shocks and thus affect macroeconomic developments, in turn it must be the case that agents’ choices about portfolio allocation (as between home and foreign equities, debt, and FDI) reflect the pattern of shocks and the hedging properties of different assets. So the portfolio positions illustrated in Table 13.1 must be endogenous to the macroeconomic environment. Thus, for instance, the choices that US households and firms make about the allocation of their net wealth over home and foreign equities, debt, and FDI reflect the pattern of uncertainty in the US economy and the ability of different assets to hedge these risks. The optimal portfolio allocation of US households and firms will change as these risks and hedging properties change. It is clear, therefore, that a full analysis

of the macroeconomics of financial globalization must take account of the valuation effects generated by large gross positions and of the endogenous determination of these gross portfolio positions. This requires that endogenous portfolio allocation is incorporated into international macro models. Until recently, however, the explicit analysis of portfolio allocation in general equilibrium models has been very difficult. Available solution techniques allowed the analysis of only a very limited range of financial structures. In effect, modeling choices were restricted to one of two extreme settings: trade in a single bond or trade in a full range of contingent assets. Neither of these structures adequately captures the nature of international asset trade. As can be seen from Table 13.1, international asset trade usually involves more than one asset (and thus does not correspond to the one-asset model) but it does not involve enough independent

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

133

GENERAL PROPERTIES OF APPROXIMATE SOLUTIONS

assets to replicate a complete market outcome (and thus does not correspond to models which assume that full risk sharing is possible). The underlying problem is the apparent difficulty in applying approximation techniques to portfolio selection problems when markets are incomplete. Portfolio choice depends fundamentally on the risk characteristics of different assets. The linearization techniques typically used in macro modeling cannot capture the risk characteristics of assets and thus cannot be used to analyze models of portfolio choice. Recently, however, Devereux and Sutherland (2010b, 2011) and Tille and van Wincoop (2010) have developed new solution methods which overcome this problem. These new techniques extend the standard approximation approach by combining higher-order approximations of the portfolio choice problem with conventional analysis of the nonportfolio parts of a model. This makes it possible to capture the risk and hedging properties of different assets and thus allows a consistent analysis of portfolio selection. These new techniques make it possible to solve for equilibrium portfolio holdings and equilibrium asset returns. This chapter describes the basic features of the new solution approach and discusses some applications and directions of research that have been made possible by the new solution methodology.

defined to be rx,t ¼ r1,t  r2,t, then the budget constraint can then be written in the form: Wt ¼ Yt  Ct þ r2;t Wt1 þ a1;t1 rx;t

ð13:3Þ

The consumption and portfolio allocation problem faced by foreign agents is assumed to be similar to that faced by home agents. Market clearing conditions in asset markets are as follows: a1;t1 þ a1;t1 ¼ 0 a2;t1 þ a2;t1 ¼ 0 where asterisks indicate foreign variables. Thus, assets are assumed to be in zero net supply. This aids exposition but implies no loss of generality. The solution approach applies equally to the case where assets are in positive net supply. Portfolio selection takes place at the end of each period. Assets are held into the following period, at which point asset returns are realized and agents receive their portfolio returns. The first-order conditions for the choice of a1,t and a2,t imply the following: h 0 i h 0 i Et u ðCtþ1 Þr1;tþ1 ¼ Et u ðCtþ1 Þr2;tþ1 ð13:4Þ A similar condition arises from foreign portfolio choices. Other aspects of the model, such as the determination of income, labor supply, prices, exchange rate, etc., are not directly relevant to describing the solution of the portfolio problem, so those details are not discussed here.

A SIMPLE EXAMPLE MODEL To understand the nature and properties of approximate solutions to portfolio holdings, it is useful to describe a simple open economy model. Consider a two-country model where the utility function of home country agents takes the form: U t ¼ Et

1 X

btt uðCt Þ

ð13:1Þ

t¼t

where C is consumption, u(Ct) is utility from consumption in period t, and b is the discount factor (where 0 < b < 1). Home and foreign households are assumed to be able to trade in two assets. The budget constraint of home households therefore has the following form: a1;t þ a2;t ¼ a1;t1 r1;t þ a2;t1 r2;t þ Yt  Ct

ð13:2Þ

where Y is income received by home agents, C is consumption of home agents, a1,t 1 and a2,t 1 are the real holdings of assets (purchased at the end of period t  1 for holding into period t), and r1,t and r2,t are gross real returns. Define Wt ¼ a1,t þ a2,t to be NFAs of home households at the end of period t (i.e., the home country’s net claims on foreign households). If the ‘excess return’ on asset 1 is

GENERAL PROPERTIES OF APPROXIMATE SOLUTIONS The following brief description of the approximated solutions for portfolio holdings and asset returns draws heavily on the more extensive discussion by Devereux and Sutherland (2010a,b, 2011). In analyzing a model of the type described above, the objective is to derive approximate solutions for portfolio holdings, a1,t and a2,t, and excess returns, rx,t. For the purposes of the following discussion, focus on a1,t. The discussion applies equally to a2,t. In general, the standard approach to analyzing a macro model is to consider the first-order accurate dynamics of variables around some point of approximation. To see what this implies for the analysis of portfolio holdings, one may consider the following approximation of a1,t up to order 1: h i
~a1 þ ^að1Þ a1;t ’ bY ð13:5Þ 1;t ðiÞ
^a1;t ¼ ða1;t 
a1 Þ=ðbYÞ,
and ^ a1;t is the where ~a1 ¼
a1 =ðbYÞ,
are the values of order-i component of ^a1;t . Here,
a1 and Y portfolio holdings and income at the approximation
is a convenient point, respectively. In this expression, bY

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

134

13. ENDOGENOUS PORTFOLIOS IN INTERNATIONAL MACRO MODELS

normalizing factor which simplifies notation and allows ~a1 and ^ a1;t to be interpreted (approximately) in terms of GDP units. By the normal definitions implied by Taylor series ð1Þ a1;t captures the approximations, ~ a1 is constant while ^ first-order variations of a1,t around the point of approximation. Thus, ~a1 can be thought of as the average, ð1Þ or steady-state, element of portfolio holdings while ^a1;t represents the (first-order) time-varying element. So ~a1 ð1Þ and ^ a1;t capture the two most important and interesting components of portfolio holdings. The standard approach to analyzing dynamic stochastic general equilibrium models is to solve a linear approximation of the model where the nonstochastic steady state of the model is used as the approximation point. This implies that the first-order component of each variable is solved using a first-order approximation of the model’s equations. This approach, however, cannot be used to solve portfolio problems. The portfolio optimality condition (13.4) shows clearly that asset returns are equalized in a nonstochastic equilibrium. Furthermore, a first-order approximation of condition (13.4) implies that expected rates of return are equalized in a first-order approximation of the model. The optimal portfolio is therefore undetermined in both the nonstochastic steady state and the first-order approximated form of the model. So neither the nonstochastic steady state nor the first-order approximation of the model provides the necessary equations to solve for optimal portfolio holdings. To obtain conditions which define equilibrium portfolios, it is therefore necessary to go to higher-order approximations of a model. The solution method described by Devereux and Sutherland (2010b, 2011) and Tille and van Wincoop (2010) uses a second-order approximation of the portfolio optimality conditions to derive solutions for ~a1 and ~a2 and a third-order approximation of the portfolio optið1Þ ð1Þ mality conditions to derive solutions for ^ a1;t and ^a2;t . The second-order approximation of the model naturally captures the second moments (i.e., the variances and covariances) of variables and thus captures the risk characteristics of different assets. This provides enough conditions to determine the equilibrium steady-state a2 . A third-order approximation portfolio, that is, ~ a1 and ~ of the model captures the way second moments vary through time and thus captures the way the risk characteristics of different assets change in response to the evolution of other macro variables. A third-order approximation of the model therefore provides enough conditions to determine the first-order dynamics of portð1Þ ð1Þ a2;t . folio holdings, that is, ^ a1;t and ^ Now consider equilibrium returns, or, more specifically, excess returns, rx,t. Again, the focus of interest is

on the steady-state value of excess returns and the approximate dynamics of this steady state. While a first-order approximation of portfolio holdings is sufficient to capture the approximate dynamics of a1 and a2, it is necessary to go to higher orders of approximation to capture the dynamics of excess returns. To see this, consider an approximation of rx,t up to order 3: i 1 h ð1Þ ð2Þ ð3Þ rx;t ’
rx þ ^rx;t þ ^rx;t þ ^rx;t ð13:6Þ b ðiÞ

where ^rx;t ¼ bðr1t  r2t Þ, ^rx;t is the order-i component of ^rx;t , and 1/b is the steady-state equilibrium value of r1 and r2. Equation (13.4) implies that in the nonstochastic steady state
rx ¼ 0, that is, when there is no risk, all assets pay the same return. hAnd ai first-order approximation of

ð1Þ (13.4) implies that Et ^rx;tþ1 ¼ 0, that is, certainty equiv-

alence implies that (up to first-order) all assets pay the same expected return. Expected excess returns therefore only differ from zero at the second order or above. The solution method outlined by Devereux and Sutherland (2010b, 2011) and Tille h and i van Wincoop ð2Þ (2010) shows that a solution for Et ^rx;tþ1 can be obtained

simultaneouslyh withisolutions for ~a1 and ~a2 . Thus, the ð2Þ solution for Et ^rx;tþ1 is derived using a second-order approximation of the portfolio optimality conditions. h i ð2Þ ^ This solution shows that Et rx;tþ1 is nontime-varying. h i ð2Þ Et ^rx;tþ1 is therefore the steady-state equilibrium expected excess return which corresponds to the a2 . steady-state equilibrium portfolio holdings, ~ a1 and ~ It is further shown by Devereux and Sutherland (2010b, 2011) and Tille and van Wincoop (2010) that a solution for h i the third-order component of excess returns, ð3Þ Et ^rx;tþ1 , can be obtained simultaneously with solutions ð1Þ

for the first-order component of asset holdings, ^ a1;t and h i ð1Þ ð3Þ ^a2;t . Thus, an expression for Et ^rx;tþ1 is derived using a third-order approximation of the portfolio optimality h i ð3Þ conditions. Et ^rx;tþ1 is the time-varying component of expected excess returns which corresponds to the firstorder time-varying component of portfolio holdings. h i ð2Þ Thus, in summary, solutions for ~a1 , ~a2 , and Et ^rx;tþ1

can be derived using a second-order approximation of the portfolio optimality condition and can be thought of as steady-state values for portfolio holdings and expected excess returns, that is, values for these variables at h i the ð1Þ ð1Þ ð3Þ approximation point. And a^1;t , a^2;t , and Et ^rx;tþ1 can

be derived using a third-order approximation of the

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

APPLICATIONS

portfolio optimality condition and can be thought of as capturing the approximate dynamics of portfolios and expected excess returns around the approximation point.

MATHEMATICAL FOUNDATIONS While the solution procedure developed by Devereux and Sutherland (2010b, 2011) and Tille and van Wincoop (2010) is novel in terms of its application to macroeconomic models, the basic mathematical foundations have already been described and analyzed in the work of Samuelson (1970) and Judd (1998). Working with a simple partial equilibrium model of a single investor, Samuelson shows that the optimal portfolio in the near neighborhood of the nonstochastic equilibrium can be identified using a mean-variance approximation of the portfolio allocation problem, while Judd (1998) uses the Bifurcation Theorem and the Implicit Function Theorem to offer an alternative interpretation and approach to deriving this approximate solution. The approach developed by Devereux and Sutherland (2010b, 2011) and Tille and van Wincoop (2010) uses a Taylor series approximation of the model but the underlying theory developed by Judd (1998) is applicable. In particular, the steady-state portfolio identified by Devereux and Sutherland (2010b, 2011) and Tille and van Wincoop (2010) technique can be interpreted as a bifurcation point in the set of possible nonstochastic equilibria.

APPLICATIONS Home Bias One of the most obvious applications of the solution techniques outlined above is to investigate the potential theoretical explanations for ‘home bias’ in equity holdings. Home bias is the observed tendency for investors to hold a higher proportion of the equity assets from their own country than would be appropriate given the share of their own country’s equity in the total value of all equity assets in the world economy. This empirical fact was first highlighted by French and Poterba (1991) and has been the subject of an active theoretical literature ever since (see for instance the analysis of Baxter and Jermann (1997) and subsequent related literature). Much of this literature has been based on simple twocountry models where agents face a choice between home and foreign equity. Prior to the development of the solution technique described above, it was necessary to confine analysis to highly restricted models where two types of equity asset are sufficient to support a complete market equilibrium. This makes it possible to obtain a

135

solution for the equilibrium asset allocation by searching for the portfolio which delivers perfect risk sharing. This approach yields many useful benchmark results, but it restricts attention to simple models where there are a strictly limited number of sources of uncertainty. The great advantage of the solution technique described above is that it can be applied to much more complex models which include multiple assets and many sources of uncertainty. And, crucially, there is no need to match the number of assets to the number of sources of uncertainty. Solutions for portfolio allocation can therefore be obtained even when perfect risk sharing is not possible. Coeurdacier et al. (2007, 2010) have applied the new solution technique to analyze equity home bias in a more general framework than has been possible to date. In particular, they consider multiple sources of uncertainty and a menu of asset which includes both equities and debt instruments.

Valuation Effects and Current Account Imbalances A second obvious application of the new solution technique is the analysis of ‘valuation effects’ and their implications for current account imbalances and the dynamics of NFAs. Gourinchas and Rey (2007) have argued that the valuation effects generated by changes in the value of the US dollar have tended to offset US current account imbalances and thus provide an important stabilizing mechanism. While a lively literature has analyzed and debated the empirical validity of this argument, a theoretical analysis of how valuation effects interact with portfolio allocation and NFA dynamics has only become possible following the development of the solution technique described above. Devereux and Sutherland (2010a) analyze valuation effects in a two-country endowment model using the new solution technique. In general terms, they find a correspondence between the valuation effects in the model and the valuation effects measured in the data. They show, however, that ‘unanticipated’ and ‘anticipated’ valuation effects have very different properties and effects. Unanticipated valuation effects can be large and can dominate the dynamics of NFA, while anticipated effects tend to be small for reasonable parameterizations of the model appear only at higher orders of approximation. There tends to be a negative correlation between unanticipated valuation effects and trade imbalances (i.e., a positive unanticipated valuation effect tends to be associated with a trade deficit, and vice versa for a negative valuation effect) so unanticipated valuation effects tend to stabilize NFA movements. Anticipated valuation effects also tend to have a stabilizing effect on NFA but the magnitude of these effects appears to be too small to be of any practical relevance.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

136

13. ENDOGENOUS PORTFOLIOS IN INTERNATIONAL MACRO MODELS

Global Imbalances and Emerging Market Portfolios A third application of the new methodology has been to investigate the increasing participation of emerging market economies in international financial markets. Recent experience for many emerging countries shows that the change in gross portfolio positions is considerably higher than the change in net positions, even where there has been substantial growth in net positions. While the focus of the debate on global imbalances has been on large current account surpluses of China and other developing countries, and particularly on the apparent counterintuitive outflow of capital from emerging market economies to developed economies, there is in fact substantial two-way capital flows. Emerging economies have been building up official reserve assets (in the form of US Treasury Bills), but simultaneously there have been large FDI and portfolio equity inflows and also inward flows into bond markets. Compared to 20 years ago, when emerging economies typically had large liability positions in the form of bank debt and short-term bonds denominated in US dollars, they now hold large asset positions in bonds and large liability positions in FDI and portfolio equities. It has been argued that these gross asset and liability positions represent a more efficient way to finance real investment in emerging market economies, while delivering a high degree of international sharing risk. Devereux and Sutherland (2009) use the newly developed solution methodology to analyze portfolio allocation for an emerging market economy. The new methodology makes it possible to capture two-way capital flows across countries and thus allows an analysis of the gross portfolio position adopted by the emerging market economy. The analysis shows that financial globalization implies that the optimal portfolio position of an emerging market economy may be a large long position in bonds and negative gross position in FDI and equity. Devereux and Sutherland further show that gross portfolio positions of this form may deliver greater risk sharing for the emerging market economy, compared to an asset structure which only allows trade in foreign currency bonds.

Monetary Policy and Financial Globalization A fourth application of the new solution technique is in the analysis of macroeconomic policy in open economies. The presence of large gross portfolio positions not only affects the way that shocks are transmitted between countries, but it also potentially affects the transmission mechanism and impact of policy changes. The recent theoretical literature has suggested that the main goal of monetary policy should be the stabilization of prices. The analysis which supports this conclusion is usually

based on closed economy models, or open economy models with very stylized asset market structures. Given the growth of cross-border gross portfolio holdings, it is important to determine whether price stability continues to be optimal in models which allow for such gross asset positions. Devereux and Sutherland (2008) use the new portfolio solution methodology to analyze how financial globalization may change the form of optimal monetary policy. They model optimal gross asset holdings under different configurations of international financial markets and show how portfolio choice interacts with monetary policy in determining macroeconomic outcomes. In particular, they show that monetary policy affects the extent of cross-country risk sharing. The link between monetary policy and risk sharing creates a distinct and new link between monetary policy and welfare which has potentially important implications for the choice of optimal monetary policy. However, Devereux and Sutherland find that the endogenous choice of portfolios does not change the form of optimal monetary policy. In fact, the link between monetary policy and risk sharing may even reinforce the case for price stability. Devereux and Sutherland show that price stability increases the capacity of nominal assets to provide risk sharing when financial markets are incomplete. Furthermore, the link between price stability and the risksharing capacity of nominal bonds exists even when nominal goods prices are perfectly flexible. The explanation is that price stabilization reduces the degree of extraneous volatility in the real returns on nominal assets and this allows nominal assets to play an enhanced role in risk sharing. While Devereux and Sutherland’s (2008) analysis yields a simple and intuitively appealing result, it must be emphasized that their work only represents an initial step in investigating the full impact of financial globalization on welfare-maximizing macroeconomic policy. Many important issues remain to be analyzed and policy prescriptions may be sensitive to aspects of financial globalization which are not captured by the simple model considered by Devereux and Sutherland (2008). In particular, large gross positions create significant potential for financial contagion following a shock or a monetary contraction in one country. This is likely to have significant implications for the optimal design of monetary policy and is likely to form an important topic of future research.

CONCLUSION Large cross-border gross portfolio positions have become a significant feature of international financial trade in the last 20 years. This creates significant new channels for the transmission of macroeconomic shocks and

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

CONCLUSION

policy changes between countries. Until recently, the standard theoretical models used in international macroeconomics could not handle these questions because it was not possible to solve for equilibrium portfolios using standard methods of analysis. New solution techniques developed by Devereux and Sutherland (2010b, 2011) and Tille and van Wincoop (2010) allow significant progress to be made in analyzing portfolio allocation in open economy general equilibrium models and open up many important new lines of research.

SEE ALSO Theoretical Perspectives on Financial Globalization: International Macro-Finance; Trade Costs and Home Bias.

Glossary Gross external assets and liabilities The gross external asset position of a country is the sum of all foreign assets owned by home residents and institutions. The gross external liability position of a country is the sum of all home assets owned by foreign residents and institutions. The NFA position of a country equals gross external assets minus gross external liabilities. Incomplete financial markets Financial markets are incomplete when there are insufficient independent financial instruments to hedge all sources of random shocks. Nonstochastic steady state The nonstochastic steady state of a model is the point of long-term equilibrium toward which the model gravitates when all sources of random shocks are removed from the model. Taylor series approximation A Taylor series approximation represents a function in terms of a polynomial in the deviations of the arguments of the function from an approximation point. The coefficients of the polynomial are obtained from the derivatives of the function where the derivatives are evaluated at the approximation point.

137

Valuation effect Valuation effects are changes in the value of net foreign assets which arise because changes in exchange rates and/or asset prices result in changes in the value of gross external assets relative to the value of gross external liabilities.

References Baxter, M., Jermann, U., 1997. The international diversification puzzle is worse than you think. American Economic Review 87, 170–180. Coeurdacier, N., Kollmann, R., Martin, P., 2007. International portfolios with supply, demand, and redistributive shocks. NBER International Seminar on Macroeconomics 2007, 231–263. Coeurdacier, N., Kollmann, R., Martin, P., 2010. International portfolios, capital accumulation, and foreign assets dynamics. Journal of International Economics 80, 100–112. Devereux, M., Sutherland, A., 2008. Financial globalization and monetary policy. Journal of Monetary Economics 55, 1363–1375. Devereux, M., Sutherland, A., 2009. A portfolio model of capital flows to emerging markets. Journal of Development Economics 89, 181–193. Devereux, M., Sutherland, A., 2010a. Valuation effects and the dynamics of net external assets. Journal of International Economics 80, 129–143. Devereux, M., Sutherland, A., 2010b. Country portfolio dynamics. Journal of Economic Dynamics and Control 34, 1325–1342. Devereux, M., Sutherland, A., 2011. Country portfolios in open economy macro models. Journal of the European Economic Association 9, 337–369. French, K., Poterba, J., 1991. Investor diversification and international equity markets. American Economic Review 81, 222–226. Gourinchas, P.-O., Rey, H., 2007. International financial adjustment. Journal of Political Economy 115, 665–703. Judd, K., 1998. Numerical Methods in Economics. MIT Press, Cambridge. Lane, P., Milesi-Ferretti, G.-M., 2001. The external wealth of nations: measures of foreign assets and liabilities for industrial and developing countries. Journal of International Economics 55, 263–294. Lane, P., Milesi-Ferretti, G.-M., 2007. The external wealth of nations mark II: revised and extended estimates of foreign assets and liabilities, 1970–2004. Journal of International Economics 73, 223–250. Samuelson, P., 1970. The fundamental approximation theorem of portfolio analysis in terms of means, variances, and higher moments. Review of Economic Studies 37, 537–542. Tille, C., van Wincoop, E., 2010. International capital flows. Journal of International Economics 80, 157–175.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

Intentionally left as blank

C H A P T E R

14 Financial Contagion R. Kollmann*†{, F. Malherbe} *ECARES, Universite´ Libre de Bruxelles, Brussels, Belgium † Universite´ Paris-Est, Paris, France { Centre for Economic Policy Research, London, UK } London Business School, London, UK O U T L I N E Introduction

139

Bank Balance Sheet Adjustments as a Channel of Contagion: The International Financial Multiplier

140

Financial Contagion Through Interbank Linkages

141

INTRODUCTION During the recent (2007–09) financial crisis, gross domestic product (GDP) growth and stock markets collapsed simultaneously in most countries around the globe. Yet, the crisis was triggered by a financial shock in the United States, namely an unanticipated fall in US house prices that led to massive mortgage loan defaults by US households, and thus impaired the health of United States and foreign banks that had invested in the US mortgage market. Earlier episodes of ‘financial contagion’ included the 1997–98 crisis, during which financial troubles that originated in Asia and Russia rapidly spread to other emerging market economies. The 1997–98 crisis spawned a sizable empirical and theoretical literature on mechanisms by which a financial crisis is strongly and rapidly transmitted across countries. That literature highlights that contagion may occur through a variety of channels (trade linkages, credit flows, changes in investor sentiment, etc.). It points out that financial frictions are needed for financial strains to disrupt real activity. For, in the theoretical world of frictionless financial markets, the financial side of the economy is a veil only, and risks are hedged efficiently. Hence, a shock to any individual may be transmitted to all others, but there is nothing wrong with this. But with

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00024-6

Bank Runs and Self-Fulfilling International Crises Glossary Further Reading Relevant Websites

142 143 143 143

financial frictions, risk-sharing is incomplete, and financial shocks spill over to real activity. This chapter discusses recent theories of ‘financial contagion’ in which balance sheets of global banks (and other financial institutions) are the key channel of international transmission. This is motivated by the fact that the 2007–09 crisis in the US mortgage market was transmitted to the rest of the world through crosscountry banking linkages. By contrast to earlier contagion episodes, linkages due to international goods trade were of secondary importance in the global spread of the 2007–09 crisis (countries with close trade links with the United States did not suffer more than countries with weaker trade links to the United States). Empirically, the bulk of bank assets is financed by short-term debt; banks’ own funds (bank capital) only account for a small fraction of total assets. Furthermore, bank assets generally have a longer maturity and are less liquid than bank liabilities. This is a source of fragility that magnifies the effect of crises. In a globalized financial system, banks hold domestic and foreign securities (stocks and bonds), and they lend to domestic and foreign households and firms; banks in different countries also lend to each other. An adverse macroeconomic or financial shock in one country that lowers the capital of global banks may thus trigger a global asset sell-off and

139

# 2013 Elsevier Inc. All rights reserved.

140

14. FINANCIAL CONTAGION

a global decrease in bank lending (credit crunch), provoking a global recession. These channels of contagion are further explained below. The section ‘Bank Balance Sheet Adjustments as a Channel of Contagion: The International Financial Multiplier’ discusses international contagion due to portfolio adjustments of global banks, in response to an asset price shock in one country. The section ‘Financial Contagion Through Interbank Linkages’ considers the role of the interbank linkage for international contagion. The section ‘Bank-Runs and Self-Fulfilling International Crises’ discusses bank runs and self-fulfilling international financial crises.

BANK BALANCE SHEET ADJUSTMENTS AS A CHANNEL OF CONTAGION: THE INTERNATIONAL FINANCIAL MULTIPLIER Financial intermediaries, henceforth referred to as ‘banks,’ make loans and invest in securities and other assets (e.g., real estate). They fund their asset holdings by taking deposits and issuing other forms of debt (mostly short term) and using the bank’s own funds (bank capital): Assets ¼ Debt þ Capital. A key constraint on bank activities is that banks have to back at least a fraction of their assets by bank capital: Capital/Assets  k, for some coefficient 0 < k < 1. Hence, only a maximum fraction 1  k of bank asset holdings can be funded by debt. A constraint of this type is known as a ‘bank capital requirement.’ It can reflect a regulatory (legal) requirement, or market pressures. Bank capital requirements protect the interests of bank creditors. For, bank capital is a buffer against a fall in the value of bank assets, and thus lowers the risk of insolvency. Also, capital requirements help to limit moral hazard by bankers. A simple story is that bankers can walk away with a fraction k of bank assets without prosecution (and start a new life next period). Banks will then only be able to borrow if bank capital does not fall below a fraction k of assets. The ratio of a bank’s capital to its assets is called the ‘capital ratio’; the inverse of the capital ratio is the ‘leverage ratio.’ If the expected return on bank assets exceeds the interest rate on bank debt, then banks have an incentive to borrow the maximum amount, and thus the capital ratio will stay close to the required capital ratio k. Empirically, the capital ratios of the major European banks and of major US investment banks have typically ranged between 3 and 5% in the period 1995–2010, while the capital ratios of US commercial banks have generally been in the range of 7–8%. As shown below, low capital ratios imply that asset price changes may trigger sizable adjustments of banks’ asset holdings and debt positions. When banks are

globally active, then local (country-specific) shocks can hence induce sharp and synchronized worldwide asset price changes. Those asset price movements can feed into real activity, and thus induce global booms and recessions. In an influential 2008 paper, Paul Krugman refers to this mechanism as the ‘international financial multiplier.’ This mechanism has become more and more powerful in recent decades, as the banking industry has become globalized. For example, external assets and liabilities of US banks (each) represented about 30% of US GDP in 2009; for Germany, France, and the United Kingdom, external bank assets and liabilities represent more than 100% of domestic GDP. Consider the following numerical example of a world with two countries of equal size, called Home and Foreign (the countries can be viewed as the United States and the European Union, respectively). There is a (representative) global bank that holds both Home and Foreign assets (loans, stocks, and bonds), and takes deposits from Home and Foreign households. The bank wishes to allocate 50% of its total assets to Home (Foreign) assets. Let k ¼ 0.05, that is, the bank has to fund at least 5% of its assets using capital. Assume that, initially, the bank holds Home and Foreign assets both worth 50, its debt is 95, and its capital is 5. The initial balance sheet of the global bank is thus Assets

Liabilities

Home: 50 Foreign: 50

Debt: 95 Capital: 5

Consider now what happens when the value of the bank’s Home assets drops by 0.5 (i.e., by 1%). This might be due to bad news about the future profits (dividends) of Home firms, which lowers the Home stock price; it could also be due to Home credit losses, that is, defaults by Home households or firms on loans received from the global bank. As a result of this shock, the bank’s capital drops to 4.5, that is, the bank’s new balance sheet becomes Assets

Liabilities

Home: 49.5 Foreign: 50

Debt: 95 Capital: 4.5

Note that bank capital falls by a much larger percentage (10%) than the fall in total assets (0.5%). The bank’s capital ratio is now 4.5/99.5 ¼ 4.52%, which is smaller than the target ratio (5%). Unless the bank’s shareholders provide new capital to the bank, the bank has to reduce her debt and her total assets by 9.5, that is, total assets and debt have to fall to 90 and 85.5, respectively – as then the capital ratio is again 5%. Assuming that the bank continues to allocate 50% of her total assets to Home assets, the adjusted balance sheet is

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

FINANCIAL CONTAGION THROUGH INTERBANK LINKAGES

Assets

Liabilities

Home: 45 Foreign: 45

Debt: 85.5 Capital: 4.5

Thus, the initial 1% fall in the value of Home assets has triggered a much bigger simultaneous reduction of the bank’s Home and Foreign asset holdings and of its debt (10%, relative to the initial balance sheet). The sale of Home and Foreign assets by the global bank is likely to lead to a (further) fall in Home and Foreign asset prices, which then can lead to an additional round of Home and Foreign asset sales, etc. This is likely to reduce Home and Foreign real activity. The spillover into the real economy may be due to the fact that the asset sell-off makes it harder for firms to fund physical investment projects and to obtain working capital; it also limits the supply of consumption and mortgage loans to households. Investment and consumption will fall thus, which lowers output. A general equilibrium model is needed for a rigorous analysis of these feedback effects (see below).1 An equivalent vicious circle of global asset sales and falling asset prices can also be triggered by an adverse shock that induces depositors (or other bank creditors) in one country to withdraw their funds from the global bank. Of course, a similar powerful effect also operates when asset values rise. In our numerical example, an initial 1% rise in the value of Home assets will, on impact, raise the bank’s capital to 5.5, and her capital ratio to 5.47%. In order to again reach a 5% capital ratio, the bank then has to increase her holdings of Home and Foreign assets and her debt by 10% (compared to the initial situation). Much recent research has been devoted to building quantitative, dynamic, general equilibrium models with the mechanisms that were just described. Devereux and Yetman (2010) present a model of a two-country world with cross-country trade in equity (claims to physical capital) and in one-period bonds. Within each country, there are patient households who save, and impatient households who invest. Investors (who resemble banks) hold domestic and foreign equity, and they face a capital requirement (i.e., maximum debt depends on investor net worth). When the capital requirement does not bind, then the international transmission of macroeconomics shocks is very limited. By contrast, with binding capital requirements, balance sheet linkages across banks generate a powerful mechanism for the international transmission of shocks. Kollmann et al. (2010) model the international transmission of defaults on bank loans. A two-country world is considered, in which global banks collect deposits from households, and lend to

141

entrepreneurs, in both countries. An unanticipated loan default in one country brings about a wealth transfer from banks to entrepreneurs; hence bank capital falls, which impairs the bank’s ability to channel funds from savers to borrowers. The deposit rate falls while the loan rate rises, in both countries, thus lowering lending, investment, and output in both countries. The important insight is that a credit loss (default) in one country triggers an immediate and identical fall in both domestic and foreign output.

FINANCIAL CONTAGION THROUGH INTERBANK LINKAGES The recent financial crisis has revealed that interbank borrowing and lending is a key channel of international contagion. Because depositors have a preference for liquidity, deposits are generally available ‘on demand’ (demand deposits). But most bank loans have a longer maturity (mortgage loans for instance). The reason why banks can engage in such a ‘maturity transformation’ is that normally not all depositors withdraw their funds at the same time (one may think that, on average, withdrawals are compensated by new deposits). Still, banks must be able to cope with unexpected, large withdrawals. For this reason, banks hold a fraction of their assets in cash or other liquid assets (i.e., in assets that can easily and rapidly be turned into cash). When choosing her holdings of liquid reserves, a bank faces the following trade-off: on the one hand, any euro held in reserves earns the bank less interest than loans; on the other hand, if the bank does not hold enough reserves to face unexpected high withdrawals, it will have to sell less liquid assets on short notice, often at ‘fire-sale’ prices. Banks may respond to this uncertainty by holding demand deposits in other banks. Such cross-holdings of deposits allow banks to reallocate cash among themselves, according to their respective liquidity needs. That mechanism improves the efficiency of the banking system: for a given total of liquid assets held by the banking system, each bank will be able to cope with larger idiosyncratic deposit withdrawals. Banks thus need lower total reserves to hedge against a given amount of withdrawal risk by households. Hence, they can issue more loans, which raises real activity. This logic holds for not only banks within a country, but also across banks in different countries, as deposit withdrawals are likely to be imperfectly correlated across countries. Imagine, for example, that there is a natural disaster in one country and that its residents

1

A general equilibrium setup is also required for an account of who buys the assets sold by banks. The buyers are likely to be less productive at managing those assets, which too contributes to the drop in aggregate output.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

142

14. FINANCIAL CONTAGION

draw on their deposits at the local banks to fund unanticipated spending. This is likely to be independent of deposit withdrawals by households in other countries. Banks in the country in which the disaster occurs can more easily meet deposit withdrawals if they hold deposits in foreign banks. International cross-holdings of deposit thus facilitate international risk-sharing, but they also imply that an unexpected withdrawal in one country may impact on real activity in other countries. For instance, a bank could be forced to cut its lending in order to fulfill its obligations toward its foreign counterparts. Also, for the reasons explained above, banks may end up holding less total liquid assets when there is risk-sharing via interbank linkages. This implies that a system of global interbank linkages is more resistant to a local increase in withdrawals, but that it may be less able to resist infrequent global shocks. Global crises might hence be magnified by cross-country interbank linkages. In this respect, the precise structure of the interbank linkages is crucial. A network in which all banks are directly linked to each other is most resilient. If setting up such a network is too costly, then it is preferable to have a series of smaller networks, rather than a ‘chain’ that indirectly connects all banks. The intuition for this is that, in a ‘chain,’ if a bank faces a negative shock, it might force the next bank in the chain to liquidate assets at fire-sale values; this, in turn, can trigger fire sales by another bank, and so on and so forth; hence, a smaller initial shock suffices to bring about the collapse of the entire system (than in a ‘complete network’). The notion of balance sheet contagion has several other applications that are explored in the next section.

BANK RUNS AND SELF-FULFILLING INTERNATIONAL CRISES The maturity mismatch of bank balance sheets exposes banks to the risk of ‘bank runs,’ which can lead to self-fulfilling international financial crises. As explained above, banks are able to borrow short term and lend long term because, in normal times, not all deposits are withdrawn at the same time. A run occurs when depositors panic and try to withdraw their money simultaneously. Imagine you are a depositor, and that you do not need your money now, and thus have no fundamental reason to withdraw. However, you know that if all depositors demand their money back at the same time, then the bank will be in trouble: it will have to sell its assets at fire-sale price, which might imply that the bank would fail. Whether you should withdraw depends thus on your expectations about the behavior of other depositors. This can lead to two possible outcomes. A no-bank-run equilibrium: expecting that no one will withdraw, you are better off not

withdrawing; this is true also for others depositors; hence, no one withdraws. There is also a bank-run equilibrium: expecting others to run, you are better off running as well – if you wait, the bank will deplete her assets, and fail, before you show up; this is of course true for other depositors too, and a ‘bank run’ can thus become a self-fulfilling prophecy. This simple story applies too to emerging market countries. Note that these countries usually borrow short term to finance longer term projects. They hence face the risk of a run. A run on a country is usually called a ‘sudden stop’ (of capital inflows). The basic story can also be enriched to study the cross-country contagion of self-fulfilling crisis. Consider for instance two countries, each with a bank, and imagine that all depositors decide to run in the first country. If depositors in the other country believe that this will lead to a run in their country too, there will be a run. However, if depositors think the opposite, there will be no run. Note that this is the case even if the two countries are autarkic. This purely self-fulfilling equilibrium is therefore consistent with both international contagion and the absence of contagion. In reality, though, bank runs (or sudden stops) usually follow the disclosure of some unfavorable news about economic fundamentals. Formal models of runs show that there is generally a threshold for the liquid reserves of the bank (or the economy) below which a run can occur. The bank-run threshold theory is useful for thinking about contagion because, when the same investors lend to banks in different countries, then a run in one country can affect the likelihood of a bank run in other countries. Assume for instance that there is a run in the first country, and that investors lose a substantial amount of money. Then, they might be more concerned about facing the risk of a run in the second one, and, as a result, a run in the second country becomes more likely. This can, for instance, happen when investors have decreasing absolute risk aversion, which implies that losses make them more risk-averse. This story applies also to speculative attacks on an exchange rate peg. Imagine that speculators successfully attack the peg of a first country and thereby make sizable profits. Then they will be more eager to attack the currency peg of another country. Note that this mechanism also works the other way around: a failed attack decreases the probability of a successful attack on another currency peg.

SEE ALSO Safeguarding Global Financial Stability: Resolution of Banking Crises; Central Banks Role in Financial Stability; Financial Supervision in the EU; Prevention of Systemic Crises; Lines of Defense Against Systemic Crises: Resolution.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

BANK RUNS AND SELF-FULFILLING INTERNATIONAL CRISES

Glossary Bank capital requirement Requirement that banks have to fund at least a specified fraction of their assets using bank capital (own funds). Financial contagion The strong and rapid spread of a financial crisis across markets, sectors, or countries. Global bank Bank that holds domestic and foreign assets, and borrows from domestic and foreign residents. International financial multiplier Mechanism by which a fall in asset values in one country triggers a global asset sell-off, due to balance sheets adjustments by global banks.

Further Reading Adrian, T., Shin, H.S., 2008. Liquidity and financial contagion. Banque de France Financial Stability Review 11, 1–7. Allen, F., Gale, D., 2001. Financial contagion. Journal of Political Economy 108, 1–33. Bacchetta, P., van Wincoop, E., 2010. Explaining sudden spikes in global risk. University of Lausanne Working Paper. Castiglionesi, F., Feriozzi, F., Lorenzoni, G., 2010. Financial Integration and Liquidity Crises. Manuscript, Tilburg University. Cetorelli, N., Goldberg, L.S., 2010. Global banks and International Shock Transmission: Evidence from the Crisis. National Bureau of Economic Research, Cambridge, MA. Working Paper 15974 www.nber.org. Claessens, S., Forbes, K. (Eds.), 2001. International Financial Contagion. Kluwer Academic, Boston, Dordrecht and London. Dedola, L., Lombardo, G., 2010. Financial Frictions, Financial Integration and the International Propagation of Shocks. European Central Bank Working Paper.

143

Devereux, M.B., Yetman, J., 2010. Leverage constraints and the international transmission of shocks. Journal of Money, Credit, and Banking 42, 71–105. Diamond, D., Dybvig, P., 1983. Bank runs, deposit insurance, and liquidity. Journal of Political Economy 91, 401–419. Dornbusch, P., Park, Y., Claessens, S., 2000. Contagion: understanding how it spreads. World Bank Research Observer 15, 177–197. Goldstein, I., Pauzner, A., 2004. Contagion of self-fulfilling financial crises due to diversification of investment portfolios. Journal of Economic Theory 119, 151–183. Kaminsky, G., Reinhart, C., Vegh, C., 2003. The unholy trinity of financial contagion. Journal of Economic Perspectives 17, 51–74. Kollmann, R., Enders, Z., Mu¨ller, G., 2010. Global Banking and International Business Cycles. Centre for Economic Policy Research, London, UK. Discussion Paper 7972 www.cepr.org. Krugman, P., 2008. The International Finance Multiplier. Princeton University Working Paper. Malherbe, F., 2010. Self-fulfilling liquidity dry-ups, National Bank of Belgium, Research Series.

Relevant Websites http://www.bis.org – Bank for International Settlements. http://www.bis.org – BIS Committee on the Global Financial System. http://www.federalreserve.gov – Board of Governors of the Federal Reserve System. http://www.ecb.int – European Central Bank. http://www.imf.org – IMF Global Financial Stability Report. http://www.imf.org – International Monetary Fund.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

Intentionally left as blank

C H A P T E R

15 Financial Development and Global Imbalances E.G. Mendoza*†, V. Quadrini{}† †

*University of Maryland, College Park, MD, USA National Bureau of Economic Research (NBER), Cambridge, MA, USA { University of Southern California, Los Angeles, CA, USA } Centre for Economic Policy Research (CEPR), London, UK O U T L I N E

Introduction

145

Financial Globalization and Financial Underdevelopment: Stylized Facts

145

Explaining Global Imbalances: Financial Globalization with Financial Underdevelopment Mendoza et al. (2009) Quantitative findings, extensions, and robustness Cabellero et al. (2008) Other Contributions with Market Incompleteness

147 148 150 150 151

INTRODUCTION The unprecedented global imbalances that have emerged between the United States, the rest of the industrial world, and the emerging economies over the past two decades are a puzzling phenomenon that has become a dominant issue in International Macroeconomics. These imbalances raise paramount questions that have proven difficult to answer. In general, we can classify these questions into two major classes. The first asks what the causes of the imbalances are. The second asks about the consequences of the imbalances: What are the macroeconomic implications? Are they consistent with external solvency? Are there adverse welfare consequences from the large global redistribution of wealth? And more recently, what role did the imbalances play in the ongoing world financial crisis? Logic dictates that the consequences of the imbalances (second class of questions) depend on the causes of the imbalances (first class of questions). Thus, before

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00020-9

Consequences of Global Imbalances Welfare Imbalances as a Source of Crisis Imbalances as a Mechanism for Amplification and Contagion of Crisis

152 152 152 153

Global Imbalances with Cross-Country Heterogeneity in Growth

153

Conclusions References

154 154

studying the possible consequences, we should first identify the causes. In this chapter, we explore mechanisms in which financial frictions and financial heterogeneity play a central role in causing global imbalances. Before describing these mechanisms, we present a review of the key stylized facts of global imbalances.

FINANCIAL GLOBALIZATION AND FINANCIAL UNDERDEVELOPMENT: STYLIZED FACTS Figure 15.1 summarizes the three key facts of global imbalances. The top panel shows the evolution of net foreign assets (NFA) positions in debt instruments and foreign reserves as a share of world gross domestic product (GDP) since the 1970s, the middle panel shows the net positions in risky assets (portfolio equity and foreign direct investment (FDI)), and the bottom panel shows the real interest rate. Combining the first two panels, the

145

# 2013 Elsevier Inc. All rights reserved.

NFA in debt and international reserves 4

Percent of world GDP

2 0 −2 −4 −6 −8

United States OECD countries except US Emerging economies

−10 1970

(a)

1975

1980

1985

1990

1995

2000

2005

1995

2000

2005

NFA in portfolio equity and FDI 4

Percent of world GDP

2 0 −2 −4 −6 −8

United States OECD countries except US Emerging economies

−10 (b)

1970

1975

1980

1985

1990

Gross interest rates on US three-month treasury bills 1.20 Nominal Ex-post real 1.15

Moving average of real interest rate (four period)

1.10

1.05

1.00

0.95 1981Q1 1983Q1 1985Q1 1987Q1 1989Q1 1991Q1 1993Q1 1995Q1 1997Q1 1999Q1 2001Q1 2003Q1 2005Q1

(c)

FIGURE 15.1

Net foreign asset positions in debt instruments and risky assets and US interest rate. (a) Net foreign assets (NFA) in debt and international reserves. (b) NFA in portfolio equity foreign direct investment (FDI). (c) Gross interest rates on US 3-month treasury bills.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

EXPLAINING GLOBAL IMBALANCES: FINANCIAL GLOBALIZATION WITH FINANCIAL UNDERDEVELOPMENT

147

FIGURE 15.2

Indices of financial openness from Chinn and Ito (2008). Index of capital account openness.

Index of capital account openness 3

Standardized measure

2.5

2

1.5

1

.5

United States

OECD countries except US All countries except US

0 1970

1975

1980

1985

1990

1995

total NFA of the United States started a secular decline in the early 1980s. At about the same time, we started to observe a shift toward a US portfolio short in riskless debt and long in risky assets, and we also observed a large and persistent decline in the real interest rate. Figure 15.2 uses the Chinn–Ito index of financial integration to show that the emergence of global imbalances occurred in tandem with the global integration of capital markets. This radical change has led to near-perfect international capital mobility across industrial countries and a substantial increase in capital mobility in emerging economies. Figure 15.3 provides rough evidence of the fact that, while financial globalization has been a global phenomenon (as Figure 15.2 shows), domestic financial development has not. The top panel uses the results of an International Monetary Fund (IMF) study to show that, even across industrial countries, the degree of underdevelopment of domestic financial markets relative to the United States has hardly changed. The bottom panel uses the indicator constructed by Abiad et al. (2008) to show that the same fact is observed comparing Organization for Economic Cooperation and Development (OECD) countries with emerging economies: There is financial development in absolute terms, but in relative terms, emerging economies are lagging behind OECD countries by about as much as they were 20 years ago. To summarize, the data show three defining features of the global imbalances: (a) the wealth redistribution fact, reflected in the enormous secular decline in NFA, that the United States has experienced since the mid 1980s, equivalent to about 1/10 of the world’s GDP; (b) the portfolio fact, shown in the stark portfolio structures of

2000

2005

external assets, with the United States going very short in riskless debt and long in risky assets, while emerging economies take opposite positions; and (c) the interest rate fact, reflected in a persistent decline in the US real interest rate.

EXPLAINING GLOBAL IMBALANCES: FINANCIAL GLOBALIZATION WITH FINANCIAL UNDERDEVELOPMENT Standard models based on complete markets of either neoclassical or neokeynesian upbringing find it difficult to account for all the three defining features of the global imbalances described above: the wealth redistribution fact, the portfolio fact, and the interest rate fact. Two recent contributions have suggested that heterogeneity in financial markets across countries could play an important role in explaining these facts. The first contribution is Cabellero et al. (2008) (henceforth CFG) and the second is Mendoza et al. (2009) (henceforth MQR). Though these two papers suggest the common role played by financial heterogeneity, they present different theoretical frameworks. In particular, MQR emphasize the role played by uncertainty and risk, while there is no uncertainty in the CFG framework. The two papers differ also in the main driving force underlying the buildup of imbalances in the presence of financial market heterogeneity. MQR argue that the imbalances started to emerge when the international capital markets were liberalized. CFG argue that the imbalances emerged as a consequence of shocks to productivity growth and/or to the financial structure.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

148

15. FINANCIAL DEVELOPMENT AND GLOBAL IMBALANCES

FIGURE 15.3 Indices of financial market heterogeneity. The index in panel (a) is from the International Monetary Fund (IMF) (2006). The index in panel (b) is from Abiad et al. (2008). (a) Financial index score for advanced economies. (b) Index of financial liberalization.

Financial index score for advanced economies 0.8

1995 2004

0.7 0.6 0.5 0.4 0.3 0.2

USA

SWE

PRT

NOR

NLD

ITA

JPN

GRC

FRA

GBR

FIN

ESP

DEU

DEN

CAN

BEL

(a)

AUT

0

AUS

0.1

Index of financial liberalization 1

0.8

0.6

0.4

0.2 OECD countries Emerging economies

0 (b)

1970

1975

1980

1985

1990

1995

2000

Mendoza et al. (2009) MQR and a more recent paper by Mendoza and Quadrini (2010) propose a framework capable of accounting for the three facts outlined above. The approach introduces a new way of thinking about International Macroeconomics that emphasizes the role of risk at the level of an individual agent (idiosyncratic) coupled with heterogeneous economic institutions across countries (financial development). The main conclusion is that the three facts outlined in Figure 15.1 resulted from the process of financial globalization among countries with heterogeneous domestic financial markets. MQR do not rely on any of the mechanisms that several studies in the literature have explored (e.g., cross-country risk diversification, growth differentials, differences or changes in aggregate output variability,

2005

and large unexpected shocks to preferences, technology, or fiscal policies). They consider a world economy made up of N countries, each inhabited by a continuum of agents. Agents in each country have identical preferences and face idiosyncratic shocks. There are no aggregate country-level shocks. Therefore, cross-country risk sharing is not an issue. The problem of risk sharing is among individual agents within a country. Financial markets heterogeneity derives from the different degrees of contract enforcement captured in the model by two parameters: one is a lower bound on debt (a limited liability constraint) and the second is a parameter that limits the ability to use state-contingent claims to insure away individual risks. Financial development means better enforcement of financial contracts, that is, lower bounds on debt and larger ranges of state-contingent claims that can be purchased in the market. In least

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

EXPLAINING GLOBAL IMBALANCES: FINANCIAL GLOBALIZATION WITH FINANCIAL UNDERDEVELOPMENT

developed markets, only contracts that are not state-contingent are available (bond economy). When the enforcement of contract is strong enough, the environment supports the full array of Arrow securities without any frictions (perfect insurance). Essentially, financial development determines the degree to which individual agents can insure the idiosyncratic risks. In fact, for any degree of financial development, and thus with some availability of statecontingent claims, the economy is equivalent to a bond economy but with a transformed (less risky) process for the exogenous shocks. This represents a convenient methodological innovation that facilitates the characterization of the model with tools that are typically used in solving models with uninsurable idiosyncratic risks. Now think of the world economy living in global financial autarky. If we simplify the model so that (i) all incomes come from an exogenous stochastic individual earnings process and (ii) there are either only nonstatecontingent debt instruments or full Arrow securities, the model collapses to one of two canonical models: the model with full Arrow securities and perfect-risksharing with an equilibrium interest rate equal to the rate of time preference; or the bond economy with only nonstate-contingent debt as in the classic Bewley (1986)– Huggett (1993)–Aiyagari (1994) setup. In the latter, incomplete markets create incentives for precautionary saving that result in a well-defined distribution of wealth across ex post heterogeneous agents, with an equilibrium rate of interest lower than the rate of time preference. Suppose now that there are only two countries in the global economy, the ‘Arrow country’ with perfect enforcement and the ‘Bewley country’ with a single nonstate-contingent bond. Under autarky, the equilibrium of each country looks like those we described above. In particular, the first country would have a higher interest rate than the second country. What happens if the two countries decide to integrate their financial markets? The new equilibrium will be characterized by a single worldwide interest rate, which is lower than the rate of time preference (if the two rates

K1(r)

r1

K2(r)

r2

(a)

were equal, Bewley citizens would accumulate an infinite amount of assets). This implies that capital markets integration leads to a decline in the interest rate of country 1 and an increase in country 2. As a result, agents in country 1 save less while agents in country 2 save more. This implies that country 1 borrows from country 2. Thus, the Arrow nation ends up with negative NFA and the Bewley country ends up with positive NFA. In short, financial globalization leads to global imbalances between the Arrow and Bewley economies, and it lowers the real interest rate relative to the autarky rate of the Arrow country. MQR provide formal proofs of the above arguments and show that the results extend to a more general (and more realistic) case in which the two countries are not at the extremes of Arrow and Bewley, but somewhere in between. In this more realistic scenario, they label the two countries as the United States, with a higher level of financial development, and the rest of the world (ROW), with a lower level of financial development. Figure 15.4 shows how equilibrium is determined. The upward-sloping, concave curves are analog to the savings curves widely used to characterize the equilibrium of closed-economy heterogeneous-agent models, but recall that in these models typically only nonstatecontingent bonds are traded. Here agents trade statecontingent claims, but with the caveat that enforcement in ROW is weaker than in the United States. This explains why the savings curve for the US sits higher than that for ROW. Moreover, compared with standard heterogeneous-agent models, the main novelty here is in deriving the open-economy implications of the analysis: financial integration reduces (increases) the risk-free rate in the US (ROW), and results in a negative (positive) NFA position for the US (ROW). Thus, the model can explain two of the three facts outlined above: the wealth fact and the interest rate fact. What about the third fact, that is, the portfolio structure? Extending the model to address this issue requires that we introduce some additional assets with risky return in addition to state-contingent claims. We assume that agents can now buy shares of an asset that can be

Supply of K

Supply of K

Autarky

K

K1(r) K2(r)

r

(b)

149

Mobility

FIGURE 15.4 Steady-state equilibria with heterogeneous financial conditions. (a) Autarky. (b) Mobility.

K

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

150

15. FINANCIAL DEVELOPMENT AND GLOBAL IMBALANCES

used in individual production. Individual production is subject to idiosyncratic shocks. The asset is in fixed supply in each country but individual agents can buy different quantities of it (up to the limit of the available aggregate supply, domestic or global, depending on the international financial regime). Thus, each agent chooses a portfolio of production assets and state-contingent claims. The key feature of this extension is that individual production is subject to the idiosyncratic shock. As the magnitude of the risk brought by this shock depends on the asset used in production, we refer to it as investment risk. Notice that agents have the choice of whether or not to face this risk by choosing their portfolio composition (whether or not to buy the productive asset). This is the key difference between the investment shock and the more standard shocks to endowments and/or earnings, because the exposure to this risk is not under the control of the agent. Reconsider now the world with the Arrow and Bewley countries. In the Arrow country, the marginal product of capital is equal to the world interest rate, and there is no equity premium, that is, the excess marginal return on the productive asset. This is because agents can fully insure the investment risk. In the Bewley country, instead, there is an equity premium because Bewley agents will demand a higher marginal return on the productive investment as the risk cannot be insured perfectly. Now let us again consider financial globalization between these two countries. As before, the interest rate falls (rises) relative to autarky for Arrow (Bewley) citizens, and the position in state-contingent claims falls (rises) for the Arrow (Bewley) nation. What about the net position in risky assets? Clearly Arrow citizens are more tolerant of the risk of holding risky assets than the Bewley citizens, so the former (latter) end up with a positive (negative) net position in risky assets. Essentially, the Arrow country purchases the productive asset from the Bewley country by borrowing from this country. Quantitative findings, extensions, and robustness In MQR, the properties of the economy discussed above are also shown quantitatively in environments with intermediate degrees of market incompleteness. The model is calibrated to the standard parameters of a heterogeneous-agent model, including an individual earnings process and profits process set to match US data, and to observed measures of the size of the US economy relative to ROW. The choice of the parameters that characterize financial frictions (the enforcement parameter and the limited liability constraint) aims at matching the relative financial underdevelopment documented earlier, under the assumption that contract enforcement is residence-based (i.e., the enforcement of contracts for agents depends on the country of residence

even on activities performed abroad). MQR also consider alternative enforcement assumptions. Comparing stationary equilibria under financial autarky and under full financial globalization, the baseline model predicts changes in line with the three facts of financial globalization emphasized above: (A) the wealth redistribution fact reflected in the enormous secular decline in NFA for the United States; (B) the portfolio fact, with the United States going short in riskless debt and long in risky assets; and (C) the interest rate fact characterized by a decline in the US real interest rate. Quantitatively, the model with full capital market integration predicts a long-term NFA of 51.4% (22.1%) of income in the United States (ROW). Also, the United States takes a short position in riskless debt equivalent to 88.8% of income and a long position in risky assets of 37.4% of income. Furthermore, the US interest rate falls by 25 basis points relative to its autarky level. The transition from financial autarky to the new steady state with financial globalization is a gradual process taking about 30 years. How robust are these results? One can point to a number of assumptions that are unrealistic or abstract from important elements of financial globalization. In order to address some of these concerns, MQE extended the model to add the following elements: (1) domestic financial heterogeneity in both enforcement and limited liability (debt or credit limits); (2) alternative environments for the enforcement of cross-border contracts (sourcebased vs. residence-based); (3) risk-sharing motives for international diversification of asset holdings; (4) growth differentials, with less financially developed countries growing faster (e.g., China); and (5) cross-country differences in individual income volatility. We also considered a three-country setup designed to represent the United States, other industrialized countries, and emerging economies. In all these scenarios, the model keeps predicting that the United States builds a large negative NFA position and converges to a lower real interest rate. Most of the experiments also keep predicting that US agents go short in riskless assets and long in risky debt, and in the case of the three-country model, we obtain portfolio allocations in line with what we see in the data. The only case in which some of the results are weakened is when the enforcement of cross-border contracts in all countries is 100% source-based. In this case, the model is still consistent with facts A (wealth redistribution fact) and C (interest rate fact) but it does not explain the international portfolio composition of the United States.

Cabellero et al. (2008) The structure of the CFG model can be described in a stylized fashion as follows. Consider an economy populated by a continuum of agents who live for two periods. Agents receive an endowment when young but they

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

EXPLAINING GLOBAL IMBALANCES: FINANCIAL GLOBALIZATION WITH FINANCIAL UNDERDEVELOPMENT

want to consume only when old. Because the endowment is not storable, the only way to consume when old is to use the endowment to buy an asset from old agents. The asset is similar to the famous Lucas tree. The problem is that trees are in limited supply and the availability of trees differs across countries (or equivalently, countries have the same quantity of trees but they produce different fruits). CFG interpret the number of available trees as characterizing the properties of the financial system. Essentially, the more the trees, the more the financial instruments that young savers can buy to transfer consumption into the future. Countries where there are more trees available are interpreted as more financially developed. Given a larger number of trees, their equilibrium price is lower and the interest rate higher. What happens when two countries with different financial development (different availability of trees) become financially integrated? Because the preliberalization price of the trees is lower in the more developed country, young agents in the less developed country will buy foreign trees, bringing cross-country equalization in prices and interest rates. This generates a trade surplus in the less developed country since young agents export part of the endowment to pay for the foreign trees. The CFG model can explain facts (A) and (C) as the MQR model does. However, the mechanism by which it does so is different. In CFG, the imbalances do not derive necessarily from differences in savings. Instead, the imbalances derive from the fact that in less financially developed countries, savers have more difficulties to transfer savings into the future because of the insufficient supply of investment instruments. In MQR, on the contrary, the imbalances derive from the higher savings of less developed countries. Countries with lower financial markets development save more because they face higher uninsurable risks. Risk is central to the analysis of MQR. In their setup, the imbalances would disappear in the absence of uncertainty. There is also a difference in terms of explanations for why the US imbalances started to grow. The view of MQR is that the ultimate factor underlying the buildup of the imbalances is the international liberalization of financial markets that started in the early 1980s and expanded in the 1990s. The view of CFG, on the other hand, is multidimensional. For example, the reason why the US imbalances grew in the 1990s is that the United States experienced higher growth than other industrialized countries. The reason why Asian current accounts surged, however, is different. In particular, CFG argue that the Asian financial crisis of 1997 was a shock that reduced the number of the region’s trees or fruits. This led these economies to demand more foreign assets from countries like the United States.

151

CFG can also capture the portfolio composition fact once the model is extended to allow for the plantation of new trees. The reason most financially developed countries will plant trees in foreign countries, which are interpreted as FDIs, is that they have the ability to plant better trees than less developed countries. Again, a better tree here should be interpreted as higher ability to create better financial instruments. In general, besides the emphasis on financial markets heterogeneity, CFG and MQR propose very different approaches both methodologically and as an explanation for the imbalances. Of course, this does not mean that these two approaches are substitutes. Quite the contrary, they can be seen as complementary approaches, which combined provide a deeper understanding of the recent trends toward the formation of large global imbalances.

Other Contributions with Market Incompleteness There are other contributions in the literature that emphasize financial market imperfections in the analysis of global imbalances. Most of the papers, however, do not address the second fact outlined above, that is, the portfolio fact. Willen (2004) studied the qualitative predictions of a two-period endowment economy with exponential utility and normal independent and identically distributed (i.i.d.) shocks. He showed that, under incomplete markets, trade imbalances emerge because of reduced savings by the agents residing in countries with ‘more complete’ asset markets. This chapter has already captured one of the mechanisms emphasized in MQR as described above, that is, the fact that countries with higher financial development accumulate negative NFA. However, unlike MQR, Willen’s model does not allow for endogenous production with ‘investment risks,’ which is the key to explaining the composition of external-asset portfolios in addition to total NFA positions. More recently, Prades and Rabitsch (2007) developed a model with incomplete markets that emphasizes the importance of cross-country financial market heterogeneity for the formation of global imbalances. One feature that differentiates this paper from the recent contributions described above is the central role played by aggregate uncertainty. In MQR and CFG, there is no aggregate uncertainty at the country level. Therefore, cross-country risk sharing does not play any role in the mechanism driving global imbalances in these papers. More specifically, Prades and Rabits show that global imbalances may arise from cross-country asymmetries in capital controls, which are modeled as lower and upper limits to the volume of international financial trades. Essentially, these limits reduce the ability to share the risk

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

152

15. FINANCIAL DEVELOPMENT AND GLOBAL IMBALANCES

across countries. Thus, countries with tighter limits save more for precautionary purposes. Cross-country limited risk sharing also plays a central role in Fogli and Perri (2006). This paper emphasizes a different source of cross-country heterogeneity, which is not financial market differences or differences in capital controls. The key mechanism in Fogli and Perri is the heterogeneity in country-level risk or aggregate uncertainty. They start from the observation that since the early 1980s, the macroeconomic volatility in the United States has declined more than in other industrialized countries. Because of this, precautionary savings have declined more in the United States than in other industrialized countries. As a result, the United States has started to borrow from other countries.

CONSEQUENCES OF GLOBAL IMBALANCES The consequences of global imbalances can be examined from more perspectives than we can focus on in this short note. Hence, we choose to focus on two that are likely to be the most relevant: the implications of global imbalances for welfare and financial crises.

Welfare One of the most obvious questions is whether financial globalization is welfare improving for the participating countries. MQR is one of the papers that addresses this question. The welfare effects are calculated as the percentage change in consumption across state and time that each individual would require to be as well off under financial integration as under financial autarky. Because of agent heterogeneity, the welfare consequences are in general different for different agents: some may gain and some may lose. Therefore, in order to report a measure at the country level, some form of aggregation needs to be adopted. MQR uses a hypothetical utilitarian planner that weighs each agent in the country equally. The welfare measure is then the percentage increase in aggregate consumption (same percentage increase for all agents) in the autarky equilibrium that would make the planner indifferent between staying in autarky or transiting to the equilibrium with capital mobility. The aggregate welfare measures indicate that the more financially developed country, the United States, gains from capital liberalization while the ROW experiences a negative welfare effect. The gains of the United States, however, could be bigger than the losses in the ROW. Therefore, a worldwide planner may still prefer globalization.

These welfare effects result primarily from the fact that, relative to autarky, the cost of borrowing falls for the United States and rises for the ROW. This mechanism has different welfare implications among agents within the same country. More specifically, because ‘poor’ agents are net borrowers, they gain in the United States but experience a loss in the ROW. The opposite is true for ‘rich’ agents who are net lenders. Because of the concavity of the utility functions, the losses of poorer (low consumption) agents are more painful and the gains more important. As a result, in the equally weighted welfare function, the welfare impact of capital liberalization is dominated by the impact on poorer agents. Thus, the aggregate consequence is positive for the United States and negative for the ROW. These results are robust to all the extensions considered in the MQR paper as well as to the introduction of capital accumulation explored by Mendoza et al. (2008). With capital accumulation, the model predicts the reallocation of capital from the ROW to the United States resulting from financial globalization. Hence, in this case, the welfare effects are amplified by the international allocation of capital and the consequent impact on wages. Wages rise in the United States and fall in the ROW, and as the poor collect a larger fraction of income from wages, the wage changes make matters worse for poorer agents in the ROW and better for the poorer agents in the United States. Should we learn from these findings that financial globalization was a bad idea and should be reconsidered, at least for some countries? Certainly not. MQR argue that these welfare results should be viewed as evidence of the importance of promoting the structural and institutional changes that enhance domestic financial development. Perhaps capital liberalization itself could be an automatic mechanism creating the incentives for institutional improvements. This is the mechanism that was considered in MQR.

Imbalances as a Source of Crisis There is a great deal of concern about the sustainability of the current imbalances. Experiences in emerging economies have shown that in many instances, the buildup of foreign imbalances has been followed by financial crises that brought the economies back to more balanced positions. It is then natural that many observers see the potential of an implosion for the United States as well leading to a severe and costly crisis. Without going into too many details, many observers have also recognized that the US imbalance is different from previous imbalances in emerging economies. One difference is the fact that most of the US foreign debt is not denominated in foreign currency while a large

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

GLOBAL IMBALANCES WITH CROSS-COUNTRY HETEROGENEITY IN GROWTH

share of US foreign assets is denominated in foreign currency. Therefore, a sudden depreciation of the US dollar is unlikely to generate the type of disruption experienced in various episodes of currency crisis in emerging economies. All this boils down to the question of whether the US imbalance is sustainable. In MQR, global imbalances are ‘sustainable’ in the sense that they represent the equilibrium outcome of global financial integration in a world where differences in domestic financial development remain unchanged. The imbalances are consistent with solvency conditions at the individual and aggregate levels in each country. The buildup of the imbalance will eventually stop, but this will happen smoothly and not require a sudden and costly reversal. Of course, this does not exclude the potential for a future crisis. However, if this is a potential threat, it must derive from some additional mechanisms that are not necessarily related to the differential incentives to save, as investigated in MQR, or the differences in available investment instruments as in CFG.

Imbalances as a Mechanism for Amplification and Contagion of Crisis Even if we think that the global imbalances are not a ‘source’ of financial crisis, there is still the issue of whether the transmission of a financial crisis is different in a globalized economy with large imbalances, in particular, whether the impact of a crisis could be amplified in the presence of imbalances and whether it could generate a contagion to other countries. This has been a central issue in the recent financial crisis. Mendoza and Quadrini (2010) investigate how the foreign imbalances accumulated by the United States have affected the current financial crisis. They extend the MQR model by adding a richer financial intermediation structure. In particular, agents in each country are now split into half ‘savers’ and half ‘producers.’ Savers are like the heterogeneous agents of the original MQR model. Producers are modeled as representative producers that do not face aggregate or idiosyncratic risk, but face a similar enforcement constraint as the savers. Banks intermediate funds between savers and producers, with three features that make the role of banking central to the analysis: First, banks cannot pay negative dividends. Second, banks face a mark-to-market capital requirement that requires ‘standard’ loans not to exceed a fraction of the market value of their capital. Third, banks can issue loans that escape the capital requirement, but in doing so, they incur a convex cost that rises as individual loans exceed a threshold price. Banks minimize costs so that at equilibrium the threshold price equals the capital requirement.

153

In the long run, the equilibrium allocation of loans is lower than the amount set by the capital requirement, and hence banking is inessential in the steady state. The main goal of the analysis is to study the offsteady-state effect of a small once-and-for-all, unanticipated shock to the balance sheet of US banks under financial autarky and under financial globalization. The calibrated version of the model shows that a shock that generates a loss of 1.5% to the stock of loans made by US banks produces a 12% asset price decline under financial globalization. Even though the loan losses are initially experienced only by US banks, the price decline is global. Thus, there are big spillovers to countries integrated with the United States. At impact, the risk-free rate drops but the financing cost rises (as banks hit the capital requirement and move to finance loans at the increasing cost), so the model predicts a sharp rise in the spread between firm-financing costs and the risk-free rate. Would the response to the shock have differed had the United States been in financial autarky? By comparing the counterfactual response in the autarky regime, Mendoza and Quadrini (2010) examine the role of financial globalization, which has three important transmission effects. First, global capital markets provide a larger market to absorb the asset sales triggered by the credit shock. This effect works to weaken the asset price response to the shock. Indeed, if the same credit shock were to hit the US economy under financial autarky, the asset price would fall twice as much than in the baseline with financial globalization. Thus, financial globalization does not act as an amplification mechanism but as a smoothing mechanism for the country where the shock originates. The second effect works through the buildup of leverages in the US economy that are produced by financial globalization with less financially developed countries. Globalization is the force that drives the growth of leverage in the United States and without it, the leverage would be significantly lower. Lower leverages imply that the effects of the crisis would be weaker. Conversely, the larger the differences in domestic financial development, the larger is the buildup of leverages and the stronger the financial crisis triggered by the credit shock. The third effect is contagion. As mentioned above, the US shock triggers a crisis in the ROW, which would not happen under financial autarky.

GLOBAL IMBALANCES WITH CROSSCOUNTRY HETEROGENEITY IN GROWTH A well-known feature of standard utility-based optimization models is that economic growth has a negative impact on savings. The intuition is straightforward.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

154

15. FINANCIAL DEVELOPMENT AND GLOBAL IMBALANCES

If I expect my income to be higher in the future, consumption smoothing requires me to increase consumption today relatively to my current income. In a closed economy, this will generate a hike in the interest rate. In an open economy, on the other hand, higher growth (relative to other countries) induces foreign borrowing. Unfortunately, the evidence does not support this basic prediction of standard models. The anecdotal example is the recent experience of China, a fast-growing economy that has been exporting capital. Gourinchas and Jeanne (2007) provide more formal evidence for emerging economies in general. So why do emerging countries with fast-growing economies save more than they invest? There have been several attempts at explaining this pattern, primarily with some type of market incompleteness. Carrol and Jeanne (2008) consider a model with uninsurable idiosyncratic risks that generates bufferstock patterns of savings. The model is in the spirit of MQR and, therefore, it shares some of the key properties. In particular, keeping the growth rate constant, higher uncertainty implies higher savings. At the same time, keeping constant the degree of market incompleteness and the riskiness of the idiosyncratic shock, higher growth implies lower savings as in the typical model with complete markets. The key to breaking the negative link between growth and savings is assuming that uncertainty is not independent from growth. Carroll and Jeanne posit that fast-growing economies are likely to experience higher uncertainty at the level of individuals. Thus, the impact on savings depends on two opposite effects: the negative effect of higher growth and the positive effect of higher uncertainty. If the increase in uncertainty is sufficiently large relative to the increase in growth, the model predicts that growing economies are net exporters of capital. This mechanism is complementary to the mechanism explored in MQR. While in MQR agents in more developed countries face a lower risk because of better insurance options, in Carroll and Jeanne these economies face a lower risk because of the different nature of the risk. Once we put these two mechanisms together, the higher savings of emerging economies could result from lower financial development and higher uncertainty if the total impact of these two mechanisms is stronger than the negative impact of growth. Another attempt at explaining the positive relation between growth and export of capitals, with special attention to China, is found in Song et al. (2009). They build a transitional growth model with two important elements: financial imperfections and reallocation among firms with heterogeneous productivity. During the transition, some firms use more productive technologies than others, but low-productivity firms survive because of better access to credit markets. Due to the financial

imperfections, high-productivity firms must be financed out of internal savings from entrepreneurs. If these savings are sufficiently large, the high-productivity sector outgrows the low-productivity sector, and attracts an increasing employment share. During the transition, low wage growth sustains the return to capital. The downsizing of the financially integrated sector forces a growing share of domestic savings invested in foreign assets, generating a foreign surplus. The model is stylized and their paper provides only qualitative predictions that are consistent with salient features of the Chinese growth experience since 1992: high output growth, sustained returns on capital investments, extensive reallocation within the manufacturing sector, falling labor share, and accumulation of a large foreign surplus. The entrepreneurial incentives in growing economies due to the need of internal finance are also studied in Sandri (2009). In this model, entrepreneurs could save more than needed to finance investment growth because of precautionary savings. Through this mechanism, the model could generate current account surpluses in fast-growing developing countries. A calibrated version of the model shows that this is a plausible outcome.

CONCLUSIONS Understanding the global imbalances phenomenon and its implications for macroeconomic stability and the stability of capital markets has become a central issue in International Macroeconomics. We reviewed recent developments in the literature that emphasize the role of market incompleteness and financial markets development for understanding the three key defining features of global imbalances: the wealth fact, the portfolio allocation fact, and the interest rate fact. The quantitative findings of this literature indicate that the heterogeneity of domestic financial markets in today’s era of financial globalization is significant for explaining the global imbalances facts. We also discussed some puzzles that have not been fully resolved, in particular, the observation that fastgrowing countries have the tendency to become net suppliers of international capital. This is another way of reformulating the Lucas (1990) question of why capital does not flow to poor countries. We have reviewed some recent studies that address this question, most of which are based on some form of financial market incompleteness.

References Abiad, A., Detragiache, E., Tressel, T., 2008. A New Database of Financial Reforms. Working Paper no. 08/266 (December). International Monetary Fund, Washington, DC.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

CONCLUSIONS

Aiyagari, S.R., 1994. Uninsured idiosyncratic risk and aggregate saving. Quarterly Journal of Economics 109 (3), 659–684. Bewley, T.F., 1986. Stationary monetary equilibrium with a continuum of independent fluctuating consumers. In: Hildenbrand, W., Mas-Colell, A. (Eds.), Contributions to Mathematical Economics in Honor of Gerard Debreu. North-Holland, Amsterdam. Cabellero, R.J., Farhi, E., Gourinchas, P.O., 2008. An equilibrium model of global imbalances and low interest rates. American Economic Review 98 (1), 358–393. Carrol, C.D., Jeanne, O., 2008. A Tractable Model of Precautionary Reserves or Net Foreign Assets. Johns Hopkins University, Baltimore, MD. Unpublished manuscript. Chinn, M.D., Ito, H., 2008. A new measure of financial openness. Journal of Comparative Policy Analysis 10 (3), 307–320. Fogli, A., Perri, F., 2006. The Great Moderation and the US External Imbalance. NBER Working Paper No. 12708. Gourinchas, P.O., Jeanne, O., 2007. Capital Flows to Developing Countries: The Allocation Puzzle. NBER Working Paper No. 13602. Huggett, M., 1993. The risk free rate in heterogeneous-agents, incomplete markets economies. Journal of Economic Dynamics and Control 17 (5/6), 953–970.

155

Lucas, R.E., 1990. Why doesn’t capital flow from rich to poor countries? American Economic Review 80, 92–96. Mendoza, E., Quadrini, V., 2010. Financial globalization, financial crises and contagion. Journal of Monetary Economics 57, 24–39. Mendoza, E., Quadrini, V., Rı´os-Rull, J.V., 2008. On the welfare implications of financial globalization without financial development. In: Clarida, R., Giavazzi, F. (Eds.), NBER International Seminar. University of Chicago Press, Chicago, IL, pp. 199–239. Mendoza, E., Quadrini, V., Rı´os-Rull, J.V., 2009. Financial integration, financial development, and global imbalances. Journal of Political Economy 117 (3), 371–416. Prades, E., Rabitsch, K., 2007. Capital Liberalization and the US External Imbalance. European University Institute, Florence. Unpublished manuscript. Sandri, D., 2009. Growth and Capital Flows with Risky Entrepreneurship. John Hopkins University, Baltimore, MD. Unpublished manuscript. Song, Z.M., Storesletten, K., Zilibotti, F., 2009. Growing like China. CEPR Discussion Paper 7149. Willen, P.S., 2004. Incomplete Markets and Trade. Working Paper Series No. 04-8, Federal Reserve Bank of Boston.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

Intentionally left as blank

C H A P T E R

16 Foreign Currency Debt M. Chamon International Monetary Fund, Washington, DC, USA O U T L I N E Introduction

157

Development of Local Bond Markets

159

Risks of Foreign Currency Debt

157

Reasons for Holding Foreign Currency Debt

158

Conclusion References

160 160

INTRODUCTION Most foreign lending to developing countries is denominated in foreign currency. As discussed in detail in this chapter, this is undesirable from a risk perspective, and creates financial fragility. The chapter then surveys some of the explanations to why foreign currency debt is chosen despite the risks involved, and the development of local bond markets.

RISKS OF FOREIGN CURRENCY DEBT As the exchange rate tends to appreciate when times are good and depreciate when times are bad, denominating liabilities in a foreign currency shift the debt burden from good to bad states. This is undesirable from a risksharing perspective (except for borrowers with a natural hedge), particularly if external financing constraints are more likely to bind during the bad times (when the exchange rate is depreciated). In the absence of foreign currency debt, an exchange rate depreciation can play a stabilizing role following adverse shocks (e.g., through traditional expenditure shifting effects that increase the demand for home goods). But if there is a large stock of liabilities in foreign currency, an exchange rate depreciation can compound an adverse shock, by weakening the balance sheet of firms. Under foreign currency debt, ☆

the greater the depreciation, the greater the resources (in domestic currency terms) that need to be transferred to foreigners, further decreasing the relative price of nontradables, which feeds back into more downward pressure on the currency. Exchange rates can be very volatile. The domestic currency may depreciate sharply following adverse terms of trade shocks, a weaker macroeconomic environment, or a contraction in global liquidity. This volatility may not be much of a concern for sectors with a natural hedge against exchange rate movements, such as exporters. But it can be very problematic for borrowers without a natural hedge. For example, an otherwise viable nontradable firm may find itself under distress if the local currency value of its liabilities were to increase substantially due to a sharp movement in the exchange rate. The risk stemming from currency mismatches can be compounded in the presence of external financing constraints. If a firm faces borrowing constraints based on the value of its net worth, an exchange rate depreciation will lower the value, in foreign currency terms, of its nontradable collateral. That will limit its ability to borrow, causing its investment to decline. If several firms are in this situation, one can have a setting like that in ‘third-generation’ currency crises models (e.g., Aghion et al., 2004; Krugman, 1999) where multiple equilibria involving a currency crisis can occur. In such a setting, if creditors cut lending, this credit crunch leads to a

The views expressed in this chapter are those of the author and do not necessarily represent those of the IMF or IMF policy.

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00013-1

157

# 2013 Elsevier Inc. All rights reserved.

158

16. FOREIGN CURRENCY DEBT

collapse in investment (through the channel described above), which leads to a decline in the demand for local goods and a currency depreciation, lowering the value of the firm’s nontradable collateral, rationalizing the credit crunch. Foreign currency debt is particularly dangerous when routed through the domestic banking system. If a bank borrows from abroad in foreign currency (or takes foreign currency deposits domestically) to lend to domestic firms in local currency, it would expose itself to a currency mismatch. If it lends in foreign currency to borrowers without a natural hedge, it may be just trading the explicit currency risk for an implicit credit risk. Given the volatility of exchange rates, particularly during crisis episodes, currency mismatches can easily wipe out a bank’s capital, and if mismatches are widespread, lead to a systemic banking crisis. And the presence of foreign currency deposits can limit the Central Bank’s ability to act as the lender of last resort during a banking crisis. The international finance literature has paid increasingly more attention to the risks posed by currency mismatches, especially after the Asian Crisis. Both academics and policy makers are quite aware of the vulnerabilities created by currency mismatches.

REASONS FOR HOLDING FOREIGN CURRENCY DEBT If foreign currency debt is so dangerous, why is it so common? The issue of foreign debt denomination was traditionally dismissed as a straightforward moral hazard problem. If the foreign debt were denominated in local currency, the government would have an incentive to inflate in order to decrease the real value of its liabilities and that of its citizens. As a result, foreigners would be unwilling to lend in domestic currency (or at least not without demanding a large premium). While simple, and plausible at first sight, there are a number of limitations to this simple explanation. First, these fears could be attenuated by inflation-indexed contracts (although lenders may still fear that the government would misreport the inflation index). Second, and more importantly, small lenders and borrowers do not internalize the effects of their actions on monetary policy. That is, while inflation may indeed be higher if foreigners lend more in local currency, an individual borrower or lender will take the expected inflation rate as given when choosing between local and foreign currency debt. Eichengreen and Hausmann (1999) propose an ‘original sin’ hypothesis. According to that view, emerging markets suffer from a fundamental incompleteness in financial markets which prevents the domestic currency from being used to borrow abroad, or to borrow

long term even domestically. In the presence of this incompleteness, financial fragility is unavoidable, not because borrowers are imprudent, but because they must either borrow in foreign currency (and hold currency mismatches) or borrow short term in domestic currency (and hold maturity mismatches). As currency mismatches became one of the preferred explanations for the vulnerability to currency crises, a number of papers proposed specific channels that could help explain borrowing in foreign currency: Uncertainty over inflation rate: Jeanne (2005) presents a model where there is limited monetary policy credibility, and there is a small probability that inflation will be very high. Local currency debt must have a nominal interest rate that is sufficiently high to compensate for that risk. As a result, the ex post real interest rate in local currency will be either too high (if inflation remains low) or too low/negative (if inflation is indeed high), making local currency debt actually riskier than borrowing in foreign currency. Note that while the traditional monetary policy credibility story focused on the level of inflation, Jeanne’s explanation relies on the variability of the inflation rate. Broader government moral hazard: There are a number of government policies that can affect the productivity of domestic firms, and as a result the prospects of repaying foreigners. Some types of liability structure may be more sensitive to government policies. For example, there are a number of policies that can affect the value of the real exchange rate, beyond monetary policy. Tirole (2003) emphasizes the role of the private borrowers’ government, with whom their investors do not contract. He shows that at the margin, local currency debt can actually be welfare reducing (through its distortion of the government’s policies). That is, the vulnerabilities created by foreign currency debt can actually play a disciplining role in government policy. While an individual borrower or lender would take the government’s actions as given, the introduction of uncertainty considerations could discourage the use of local currency debt (along the lines discussed in Jeanne (2005)). Bailout guarantees: Schneider and Tornell (2004) and Burnside et al. (2001) emphasize the role of bailout guarantees. In these models, when foreign currency debt is widespread, agents anticipate a government bailout in the event of a crisis (to prevent widespread bankruptcy), causing them to choose cheaper foreign currency debt over the safer (and more expensive) local currency debt (because they reap the benefits of the cheaper financing while the government bears much of the cost if there is a crisis). Moreover, it is precisely these currency mismatches that create the fragilities that allow a self-fulfilling currency crisis to happen. Disciplining device: Jeanne has written a series of papers where short-term foreign currency debt serves

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

DEVELOPMENT OF LOCAL BOND MARKETS

as a commitment device. In Jeanne (2009), the risk of a crisis if creditor confidence falters induces the government to pursue a necessary (but costly) policy reform. Dilution: Chamon (2003) presents a model where the interplay between credit and currency risk creates an opportunity to dilute local currency debt. It assumes that in the event of a default, creditors are repaid proportionately to the face value of their claims (pari passu). As defaults are associated with depreciation, local currency debt will tend to fare worse after a default. This creates a dilution opportunity, whereby a borrower with local currency debt can take additional foreign currency debt at favorable terms (because this new debt will fare better in the event of a default at the expense of the local currency debt). If the borrower cannot commit not to pursue this dilution strategy, it may end up with excessive foreign currency debt. Debt seniority clauses could prevent such dilution, and therefore this channel seems more relevant for sovereign debt than for private debt (where seniority could in principle be enforced by courts). Broda and Yeyati (2006) make a similar point in the context of dollarization of bank deposits. de la Torre and Schmukler (2004) present a model where after a crisis and default, local currency debt holders are more time-pressed to settle because the higher inflation (in the aftermath of the crisis) erodes the real value of their bonds. This will put them in a weaker bargaining position, and to compensate for that risk they demand a higher interest rate. Interplay between monetary policy and liability denomination: Chamon and Hausmann (2005) present a model where private borrowers choose between short-term local currency debt and foreign currency debt (i.e., choose between interest rate and exchange rate risk). If most agents borrow in foreign currency, the Central Bank may decide to stabilize the exchange rate at the expense of higher volatility in the interest rate, at least over some parameter range, in order to minimize costly bankruptcy of those borrowers. This can lead to multiple equilibria, whereby the more people borrow in foreign currency, the more monetary policy will try to protect them from swings in the exchange rate, further encouraging foreign currency debt. Credit market frictions: Caballero and Krishnamurthy (2003) present a real model where excessive dollar debt is the result of domestic financial constraints that lead firms to undervalue the social benefit of borrowing in local currency. Natural hedge: Perhaps the best reason for holding foreign currency debt is to have a natural hedge. Bleakley

159

and Cowan (2008) study a sample of Latin American firms in the 1990s and find that the investment of firms holding dollar debt is not adversely affected following a depreciation vis-a`-vis their peso-indebted counterparts. This result is attributed to firms matching the balance sheet effect to the exchange rate sensitivity of their profits. While interesting, their findings cannot be used to dismiss ‘original sin’ concerns, as a number of firms in the nontradable sector (without a natural hedge) likely faced limited financing opportunities vis-a`-vis the firms that could more safely borrow in foreign currency.

DEVELOPMENT OF LOCAL BOND MARKETS Eichengreen et al. (2005) show that macroeconomic and institutional variables cannot explain why emerging markets are not able to issue abroad in their own currencies. This result is at least partly driven by the limited variation in international issues in domestic currency. While foreign bond issues in emerging market currencies remain relatively rare, the last decade has seen a rapid development of local bond markets in several key emerging markets.1 Studies that have considered domestic bond markets typically do find that macroeconomic and institutional variables can explain the depth and development of local bond markets, for example, Burger and Warnock (2006) and Claessens et al. (2007). While in some countries, domestic local currency debt tends to be issued either short-term or in indexed forms (what Eichengreen et al. refer to as ‘domestic original sin’), many countries have been able to gradually lengthen the maturity of their local currency nominal bonds. For a database with detailed breakdown of domestic and foreign public debt by maturity and denomination, please refer to Guscina and Jeanne (2006). While indexed debt is not as desirable from a risk-sharing perspective as nominal long-term local currency debt, inflation-indexed debt is a very close substitute. That is, inflation-indexed debt is much closer to nominal local currency debt from an international risk-sharing perspective than to foreign currency debt (as large swings in the nominal exchange rate often translate into larger swings in the real exchange rate due to incomplete pass-through). On the other hand, debt indexed to the interest rate can resemble foreign currency debt, if the interest rate tends to rise when the currency is under pressure. Borensztein et al. (2005) present a simple back-of-the-envelope exercise where they show that inflation-indexed debt may even be a better hedge for

1

While there have been some sizable issues (e.g., Brazil has about USD 5 billion of long-term Real-denominated bonds issued in global markets), they remain relatively small compared to the size of local bond markets.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

160

16. FOREIGN CURRENCY DEBT

emerging markets than nominal debt (through the channels described in Jeanne (2005)). Of course, in order for inflation-indexed debt to be attractive to investors, a credible statistical framework must be in place for measuring inflation. Many emerging markets have amassed large amounts of foreign exchange reserves. This tends to reduce their aggregate currency mismatches, even if the government or private firms continue to borrow externally in foreign currency. Goldstein and Turner (2004) argue that a measure of aggregate currency mismatches is more informative about fragilities than ‘original sin.’ But ideally, emerging markets would be net borrowers in their own currency (which would require foreigners to be more willing to hold emerging market currency risk). Moreover, how the currency risk is distributed within an economy (for a given aggregate level of mismatch) can also be an important determinant of vulnerability. An individual firm should be indifferent between borrowing in domestic currency from a resident and from a foreigner. But for the domestic economy as a whole, it would be preferable for that debt to be held by a foreigner, as the smaller the quantity of tradable goods that needs to be transferred to foreigners during bad times, the smaller is the decline in the relative price of nontradables during those bad times. The same is true for the transfer during good times, so an aggregate net exposure to foreigners in domestic currency contributes to relative price and exchange rate stability.2 Foreign purchases of local currency debt through participation in local bond markets have the same risksharing benefits as purchases of local currency debt offered through international bond markets. And there may be a case to be made for focusing on the development of a deep and liquid domestic local currency market (e.g., it may be better for foreigners to come to Sa˜o Paulo to buy local currency bonds in a deep and liquid market than for Brazilian firms to try to issue in Reais in a shallow global market). It is difficult to measure the full extent of foreign participation in local bond markets. The Treasury Department publishes data on US portfolio holdings of foreign securities (available at http://www.ustreas.gov/tic/fpis.shtml). The holdings are quite significant for some countries (e.g., roughly 20 billion dollars at the end of 2008 for Brazil, 8.5 billion of which were in long-term domestic currency instruments). Ideally, more data should be available (e.g., covering holdings by other countries) to determine the full extent of foreign participation in local markets.

2

CONCLUSION While a number of channels have been proposed to explain ‘excessive’ foreign currency debt, as surveyed in this chapter, it is difficult to quantify their relative importance, and there is no consensus in the literature on which is perceived to be the root cause of the problem. Perhaps the implications of currency mismatches were not as well understood in the past as they are now, and their risk may have been substantially underestimated. Most countries that suffered from crises in the late 1990s/early 2000s took advantage of the favorable global conditions during 2002–07 to reduce their exposure to risky debt structures, gradually shifting toward local currency debt, and lengthening the maturity of their bonds. This has certainly contributed to the striking resilience emerging markets have shown during the current global financial crisis.3 While the extent of foreign lending in domestic currency remains far lower than its theoretically ideal level, one could see the glass as half full, and expect countries that pursue sound macroeconomic policies to continue to gradually move toward safer liability structures.

References Aghion, P., Bacchetta, P., Banerjee, A., 2004. A corporate balance-sheet approach to currency crises. Journal of Economic Theory 119 (1), 6–30. Bleakley, H., Cowan, K., 2008. Corporate dollar debt and depreciations: much ado about nothing? The Review of Economics and Statistics 90 (4), 612–626. Borensztein, E., Chamon, M., Jeanne, O., Mauro, P., Zettelmeyer, J., 2005. Sovereign Debt Structure for Crisis Prevention. International Monetary Fund, Washington, DC. IMF Occasional Papers: 237. Broda, C., Levy-Yeyati, E., 2006. Endogenous deposit dollarization. Journal of Money, Credit, and Banking 38 (4), 963–988. Burger, J., Warnock, F., 2006. Local currency bond markets. IMF Staff Papers 53, 115–132 (special issue). Burnside, C., Eichenbaum, M., Rebelo, S., 2001. Prospective deficits and the Asian currency crisis. Journal of Political Economy 109 (6), 1155–1197. Caballero, R., Krishnamurthy, A., 2003. Excessive dollar debt: financial development and underinsurance. Journal of Finance 58 (2), 867–893. Chamon, M., 2003. Why can’t developing countries borrow from abroad in their currency? Mimeo. Available through SSRN. Chamon, M., Hausmann, R., 2005. Why do countries borrow the way they borrow? In: Hausmann, R., Eichengreen, B. (Eds.), Other Peoples Money: Debt Denomination and Financial Instability in Emerging Market Economies. University of Chicago Press, Chicago and London, pp. 218–232. Claessens, S., Klingebiel, D., Schmukler, S., 2007. Government bonds in domestic and foreign currency: the role of institutional and

Of course, assuming that this liability structure does not distort monetary policy decisions.

3

While a number of emerging markets have been hit pretty hard by the current crises, it is truly impressive that the damage has been relatively limited given the severity of the crisis in advanced economies. A couple of years ago, no one would have believed such an outcome to be possible.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

CONCLUSION

macroeconomic factors. Review of International Economics 15 (2), 370–413. de la Torre, A., Schmukler, S., 2004. Coping with risks through mismatches: domestic and international financial contracts for emerging economies. International Finance 7 (3), 349–390. Eichengreen, B., Hausmann, R., 1999. Exchange rates and financial fragility. NBER Working Paper No. 7418. Eichengreen, B., Panizza, U., Hausmann, R., 2005. The mystery of original sin. In: Hausmann, R., Eichengreen, B. (Eds.), Other People’s Money: Debt Denomination and Financial Instability in Emerging Market Economies. University of Chicago Press, Chicago and London, pp. 233–265. Goldstein, M., Turner, P., 2004. Controlling Currency Mismatches in Emerging Markets. Institute for International Economics, Washington, DC, pp. xvi, 164. Guscina, A., Jeanne, O., 2006. Government Debt in Emerging Market Countries: A New Data Set. International Monetary Fund. IMF Working Papers: 06/98, p. 31.

161

Jeanne, O., 2005. Why do emerging economies borrow in foreign currency? In: Hausmann, R., Eichengreen, B. (Eds.), Other People’s Money: Debt Denomination and Financial Instability in Emerging Market Economies. University of Chicago Press, Chicago and London, pp. 218–232. Jeanne, O., 2009. Debt maturity and the international financial architecture. American Economic Review 99 (5), 2135–2148. Krugman, P., 1999. Balance sheets, the transfer problem, and financial crises. In: International Finance and Financial Crises: Essays in Honor of Robert P. Flood, Jr.. International Monetary Fund, Washington, DC, pp. 31–44. Schneider, M., Tornell, A., 2004. Balance sheet effects, bailout guarantees and financial crises. Review of Economic Studies 71 (3), 883–913. Tirole, J., 2003. Inefficient foreign borrowing: a dual- and commonagency perspective. American Economic Review 93 (5), 1678–1702.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

Intentionally left as blank

C H A P T E R

17 International Trade and International Capital Flows K. Jin London School of Economics, London, UK O U T L I N E Introduction Specialization and Capital Flows The ‘Composition Effect’ Implications Trade and Capital Flows with Frictions

163 164 165 166 166

INTRODUCTION Trade and capital flows are both integral processes of globalization, yet until recently, there has been little study on how they interact. The conventional separation of international trade and international macroeconomics excludes the impact macroeconomic dynamics exert on the structure of trade and the aggregate feedback effect of commodity trade. In the standard workhorse, openeconomy, macroeconomic framework, either only intertemporal trade is present or an exogenously rigged structure of trade is assumed.1 Recent works have suggested that the separation is not always innocuous. Trade and capital flows jointly determine the global allocation of capital, and give markedly different predictions on the way shocks impinge on the world. The interest in the relationship between commodity trade and capital flows harks back to the fundamental insights of the Heckscher–Ohlin framework. Under certain conditions, in a two-country, two-factor model, trade and capital flows are perfect substitutes: Commodity trade is sufficient to ensure factor price equalization, and factor price equalization is sufficient to ensure commodity price equalization.2 In other words, the ability to engage in

Conclusion Glossary References

commodity trade can eliminate the need for capital to flow from the capital-abundant countries to the capital-scarce countries, as the rate of return differences can be eliminated through trade alone. The implication is that trade liberalization reduces the need for capital to flow toward developing countries characterized by capital scarcity. Mundell (1957) puts this substitutability of trade and capital flows in a different way: An increase in trade impediments stimulates factor movements, and an increase in impediments to factors stimulates trade. An example helps illustrate these points. Suppose that Home is capital abundant and Foreign is labor abundant. If factors (labor and capital) are internationally immobile but trade impediments are absent, Home exports the capital-intensive good and Foreign the labor-intensive good. With factor price equalization, no capital flows will take place once barriers to capital mobility are removed. But now suppose that Foreign imposes a tariff on the capital-intensive good, causing its relative price to rise. Factors will move out of the labor-intensive sector and into the capital-intensive sector. At constant factor prices, the shift in production structure creates an excess supply of labor and an excess demand for capital. Consequently, the marginal product of capital (MPK) rises

1

One common example is that each country specializes in differentiated goods.

2

This chapter focuses on the case where countries incompletely specialize in goods.

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00023-4

167 168 168

163

# 2013 Elsevier Inc. All rights reserved.

164

17. INTERNATIONAL TRADE AND INTERNATIONAL CAPITAL FLOWS

in Foreign compared to Home, and capital flows from Home to Foreign. This simple example points to the possible interaction between trade and capital flows: changes in the structure of trade, in this case led by an exogenous increase in tariffs, create incentives for capital to move across borders. While insightful, the Heckscher–Ohlin–Mundell framework (henceforth HOM) remains limited in its capacity to provide a comprehensive analysis of trade and capital flows and their interaction under a rich set of scenarios. One important drawback is that capital mobility in the static two-country, two-factor, two-sector (2  2  2) framework is confined to the allocation of capital across countries, for a fixed level of world capital stock. A dynamic model in which capital flows are driven by the allocation of savings across countries becomes a relevant and important extension. In other words, to scrutinize the relationship between trade and capital flows requires the marriage of an endogenous structure of trade and a rigorous macroeconomic framework in which intertemporal decisions are made. It is also imperative to allow for simultaneous trade and capital mobility. The Heckscher–Ohlin (HO) model allows for only trade mobility, and arguments made for capital flows amount to simple comparisons of the return to capital across financially insulated economies.3 This chapter examines new theoretical predictions arising from the integration of a factor proportions-based structure of trade and macroeconomic dynamics. It focuses on isolating the impact of trade and specialization patterns as drivers of capital flows in a realistic, macroeconomic setting. Under certain circumstances, capital can flow out of countries experiencing a permanent increase in labor force/productivity, and it can also flow ‘upstream’ from capital-scarce to capital-abundant countries when economies integrate. These set of results are analyzed in Jin (2009). The chapter then discusses the relationship between trade and capital flows when they can interact with financial market friction. When comparative advantages across countries are determined not only by factor endowments but also by financial heterogeneity, the HOM results of substitutability can turn into one of complementarity, as forcefully shown in Antras and Caballero (2009). Ju and Wei (2007) put labor market frictions at center stage and examine the impact of labor market rigidity on current account adjustments. Their results generalize the

‘intertemporal approach to current account’ by allowing for an additional, intratemporal channel of adjustment4.

Specialization and Capital Flows In the HOM setting, trade integration reduces the need for capital to flow from capital-abundant to capital-scarce countries. Likewise, there is less incentive for capital to flow toward countries experiencing a labor force/productivity boom if bestowed with the ability to trade. How, under certain conditions, capital flows can be entirely reversed–away from capital-scarce countries and out of countries with the labor force/productivity boom – precisely economies with a relative low capital– labor ratio. This comes out of a realistic departure from the standard Heckscher–Ohlin framework. First, it allows for simultaneous mobility in trade flows and capital flows. The assumption of the existence of capital adjustment costs breaks factor price equalization temporarily and pins down the country-level capital stock. Second, a typical macroeconomic setup requires capital to adjust after one period; capital in each sector is augmented by investment, rather than through the reallocation of capital from other sectors. These features make capital effectively ‘sector specific.’5 Third, the world capital stock is no longer fixed but can grow, and capital flows are determined based on the allocation of savings. The setup can be seen as a minimal departure from a standard, two-country growth model, augmented by multiple-tradable goods that feature factor intensity differences. Consider now, two initially symmetric, open economies, Home and Foreign. Suppose that Foreign experiences an unexpected, permanent increase in the labor force. Now labor abundant, Foreign specializes in labor-intensive goods and becomes a net importer of capital-intensive goods, and Home, a net importer of labor-intensive goods. The change in the structure of trade leads to attendant changes in the relative demand for capital in each country: Home sees a rise in the share of capital-intensive goods in total domestic production and hence a rise in its investment share of output. Note that in a setting where sectors are distinguished by their factor intensity, investment demand depends not only on the scale of the economy,6 but also on the composition of production. Industrial structures tilted

Under factor price equalization in a (2  2  2) world, capital is indeterminate at the country level if there are no barriers to capital flows. Capital earns the same returns everywhere and can be located anywhere. One way of breaking factor price equalization and the resultant indeterminacy result is to assume some form of market friction. For instance, Mundell (1957) assumes impediments to trade. Jin (2009) assumes the existence of capital adjustment costs. 3

4

See Obstfeld and Rogoff (1996): Chapter 1 and 2.

5

See the specific factors models of Neary (1978), Jones (1971), and Mussa (1978).

6

By this it is meant that countries that expand, either through size or productivity, will see greater investment demand.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

165

INTRODUCTION

toward capital-intensive sectors tend to see a higher investment demand, all else equal. Ultimately, whether this trade-induced impact on Home’s investment will result in a net capital inflow depends on the relative strength of two effects that underly a unified framework of trade and capital flows. The first effect is driven by commodity trade, which, to recapitulate, induces capital to flow toward the country becoming more specialized in capital-intensive goods, a ‘composition effect.’ The second effect is the standard impetus to capital flows in the basic macroeconomic framework, whereby capital tends to flow toward where it is relatively scarce, a ‘convergence effect.’ The dominant one of the two effects determines the direction of the flow. The neoclassical one good case thus becomes a special case in this integrated framework. It is the case where sectors do not feature factor intensity differences and only the ‘convergence effect’ is present. The next section discusses the special case that isolates the ‘composition effect’ and demonstrates its disparate impact on capital flows.

The ‘Composition Effect’ Consider the same multiple sector setup, except that the most labor-intensive sector uses only labor as an input to the production technology. Assume, for simplicity, that there are no productivity differences across countries, and that labor is perfectly mobile across sectors. The wage for any country j, wj, then, is pinned down by the price of the most labor-intensive good, p1. Since p1 is equalized across countries under free trade, wages are equalized across countries at any point in time. Labor reallocation across sectors alone, within a country, is

sufficient to equalize capital–labor ratios in each sector, across countries. Normally, both labor reallocation and the adjustment of capital stocks are necessary to bring about equalization of capital–labor ratios, and hence, factor prices. But in this special case, the need for ‘convergence’ is effectively eliminated. How does Foreign with the labor force/productivity boom allocate its marginal unit of savings? Since returns to capital are equalized across countries, it is allocated to both countries. Exactly how much is apportioned to each country is pinned down by adjustment costs: If countries were originally symmetric, one half of saving would be allocated domestically and one half, abroad; if a country started out with a higher capital stock, it would receive proportionately more investment for the reason that lower adjustment costs are incurred there per unit of investment. The important result is that investment always comoves across countries. An alternative way of understanding this result, not based on simple comparisons of the rate of return to capital, is examining a country’s investment demand relative to its savings capacity. The following graphical illustration can help build intuition for the impact effect of shocks on capital flows. Figure 17.1 graphs the investment–gross domestic product (GDP) ratio and the savings–GDP ratio against the relative capital–labor ratio of country j – defined as the ratio of j’s capital–labor ratio to that of the world. When j’s relative capital–labor ratio is 1, no comparative advantage differences exist across countries. In the first panel, featuring the multisector special case, country j’s investment share of GDP rises with its relative aggregate capital–labor ratio. The reason is that greater comparative advantage in capital, manifested by a high relative capital–labor ratio, bids Home to specialize more in capital-intensive sectors. As a result of the Multiple sector case (Composition effect only)

One sector model

I/GDP S/GDP

1 Relative capital labor ratio

S/Y I/Y

1 Relative capital labor ratio FIGURE 17.1 Savings to GDP ratio and investment to GDP ratio as a function of the country’s capital–effective–labor ratio. The left panel is the two–country OLG model with one sector, and the right panel is a special and in the two-good case. Source: Jin (2012).

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

166

17. INTERNATIONAL TRADE AND INTERNATIONAL CAPITAL FLOWS

changing industrial structure, the relative share of output attributed to investment increases, while the share of GDP attributed to labor income is reduced. As in Jin (2009), that the saving–GDP ratio is a downward sloping function comes out of a Diamond-type overlapping generations model, where saving derives solely from wage income of the young cohort. The important point is that investment demand can rise above its capacity to save as a country becomes more capital intensive in production, inducing a net capital inflow from abroad. At the point where countries’ capital–labor ratios are equalized, domestic savings is just enough to serve their domestic investment needs, and no net capital flows needs to occur between countries. The one-sector case shows the opposite result. The investment–GDP curve is downward sloping, as drawn in the second panel, since lower capital-effective-labor ratio in any country requires greater investment in order for capital to ‘scale up’ with labor.7 The impact effect of a fall in capital-effective labor ratio in Foreign is thus a net capital inflow, in contrast to the net outflow in the multisector case. The striking difference between the two special cases is the slope of the investment demand curve, which is negative in the one-sector case but positive in the multisector case. These cases, of course, display the results of the composition and convergence effect in isolation, but when they coexist in the general setting, the investment– GDP curve lies somewhere in between – positively sloped when the composition effect is stronger and negatively sloped when the convergence effect is stronger. The main action stems from the behavior of investment rather than savings, which may exhibit different patterns for different assumptions of the structure of the economy and parameters.8

Implications An important implication is that one cannot resort to comparing the autarky steady-state MPK to infer the direction of capital flows once the economies undergo financial and trade liberalization. The reason is that opening up to trade can have an impact effect on the MPK in both countries as each undergoes industrial restructuring. This relates to the point originally made by Lucas (1990) that large differences in capital–labor ratios may not imply vast differences in the MPK, as poor countries also have lower endowments of factors complementary with

physical capital, such as human capital and total factor productivity. As such, it is plausible that very little capital flows from rich to poor countries. Yet, one cannot draw these conclusions when commodity trade is allowed. In integrating with the rest of the world, both in terms of trade and financial flows, capital-abundant, advanced economies become net exporters of capitalintensive goods and see a rise in the MPK. Under certain conditions, part of the saving derived from trade liberalization in developing economies can be allocated also to rich countries, reversing the flow of capital from capital-poor to capital-abundant economies. Another implication is that the sequencing of liberalization can also have differential impact on developing countries in such a process. While simultaneous liberalization may lead to a capital outflow in South, trade liberalization preceding financial liberalization may prevent such an outflow.

Trade and Capital Flows with Frictions Antras and Caballero (2009) explore the relationship between trade and capital flows in the presence of financial heterogeneity. Motivated by empirical evidence, the paper assumes that countries differ in financial development and sectors differ in financial dependence. A borrowing constraint in one sector limits the amount of capital allocated to that sector. The tightness of the borrowing constraint can vary across countries. In the benchmark case, which consists of a two-factor, two-sector, general equilibrium model, sectors are initially symmetric (no factor intensity differences) except for the presence of a borrowing constraint in the financially constrained sector, and countries are also initially symmetric except for the tightness of the borrowing constraint. One can think of a developing ‘South’ as having a lower level of financial development, and advanced ‘North’ as having a higher level of financial development. It is useful to first consider the autarkic equilibrium, in which goods and factor markets have to clear domestically. In South, with the worse financial institutions, disproportionately more capital is apportioned to the unconstrained sector, whose output becomes oversupplied and its relative price depressed.9 The high capital–labor ratio in the unconstrained sector leads to

7 On the other hand, the savings to GDP ratio is a constant. This comes from the assumption of log utility in the two-period OLG model. See Jin (2009) for a discussion. 8

It is for this reason that the assumption of an overlapping generations model is not crucial, although most likely to be quantitatively important. The amount of capital flows that take place depends partly on the amount of saving generated by the country with the boom. In a representative agent model, the saving generated would depend on the persistence of the shocks. Highly persistent labor/productivity shocks can lead to an initial fall in saving.

9

The financial friction, however, does not distort the allocation of labor across sectors.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

167

CONCLUSION

relatively depressed wages and rental rates of capital in South. If international capital flows are now allowed, capital will move away from South toward North with the better financial development. The implications on capital flows change when international trade in goods is allowed. When South opens itself up to trade, it will see an increase in the price of the unconstrained sector’s output, causing it to specialize (incompletely) in the unconstrained sector, in which it has a comparative advantage. It becomes a net importer of the financially dependent sector’s output. Trade integration thus allows South to allocate a disproportionate number of workers in the unconstrained sector, thereby increasing the MPK and its equilibrium rental rate. By allowing South to specialize in a sector with lower financial frictions, international trade reduces the negative impact of financial underdevelopment on the rental rate of capital. In fact, the rate of return to capital becomes higher in South – trade integration not only reduces the difference in the real return to capital in North and South, but actually overturns it. The implication is thus that, in the presence of financial differences, trade integration increases capital flows from the North to South. These results carry over to the extension with HO determinants of trade and when countries feature differences in aggregate capital–labor ratio.10 Thus, the complementarity between trade and capital flows is drawn into stark contrast with the substitutability nature featured in Mundell (1957), whereby a process of trade integration necessarily lowers the rental rate of capital in capital-scarce countries, as these countries become more specialized in labor-intensive goods and sees a fall in the demand for capital. Unlike Antras and Caballero (2009), Ju and Wei (2007) put labor market rigidity at center stage in a dynamic Heckscher–Ohlin framework. The setting is the same two-country, two-sector, general equilibrium model in which added to labor adjustment costs are trade costs and costs to capital flows. Unlike Jin (2009), capital can be costlessly and instantaneously adjusted across sectors, within a country. Labor, rather than capital, becomes the ‘specific factor.’ Ju and Wei (2007) show that the degree of labor market frictions in a country affects the size of the current account response to shocks and the speed of adjustment to its long-run level. When there is some degree of mobility in both trade and capital, an economy’s adjustment to shocks involves a combination of a change in the composition of goods trade (intratemporal trade) and in the current account (intertemporal trade). In the extreme

case that labor is immobile across sectors, all adjustment to shocks takes place through intertemporal trade.11 Thus, a relatively more rigid labor regulation induces a larger response of the current account, and slows down the speed of adjustment of the current account toward the long-run equilibrium. They provide three pieces of supporting empirical evidence: a rigid labor market is associated with a lower churning of its trade structure; a higher rigidity of the labor market reduces the speed of convergence of the current account to its long-run equilibrium; and a country with a rigid labor market is likely to exhibit a higher variance of current account to total trade.

CONCLUSION This chapter has highlighted some recent progress in understanding the relationship between trade and capital flows under a set of more realistic assumptions and a rich set of environments. Incorporating an endogenous structure of commodity trade into rigorous macroeconomic settings can lead to markedly different theoretical implications. With deepening trade and financial ties, the necessity for integrating these two dimensions of globalization becomes apparent, and moreover is reinforced by the fact that it provides new, and often surprising, insights. Not only do these predictions extend to the widely debated issues of global imbalances, there are also profound implications for asset prices, risk sharing, international business cycles, and policy and welfare implications for each type, and the timing, of liberalization. The joint analysis of trade and macroeconomic dynamics is particularly pertinent when shocks or structural changes under investigation fundamentally change a country’s comparative advantage, and therefore its structure of trade, which can affect macroeconomic aggregates, as reviewed in this chapter. Demographics, globalization, productivity changes, and financial market developments are best examples of such scenarios, and in the world one observes today, incontrovertibly relevant. These issues, until now, remain principally interesting theoretical inquisitions. The empirical relevance of the interaction between trade and capital flows, and between trade and market frictions, has yet to be tested. An obviously important question is to what extent and through what channel can trade affect capital flows? Can differing experiences and the timing of trade and financial liberalization across countries, and various other

10

More precisely, the result that trade integration raises the rental rate of capital in South even when there are factor intensity differences across sectors, holds as long as South continues to command a comparative advantage in the unconstrained sector. 11

This result resonates with the HOM prediction in the context of prohibitive trade.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

168

17. INTERNATIONAL TRADE AND INTERNATIONAL CAPITAL FLOWS

shocks, be exploited to test these theories? The literature that bridges trade and macroeconomics promises to be a fertile ground for further research, but in the absence of parallel advances in empirics, its scope remains limited and its impact conceivably bounded by the perimeters of a purely theoretical construct.

exporting goods in the future (running a future current account surplus). Lucas puzzle (paradox) The observation that capital does not flow from capital-abundant, developed countries to capital-scarce, developing countries.

Glossary

Antras, P., Caballero, R.J., 2009. Trade and capital flows: a financial frictions perspective. Journal of Political Economy 117 (4), 701–744. Jin, K., 2009. Industrial Structure and Capital Flows. Harvard University, Mimeo. Jin, K., 2012. Industrial structure and capital flows. American Economic Review 102 (5), 2111–2146. Jones, R., 1971. A three-factor model in theory, trade, and history. In: Bhagwati, J.N. et al., (Ed.), Trade, Balance of Payments, and Growths: Essays in Honor of Charles P. Kindleberger. NorthHolland, Amsterdam. Ju, J., Wei, S., 2007. Current Account Adjustment: Some New Theory and Evidence. NBER, Working Paper No. 13388. Lucas, R., 1990. Why doesn’t capital flow from rich to poor countries? American Economic Review 80 (2), 92–96. Mundell, R., 1957. International trade and factor mobility. American Economic Review 47, 321–335. Mussa, M., 1978. Dynamic adjustment in the Heckscher–Ohlin– Samuelson model. Journal of Political Economy 86 (5), 775–791. Neary, P.J., 1978. Short-run capital specificity and the pure theory of international trade. The Economic Journal 88, 488–510. Obstfeld, M., Rogoff, K., 1996. Foundation of International Macroeconomics, 1st edn., Volume 1. MIT press.

Comparative advantage Lower relative cost of production of a good compared to another economy’s relative cost of production of the same good. The law of comparative advantage says that two economies can both gain from trade even if a country is more efficient in the production all goods, as long as the more- and less-efficient economies have different relative efficiencies of production. Current account One of the two primary components of the balance of payments. It is the difference between domestic saving and domestic investment in an economy. A country runs a current account surplus if it is a net lender and a current account deficit if it is a net borrower. Factor endowments The amount of resources, such as land, capital, labor, that a country possesses and can exploit for production. Factor price equalization The equalization of the relative prices for factors of production, such as rate of return to capital and wages. Factor proportions trade (Heckscher–Ohlin theory) The theory that countries should produce and export goods that use intensively the country’s abundant factor, such as capital, labor, or land. Intertemporal trade Trade across time, whereby an economy imports goods today (by running a current account deficit) and in return,

References

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

C H A P T E R

18 International Macro-Finance A. Pavlova*, R. Rigobon† †

*London Business School and CEPR, London, UK Sloan School of Management, MIT and NBER, Cambridge, MA, USA O U T L I N E

Introduction

169

The Workhorse Model Log-Linear Preferences

171 172

INTRODUCTION Financial markets and their role in international risk sharing have inspired a vast body of theoretical literature. Over the past 40 years, international finance and economics has evolved into a vibrant field spreading from the basic international version of the Capital Asset Pricing Model (CAPM) to some of the most sophisticated dynamic stochastic general equilibrium (DSGE) models. Interestingly, however, the research effort in economics has evolved almost in parallel with that in finance. In economics, the main focus has been on real quantities and international relative prices such as consumption, investment, current account, terms of trade, and exchange rates. International portfolio choice and international equity markets have been largely overlooked. Indeed, the asset structure of these models has been mostly of two types: either the only asset is an international bond and markets are incomplete or there is a full set of Arrow–Debreu securities and markets are complete. Both approaches have been very useful, but they cannot address many questions pertaining to portfolio problems and to the international equity markets. Finance, on the other hand, has focused more on crosscountry portfolio allocations and asset prices. Terms of trade and hence exchange rates have been largely overlooked because the majority of the models featured a single-good framework, in which forces of arbitrage equate terms of trade to unity. Models with endogenous portfolio selection, equity prices, and time-varying terms

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00022-2

Log-Linear Preferences with Demand Shocks Next Steps References

174 175 176

of trade and exchange rates in a single framework have been quite rare. Although a tremendous amount has been learned from this research, in recent years two main phenomena have required a redefinition of the agenda behind the theories of financial markets in the international context: contagion among developed and relatively unconnected countries, and the role that asset prices and exchange rates play in the global imbalances. Contagion refers to the transmission of crises from one country to another. Prominent examples of this phenomenon include the 1997 Asian crisis, the 1998 Russian crisis, and the subprime mortgage crisis of 2007–2008. Contagion has attracted attention of academics in the aftermath of the 1994 Mexican crisis, and from the very beginning crises were thought to be transmitted through trade relationships driven by competitive devaluations. Also, contagion appeared to be mostly an emerging markets problem. In recent years, however, these views have shifted. Today, one tends to think that turmoil is transmitted, most likely, through the financial sector and that contagion is a global problem, affecting all countries, regardless of their level of financial development. The emphasis on the financial sector as the main transmission mechanism started shortly after the 1997 Asian and the 1998 Russian crises. During these crises shocks propagated through international banks and other financial intermediaries. The transmission had little to do with competitive devaluations and their impact on trade, but rather with the impact of large swings in exchange rates on asset values, balance sheets, and

169

# 2013 Elsevier Inc. All rights reserved.

170

18. INTERNATIONAL MACRO-FINANCE

interest rates. For example, an analysis of the 1997 Asian crisis shows that Thailand’s 1997 currency downfall led to capital losses for Japanese banks, forcing them to curb their lending to other Asian countries. It is thus becoming increasingly apparent to both academics and policymakers that no analysis of contagion is complete without a thorough understanding of how shocks are transmitted internationally through financial markets and intermediaries. The second significant development that has influenced the literature was the unprecedented rise in external deficits in many developed nations, which sparked a discussion about sustainability and the possible dramatic unraveling of global imbalances (with Nouriel Roubini making particularly striking predictions). The rise in external deficits, however, came hand in hand with the explosion in cross-border risky asset holdings. Before 1985, the United States held virtually no foreign equities; nowadays, foreign equities account for a large and growing part of the country’s assets. Following influential work of Lane and Milesi-Ferretti (2001) and Gourinchas and Rey (2007), it has become clear that (unrealized) capital gains on these equity positions are missing from national accounts. The alarming current accounts deficits worldwide may then be simply due to this misreported income from equity positions. Today an extremely active literature, both empirical and theoretical, is trying to better understand how the capital gains on foreign equity positions, or the so-called valuation effects, affect thinking about the current account and the external adjustment process. Unfortunately, most of the existing international macro models are not well suited for dealing with these issues because they are missing equity markets and portfolio choice. A new and rapidly growing strand of literature, commonly known as International Macro-Finance, is trying to fill this gap. This new generation of macro models provides a redefinition of the current account (adjusted for capital gains on equity holdings) and modifies the standard theories of the current account. More generally, this research program focuses on the interaction between the financial sector and the real economy, and as such can address a wide spectrum of issues such as contagion, valuation effects, and others. The literature in international economics and finance can be split along several paradigms: small open economy (partial equilibrium) versus general equilibrium;

pure exchange versus production economies; singleversus multiple-good models; complete versus incomplete markets. Because the format of this chapter imposes a limit on the number of cited references, it is impossible to summarize the research that has been done in all these areas.1 This survey concentrates on a framework that has become the core of international macrofinance: general equilibrium asset-pricing models with multiple goods. The richness of this framework comes at a cost: most of the macro-finance models are quite complex. This literature can be split into several parts based on their approaches. The first approach relies on traditional approximation methods. For example, Kollmann (2006) computes portfolios and changes in net foreign assets using standard first-order approximations around a deterministic steady state. The second approach makes use of higher-order approximations to analyze countries’ portfolios and the evolution of external accounts. These methodologies have been developed by Engel and Matsumoto (2006), Devereux and Sutherland (2011), and Tille and van Wincoop (2010). A disadvantage shared by these two approaches is that to this day little is known about the behavior of these economies away from the deterministic steady state, where the underlying volatilities are not small.2 The third strand of the macro-finance literature simplifies the models and seeks to find exact solutions. The main advantage of this approach is that the economy can be analyzed away from the steady state, but the disadvantage is that solutions only exist in few special cases. An early work that presents one such special case is that of Helpman and Razin (1978). Their setup has been developed further by a number of authors, including Cole and Obstfeld (1991) and Pavlova and Rigobon (2007). These papers consider pure exchange economies in which a representative agent in each country has loglinear preferences. In recent years, the literature has made progress extending the setup beyond log-linear preferences.3 The solution is especially simple in complete markets – and this solution is discussed in detail in section ‘The Workhorse Model’ – but it is also possible to introduce market frictions (see Pavlova and Rigobon, 2011, for references). It is believed that models incorporating market incompleteness and institutional frictions are the most promising step toward understanding the phenomena of contagion, world systemic risk, and the proper definition of external sustainability.

1

The two strands that are the closest to the subject here are international real business cycles in Economics and international asset pricing in Finance.

2

More generally, the accuracy and performance of numerical algorithms for solving multicountry DSGE models remain the subject of an ongoing debate. The January 2010 special issue of the Journal of Economic Dynamics and Control on numerical methods for multiagent DSGE models provides an excellent reference for the current state of this debate. 3

See a longer version of this survey, Pavlova and Rigobon (2011), for references.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

171

THE WORKHORSE MODEL

In terms of the economic applications, the two questions that the theoretical literature on macro-finance has mainly focused on so far are (i) the composition of international portfolios and (ii) the valuation effects and global imbalances. These strands of literature are discussed in more detail by Pavlova and Rigobon (2011). In January 2010, the Journal of International Economics ran a special issue on international macro-finance. This collection of works gives an excellent overview of the latest contributions to this field and provides further references.

THE WORKHORSE MODEL This section develops a simple multigood multicountry general equilibrium asset-pricing model that allows for closed-form solutions. While the setup here is certainly simplified, it possesses many elements that form the core of international macro-finance. It is believed that it is this kind of models that should lead future research in this area. A discrete-time pure exchange economy is considered. Let time t run from 0 to T. There is a finite number of trading dates, but one can also look at the infinitehorizon version of the economy by adding an appropriate transversality condition and taking the limit as T ! 1. There are two countries in the world: Home and Foreign. Each country is endowed with a Lucas tree producing a country-specific perishable good. In the baseline model, the uncertainty about the output of the trees is the only source of risk in the economy. The state ot, an element of the set Ot, is the history of the economy up to time t. This history occurs with probability p(t, ot). In the notation hereafter, the second argument, ot, is suppressed unless necessary for clarity. The statedependent outputs of the Home and Foreign trees are denoted by Y(t) and Y*(t), respectively, and the corresponding prices of the goods by p(t) and p*(t). The terms of trade, ToT, are defined as the price of the Home good relative to that of the Foreign good: ToT  p/p*.4 The world numeraire basket is fixed to contain a 2 (0, 1) units of the Home good and (1a) units of the Foreign good, and the price of this basket is normalized to be equal to unity. Available for investment are one-period riskless bonds, with prices B(t) and B*(t), paying out in units of Home and Foreign goods, respectively, and claims to the Home and Foreign trees (stocks), with prices S(t) and S*(t). Additionally, agents can trade a complete set of Arrow–Debreu securities. In what follows, the

particular interest is in determining equilibrium prices of the stocks and bonds. Due to space limitations, the countries’ portfolios are not examined, and so more assets than what is needed are allowed for to ensure market completeness. The surplus assets will simply be priced by no-arbitrage. In the presence of redundant assets, of course, the countries’ portfolios are indeterminate. To make the model a meaningful model of portfolio allocation, the assumption that the Arrow–Debreu securities are available for trade is simply dropped and instead the markets are (dynamically) completed with the countries’ stocks, bonds, and possibly other assets. One, of course, has to ensure that there are at least as many assets as there are states of the world at each node; otherwise, the solution approach needs to be modified to explicitly account for market incompleteness. Each country i is populated by a representative agent with preferences characterized by the expected utility function " # T X t b ui ðCi ðtÞ; Ci ðtÞÞ ; i ¼ fH; Fg ð18:1Þ E t¼0

with 0 < b < 1. In particular, the agents in each country derive utility from consuming both the domestic and the foreign good. Since it is assumed that markets are complete, the budget constraints of the agents can be written in their Arrow–Debreu form: " # T X   qðtÞðpðtÞCi ðtÞ þ p ðtÞCi ðtÞÞ ¼ Wi ; i ¼ fH; Fg E t¼0

ð18:2Þ where q(t, ot) is the Arrow–Debreu state price of ot divided by the probability of that state p(t, ot). The quantity Wi denotes the initial wealth of country i; here, WH ¼ S(0) and WF ¼ S*(0). This form of a budget constraint is widely used in the asset-pricing literature; it is often called a ‘static’ budget constraint, implying that all asset trades take part at time 0 and no trades take place afterward. The ability to do so, of course, requires market completeness. If markets are incomplete or there are other frictions, one can still write the budget constraint in its static form, but the state prices q(t) entering the budget constraint will be country-specific. A competitive equilibrium in this economy consists of state-contingent allocations (Ci, Ci*), i 2 {H, F} and prices (ToT, q) such that (i) both countries maximize their utility subject to the budget constraint and (ii) goods and asset markets clear. It is convenient to solve for this

4

The presence of time-varying terms of trade is what makes this framework different from standard international asset-pricing models in Finance. Most of these models feature a single good, and therefore, by construction, the terms of trade and the real exchange rate are equal to unity. Nontrivial implications for terms of trade or exchange rates in single-good models have been obtained by either introducing shipping costs into a real model or by exogenously specifying a monetary policy and focusing on the nominal exchange rate.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

172

18. INTERNATIONAL MACRO-FINANCE

equilibrium using the Negishi method. First, the social planner’s problem is solved, for a set of utility weights {li}, and then the weights lis and the competitive equilibrium prices are pinned down. The planner’s problem is as follows: max

fCH ;CH ;CF ;CF g

E

T X

1 Et ½qðt þ 1Þpðt þ 1Þ; qðtÞ 1 Et ½qðt þ 1Þp ðt þ 1Þ B ðtÞ ¼ qðtÞ BðtÞ ¼

ð18:9Þ

ð18:3Þ

where Et[] denotes the expectation conditional on the information available up to time t. In the baseline model, the bonds mature in one period. To price s-period zerocoupon bonds, one obviously needs to replace t þ 1 by s in Eq. (18.9).

ð18:4Þ

Log-Linear Preferences

ð18:5Þ

A well-known special case of this model will now be examined which serves as an important benchmark in international finance. In this benchmark, stock prices can be computed in closed form. Suppose that the countries’ utilities are log-linear. That is,

bt ½luH ðCH ðtÞ; CH ðtÞÞ þ uF ðCF ðtÞ; CF ðtÞÞ

t¼0

with multipliers s:t: CH ðtÞ þ CF ðtÞ ¼ YðtÞ; CH ðtÞ þ CF ðtÞ ¼ Y ðtÞ;

ðtÞ  ðtÞ

where, without loss of generality, the weight on the Foreign country has been normalized to be equal to 1. This problem can be broken down into a series of state-bystate maximizations of the quantity inside the summation, weighted by the probability of the corresponding state p(t, ot), subject to the resource constraints (18.4) and (18.5) for that state. The multipliers from that maximization allow the computation of the state prices prevailing in the decentralized equilibrium. The multiplier on Eq. (18.4), (t, ot), is the price of one unit of the Home good to be delivered at time t in state ot. Therefore, the price of one unit of the numeraire to be delivered in state ot, scaled by the probability of that state, is ðt; ot Þ qðt; ot Þ ¼ pðt; ot Þpðt; ot Þ

ð18:6Þ

ð18:10Þ This specification first appears in Helpman and Razin (1978) and is subsequently used in an influential paper by Cole and Obstfeld (1991). For this economy, the expressions that have been presented above simplify significantly. First, the multipliers are computed on the resource constraints of Home and Foreign (Eqns. (18.4) and (18.5)). They are ðtÞ ¼ bt pðtÞ

laH þ aF lð1  aH Þ þ 1  aF and  ðtÞ ¼ bt pðtÞ YðtÞ Y ðtÞ ð18:11Þ

In Finance, this quantity q is typically referred to as the stateprice deflator. It is, of course, nothing else than the marginal utility of the representative agent evaluated over the consumption index (over the two goods) at the optimum. The terms of trade are simply the marginal rate of substitution between the Home and Foreign goods, or, equivalently, the ratio of the multipliers on the resource constraints: ðtÞ ToTðtÞ ¼   ðtÞ

ui ðC; C Þ ¼ ai log C þ ð1  ai Þlog C ; ai 2 ð0; 1Þ; i 2 fH; Fg

ð18:7Þ

The planner’s weight l is determined by substituting the solution to Eqns. (18.3)–(18.5) into either country’s budget constraint (18.2). One is now ready to price financial assets in the model. By no-arbitrage, the prices of stocks and bonds are given by their state-contingent payoffs, discounted with the state-price deflator q: " # T X 1 qðsÞpðsÞYðsÞ ; Et SðtÞ ¼ qðtÞ s¼tþ1 " # ð18:8Þ T X 1    qðsÞp ðsÞY ðsÞ Et S ðtÞ ¼ qðtÞ s¼tþ1

Hence, the state-price deflator is qðtÞ ¼ bt

laH þ aF pðtÞYðtÞ

ð18:12Þ

Substituting Eq. (18.11) into Eq. (18.7), the terms of trade obtained are ToTðtÞ ¼

laH þ aF Y ðtÞ lð1  aH Þ þ 1  aF YðtÞ

ð18:13Þ

The reason why it is typically quite difficult to solve for asset prices and portfolios in macro-finance models is that the conditional expectations in Eqns. (18.8) and (18.9) cannot be evaluated explicitly. Numerical evaluation is not straightforward because future payoffs of the trees and the state prices are endogenously determined in equilibrium. Solving for optimal portfolios, accordingly, is also complicated because portfolio positions depend on the expected returns on the assets. To this day, the main methods that exist in macro-finance for solving these types of problems are the methods involving approximations around a deterministic steady state. This literature has been reviewed in the introduction. The advantage of the setup here is that there is a helpful

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

173

THE WORKHORSE MODEL

simplification and the conditional expectations in the expressions for stock prices admit closed-form representations. In particular, the Home stock’s price turns out to be " # " # T T X X 1 1 qðsÞpðsÞYðsÞ ¼ bt ðaH þ laF Þ Et Et SðtÞ ¼ qðtÞ qðtÞ s¼tþ1 s¼tþ1 ¼

bð1  bTt Þ bð1  bTt Þ ToTðtÞ pðtÞYðtÞ ¼ YðtÞ 1b 1  b aToTðtÞ þ 1  a ð18:14Þ

where in the last equality the price normalization ap(t) þ (1a)p*(t) ¼ 1 has been used. Similarly, the Foreign stock’s price is S ðtÞ ¼

bð1  bTt Þ 1 Y ðtÞ 1  b a ToTðtÞ þ 1  a

ð18:15Þ

Remark 1 Single consumption good The majority of models of international asset-pricing consider single-good economies. Interestingly, multigood frameworks can be more tractable than single-good ones, despite the fact that in single-good economies one does not need to solve for the countries’ terms of trade. To see why, consider a variation of this model in which both trees pay off in the same (Home) good and the countries derive utility only from that good. In particular, the value can be set as aH ¼ aF ¼ 1 in Eq. (18.10). This is the model considered in Cochrane et al. (2008). In this model, ToT(t) ¼ p(t) ¼ 1, with the latter equality occurring because the consumption good is the numeraire. The state-price deflator is now qðtÞ ¼

1 ðl þ 1ÞðYðtÞ þ Y ðtÞÞ

ð18:16Þ

Note that the world’s total output of the good is now Y þ Y*, and the denominator in Eq. (18.16) reflects this adjustment. The argument for the Foreign stock is analogous; therefore, the focus is again on the Home stock. The price of the Home stock is given by the same formula as in Eq. (18.8), but with p(t) ¼ 1: " # T X 1 Et qðsÞYðsÞ SðtÞ ¼ qðtÞ s¼tþ1 " # ð18:17Þ T X 1 YðsÞ t ¼ Et b qðtÞ YðsÞ þ Y ðsÞ s¼tþ1 In a closed economy, Y* ¼ 0, and so one can again readily evaluate the conditional expectation in Eq. (18.17). In an 5

international setting, the task is far more complex. Even under the most tractable distributional assumptions for the output processes used in asset pricing – lognormality – the expectation in Eq. (18.17) is not straightforward to compute. This is because the sum of lognormally distributed random variables is not itself lognormally distributed. Cochrane, Longstaff, and Santa-Clara are able to evaluate the conditional expectation of the quantity in Eq. (18.16) only (i) under i.i.d. Y(t) and Y*(t), distributed lognormally, (ii) for specific values of the discount factor and the volatility of output, and (iii) in continuous time.5 Remark 2 Real exchange rate The real exchange rate, e, is defined as e(t) ¼ PH(t)/PF(t), where PH and PF are Home and Foreign price indexes, respectively. For the preferences considered here, the price indexes are     pðtÞ aH p ðtÞ 1aH ; PH ðtÞ ¼ aH 1  aH     pðtÞ aF p ðtÞ 1aF PF ðtÞ ¼ aF 1  aF Hence, the real exchange rate, expressed as a function of the terms of trade, is eðtÞ ¼ ToTðtÞaH aF

ð1  aF Þ1aF aaFF ð1  aH Þ1aH aaHH

It is more convenient to work with the terms of trade rather than the exchange rate. In this analysis, it is kept in mind that there is a one-to-one mapping between the exchange rate and the terms of trade. If it is assumed further that aH  aF > 0 – an inequality that will be satisfied once home bias is assumed in consumption – the real exchange rate is an increasing function of the terms of trade. The explicit computation of bond prices and interest rates requires additional distributional assumptions on the output processes Y and Y*. These assumptions are not made at this point, because, as it turns out, the setup here is not a good one for studying asset prices. The reason for that is apparent from examining Eqns. (18.14) and (18.15): SðtÞ laH þ aF ¼  S ðtÞ lð1  aH Þ þ 1  aF

ð18:18Þ

where the last equality follows from Eq. (18.13). The righthand side of Eq. (18.18) is constant, and hence the two stock markets are perfectly correlated! Cole and Obstfeld show further that the existence of financial markets does not matter in this model at all: even with no investment

The recent literature provides important extension involving general utility and N trees (see Pavlova and Rigobon 2011 for references).

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

174

18. INTERNATIONAL MACRO-FINANCE

opportunities available, the countries are able to reach a Pareto-efficient allocation through international trade (in goods) alone; there are no benefits to investing internationally. Even if the countries’ investment opportunities are restricted to stocks alone, their portfolios are indeterminate because the two stocks represent the same investment opportunity. This result has had big impact in the international finance literature. While quite stark, however, the result is not robust: it holds only for asset-market economies with log-linear consumer preferences in which all goods are tradable. Any departure from that setup leads to an economy with regular equilibria. One such departure that maintains the tractability of the setup but breaks the perfect correlation among international stock markets has been suggested by Pavlova and Rigobon (2007), who introduce demand shocks.

Log-Linear Preferences with Demand Shocks Consider the following modification of the preferences specified in Eq. (18.10): ui ðC; C Þ ¼ yi ðai log C þ ð1  ai Þ logC Þ;

i 2 fH; Fg ð18:19Þ

where yi is a state-dependent quantity representing a country’s demand shock. It is further required that each yi be a martingale; that is, Et[yi(s)] ¼ yi(t). A demand shock creates shifts in the countries’ demand schedules which may or may not be related to supply. An example of a demand shock is news about weather. This news is unrelated to supply news, but it does affect agents’ demands (e.g., for heating oil). The empirical evidence indicates that demand shocks are important for reproducing the real-world dynamics; supply shocks alone are typically not sufficient. The literature provides a number of estimates for the size of the supply shocks relative to demand shocks, and it is not uncommon to see numbers in excess of 85% for the size of demand shocks relative to supply shocks. The solution to this model follows similar steps to the ones outlined above. The terms of trade, now reflecting demand shocks, are given by ToTðtÞ ¼

lyH ðtÞaH þ yF ðtÞaF Y ðtÞ lyH ðtÞð1  aH Þ þ yF ðtÞð1  aF Þ YðtÞ

ð18:20Þ

The stock prices also have a simple closed-form representation, the same one as that presented above: SðtÞ ¼

bð1  bTt Þ ToTðtÞ YðtÞ and 1  b a ToTðtÞ þ 1  a

bð1  bTt Þ 1 S ðtÞ ¼ Y ðtÞ 1  b a ToTðtÞ þ 1  a

ð18:21Þ

Now the ratio S(t)/S*(t) is stochastic, and so the countries’ stock markets are no longer perfectly correlated. Bond prices and interest rates are still difficult to compute in this model, even under additional distributional assumptions. One way to resolve this technical difficulty is to cast the model in continuous time. To convey the economic mechanisms behind the formulas, one needs to make the following assumption: Assumption 1 (Home bias in consumption): aH(1  aF)aF(1aH)> 0 For this assumption to be satisfied, it is sufficient that aH > 1  aH and 1  aF > aF, or, in words, the expenditure shares on the domestic good for the Home and Foreign country, respectively, exceed the expenditure shares on the foreign good. The following simple table summarizes how the terms of trade and the stock respond to movements in the underlying state variables and some important comparative statics. Boldface in the table means that the sign obtains unambiguously; otherwise, the sign obtains if and only if Assumption 1 is satisfied. Effects of

Y

Y*

uH

uF

l

On the terms of trade ToT On the Home stock S On the Foreign stock S*

2 1 1

1 1 1

þ þ 

  þ

þ þ 

A positive output shock at Home (an increase in Y) raises the dividend on the Home stock and so Home’s stock price increases. At the same time, it increases the supply of the Home good in the world. As the good becomes less scarce, its price falls relative to that of the Foreign good. Hence, Home’s terms of trade deteriorate and Foreign’s terms of trade improve. This effect of an output shock on the terms of trade is known as the Ricardian effect (named after David Ricardo). The improvement of Foreign terms of trade increases the value of Foreign’s output and hence the Foreign stock goes up. The response following a Foreign output shock (a shock to Y*) is analogous. So, the stock markets always comove in response to an output shock. A positive demand shock at Home (an increase in yH) creates an excess demand in the world for both goods. Since Home has a preference bias for the domestic good, however, the demand for the Home good goes up by more. This pushes up its price relative to that of the Foreign good and therefore improves Home’s terms of trade. This effect is best known as the dependent economy effect, highlighted by Dornbusch (1980, Chapter 6). The value of Home’s output (dividend) increases while that of Foreign’s decreases. Hence, Home’s stock market goes up and Foreign’s stock market falls. Demand shocks thus cause divergence in the international financial markets.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

NEXT STEPS

A positive shift in l represents an increase in the weight of the Home country’s utility in the representative agent. This weight reflects the initial wealth distribution in the economy. This is left as an exercise to the reader to show that WH/WF ¼ lyH(0)/yF(0). An increase in l is then akin to a wealth (income) transfer from Foreign to Home. To develop an intuition for the effect of the wealth transfer on the terms of trade, it is useful to recall the classic Transfer Problem.6 A wealth transfer to Home raises Home’s demand for both goods. But in the presence of home bias in consumption, demand for the Home good goes up by more. Hence, the relative price of the Home good increases, that is, Home’s terms of trade improve, just as in the Transfer Problem. Since Home’s terms of trade improve, the value of its output (dividend) goes up, and hence Home’s stock price increases. As Foreign’s terms of trade deteriorate, the value of that country’s output goes down and the price of its stock falls. One of the reasons for highlighting the effect of a change in the planner’s weight l here is its relevance for models with financial market frictions such as, for example, portfolio constraints or incomplete markets. One solution method, discussed by Pavlova and Rigobon (2010), involves solving for a competitive equilibrium using a representative agent with stochastic weights. Here, weights in the planner problem are constant and the allocation is Pareto-efficient. A device involving stochastic weights is employed for solving for an inefficient allocation, occurring in models with frictions. The solution technique follows much the same steps as the ones in the exposition above, except that a stochastic weight l emerges in place of the constant l that is used here. This stochastic l becomes a new (endogenous) state variable in the model, and much of interesting dynamics are due to movements in this state variable (i.e., endogenous wealth transfers). The final step – solving for l – is more complex than in complete markets models, but still feasible.

NEXT STEPS The models referenced so far offer important insights on how elements of international asset pricing, international trade, and open economy macroeconomics can be combined within one framework and how they interact with each other. The literature is evolving in several directions, and future research in this area is going to be active and fruitful. Although a great deal has been learned, many questions remain open. In order to tackle

175

more ambitious questions raised by the data and current events, the existing models certainly require improvements along several dimensions. First, the discussion of global imbalances, current account sustainability, and, more generally, of international portfolios and risk sharing requires the introduction of market incompleteness into the story. This direction is important not only because markets are generally believed to be incomplete but also because there is no role for policy under complete markets (an allocation is already Pareto-efficient). Having incomplete markets adds a layer of methodological complexity. In his Ohlin lecture, Maurice Obstfeld remarks that ‘portfolio choice under incomplete markets is largely terra incognita.’ Developing such models and understanding their workings constitute frontier research in international macroeconomics these days. In the standard textbook models, market incompleteness is due to the inability to trade any asset other than an international bond. This is one form of incompleteness that is certainly relevant, but in order to understand portfolio choices and how portfolio income contributes to the current account one needs to consider a broader menu of financial assets. Otherwise, one can no longer address the question ‘What makes a country’s current account path sustainable?’ – the question that continues to be at the core of international macroeconomics for more than 150 years now. Second, many models that have been developed in international macro-finance so far feature pure exchange economies. This view of production is too simplistic. The natural next step is to include factors of production into these asset-pricing models. Labor market considerations such as effort and unemployment, as well as investment, are important elements through which the real economy and financial markets interact with each other. Third, these models are missing a full-fledged financial sector, the importance of which has been underscored by a series of recent contagious crises. The first step could be to model the financial inefficiencies stemming from the organizational structure of the financial sector in reduced form – for example, in the form of financial constraints on certain market participants (margin constraints, VaR considerations), which may prevent them from supplying liquidity at times when it is needed the most. The next step would then be to endogenize these constraints – that is, to model the agency problems that give rise to the need to restrict traders to take unlimited asset positions and unlimited risk. The fact that constraints on traders that are observed in the real world tend to bind at the same time,

6

The original ‘Transfer Problem’ was the outcome of a debate between Bertil Ohlin and John Maynard Keynes regarding the true value of the burden of reparations payments demanded of Germany after World War I. Keynes argued that the payments would result in a reduction of the demand for German goods and cause a deterioration of the German terms of trade, making the burden on Germany much higher than the actual value of the payments.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

176

18. INTERNATIONAL MACRO-FINANCE

normally in bad times, can emerge as one of the leading explanations of contagion and as a channel of propagation of systemic risk. Another set of interesting frictions includes enforcement problems (as in, e.g., Kehoe and Perri, 2002 ) and financial market deepness (as in Caballero et al., 2008). All these frictions are important and complementary, and exploring their role constitutes a promising direction of future research. Finally, so far closed-form solutions have been obtained only in models in which agents have log-linear preferences. Although some promising work has found the constant elasticity of substitution (CES) preferences to be tractable as well, future research should continue extending the workhorse model to include utility functions that generate more realistic price/dividend ratios, equity premia, and other asset-pricing moments. Because the format of this survey imposes a limit on the number of cited references, this literature is reviewed in Pavlova and Rigobon (2011), with regular posting of further updates. Of course, there is a natural limit to a set of models that admit closed-form solutions; for the remaining, more general, models, the literature will need to rely on numerical methods. Problems involving portfolio choice are particularly difficult to analyze because for these problems standard first-order approximation methods cannot deliver the desired results (see Devereux and Sutherland, 2011; Tille and van Wincoop, 2010 ). The literature is now testing the appropriateness of higher-order approximation methods, with the approximations taken around a deterministic steady state. Perhaps even more complex methods (finite-element methods or projection methods) are what is required. The field of international macro-finance is a new and active area of research. There are many ways in which one can push its frontier. Here, several possible promising directions have just been highlighted. Certainly there are many more.

SEE ALSO Theoretical Perspectives on Financial Globalization: Endogenous Portfolios in International Macro Models; Valuation Effects, Capital Flows and International Adjustment.

References Caballero, R.J., Farhi, E., Gourinchas, P.-O., 2008. An equilibrium model of global imbalances and low interest rates. American Economic Review 98, 358–393. Cochrane, J.H., Longstaff, F.A., Santa-Clara, P., 2008. Two Trees. Review of Financial Studies 21, 347–385. Cole, H.L., Obstfeld, M., 1991. Commodity trade and international risk sharing. Journal of Monetary Economics 28, 3–24. Devereux, M.B., Sutherland, A., 2011. Country portfolios in open economy macro models. Journal of the European Economic Association 9, 337–369. Dornbusch, R., 1980. Open Economy Macroeconomics. Basic Books, Inc. Publishers, New York. Engel, C., Matsumoto, A., 2006. Portfolio Choice in a Monetary OpenEconomy DSGE Model. University of Wisconsin Working paper. Gourinchas, P.-O., Rey, H., 2007. International financial adjustment. Journal of Political Economy 115, 665–703. Helpman, E., Razin, A., 1978. A Theory of International Trade under Uncertainty. Academic Press, San Diego. Kehoe, P.J., Perri, F., 2002. International business cycles with endogenous incomplete markets. Econometrica 70, 907–928. Kollmann, R., 2006. A dynamic general equilibrium model of international portfolio holdings: comment. Econometrica 74, 269–273. Lane, P.R., Milesi-Ferretti, G.M., 2001. The external wealth of nations: measures of foreign assets and liabilities for industrial and developing countries. Journal of International Economics 55, 263–294. Pavlova, A., Rigobon, R., 2007. Asset prices and exchange rates. Review of Financial Studies 20, 1139–1181. Pavlova, A., Rigobon, R., 2010. Equilibrium Portfolios and External Adjustment under Incomplete Markets. London Business School, London Working paper. Pavlova, A., Rigobon, R., 2011. International Macro-Finance. London Business School, London Working paper. Tille, C., van Wincoop, E., 2010. International capital flows. Journal of International Economics 80, 157–175.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

C H A P T E R

19 Monetary Policy and Capital Mobility M.M. Spiegel Federal Reserve Bank of San Francisco, San Francisco, CA, USA O U T L I N E Introduction

177

Growth in International Capital Mobility and Monetary Policy Measures of Capital Mobility Global Imbalances Convergence in Yields Monetary Policy Responses

178 178 179 180 180

Arguments for Improved Monetary Policy Under Increased Capital Mobility Increased International Asset Substitutability Increased Wage and Price Flexibility

181 181 182

INTRODUCTION This chapter examines the relationship between capital mobility and monetary policy. By all measures, the previous decade was one of dramatic increase in the magnitude of global capital mobility. Measures of international cross-holdings of assets, examined in more detail below, demonstrate that there was an explosion in international trade in assets prior to the onset of the 2008 global financial crisis. Even after that event is taken into account, it is clear that the world is more integrated than it was at the start of the decade. The chapter begins with a brief review of recent changes in capital account openness and inflation outcomes. The data show that with the use of either de facto measures of financial openness, such as the ratio of gross international asset positions as a share of gross domestic product (GDP), or de jure measures of capital account policies, there has been a high growth in capital mobility and a substantial drop in average inflation rates for a broad cross-section of emerging market economies. A marked convergence in bond spreads across countries was also observed. This suggests that debt

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00016-7

Reduced Public Pressure for Output Stabilization

182

Theoretical Arguments for Reduced Monetary Policy Quality 183 Reduced Policy Effectiveness 183 Increased Exposure to Global Shocks 183 Empirical Evidence

184

Conclusion Glossary References

185 186 186

obligations across countries are considered to be more substitutable than they have been in the past. Of course, at some point in this decade, risk tolerance became excessive, culminating in the global financial crisis. Some of the convergence in spreads was probably excessive as well, as investors bid down spreads on debt issued from more risky economies in an effort to ‘chase yield.’ Still, by the end of 2009, bond spreads were substantially below the levels that prevailed at the start of the decade. The fact that these developments took place simultaneously raises the question whether there is any linkage between them. In the following paragraphs, these arguments concerning the impact of increased capital mobility on monetary policy outcomes are reviewed. The arguments for improved monetary policy outcomes usually concentrate on the potential for increased policy discipline under increased capital mobility. These arguments usually are made for an economic environment where optimal monetary policy is not time consistent, exhibiting a bias toward output stabilization relative to attention to maintaining price stability. Given this bias, there are a number of potential channels through which increased international capital mobility

177

# 2013 Elsevier Inc. All rights reserved.

178

19. MONETARY POLICY AND CAPITAL MOBILITY

can raise the attention directed toward maintaining price stability, moving monetary policy closer to the optimal mix. However, it is also possible that the quality of monetary policy may deteriorate under increased capital mobility. First, for the same reason that increased capital mobility may improve the capacity for commitment, namely, deterioration in the output–inflation trade-off, it may also reduce the capacity for undertaking desirable discretionary policy. Second, increased capital mobility may leave a nation more exposed to global shocks. This is particularly true for emerging market economies. Given that the predicted impact of increased capital mobility on monetary policy outcomes is ambiguous, it is not surprising that empirical studies along these lines have yielded quite mixed results. Overall, while there is some evidence of a positive correlation between international capital mobility and the quality of policy outcomes, the results are quite fragile. Moreover, it would not be surprising if the results were reversed in future studies that include the current financial crisis in their samples. Finally, it should be remembered that the rate of inflation that is usually used as a measure of the quality of monetary policy outcomes is not directly tied to welfare, and therefore represents an intermediate policy goal. As such, price stability should be considered an ‘intermediate target’ of value because hitting that target has been shown to be associated with superior overall economic performance. It follows that it might be preferable to examine the relationship between financial openness and long-term economic growth directly. However, as discussed below, the evidence along these lines is also very mixed. Moreover, it is quite difficult to isolate the component of growth performance that is attributable to monetary policy, as opposed to, say, fiscal policy. The remainder of this chapter is divided into six sections. The section ‘Growth in International Capital Mobility and Monetary Policy’ reviews the changes in capital mobility and inflation over the previous decade. The section ‘Arguments for Improved Monetary Policy Under Increased Capital Mobility’ examines the theoretical arguments in favor of a positive impact from increased capital mobility on monetary policy outcomes. The section ‘Theoretical Arguments for Reduced Monetary Policy Quality’ looks at arguments on why monetary policy may deteriorate under increased capital mobility. The section ‘Empirical Evidence’ examines empirical evidence concerning international financial integration and the quality of monetary policy outcomes. The last section concludes the chapter.

1

GROWTH IN INTERNATIONAL CAPITAL MOBILITY AND MONETARY POLICY Measures of Capital Mobility The author follows the literature in defining de facto financial openness as gross international asset positions, measured as the sum of stocks of external assets and liabilities of foreign direct investment (FDI) and portfolio investment, as a share of GDP. It is well documented that financial globalization according to this measure took off in both industrial and emerging market countries in the latter half of the 1990s.1 For example, see Figure 19.1, which plots average levels of de facto capital account openness against average inflation rates for a cross-section of emerging market economies, beginning immediately after the 1997 Asian financial crisis. It can be seen that there is a rapid increase in the index of average de facto capital account openness over this period, from 24.0 in 1998 to 56.8 in 2007. Over the same period, average inflation rates for these countries fell from 18.6% to 5.6%.

De facto capital account openness and average inflation (EMEs 1998–2007) 20

60

18 50

16 14

40

12 10

30

8 20

6 4

10

2 0

0 98 99 00 01 02 03 04 05 06 07 19 19 20 20 20 20 20 20 20 20

de facto KA open Average inflation

FIGURE 19.1 Average levels of de facto capital account openness, measured as the sum of stocks of external assets and liabilities of FDI and portfolio investment, as a share of GDP, and average levels of consumer price index (CPI) inflation for a cross-section of emerging market economies. Sources: Lane and Milesi-Ferretti EWN II data set and World Development Indicators.

See Lane and Milesi-Ferretti (2003) and Prasad et al. (2003).

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

179

GROWTH IN INTERNATIONAL CAPITAL MOBILITY AND MONETARY POLICY

2

See Wynne and Kersting (2009) for more details.

3

See Chinn and Ito (2008) for more details.

4

For example, see Lucas (1990).

De jure capital account openness and EME inflation (1998–2007) 0.8

18

0.7

16

0.6

14

0.5

12

0.4

10

0.3

8

0.2

6

0.1

4

0

2

−0.1

0

−0.2 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07

20

19

The de facto capital account openness index for developed countries also rose at an impressive rate, from 72.7 in 1998 to 162.3 in 2007. There was no corresponding falloff in inflation, as developed economies had already achieved low average inflation levels in the 1980s, but inflation remained well contained over this period. There are a number of reasons for this dramatic upturn in international capital movements: Technological progress has reduced the cost of acquiring and managing holdings of foreign assets, and thereby increased investors’ demand for internationally diversified portfolios. In addition, innovations in finance have increased the capacity for hedging investment positions, leading to a proliferation of available international investment vehicles. This phenomenon has led to a reduction, but in no sense an elimination, of the so-called home bias effect, whereby economic agents are found to hold a disproportionate share of their asset portfolios in assets originating from their home country. There has been a modest reduction in overall international cross-holdings of assets subsequent to the advent of the global financial crisis. However, this is likely attributable to the dramatic falloff in international trade in goods subsequent to the onset of the crisis. In its October 2009 World Economic Outlook, the International Monetary Fund (IMF) projects that overall trade in 2009 will decline by 11.9% relative to the previous year.2 Cross-holdings of assets are usually positively correlated with the volume of goods trade, because hedging risk associated with goods trade is an important component of the motivations for holding foreign assets. With less trade, there is less risk associated with trade to be hedged. Alternatively, one can characterize the degree of a country’s de jure capital mobility, that is, the degree to which a country pursues policies associated with encouraging international capital mobility. Figure 19.2 plots the same results for such a series, the Chinn-Ito de jure measure of capital account openness.3 The results are quite similar. There has been a dramatic increase in capital account liberalization over the period among emerging market economies and this coincided with a dramatic decline in average inflation among emerging market economies. The developed countries as a group did not experience a similar increase in de jure capital account openness over this period according to the Chinn-Ito data set, but this is attributable to the fact that they entered the period with liberalized capital accounts.

Average inflation de jure KA open

FIGURE 19.2 Average levels of de jure capital account openness policies (see Chinn and Ito 2008, for details), and average levels of CPI inflation for a cross-section of emerging market economies. Sources: Chinn and Ito (2008) data set and World Development Indicators.

Global Imbalances The increased capital flows documented above have had a number of important impacts on the international economy: Emerging market economies have become net creditors as a group, which has allowed some developed economies, particularly the United States, to finance large current account imbalances at relatively favorable rates. This pattern of capital flows is generally considered to be nonstandard for a number of reasons: First, standard theory suggests that capital scarcity in developing countries leaves their marginal products of capital higher than the developed countries as a group. Holding all else equal, one would expect capital to flow from countries that are rich in capital to those that are poor in capital.4 Second, higher expected future incomes in the rapidly growing emerging market economies provide them with an incentive to run current account deficits now to smooth consumption. Much work has gone into explaining this paradoxical investment pattern. One theory focuses on differences in the quality of financial intermediation between developed and emerging market economies, where portfolio capital moves from south to north, to return as FDI.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

180

19. MONETARY POLICY AND CAPITAL MOBILITY

Alternatively, some have suggested that net outflows from China may serve as collateral against future opportunistic behavior. A third approach articulated by the then Governor Ben Bernanke prior to his becoming the Federal Reserve Board Chairman argues that poor investment opportunities in Asia have resulted in a global ‘savings glut’ that has freed up capital for lending to developed economies.5

Convergence in Yields The increased volume of trade in financial assets has had a significant impact on international borrowing terms. For example, see Figure 19.3, which displays the blended spread on the J.P. Morgan Emerging Market Bond Index since the beginning of the decade. From the start of the decade to the onset of the global financial crisis that began in 2008, spreads on emerging market bonds decreased markedly. This period was marked by a growing tolerance for risk that reduced yields and induced global investors to venture more deeply into foreign exposure. These improved borrowing terms led some firms to borrow sufficiently to leave themselves vulnerable when the global crisis began. As is quite apparent from the graph, the onset of the crisis saw a spike in the J.P. Morgan Index. This spike actually reached levels higher than those that prevailed subsequent to the 1997 Asian financial crisis. However, when liquidity was restored to these markets, these spreads quickly came down. While the spreads are still elevated relative to levels seen in 2007, they are about 200 basis points below those that prevailed at the start of the decade.

The marked reduction in spreads prior to the crisis suggests that debt obligations across countries are being treated (at least under tranquil conditions) as more substitutable than they have been in the past. One inference that might be drawn from this pattern is that international financial markets are more integrated. If international assets actually are viewed by investors as more substitutable, the sensitivity of domestic and foreign investors to interest rate differentials would be increased.

Monetary Policy Responses The additional discipline placed on monetary authorities from the increase in financial integration over the decade has led to increased focus by monetary authorities on targeting the inflation rate. For example, as of 2007, nearly half of the OECD countries formally targeted inflation, along with ten emerging market economies. Moreover, the European Monetary Union (EMU) announces an inflation target, while the United States cites ‘price stability’ as one of its two monetary policy goals. Inflation-targeting regimes have also been durable. These regimes have existed for close to 20 years but have been left only by those countries that abandoned their individual inflation targets to join the EMU which itself targets price stability.6 The variability of inflation has also declined markedly in emerging market economies. Standard economic theory suggests that the variability of inflation, rather than its level, is the key to determining output volatility. In practice, high inflation tends to coincide with variable

JPMorgan EMBI+ brady broad blended spread

Basis points

FIGURE 19.3 Average levels of JP Morgan EMBI and Brady Bond Broad Blended Spreads, 2000–present. Source: Data from Bloomberg.

1400 1200 1000 800 600 400 200

1/ 20 02 03 /0 1/ 20 03 03 /0 1/ 20 04 03 /0 1/ 20 05 03 /0 1/ 20 06 03 /0 1/ 20 07 03 /0 1/ 20 08 03 /0 1/ 20 09

20 01

03 /0

01 /

03 /

03 /0

1/ 20 00

0

5

Bernanke (2005).

6

See Rose (2007). The countries in question are Finland and Spain.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

ARGUMENTS FOR IMPROVED MONETARY POLICY UNDER INCREASED CAPITAL MOBILITY

inflation, which is why keeping the rate of inflation under control will help in controlling its variability as well. Over the preceding 10 years, as average inflation rates fell in emerging market economies, the variability of inflation in those countries has also fallen. Prior to the onset of the crisis, the renewed focus on controlling inflation and inflation expectations had further improved conditions in capital markets. Emerging market economies moved from bank borrowing in external so-called hard currencies toward external borrowing in bonds denominated in their domestic currencies with relatively long maturities and fixed interest rates. Korea and Thailand introduced 10-year domestic currency bonds in the 1990s, while by the year 2000, Brazil, Chile, Colombia, Indonesia, Mexico, and Russia had also issued domestic currency bonds.7 As these instruments have become more standard, their yields have decreased, allowing these countries to borrow at favorable terms. This shift has achieved a number of desirable effects. First and foremost, currency risk has shifted from the borrower to the lender. The desirability of this shift was dramatically demonstrated in the recent global financial crisis, where countries with substantial outstanding debts denominated in foreign ‘hard’ currencies experienced substantial debt-servicing difficulties associated with ‘currency mismatch,’ the balance sheet deterioration which is experienced subsequent to domestic currency devaluation when assets are denominated in domestic currency and liabilities are denominated in foreign currencies. In addition, the fixed interest rates associated with long-term issues have shifted interest rate risk to creditors as well. Third, the longer maturities have reduced the risk of disruptive ‘sudden stops’ in credit, resulting in costly failures in the past. Fourth, government issues in local currency have helped encourage the development of local bond markets by providing ‘benchmark’ yield curves for pricing private debt. Finally, when defaults do take place in bond markets, contagion is limited by the wide dispersion of creditors. Still, the hindsight now afforded by the global financial crisis suggests that a part of the ability of issuers to shift this risk onto creditors was attributable to the actions of the global rating agencies who exaggerated the safety of emerging market securities (and the securities of developed economies as well) prior to the crisis. 7

181

This led lenders to ignore standard due diligence practices when extending and underwriting debt. As such, the increase in overall borrowing levels left the system in a more precarious situation.

ARGUMENTS FOR IMPROVED MONETARY POLICY UNDER INCREASED CAPITAL MOBILITY Most analyses that predict improved monetary policy outcomes under increased capital mobility concentrate on channels where increased capital mobility deteriorates the ‘output–inflation trade-off’ faced by a monetary authority. This can improve policy outcomes when optimal monetary policy is not ‘time consistent,’ that is, where in the absence of binding rules, monetary policymakers choose a suboptimally high level of inflation. This phenomenon is referred to as time inconsistency because following a certain policy rule, say an inflation target, may be optimal over the long run, but over shorter horizons, policymakers are unable to resist the temptation to deviate to achieve some short-term objective. In the literature, the central bank typically deviates in the form of choosing a higher than optimal level of inflation in the attempt to temporarily stabilize output levels.8 Given this inflationary bias, changes in the environment that would induce a monetary policymaker to choose lower inflation levels can be welfare improving. Of course, not all of these changes are associated with the external sector, much less international capital mobility. The optimal monetary policy may be achieved, for example, by choosing a ‘conservative’ central banker who is even more averse to inflation than the populace on average.9 However, this section concentrates on the channels that argue that increased capital account openness can lead to superior monetary policy outcomes.

Increased International Asset Substitutability First, increased international capital mobility can enhance discipline among policymakers who might be tempted to exploit a captive domestic capital market that is closed internationally.10 In the presence of open capital markets, a monetary authority cannot successfully pursue exchange rate and monetary policy goals that are

Kroszner (2007).

8

The seminal article on this topic is Kydland and Prescott (1977). In the model in this article, policymakers face an irresistible short-term opportunity to increase output by raising inflation. However, the inability of policymakers to pursue this higher level of inflation is anticipated by market participants, so that in equilibrium the only change achieved under discretion is higher inflation. 9

See Rogoff (1985).

10

For example, see Obstfeld (1998).

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

182

19. MONETARY POLICY AND CAPITAL MOBILITY

inconsistent. This situation is commonly referred to as the ‘impossible trinity,’ as with free international capital mobility, domestic and foreign speculators would respond to an inconsistency between monetary policy and an announced exchange rate goal by attacking the inconsistent exchange rate. As such, a policymaker in an environment of open capital markets must consider the exchange rate response when deciding on monetary policy. To a country pursuing an exchange rate peg, or a similar narrow exchange rate target, this limits the degree of monetary policy discretion, a benign result in an environment where additional commitment is desired. However, even without any explicit exchange rate commitment, the knowledge that a volatile monetary policy is likely to lead to exchange rate disruptions may also be helpful as a disciplining device. If assets are imperfect substitutes internationally, there may be a limited scope for authorities to resist an attack on an inconsistent exchange rate peg by ‘leaning against the wind,’ in the form of foreign exchange intervention. For example, if the Japanese government is determined to keep the yen at a lower exchange rate relative to the dollar while its central bank pursues monetary policies consistent with appreciation, it could drive down the value of yen-denominated assets to some extent, and hence the exchange rate value of the yen itself, by selling yen-denominated assets and purchasing dollar-denominated assets. However, as this exercise relies fundamentally on the degree of imperfections in asset substitutability, the scope for successful foreign exchange intervention along these lines is reduced as markets become more integrated. With increased capital mobility, this will be the case in international bond markets. In this environment, longer term real interest rates are likely to be less sensitive to transitory movements in monetary policy rates. Increased capital mobility may therefore leave interest rates less sensitive to monetary policy and reduce the effectiveness of the inflation tax. Holding all else equal, in equilibrium, governments should respond to these changes by relying less on this revenue-raising instrument.11

Increased Wage and Price Flexibility It has also been argued that increased capital mobility may provide monetary policy discipline by increasing price and wage flexibility.12 In most standard models with output–inflation trade-offs, pricing power enjoyed 11

See Spiegel (2007).

12

See Rogoff (2004, 2006).

13

For example, see Razin and Loungani (2007).

by domestic producers and labor unions allows them to set the sticky prices and wages that yield a trade-off to monetary policymakers between output and inflation, commonly referred to as the ‘Phillips curve,’ which represents this trade-off as the slope of the relationship between output and inflation. As the steepness of the Phillips curve represents the terms of this trade-off, a steeper Phillips curve implies that policymakers need to generate a larger increase in inflation to achieve a given positive output response. Increased capital mobility can increase the steepness of this curve, deteriorating the output ‘payoff’ monetary authorities receive from a given increase in inflation. This is achieved through deterioration in the monopoly power enjoyed by firms and labor unions, leading to greater wage and price flexibility. As a result, monetary policymakers will be less inclined to exploit this trade-off in an attempt to raise output levels, and policy commitments may be more credible and endurable.

Reduced Public Pressure for Output Stabilization Finally, when one considers the political-economy constraints faced by a monetary authority, it becomes quite likely that the weight placed by the central bank on stabilizing output relative to the weight it places on inflation reflects the desires of the population at large. These political-economy considerations provide additional channels by which international capital mobility can influence the quality of monetary policy outcomes.13 To pursue these arguments, one must grant the assumption that the central bank lacks complete monetary policy independence. While many studies argue that more independent central banks produce superior policy outcomes, complete policy independence may be unachievable in practice. Even formally independent central banks assigned verifiable inflation targets are likely to feel the pressure to respond to some degree to public pressure concerning a desired policy. As such, a model with some deviation from complete policy independence is likely to be more realistic. Given an incomplete policy independence, increased capital mobility may affect monetary policy outcomes by affecting the weight that the public places on stabilizing output. Developments that reduce the weight on output stabilization held by the public then correspond to the reduced weight on output stabilization in the desired outcomes of the central bank, and thereby increase the weight placed on reducing inflation and maintaining price stability.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

THEORETICAL ARGUMENTS FOR REDUCED MONETARY POLICY QUALITY

How can increased capital mobility yield this outcome?14 Consider the situation where some degree of price rigidity allows for a positive inflation–output trade-off, and monetary policy is to some extent responsive to the desires of the public at large because the central bank lacks complete policy independence. In this environment, changes in the weight placed by consumers on stabilizing output and controlling inflation would correspond to changes in the degree to which the monetary authority pursues these policy goals. In particular, the weight placed by consumers on stabilizing output would be increasing in the degree to which output stabilization reduces variability in consumption, the variable that the public really cares about. Increases in capital mobility can weaken this relationship by allowing consumers to hedge against domestic output risk through the purchases of assets whose value is based on output realizations in foreign countries. By adding foreign asset holdings to their portfolio, consumption can be less correlated with domestic output. As a result, given an increased capital mobility, consumers would be less exposed to fluctuations in domestic output. Their desired mix of domestic output and inflation would therefore be closer to the true optimal mix than the mix they would choose under a reduced capital mobility. As their desired mix is to some extent reflected in actual monetary policy, this results in enhanced monetary policy discipline.

THEORETICAL ARGUMENTS FOR REDUCED MONETARY POLICY QUALITY Reduced Policy Effectiveness With the exception of the turbulence attributable to the recent global financial crisis, the increased capital mobility over the previous decade led to a convergence in yields. This convergence suggests that in the steady state, assets are closer substitutes worldwide, and domestic interest rates are likely to be influenced by global factors. In this type of environment, longer term interest rates are likely to be less sensitive to transitory movements in the Federal Funds rate, the interest rate targeted by the United States Federal Reserve. This lack of sensitivity has led some to conclude that financial globalization leaves interest rates less sensitive to monetary policy and reduces the effectiveness of monetary policy. 14

Ibid (2005).

15

Rogoff (2006).

16

For an example, see Woodford (2010).

17

See Rogoff (2004, 2006).

18

See Eichengreen and Leblang (2003).

183

Monetary policy may be left less effective in this environment because even the largest central banks “. . . have less direct impact on medium and long-term interest rates than might once have been the case.”15 Ironically, this is the same development that was used earlier to argue for improvement in monetary policy when greater commitment was desired. That argument was based on the assumption that there was an inflationary bias in discretionary monetary policy. However, there may be situations where expansionary monetary policy is desirable, and changes in the environment that deteriorate the output–inflation trade-off can impair policy outcomes. This does not mean that discretionary stabilization policy is unattainable. It has been shown for a variety of models that if monetary policy can influence prices and output under capital immobility, it will retain these powers when assets are highly substitutable.16 With high asset substitutability, a greater nominal expansion is required to achieve a given output expansion, but the monetary authority should retain the ability to control the domestic price level. Moreover, the fact that an individual central bank has lost some of its short-term influence over real interest rates would not imply that central banks as a group have lost the ability to act in concert and influence rates over the short term. Central banks acting in concert, such as the recent move by a number of banks to inject liquidity into the financial system in response to the global financial crisis, can still have a substantial impact.17 Still, it may complicate matters. Bernanke (2007) notes that financial globalization may make an analysis of financial and economic conditions more complex, arguing that increased foreign demand for US assets had contributed to recent inversions of the yield curve. Kohn (2008) acknowledges that asset price determination is more dependent on worldwide financial conditions in the wake of financial globalization, reducing the correlation between the federal funds rate, which is directly controlled by the Federal Reserve, and longer term treasury bills.

Increased Exposure to Global Shocks Increased capital mobility is likely to lead to increased financial integration with the rest of the world. Under increased financial integration, countries may have increased exposure to global shocks and therefore greater susceptibility to financial crises and difficulties associated with currency mismatch.18

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

184

19. MONETARY POLICY AND CAPITAL MOBILITY

As a result, increased financial integration can hinder monetary policy outcomes, as global shocks may confront monetary policy authorities with exacerbated exchange rate volatility. Given the desire to pursue some exchange rate goal, monetary authorities may deviate from policies consistent with achieving price stability. Alternatively, monetary authorities who concentrate on a price stability target may find themselves constrained to endure other undesirable results. This was the case prior to the onset of the global financial crisis where the Bank of Korea, a central bank that operated under a formal inflationtargeting mandate, was forced to increase interest rates in response to rising domestic prices. This policy led to surges in capital inflows that exacerbated that country’s difficulties during the financial crisis. However, recent studies have found that countries that target inflation experience no more exchange rate volatility on average than do countries that do not target inflation.19 There are also concerns that financial globalization may be more disruptive in emerging market economies because of the relative lack of development of their domestic financial sectors. Opening up to international markets may increase consumption volatility if domestic financial markets are relatively undeveloped and agents within the economy have heterogeneous access to external financial markets.20 The reason is that risk sharing within the domestic economy can deteriorate if a subset of domestic agents face increased external risk-sharing opportunities not available to all.

EMPIRICAL EVIDENCE The chapter focuses next on the existing empirical evidence concerning the relationship between capital mobility and monetary policy. As discussed in the ‘Introduction,’ there is less direct evidence concerning the relationship between capital mobility and monetary policy than exists between capital account openness and overall macroeconomic volatility. With regard to the latter measure, there is a large volume of literature that documents the weakness of this empirical relationship. In a series of recent papers, a group of researchers at the IMF have documented

19

Rose (2007).

20

See Levchenko (2004).

21

For example, see Kose et al. (2003a, 2003b, 2007).

22

E.g., Prasad et al. (2003).

23

See Bekaert et al. (2006).

24

See Rose and Spiegel (2009).

25

For example, see Eichengreen and Leblang (2003).

the fragility of the evidence of macroeconomic benefits of financial integration. These studies have shown that the ratio of consumption volatility to income increased during the 1990s for more financially integrated economies. Moreover, these studies find little evidence that increased capital mobility has led to increased risk sharing across countries.21 These results have been shown to be relatively invariant to the measure of macroeconomic volatility used. As discussed in the ‘Introduction,’ it may be preferable to examine growth outcomes directly, rather than looking at average inflation levels. Average inflation levels represent an intermediate target only of interest to consumers to the extent that it affects economic growth. With the caveat that it is difficult to calculate the shares of discrepancies in growth performances that are attributable to differences in monetary policies, at least one is examining a phenomenon that directly impacts welfare. Unfortunately, however, studies that look at financial integration and growth have also failed to demonstrate a robust statistically significant relationship.22 There have been exceptions. Some studies explicitly looking at the effects of liberalization have found evidence that liberalizing the capital account reduces consumption volatility.23 Other studies have found stronger evidence of a relationship between financial remoteness and macroeconomic volatility after instrumenting for the intensity of financial integration, which is likely to be endogenous.24 The limited empirical evidence supporting a positive impact of capital account liberalization on growth is not that surprising given the ambiguity in the predictions of theory. Moreover, the relationship may not be time invariant. Capital account liberalization may enhance growth by improving the allocation of resources and government policies in tranquil periods, but it may also expose countries to global macroeconomic shocks. Efforts to allow for the relationship to be contingent on country characteristics do purport to identify a positive relationship between liberalized capital accounts and growth performances during tranquil periods.25 However, if these gains are achieved only at the expense of greater vulnerability to global shocks, the policy implications are left unclear.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

185

CONCLUSION

There is less direct evidence on the relationship between capital mobility and the quality of monetary policy outcomes. One recent study26 examines the relationship between de facto financial openness and monetary and fiscal discipline. To address the issue of endogeneity in the determination of financial openness, financial openness in neighboring countries in the same geographic region, weighted by distance from the country in question, is used as an instrument for national financial openness. The use of this instrument is motivated from the observation that countries from some regions of the world, such as Latin America, have a disproportionate amount of financial interaction with common countries, in this case, the United States. Using this methodology, researchers have found that financial openness is negatively related to average inflation but has no measurable effect on the government budget deficit. A potential problem with this identification strategy is that some shock, such as a huge change in the price of oil, could have common implications for capital flows within a region, as well as macroeconomic policies within that region, depending on whether the region is a net oil importer or exporter. This could be problematic for the instrumenting methodology used. More recent research has used financial remoteness as a plausibly exogenous instrument for the level of financial integration.27 Raw distance has been shown to be a surprisingly good predictor of cross-holdings of international assets.28 This instrument is not time-varying, but it has a strong claim to plausible exogeneity. This research confirms a negative relationship between de facto financial openness and inflation for a univariate specification with or without instrumenting, but these findings do not appear to be robust to small perturbations in model specification. In the end, then, the evidence that capital mobility improves monetary policy performance must be considered weak. There are a large number of studies that examine the relationship between measures of financial integration and broader measures of macroeconomic performance, but these results can be characterized as mixed at best. In the limited number of studies that directly examine the relationship between financial openness and monetary policy outcomes, the results do suggest that there is a positive univariate relationship between financial integration and monetary stability, and indeed one that appears to stand up to instrumenting to address endogeneity issues, but the importance of this relationship is also sensitive to model specification.

26

Tytell and Wei (2005).

27

Spiegel (2009).

28

For example, see Portes and Rey (2005).

CONCLUSION While the global financial crisis has reduced the pace of growth in trade in both goods and assets, there can be no doubt that the previous decade was one of marked growth in international financial integration. What is more in doubt, however, is the implications of this development for the quality of monetary policy outcomes. Theory is ambiguous with regard to whether one should expect the quality of monetary policy outcomes to improve or deteriorate under increased capital mobility. On the one hand, increased financial integration can increase monetary policy discipline and thereby reduce a policy bias toward inflation. On the other hand, capital account mobility can leave a nation more exposed to global shocks and hinder a monetary authority’s capacity to pursue a desirable countercyclical policy. Given the ambiguity in theory, it is not surprising that the empirical evidence concerning the impact of increased capital mobility on the quality of monetary policy outcomes, or on other measures of macroeconomic performance, is quite mixed. It is difficult to find any study demonstrating robust evidence of a positive or negative impact. This is probably due to fact that the trade-off of increased discipline and increased exposure to foreign shocks varies over time and across countries. Moreover, both financial integration and monetary stability appear to be characteristics of well-functioning economies, but so are a myriad of other factors examined in the literature, such as the level of development of the domestic financial sector, the quality of institutions, or indeed, simply an economy’s level of GDP per capita. The prospects of isolating the role of financial globalization empirically once and for all do not seem promising. Going forward, it will be interesting to see how these empirical results change when long-term studies that include data from the global financial crisis begin to emerge. One would expect that exposure to difficulties associated with currency mismatches experienced by a number of relatively open emerging market economies would tip the evidence against the desirability of financial integration. However, the crisis experience may also lead to improved regulatory environments in countries with open capital accounts. It may therefore be necessary to include both the crisis years and the years that reflect the subsequent policy responses before one can adequately update the evidence to include the impact of the current global financial crisis.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

186

19. MONETARY POLICY AND CAPITAL MOBILITY

SEE ALSO Theoretical Perspectives on Financial Globalization: Capital Mobility and Exchange Rate Regimes.

Glossary Commitment The ability to pursue a stated path of action regardless of prevailing conditions. De facto capital account openness The extent of international trade in financial assets. De jure capital account openness The degree to which national policies are conducive to capital mobility. Impossible trinity The inability to achieve inconsistent inflation and exchange rate goals under capital mobility. Phillips curve A graphical representation of output–inflation tradeoff. The slope of the curve represents the amount by which inflation must be increased to achieve a given increase in output. Time consistency The ability to pursue a stated policy under prevailing conditions.

References Bekaert, G., Harvey, C.R., Lundblad, C., 2006. Growth volatility and financial liberalization. Journal of International Money and Finance 25 (3), 370–403. Bernanke, B., 2005. The Global Savings Glut and the U.S. Current Account Deficit. the Homer Jones Lecture, St. Louis Missouri. April 14. http://www.federalreserve.gov/boarddocs/speeches/ 2005/200503102/default.htm. Bernanke, B., 2007. Globalization and Monetary Policy, speech at the Fourth Economic Summit. Stanford Institute for Economic Policy Research, Stanford, California. March 2. http://www. federalreserve.gov/newsevents/speech/Bernanke20070302a.htm. Chinn, M.D., Ito, H., 2008. A new measure of financial openness. Journal of Comparative Policy Analysis 10 (3), 309–322. Eichengreen, B., Leblang, D., 2003. Capital account liberalization and growth: Was Mr. Mahathir right? International Journal of Finance and Economics 8 (3), 205–244. Kohn, D.L., 2008. Implications of Globalization for the Conduct of Monetary Policy. International Symposium of the Banque de France, Paris France, March 7. Kose, A., Prasad, E., Rogoff, K., Terrones, M.E., 2007. How does financial globalization affect risk-sharing? Patterns and channels. Institute for the Study of Labor, Bonn, Germany. IZA Discussion Paper No. 2903. Kose, A., Prasad, E., Rogoff, K., Wei, S.-J., 2006. Financial globalization: a reappraisal. International Monetary Fund, Washington, DC. IMF Working Paper 06/189. Kose, A., Prasad, E., Terrones, M.E., 2003a. Special Issue: Financial integration and macroeconomic volatility. IMF Staff Papers 50, 119–142 (International Monetary Fund, Washington).

Kose, A., Prasad, E., Terrones, M.E., 2003b. How does globalization affect the synchronization of business cycles? American Economic Review 93 (2), 57–62. Kroszner, R.S., 2007. Globalization and capital markets: implications for inflation and the yield curve, at the Center for financial Stability (CEF), Buenos Aires, Argentina. http://www.federalreserve.gov/ boarddocs/Speeches/2007/20070516/default.htm. Kydland, F.E., Prescott, E.C., 1977. Rules rather than discretion: the inconsistency of optimal plans. Journal of Political Economy 85 (3), 473–492. Lane, P.R., Milesi-Ferretti, G.M., 2003. International financial integration. IMF Staff Papers 50, 82–113 Special issue. Levchenko, A.A., 2004. Financial liberalization and consumption volatility in developing countries. IMF Staff Papers 52 (2), 237–259. Lucas, R.E., 1990. Why doesn’t capital flow from rich to poor countries? American Economic Review 80 (2), 92–96. Portes, R., Rey, H., 2005. The determinants of cross-border equity flows. Journal of International Economics 65 (2), 269–296. Prasad, E.S., Rogoff, K., Wei, S.J., Kose, M.A., 2003. Effects of financial globalization on developing countries: some empirical evidence. International Monetary Fund, Washington, DC. IMF Occasional Paper 220. Razin, A., Loungani, P., 2007. Globalization and inflation-output tradeoffs. In: Frankel, J., Pissarides, C.A. (Eds.), NBER International Seminar on Macroeconomics. MIT Press, Cambridge, MA, pp. 171–192. Rogoff, K.S., 1985. The optimal degree of commitment to an intermediate monetary target. Quarterly Journal of Economics 100 (4), 1169–1189. Rogoff, K.S., 2004. Globalization and global disinflation. In: Monetary Policy and Uncertainty: Adapting to a Changing Economy, Papers and Proceedings of the 2003 Jackson Hole Symposium, Federal Reserve Bank of Kansas City, 77–112. Rogoff, K.S., 2006. Impact of Globalization on Monetary Policy. In: Proceedings of the Federal Reserve Bank of Kansas City Symposium on The New Economic Geography: Effects and Policy Implications, Jackson Hole, Wyoming, 265–305. Rose, A.K., 2007. A stable monetary system emerges: inflation targeting is Breton Woods reversed. Journal of International Money and Finance 26 (5), 663–681. Rose, A.K., Spiegel, M.M., 2009. International financial remoteness and macroeconomic volatility. Journal of Development Economics 89 (2), 250–257. Spiegel, M.M., 2007. Financial globalization and monetary policy. FRBSF Economic Letter 2007-34, November 23. Spiegel, M.M., 2009. Financial globalization and monetary policy discipline: a survey with new evidence from financial remoteness. IMF Staff Papers 56 (1), 198–221. Tytell, I., Wei, S.J., 2005. Global Capital Flows and National Policy Choices. International Monetary Fund, Washington, DC (unpublished). Woodford, M., 2010. Globalization and monetary control. In: Gali, J., Gertler, M.J. (Eds.), International Dimensions of Monetary Policy. University of Chicago Press, Chicago, IL, pp. 13–77. Wynne, M.A., Kersting, E.K., 2009. Trade, globalization and the financial crisis. Federal Reserve Bank of Dallas Economic Letter 4 (8).

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

C H A P T E R

20 Theory of Sovereign Debt and Default Mark L.J. Wright Federal Reserve Bank of Chicago, Chicago, IL, USA, University of California, Los Angeles, CA, USA National Bureau of Economic Research, Cambridge, MA, USA O U T L I N E Introduction

187

Why Do Countries Repay Their Debts? Sovereign Immunity, Legal Sanctions, and Direct Punishments Restrictions on Financial Market Access Domestic Costs of Default

187

Why Do Countries Borrow So Much? A Benchmark Model

189 189

187 188 189

INTRODUCTION The absence of a supranational legal system that can enforce contracts exposes financial transactions between countries to greater risk of nonpayment than financial transactions within a country. This risk is particularly severe when the financial assets involved are the debts of a sovereign government that possesses few assets within foreign legal jurisdictions that can be seized, which can deny access to legal remedies within its own jurisdiction, and which may not be responsive to the demands of foreigners through its domestic political system. In this chapter, we review the theoretical literature on the effect of this sovereign risk on the market for sovereign debt, with a focus on the extent to which sovereign risk acts as a constraint on the process of financial globalization. To understand sovereign risk, it is necessary to understand the incentives for sovereign borrowers to repay their debts and hence also the incentives for creditors to lend to sovereigns in the first place. We begin by examining the early theoretical literature which focused on providing a qualitative understanding of the ways in which sovereign risk constrained the market for sovereign debt financial globalization and the forces that enforced the repayments of some debt. We then turn to the emerging

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00014-3

Evaluating and Extending the Benchmark Model

190

Policy and Welfare Rollover Risk and Self-Fulfilling Debt Crises Debt Dilution and the Maturity of Sovereign Debts Collective Action Problems in Debt Restructuring

191 191 191 192

Conclusion Glossary Further Reading

192 193 193

quantitative theoretical literature on sovereign default that assesses the ability of theories of the effect of sovereign risk to match the facts about sovereign debt and default. We conclude with a discussion of institutional changes and policy reforms that might act to reduce the impact of sovereign risk and hence strengthen the process of financial globalization.

WHY DO COUNTRIES REPAY THEIR DEBTS? Sovereign Immunity, Legal Sanctions, and Direct Punishments When a private borrower defaults on a domestic debt contract, the primary costs of default to that borrower are determined by the legal system of the country and its institutions governing bankruptcy. When a sovereign borrower defaults on a debt contract, however, the availability of legal remedies is limited by the doctrine of sovereign immunity, which precludes a lawsuit against a sovereign without that sovereign’s consent. As applied to lawsuits within the sovereign’s own legal system, this doctrine is based on the intuitive idea, extending back at

187

# 2013 Elsevier Inc. All rights reserved.

188

20. THEORY OF SOVEREIGN DEBT AND DEFAULT

least to Hobbes’s Leviathan, that the agent that makes the laws is not bound by those laws. As a practical matter, this limits the ability of the sovereign’s creditors to seek enforcement of a contract through the courts of the sovereign country itself. With regard to foreign borrowing, this right has typically been extended to foreign governments on the basis of international comity among nations, which as a consequence has limited the ability of foreign creditors to seek redress through their own and other countries’ courts. Over time, this so-called absolute doctrine of sovereign immunity has been weakened. In response to increased government participation in commercial activities in the postwar period, driven in part by the rise of socialist and communist countries, a more restrictive doctrine of sovereign immunity was adopted. Codified in the United States with the passage of the Foreign Sovereign Immunity Act of 1976, and in the United Kingdom by the State Immunity Act of 1978, the restrictive doctrine recognizes the immunity of a sovereign with regard to acts of state but not with respect to its private acts including its commercial activities. With debt issuance widely recognized as a commercial act, foreign creditors now have the ability to bring suit against a sovereign in default on its debts at least in their own and other foreign jurisdictions. However, this ability is of value only to the extent to which the assets of the sovereign can be attached, and a number of recent court cases have suggested that it is difficult to seize the small stocks of assets held abroad by the average debtor nation. A particularly well-known case concerns the mostly unsuccessful efforts of the Swiss company Noga to enforce contracts with Russia by seizing embassy bank accounts, Russian properties in France, naval ships, fighter jets, uranium shipments, and fine art (see the discussion in Wright, 2001). In the absence of legal remedies for private creditors, a sovereign debtor in default might be punished directly by a creditor country government. Mitchener and Weidenmier (2010), for example, present evidence suggesting that capital market participants viewed the threat of military intervention by creditor country governments as an effective deterrent against default by some countries throughout history. However, these findings are controversial. For example, Tomz (2007) in his study of sovereign borrowing across three centuries finds little evidence for the use of threatened military intervention to support repayment of debt, although it might have been used to protect the interests of foreign direct investors. Whether or not such punishments were used in the past, there is widespread agreement that they are not significant today. In the absence of legal remedies or direct punishments by creditor country governments, why do countries ever repay their debts? The costs associated with default are both hard to observe and difficult to quantify. As a consequence, there is a continuing debate about the

relative importance of the various costs that we survey in this section.

Restrictions on Financial Market Access Market participants commonly refer to the loss of normal financial market access as the primary consequence of a country’s decision to default. There are at least three reasons why access to financial markets might be lost or restricted after a default. One approach emphasizes the role of legal sanctions in blocking credit market access. As noted above, the ability to seize the assets of a sovereign is limited by the fact that most of these assets are not held in creditor jurisdictions. However, one asset that can be seized is the funds associated with servicing new loans to the country, which inevitably flow through creditor country jurisdictions. This approach has been adopted in recent litigation concerning sovereign default, including the well-known case of Elliott Associates versus Peru (see Pitchford and Wright, 2012 for details). If the funds servicing new loans can be seized, creditors will be deterred from making new loans, and hence the country will be effectively cut off from credit markets. Models with this feature include Benjamin and Wright (2008) and Pitchford and Wright (2012). A second approach, emphasized in early work by Eaton and Gersovitz (1981), focused on threats by creditors to retaliate against a country in default by denying them access to new credit. Exclusion from financial markets, however, leaves unexploited potential gains from trade in financial assets. To the extent that creditors can gain by lending to, or taking deposits from, a country in default, they might be tempted to deviate from a retaliatory punishment. That is, there is a sense in which a threat to exclude a country from financial markets may not be credible. Bulow and Rogoff (1989) demonstrated a particularly strong version of this argument by establishing conditions under which a country could default, take the payments it would have made to foreign creditors, and invest them with foreign financial institutions to generate a higher level of welfare than they could obtain from future borrowing. That is, to avoid the costs of default, a country need only be able to save abroad using a rich-enough menu of assets; it need not have future access to borrowing, and hence the threat of exclusion from future borrowing is not sufficient to enforce repayments of debts. A large literature has established the limits of the Bulow–Rogoff critique of retaliatory punishments. Kletzer and Wright (2000) showed that if creditors cannot commit to repaying deposits, the threat of exclusion from financial markets can work to support borrowing. Wright (2001) shows that even if a group of competitive creditors (in the sense of earning zero profits) can commit to repaying deposits, they can coordinate to credibly

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

WHY DO COUNTRIES BORROW SO MUCH?

exclude a defaulter from access to credit markets. Amador (2004) adds that leaders who care about the long-run ability of a country to borrow, but are concerned about the ability of rival political parties to spend any foreign savings by the country in the short run, have no incentive to use these savings opportunities and hence have no incentive to default. An alternative reason for the loss of credit market access following a default is that the decision to default reveals something about the country’s credit worthiness leading creditors to reduce or cut off lending to them. We refer to this cost of default as the loss of a country’s reputation, noting that the literature often uses the term to also describe retaliatory punishments for default. For example, if there is incomplete information about the gains from sovereign borrowing – perhaps because the country’s value of future lending or the costs of default are unknown – a default will lead foreigners to infer that the country is a ‘bad type.’ Future loans will not be forthcoming because creditors believe those loans will lose money and not as the result of a coordinated retaliatory embargo on loans (for a version of this approach, and a survey of related approaches, see Tomz, 2007). Other variants of this approach postulate that default reveals information about underlying investment opportunities in the country or the likelihood that the country will cooperate in other areas such as diplomatic relations. The empirical finding that countries in default have issued very little debt (Tomz, 2007) has led most quantitative theoretical work on sovereign debt to focus on the loss of normal market access as being one of the main costs of default.

Domestic Costs of Default A final class of costs of default identified by the literature concern the impact of a default on the domestic economy and political system of the country. There are numerous mechanisms through which this might occur. As noted above, the primary benefit from default is that a country can keep resources that it would have otherwise paid to foreigners. If, however, a country cannot discriminate between debts owned by foreigners and debts owned by its own citizens, a default will impose costs on the country’s citizens. Broner et al. (2010) argue that secondary markets may serve to reallocate bond holdings in such a way as to deter default by making the costs of default fall primarily on domestic residents. A default may also impose direct costs of the economy of the defaulting country. For example, if default damages the domestic financial system by inducing a domestic banking crisis, domestic output will fall. Another mechanism through which the domestic economy may be affected by a default is through its effects on international trade. There is some empirical evidence that

189

countries in default experience a significant decline in foreign trade, which may indicate the imposition of trade sanctions, either explicitly or sub rosa, or the loss of access to trade credit facilities. Once again, this view remains controversial: other authors have argued that these trade declines are unrelated to the pattern of debt holdings and hence might be due to forces other than trade sanctions by creditor country governments. Although the source of such domestic costs remains controversial, the perception that such costs exist is sufficiently widely accepted that the quantitative theoretical literature typically combines an ad hoc output loss with the loss of market access when modeling the costs of default. We discuss this literature in the next section.

WHY DO COUNTRIES BORROW SO MUCH? The early literature on sovereign debt and default posited a number of mechanisms that might explain the existence of sovereign debt. More recently, a quantitative theoretic literature has arisen that evaluates the extent to which these mechanisms can explain the amount of sovereign lending that is observed. We begin with a description of the benchmark model of sovereign borrowing and default.

A Benchmark Model The benchmark model is due to Eaton and Gersovitz (1981). In its simplest version, the model captures the decisions of a small open economy that accesses international capital markets both to smooth its consumption and to shift its consumption profile forward in time. The sovereign borrows by issuing bonds into a market populated by a large number of risk-neutral creditors. The bonds are state noncontingent except for the possibility that the sovereign may choose not to repay its debts. Default disrupts the country’s future capital market access and the country’s domestic economy. Default, therefore, acts a form of costly insurance for the sovereign against adverse economic outcomes. Specifically, consider the problem of a country that has borrowed through the issuance of b zero-coupon bonds in the previous period. We formulate the problem of the country recursively. At the beginning of the period, the country observes the state of the world s, a Markov process that governs the country’s output y(s). The Markov structure allows us to write the problem of the country recursively. The first decision that country must make each period is whether or not to repay its debts. If we let V(b, s) denote the value function of the country at the start of a period, as a function of its debt b and state s, it must satisfy

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

190

20. THEORY OF SOVEREIGN DEBT AND DEFAULT

Vðb; sÞ ¼ maxfV D ðsÞ; V R ðb; sÞg D

Here, V (s) is the value the country gets if it defaults which, under the assumption that all debts are extinguished upon default, is a function only of the state of the economy. The default value function encodes the two primary costs of default. First, the country is assumed to lose market access for some period of time: in the simplest version of the model, it is assumed that the country loses access to capital markets for one period and each period thereafter has a probability p of being able to reaccess capital markets with zero debt. Second, the country experiences some economic dislocation while in default: the country receives a lower default level of output yd(s). Hence, the value function in default satisfies V D ðsÞ ¼ uðyd ðsÞÞ þ bE½ð1  pÞVð0; s0 Þ þ pV D ðs0 Þjs If the country repays its debts, it retains access to capital markets and is able to borrow again in the same period. Hence, its value function in the event of repayment is given by uðcÞ þ bE½Vðb0 ; s0 Þjs V R ðb; sÞ ¼ max 0 c;b

subject to c  qðb0 ; sÞb0  yðsÞ  b The budget constraint states that consumption, less revenues gained by issuing new debts b0 at a price q(b0 , s) that depends upon the level of debt issuance, can be no greater than the country’s income net of debt repayments this period. In deciding how much new debt to issue, the country takes into account the fact that an extra unit of debt reduces the price of all of the debt that it issues. The price of debt is determined by competition among risk neutral creditors who face an opportunity cost of their funds given by the world interest rate r that drives the expected profits from lending to zero. If we let p(b0 , s) denote the probability that the country defaults next period, given that they issue b0 bonds today and given today’s state s, the bond price satisfies qðb0 ; sÞ ¼

1  pðb0 ; sÞ 1þr

That is, movements in the price of bonds reflect one-for-one changes in default probabilities. The inverse of this bond price is the gross rate of interest on sovereign borrowing.

Evaluating and Extending the Benchmark Model Early papers, such as Arellano (2008), found that versions of the benchmark model were inconsistent with the data on emerging market sovereign borrowing in at least two crucial respects: levels of sovereign borrowing, and

the interest rates charged for such borrowing, were orders of magnitude too small. Moreover, although it is possible to increase the model’s predictions for borrowing by increasing the costs of default (i.e., lowering the probability of reentering markets p or increasing the output cost by reducing yd(s)/y(s) ), and although this has an ambiguous effect of default probabilities (and hence sovereign borrowing rates) in theory, in these calibrated models default probabilities always fell thus worsening the model’s predictions for interest rates on sovereign borrowing. This finding remained true even under quite different assumed processes for the country’s output. The model’s inability to match these facts is driven by three of its features. First, for reasonable values of the variance of output, once the level of borrowing reaches the point where the country would default tomorrow in some state of the world, only a small increase in borrowing is needed to make default occur in all states of the world. That is, there is a level of borrowing after which the probability of default rises quickly from zero to one. Second, default is very costly to creditors as they lose the entire value of their investment, and hence the interest rates required to compensate creditors for this default risk rise very quickly as default probabilities rise. Putting this together with the first point, there is a level of borrowing after which the interest rate on sovereign debt rises quickly from r to a prohibitive level. Third, from the perspective of the country, a marginal increase in borrowing that raises the probability of a future default is very costly as it raises the cost of inframarginal borrowing. Hence, the country chooses to borrow at a low-enough level that the probability of default remains close to zero and interest rates on their borrowing remain close to r. In response to these quantitative failures of the benchmark model, a substantial literature has arisen proposing various modifications of this basic framework aimed at matching the facts on borrowing levels and borrowing rates. Some authors reduce the extent to which default probabilities rise with borrowing by changing the way the penalty of default varies with output. For example, Arellano (2008) postulates that the cost of default, in terms of lost output, is greater the higher the level of output. As the cost of default rises with output, it takes relatively larger levels of borrowing to induce default in the best states of the world, and hence the probability of default tomorrow rises more slowly with borrowing levels. Other authors have presented models in which output costs with this form arise endogenously. Some authors have proposed modifications that weaken the link between the probability of default in the next period and the interest rates that prevail today. One way to weaken this link is to vary the value creditors

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

POLICY AND WELFARE

place on resources in different states correlated with default (Arellano, 2008). Another way to weaken this link is to introduce longer maturity debt into the model, so that bond prices and interest rates today depend also on the forces determining bond prices tomorrow including expectations of long-term future default probabilities. Yet another mechanism for reducing the sensitivity of interest rates to default probabilities is to reduce the cost of default to creditor by allowing for the positive recovery rates on defaulted sovereign debt which are observed in practice. Benjamin and Wright (2008) study bargaining between a sovereign in default and its creditors over the recovery rate and show that their model can support much larger levels of debt. Still other authors limit the extent to which the country takes into account the effect of its marginal borrowing on the average cost of borrowing. This might be because the costs of repayment and default will be borne by different agents (for example, a different political party) or because borrowing decisions are made by many small agents who individually have little effect on the borrowing rate and hence do not take this feedback into account. The new literature of quantitative theoretical models of sovereign debt is large and growing, and the above survey touches upon only a small fraction of the issues raised by that literature. Work continues on extending the benchmark model to include production and on endogenizing both the output cost and market exclusion penalties for default. Far less work has been done assessing the quantitative performance of models in which trade sanctions are the punishment for default or on understanding the forces that lead contracts to be incomplete so that default occurs in the first place, and we speculate that these will be active research areas in the future.

POLICY AND WELFARE The quantitative theoretical literature has shown that sovereign risk alone can act as a very severe constraint on the amount of sovereign debt that can be issued. In this section, we briefly discuss the potential for self-fulfilling debt crises to arise as a further limitation on the market for sovereign debt, before turning to two policy debates that have recently been studied by the theoretical literature: the desirability of issuing longer maturity debts and the need for a mechanism to coordinate creditors in negotiations to restructure defaulted debts. Before proceeding, we stress two important caveats. First, an understanding of the costs of default, and hence of the incentives of a country to borrow appropriately and avoid default, is essential to any discussion of policy and welfare. For without an understanding of these incentives, any proposed policy for changing the market

191

for sovereign debt may be either infeasible or undesirable for the very same reasons that prevented market institutions from implementing the change in the first place. As a consequence of the ongoing debate as to the precise costs of default, policy proposals that rely on a specific interpretation of these costs should be treated with caution. Second, it is important to acknowledge the welfare trade-off between minimizing the costs of default ex post, with the need to give countries an incentive to borrow appropriately and avoid default ex ante.

Rollover Risk and Self-Fulfilling Debt Crises The possibility that debt crises could be self-fulfilling has been acknowledged for a long time and derives from the following logic: If investors believe a default to be an unlikely, they will demand a small default risk premium, and debtors, faced with a low interest rate, will have little incentive to default. Conversely, if investors think a default is likely, they will demand higher interest rates perversely giving the country a greater incentive to default. The possibility of such a self-fulfilling debt crisis is further strengthened when a country has a large amount of short-term debt falling due that it would like to roll over: that is, it would like to repay these debts by issuing new debts. In such a world, a debtor may be willing to repay that portion of the debt that is falling due if it knows it can roll over these debts by issuing new ones and unwilling otherwise. As a result, if creditors believe that other creditors will not purchase new debt issued by the sovereign, they themselves will not purchase these debts leading to a self-confirming default. A number of authors have argued that such selffulfilling rollover risk played an important role in the Mexican crisis of 1994–95, and a model of this phenomenon has been presented by Cole and Kehoe (2000). One institutional response to this problem is to use loan contracts that allow potential creditors to commit to purchasing bonds only on the condition that other investors also participate. Another potential policy implication of these models is that sovereign countries should avoid a reliance on short-term debt and should stagger longterm contracts so as to minimize the amount of debt that has to be rolled over at any one point in time.

Debt Dilution and the Maturity of Sovereign Debts Work on self-fulfilling debt crises has pointed to a possible over reliance on short-term borrowing by developing countries, without explaining why short-term lending is so predominant in the first place. One explanation for inefficiently short maturities lies in the lack of an explicit and enforceable seniority structure among

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

192

20. THEORY OF SOVEREIGN DEBT AND DEFAULT

sovereign debts. For example, if there is no seniority so that all debt holders expect to be treated equally in the event of a debt restructuring, and if the total amount to be paid to creditors is relatively fixed, new debt issues have the effect of reducing the amount that existing debt holders expect to recover and hence reduce the value of existing debt claims. This phenomenon, a form of the ‘debt dilution’ problem studied with regard to domestic corporate debt by Fama and Miller (1972), has been applied to sovereign debt by a number of recent authors. Bolton and Jeanne (2009), for example, show how, in the absence of an explicit seniority structure of lending, early creditors may attempt to make their debt implicitly senior by making it more difficult to restructure in the event of a default. They also show that this can lead to a debt structure that is excessively difficult to restructure ex post. These theoretical papers have contributed to a policy debate on the desirability of a procedure to enforce seniority in the event of a default to solve these problems. However, the precise details as to how such a system could be implemented, given the nature of sovereign risk and the difficulty of enforcing any sovereign debt contract, are still unclear.

More recently, in the aftermath of successful litigation by minority creditors against sovereigns in default such as the Elliott Associates case studied above, recent research has focused on the incentive of some creditors to engage in what has been termed strategic holdout in which a subset of creditors does not participate in a restructuring agreement in order to engage in later litigation. If such litigation is able to hold up the restructuring, such creditors might be able to extract more generous terms from the creditor, which is likely to slow down the restructuring process more generally. Formal models of the strategic holdout incentive have been developed by Pitchford and Wright (2012) who use the model to study the effect of changes in the contractual form of sovereign debts, including the introduction of collective action clauses in international bonds designed to bind holdout creditors to accept a majority settlement. They find that the implementation of such clauses reduces the cost of default ex post but may nonetheless raise the welfare of borrowing countries ex ante despite these adverse effects on the incentive of a country to default by reducing the socially wasteful costs of default.

CONCLUSION Collective Action Problems in Debt Restructuring Both of the above sets of reforms are associated with particular problems in the way in which creditors coordinate to purchase bonds or extend loans. There also exists a substantial literature advocating mechanisms to coordinate creditors after a default as debts are restructured. During the debt crisis of the 1980s, much of the focus of this literature was on the possibility of free riding by creditors in regard to the debt overhang problem, which is modeled by Krugman (1988) among others. A debt overhang refers to a situation in which a country’s debt level is too large to be repaid in full, which can lead to suboptimal decisions by both the country and its creditors. From the perspective of the country, if creditors are able to extract the bulk of any future increase in revenues, the country will have no incentive to make investments that increase its future income and hence also the value of the creditors claims. This problem can be removed if creditors write down their debt to the level where the country retains enough of the extra income to be persuaded to make the investment but still leaves creditors with a more valuable settlement. However, if creditors cannot coordinate in writing down their debts, then individual creditors have an incentive to free ride on the write downs of other creditors. This has been interpreted as an argument in favor of some sort of centralized debt restructuring mechanism.

This chapter has reviewed the theoretical literature on the role of sovereign risk on the market for sovereign debt and hence as a constraint on the process of financial globalization more generally. This literature is far from settled: there remains substantial disagreement on the precise costs faced by a country that decides to default, and although there is widespread agreement as to form the benchmark model of sovereign borrowing and default will take, existing versions of that benchmark model are at variance with the data in a number of important respects. As a consequence, the profession is unable to predict with confidence the effects of implementing one of the many different proposals for reforming the process of sovereign debt restructuring that circulated by policymakers in recent years. Also as a consequence, the topic remains an exciting one for researchers. Theoretical work toward formulating a benchmark model that is consistent with the facts about sovereign borrowing and that can be used as a vehicle for assessing the positive impact of alternative policy proposals is expected to pay large dividends in the near future. Likewise, there remains a great need for convincing empirical work to discriminate between alternative theories of the costs of default as well as quantify their relative importance. Progress toward this empirical goal is surveyed in other chapters in this volume.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

CONCLUSION

SEE ALSO Theoretical Perspectives on Financial Globalization: Foreign Currency Debt.

Glossary Collective action clauses Clauses in bond contracts allowing a majority of bondholders to change the payment terms of those bonds. Often used as part of a sovereign debt restructuring. Debt dilution The reduction in value of preexisting debts caused by the issuance of new debts. Debt overhang A situation in which the indebtedness of a country is so large that it prevents the country from undertaking profitable investments. Doctrine of sovereign immunity The legal doctrine that the government of a sovereign country is not bound by the laws that it itself creates. Such immunity has traditionally been extended to the governments of friendly foreign nations. Sovereign debt The debts issued by, or guaranteed by, a sovereign country. Sovereign debt restructuring The process by which a sovereign country in default, or contemplating default, alters the timing and/or quantity of its existing stock of debt. Sovereign default Narrowly, a violation of the contractual terms of a sovereign debt such as a failure to make a payment within the specified grace period. Often used to describe events in which creditors are coerced into accepting a reduction in the value of their debts even when contractual terms are not breached. Strategic holdout In sovereign debt restructuring, a situation in which some creditors do not participate in a proposed sovereign debt restructuring in the expectation of more generous terms in the future.

193

Arellano, C., 2008. Default risk and income fluctuations in emerging economies. American Economic Review 98 (3), 690–712. Benjamin, D., Wright, M.L.J., 2008. Recovery Before Redemption: A Theory of Delays in Sovereign Debt Renegotiations. University of California at Los Angeles. Unpublished Paper. Bolton, P., Jeanne, O., 2009. Structuring and restructuring sovereign debt: The role of seniority. Review of Economic Studies 76 (3), 879–902. Broner, F., Martin, A., Ventura, J., 2010. Sovereign risk and secondary markets. American Economic Review 100, 1523–1555. Bulow, J., Rogoff, K., 1989. Sovereign debt: Is to forgive to forget? American Economic Review 79 (1), 43–50. Cole, H.L., Kehoe, T.J., 2000. Self-fulfilling debt crises. Review of Economic Studies 67 (1), 91–116. Eaton, J., Gersovitz, M., 1981. Debt with potential repudiation: Theoretical and empirical analysis. Review of Economic Studies 48 (2), 289–309. Fama, E.F., Miller, M.H., 1972. The Theory of Finance. Holt, Rinehart and Winston, New York. Kletzer, K.M., Wright, B.D., 2000. Sovereign debt as intertemporal barter. American Economic Review 90 (3), 621–639. Krugman, P., 1988. Financing vs. forgiving a debt overhang. Journal of Development Economics 29 (3), 253–268. Mitchener, K.J., Weidenmier, M.D., 2010. Supersanctions and sovereign debt repayment. Journal of International Money and Finance 29, 19–36. Pitchford, R., Wright, M.L.J., 2012. Holdout creditors in sovereign debt restructuring: A theory of negotiation in a weak contractual environment. Review of Economic Studies 79, 812–837. Tomz, M., 2007. Reputation and International Cooperation: Sovereign Debt Across Three Centuries. Princeton University Press, Princeton. Wright, M.L.J., 2001. Reputations and Sovereign Debt. Stanford University, Stanford, Working Paper.

Further Reading Amador, M., 2004. A Political Economy Model of Debt Repayment. Stanford University, Stanford, Working Paper.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

Intentionally left as blank

C H A P T E R

21 Tax Systems and Capital Mobility M.P. Devereux, C. Fuest Oxford University Centre for Business Taxation, Oxford, UK O U T L I N E Introduction: Implications of Globalization for Tax Systems 195 National Taxation and International Mobility Residence-Based Taxation Destination-Based Taxation

196 197 197

INTRODUCTION: IMPLICATIONS OF GLOBALIZATION FOR TAX SYSTEMS The main purpose of taxation is to raise revenue. But the adverse impact of taxation on economic efficiency and growth should be minimized and the distribution of the tax burden should be in line with principles of distributional justice. In addition, taxes may be used to discourage certain undesirable activities such as environmental pollution. Last but not least, the tax system should be administratively simple. As these are partly conflicting aims, tax policy involves various trade-offs. For instance, from a distributional perspective, progressive income taxes are widely seen to be advantageous because taxpayers with higher incomes and, therefore, presumably a higher ‘ability to pay,’ are taxed at higher rates. But from an efficiency perspective, higher marginal tax rates are problematic because they may discourage labor supply or investment and induce tax evasion. Indirect taxes like the value-added tax (VAT) or excise taxes cannot easily take into account ability to pay, but they make tax evasion more difficult and their efficiency costs can be smaller. Therefore, the weight of direct and indirect taxes in the overall tax system involves trade-offs between different economic objectives. As economic integration proceeds and international economic activity gains importance, the various tradeoffs in the design of tax systems change. In a closed economy, the tax system affects domestic households and

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00021-0

Source-Based Taxation

198

International Tax Coordination

200

Summary and Conclusions Further Reading

200 200

firms and their interaction in domestic markets only. In an open economy, a number of additional issues arise. First, domestic assets may now be owned by foreigners, so that taxes on income generated domestically may fall on foreign residents. At the same time, domestic residents may have foreign source income and pay taxes abroad. Second, domestic as well as foreign taxes may affect the decision of firms and households regarding where to consume, work, or invest. The implications of internationalization are particularly visible in the corporate world, where multinational enterprises become increasingly important. Often, these firms operate in many countries and are owned by residents of many countries. They respond to international tax differences by shifting production from one country to another or by concentrating intangible assets and equity financing in low-tax countries in order to reduce their tax burden. Capital is usually more mobile than people, but some people may still react to taxation by changing their country of residence or by commuting across borders to work abroad. They may engage in cross-border shopping if there are tax differences which can be exploited. These issues have far-reaching implications for national tax policies. Given that capital is more mobile than labor, incentives arise to treat income arising from these two sources differently. Countries with high taxes on mobile economic activities face the risk of losing investment and jobs, so that pressures arise to cut taxes or to shift the tax burden to less mobile activities. At the same

195

# 2013 Elsevier Inc. All rights reserved.

196

21. TAX SYSTEMS AND CAPITAL MOBILITY

time, the burden of domestic taxes may partly fall on foreigners. This creates incentives to increases taxes. In many OECD countries, the need to adjust domestic tax systems to the increasingly globalized economy is seen as a key factor driving domestic tax reforms. At the same time, the view is widespread that economic globalization needs to be matched by more political coordination at the international level. This includes, for instance, proposals to limit tax competition by introducing minimum tax rates or new rules for the exchange of information between tax authorities of different countries.

NATIONAL TAXATION AND INTERNATIONAL MOBILITY Border-crossing flows of income, and exports and imports of goods and services, can be taxed in different ways, and one important dimension is the location of taxation. Broadly, the location can be based on residence, source, or destination. To understand these, and the rationale for using each, consider a simple example. An individual resident in the United Kingdom owns a company which has a legal residence in France. This company has a wholly owned subsidiary which is a company resident in Italy. The Italian company uses labor and capital inputs to produce a final good. This good is exported to Germany where it is purchased by a German consumer. The price paid by the consumer is income for the Italian company which generates a profit for that company. The Italian company pays the profit as a dividend to the French parent company, which in turn pays it as a dividend to the UK owner. In this example, it is common to refer to the United Kingdom as the country of residence of the shareholder, France as the country of residence of the multinational company, Italy as the source country where economic production takes place, and Germany as the destination country where the final consumption takes place. It turns out that each of these locations would typically be a place of taxation for one form of tax or another. Personal tax on capital income is generally applied under the residence principle, which means that the UK shareholder would be taxed in the United Kingdom on his/her worldwide capital income. Under the pure residence principle, individual investors cannot exploit international differences in taxation by moving their investment from one country to another because the tax paid does not depend on the location of the investment but only on the residence country of the investor. This also implies that, as long as individual investors are immobile, countries cannot attract investment from other countries by cutting their residence-based taxes.

At the other extreme, taxes that are generally considered to be levied on consumption of goods and services – including VAT, sales taxes, and excise taxes like the gasoline tax or taxes on tobacco and alcohol – are usually levied according to the destination principle. In this example, this means that Germany would tax the expenditure of the German consumer. The mechanism by which this is achieved under VAT is that exports are exempt from tax in the exporting country (Italy), but are subject to tax in the importing country (Germany). This implies that a VAT increase in one country would not increase the production costs of firms located in that country, relative to firms located in other countries, and so countries cannot increase the competitiveness of their industries by cutting destination-based taxes. It also implies that all firms competing in, say, the German market would face the same tax rate, and so competition between producers is not distorted by destination-based taxation. The intermediate case is the source principle, which is the most common form of taxation of corporate profit. This is harder to define, but is essentially the place where productive activity takes place. In the example, the Italian subsidiary would be taxed on its profits earned in Italy (including the income generated by sales in Germany). As the tax burden now depends on the location of the investment, companies can exploit tax differences by relocating from high-tax countries to low-tax countries, and individual countries may attempt to attract investment from other countries by undercutting their neighbors. The remaining possible place of taxation in the example discussed is France. Until recently, several countries applied a residence principle to companies as well as to individuals. In this example, this would mean that France would tax the worldwide profit of the parent company located in France, even if all of the productive activity took place elsewhere. In analyzing these concepts, the economics literature has generally considered the residence and source principles to be the two relevant options for taxing capital income, including corporate profit. And it has considered the source or destination principles to be the relevant options for taxing consumption (although note that this literature, somewhat confusingly, refers to tax in the country of production as the origin principle). There is a natural rationale for a residence-based approach, as governments seek to tax the worldwide income of their residents, who take advantage of the domestic public goods and services provided. It may also seem natural to extend this to a residence-based approach to tax the worldwide income of domestic companies, if corporation tax is seen as a proxy for a tax on the income of the domestic shareholders of the company. But this would only be convincing if all of the company’s shareholders were resident in the same country as the

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

NATIONAL TAXATION AND INTERNATIONAL MOBILITY

company. Of course, this is not generally true, and so the rationale for taxing the worldwide income of companies is weak. As of 2010, the only major country operating a tax system along these lines is the United States. But then the rationale for taxing corporate profit on a source basis is not very convincing either. The tax may be partly borne by the owners of the company, who may not be residents, and who do not benefit from domestic public goods. However, there may of course be political attractions in attempting to tax nonresidents. But, as discussed below, it is also likely that at least part of the tax is passed onto domestic residents in the form of lower wages and higher prices: in this case, it is typically more efficient to tax the domestic residents directly. And in any case, the distinction between taxes on income and consumption may seem natural, but is in fact largely arbitrary. To see this, consider a VAT. This is, as its name suggests, as tax on the value added by a company, defined as the sales proceeds of its output less the costs of its inputs, including capital expenditure. But value added is also equal to labor income plus the economic profit earned on behalf of the shareholders. So VAT could be reproduced with two taxes on income. First, there would be a tax on the labor income earned by the company’s employees. Second, there would be a tax on the profit earned by the firm for its owners (strictly this would be a cash flow tax on economic rent). If these two taxes were levied at the same rate as the VAT, then they would be identical, apart from possibly the administrative and compliance costs of their collection. This equivalence has implications for the location of taxation of capital income and consumption. For example, in principle, at least, it would be possible to tax capital income on a destination basis. These locations of tax are now considered in a little more detail.

Residence-Based Taxation The income tax laws of most countries imply that, in principle, domestic residents are subject to income taxation with their worldwide income. But for a number of reasons, residence-based taxation is not implemented stringently. First, taxes paid abroad are typically credited against domestic taxes. If foreign taxes were lower than the home country tax, then this would result in the overall tax payment being the same as if the income had only been taxed at home, although clearly the recipients of the tax revenue would be different. But if foreign tax rates were higher, then typically the credit offered by the home country would be limited to the home country’s tax rate, implying that the overall tax rate may vary

197

depending on where the capital is invested. Second, and especially for corporation tax, many countries that apply a residence-based system actually only tax the income when it is repatriated to the home country. This means that income arising in a low-taxed country can be sheltered from domestic tax until it is repatriated. Third, there is often little or no integration between taxation at the personal level and taxation at the corporate level. For instance, if individuals receive dividends from a domestic corporation, they also receive a tax credit. But dividends received from foreign corporations (or from domestic corporations, but funded from foreign profits) may not receive this credit.1 Fourth, it is generally difficult to enforce residence-based taxes on portfolio income because it is difficult for national tax authorities to monitor foreign bank accounts and other assets held by domestic residents. Recently, efforts have been made to improve and extend information exchange between tax authorities, but whether they will be effective is an open question.

Destination-Based Taxation Under the destination principle, goods are taxed in the country where they are ultimately consumed. As noted above, this is generally achieved in practice by exempting exports from tax, but applying a tax to imports. This requires some record keeping of trade, which has in the past been performed at the border of each country. However, as the EU requires its members to have a VAT, and as customs controls between many countries have now been abolished, identifying traded goods is achieved through invoicing. A further problem arises with cross-border shopping. If the tax rate is lower in the neighboring country, then it may be worthwhile for individuals crossing the border to make purchases at the lower tax rate. This is an exception to the pure destination principle in that goods purchased for one’s own consumption in a neighboring country can be taxed at the rate of that country. However, if the goods are purchased for resale at home, then in principle the home country rate should apply. The cross-border shopping problem arises essentially because individuals are not completely immobile. However, as noted above, a destination-based tax has many attractive properties. It generally does not distort decisions as to the location of production. It only affects the location of sale in the case of cross-border shopping, or if individuals actually move their country of residence to take advantage of a lower tax rate. For these reasons, indirect taxes which are levied according to the destination principle are an attractive revenue-raising

1

Within the European Union, this has been deemed to be discriminatory by the European Court of Justice, a view which has resulted in considerable change in the nature of integration systems.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

198

21. TAX SYSTEMS AND CAPITAL MOBILITY

7 6.5 6 5.5 5 4.5 4 3.5

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

3

Large OECD countries

FIGURE 21.1

Small OECD countries

Revenue from general taxes on goods and service in percentage of gross domestic product (GDP). Source: OECD Revenue

Statistics.

instrument in open economies. Figure 21.1 shows the trend in tax revenues from general taxes on goods and services, which mainly include VATs and general sales taxes (but excludes tariffs and excise taxes), since 1970. In general, revenue from these taxes has gained importance over time. Small countries tend to raise more revenue through this type of taxes than large countries. One explanation for this may be that small countries tend to be more open, and hence more exposed, to tax competition over source-based taxes, which is discussed below. This would mean that small countries are forced to rely more on destination-based consumption taxes, which are less vulnerable to tax competition. However, this explanation would also imply that smaller and more open economies should collect less revenue from corporate taxation, which, as noted below, does not appear to be the case.

Source-Based Taxation Source-based capital income taxation, especially in the form of corporate income taxes, is the dominant form of capital income taxation in open economies. Taxing capital income at source, however, faces a number of difficulties. First, in cases where firms operate in more than one country, it is frequently difficult – if not impossible – to determine the ‘true’ geographical distribution of the company’s overall profit. If there are tax rate differences across countries, companies have incentives to arrange

their financial and legal structures and to price intrafirm transactions so that profits are recorded in low-tax jurisdictions. Therefore, source-based taxation can easily lead to conflicts between taxpayers and tax authorities as well as between tax authorities of different countries. A second drawback of source-based taxes on capital income is that these taxes induce capital to flow to other countries. If the supply of capital is fully elastic, the aftertax return to capital is fixed, so that a source-based capital income tax drives the before-tax return up to the point where the tax burden is fully shifted to immobile factors of production. Therefore, source-based taxes on capital income are unattractive for small open economies. This suggests that source-based capital income taxes should be eroded as economic integration proceeds. The development of corporate income tax rates over time seems to support this view. Figure 21.2 shows that the average corporate income tax rate has declined dramatically over the last two decades. Large countries have maintained higher rates than small countries, but the general trend has been toward lower tax rates. However, considering statutory tax rates only overstates the tendency toward corporate tax rate reductions. Interestingly, as Figure 21.3 shows, the fall in corporate tax rates has not led to an equivalent decline in corporate income tax revenues. Corporate income tax revenues have been relatively stable over the last few decades. Figure 21.3 distinguishes between large and small OECD countries.2 While the share of corporate tax revenue in GDP is more or less constant for the group

2

The group of large countries includes France, Germany, Italy, Japan, the United Kingdom, and the United States. The group of small countries includes Austria, Belgium, Australia, Canada, Denmark, Finland, Greece, Ireland, Luxemburg, The Netherlands, New Zealand, Norway, Spain, Sweden, and Switzerland. For each group, we report the unweighted averages.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

199

NATIONAL TAXATION AND INTERNATIONAL MOBILITY

FIGURE 21.2 Corporate income tax rates in the OECD 1986–2008. Source: Oxford University Centre for Business Taxation.

49 44 39 34 29 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

24

OECD Average

Large Countries (G7)

Small Countries

4.5 4 3.5 3 2.5 2 1.5

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

1

Large OECD Countries

FIGURE 21.3

Small OECD Countries

OECD Total

Corporate tax revenue in percentage of GDP. Source: OECD Revenue Statistics.

of large countries, it has even increased in the smaller countries, even though they have cut their tax rates more than the large countries. There are several potential explanations for the observed stability of corporate tax revenue. First, many countries have combined cuts in statutory tax rates with reductions in depreciation allowance, loss offset restrictions, and limitations for the deductibility of interest payments and other expenses. This broadening of the tax base stabilizes the effective tax burden on corporate income. Second, it may be the case that corporate profits have increased, possibly as a consequence of new economic opportunities offered by globalization. Third, as corporate income tax rates decline, more businesses may incorporate, so that the corporate sector expands. Moreover, within existing corporations, income may be shifted from the personal income tax sphere to the corporate tax sphere. This would suggest that the stability of corporate income tax revenue comes at the cost of a decline in revenue from personal income taxation.

Irrespective of the observed stability of corporate tax revenue, the downward trend in corporate tax rates raises the question of whether corporate taxation is sustainable in the long run. If the current trend toward lower rates continues, tax rates will approach zero within a decade or two. While such an outcome cannot be excluded, there are some factors which may prevent this erosion. First, there are economic rents which are bound to certain locations. For instance, countries which offer access to large markets or other agglomeration advantages are attractive as locations for investment. Countries have incentives to use corporation taxes to tax these rents. Taxes on economic rents may also be borne by foreign resident owners of domestic companies. A second explanation for the survival of sourcebased capital income taxes is that it is sometimes difficult to distinguish between labor and capital income. For instance, in small and medium-sized firms, where the manager is frequently identical with the owner of the firm, it is relatively easy to shift income from the

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

200

21. TAX SYSTEMS AND CAPITAL MOBILITY

profit tax base to the personal income tax base. One way of doing so is to adjust the manager’s wage. Given this, corporate taxes may serve as a backstop to the personal income tax. Whether these factors are sufficient to bring the downward trend in corporate tax rates to an end remains to be seen.

INTERNATIONAL TAX COORDINATION The general increase in international economic activity is partly a result of policy changes such as the liberalization of international capital movements and the reduction of barriers to trade. While some of these policy changes have been implemented unilaterally, international policy coordination has played an important part. In the area of taxation, the most prominent example for this type of policy coordination is probably the reduction of import tariffs. But there are various other tax policy areas where international coordination plays an important role. Somewhat ironically, within the European Union (EU), the strongest form of international coordination is in VAT, where agreements on minimum and maximum rates have been in force for some time, and where the minimum rate is 15%. Yet the benefits of such agreements are limited – VAT is a destination-based tax which is generally unaffected by the growth of capital mobility. Harmonized would end the incentive for cross-border shopping, but the mobility of consumers is much smaller than the mobility of capital. Here therefore two issues relevant for the taxation of corporate profit are discussed, which clearly depends in important ways on the mobility of capital. Several proposals have been made to introduce minimum corporate income tax rates in the EU. For instance, in 1992, the European Commission proposed to introduce an EU-wide minimum corporate tax rate of 30% and a maximum rate of 40%. In addition, the proposal included common rules for the determination of the corporate tax base. However, these measures did not find the unanimous support of the EU member states which would be required for this type of coordination. More recently, the European Commission has made another proposal for coordination of corporation tax in the EU: the introduction of a Common Consolidated Corporate Tax Base (CCCTB) with formula apportionment. Under the current tax system, firms operating in the European Internal Market calculate the profits generated in each country separately, according to the national tax-accounting rules. Under the CCCTB, companies would use a common set of rules to determine their EU-wide profit. The right to tax this profit would then

be allocated to individual member states on the basis of a formula. For instance, if a multinational firm has two thirds of its labor force, assets, and sales in France and one-third in the United Kingdom, two thirds of the firm’s overall profits would be taxed by France, at the French tax rate, and one-third by the United Kingdom, at the UK tax rate. The advantage of this proposal is that the complications caused by having to determine taxable profits for each EU member country separately, according to the different national tax systems, can be avoided. But the new system would give rise to new complexities, including those regarding the transition from the old to the new system. In the light of these issues, support of the member states for the new proposal has so far been limited.

SUMMARY AND CONCLUSIONS The growing importance of international economic activity has far-reaching implications for national tax systems. Tax policy has to take into account the impact of tax structures on border-crossing flows of capital, goods, and people. The view is widespread that the international mobility of tax bases may erode the taxing powers of governments, in particular in the area of corporate income taxation. The strong downward trend in corporate tax rates seems to confirm this. However, revenue from corporate taxation is surprisingly stable. Whether this will continue to be the case if tax rates continue to drop remains an open question, though. Other taxes seem to be less vulnerable to the pressures of tax competition. Destination-based taxes, in particular, broad-based consumption taxes like the VAT, have gained importance during the last decades. The pressures of tax competition and various other issues including double taxation or tax avoidance and evasion call for more international tax coordination. But reaching agreement on tax coordination has proved difficult so far. Given this, it is likely that, at least in the near future, the challenges of globalization for taxation will primarily have to be met by national tax policies and tax reforms.

Further Reading Auerbach, A., Devereux, M.P., Simpson, H., 2010. Taxing corporate income. In: Mirrlees, J. et al., (Ed.), Dimensions of Tax Design. Oxford University Press, Oxford. Fuest, C., Huber, B., Mintz, J., 2005. Capital mobility and tax competition. Foundations and Trends in Microeconomics 1, 1–62. Hines, J.R., Summers, L.H., 2008. How globalization affects tax design. Tax Policy and the Economy 23, 123–157.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

C H A P T E R

22 Trade Costs and Home Bias N. Coeurdacier Sciences Po and CEPR, Paris, France O U T L I N E Introduction

201

The Equity and Consumption Home Biases: Facts and Figures The Equity Home Bias The Consumption Home Bias and the Size of Trade Costs Equity and Consumption Biases: Are they Empirically Related?

202 202 202 202

Why Investors Would Hold Different Equity Portfolios?

203

Home Bias in Equities and the Hedging of Real Exchange Rate Risk From Partial Equilibrium . . . . . . To General Equilibrium

204 204 204

INTRODUCTION Standard finance theory would predict that investors should hold a diversified portfolio of equities across the world if capital is mobile across borders. Because foreign equities provide great diversification opportunities, falling barriers to international trade in assets over the last 20 years should have led investors across the world to rebalance their portfolio away from national assets toward foreign assets. The process of ‘financial globalization’ fostered by capital account liberalizations, electronic trading, increasing exchanges of information across borders, and falling transaction costs has certainly led to a large increase in cross-border asset trade. However, investors are still reluctant to reap the full benefits of international diversification and tend to hold a disproportionate share of local equities: the ‘home bias in equities.’ Since the seminal paper of French and Poterba (1991), the home bias in equities constitutes one of the

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00011-8

Are Equities Empirically a Good Hedge against Real Exchange Rate Fluctuations? Home Bias in Equities and the Hedging of Nontradable Risk The International Diversification Puzzle is Worse than we Think . . . Or Better Than We Think . . .

205 205 205 205

Trade Costs and Portfolio Home Bias: Alternative Stories The Role of Expropriation/Sovereign Risk The Role of Information and Behavioral Biases

206 206 206

Conclusion Acknowledgments References

206 207 207

major puzzles in international finance. Despite better financial integration, it has not decreased sizeably: in 2007, US investors still held more than 80% of domestic equities and the home bias in equities is observed in all developed countries. As people hold mostly local assets, they also consume mostly locally produced goods: the ‘home bias in consumption.’ There is now quite a consensus that the home bias in consumption can be explained by the presence of international trade costs understood in a broad sense. According to the survey of Anderson and Van Wincoop (2004), international trade costs are very large, roughly 70% of production costs. Trade costs encompass various barriers to international trade in goods such as transport costs, tariffs, and other policy induced restrictions to trade, and various barriers at the borders. In an influential contribution, Obstfeld and Rogoff (2000) argues that home bias in equities might also be due to shipping costs in international goods markets

201

# 2013 Elsevier Inc. All rights reserved.

202

22. TRADE COSTS AND HOME BIAS

rather than frictions in financial markets: people hold local assets because they mostly consume local goods due to the mere presence of trade costs. The home bias in equities would be the mirror of the home bias in consumption. In this chapter, we explore the validity of this argument from an empirical and theoretical perspective and describe the key mechanisms behind such a hypothesis. We will first review the extent of the equity and consumption home biases across countries and discuss the empirical evidence showing a link between the two biases. Then, we will discuss theoretically the possible channels through which trade costs in goods markets can affect portfolio decisions.

THE EQUITY AND CONSUMPTION HOME BIASES: FACTS AND FIGURES The Equity Home Bias The proportion of local equity in investors’ portfolios is not fully indicative of the intensity of the home bias. In a world with homogenous investors and perfectly integrated financial and goods markets, the proportion of local equities in investors’ portfolios should be equal to the share of the country considered in the world market capitalization. Hence, to illustrate the equity home bias, Table 22.1 shows the difference between the proportion of local equities in a country’s portfolio and the relative market capitalization of that country at the end of 2005. We measure the degree of home bias in country i as: share of foreign equities in country i equity holdings HBi ¼ 1  share of foreign equities in the world market portfolio

ð22:1Þ

where HBi is between 0 and 1; HBi ¼ 1 if full equity home bias and HBi ¼ 0 if no equity home bias. Data on portfolio holdings are from the Coordinated Portfolio Investment Survey held by the International Monetary Fund (IMF). Table 22.1 shows that all countries have a significant equity home bias, Netherlands having the lowest degree of home bias while Japan and Greece have the largest.

The Consumption Home Bias and the Size of Trade Costs While investors hold mostly local equities, consumers have a strong bias toward locally produced goods: the ‘consumption home bias.’ Looking into consumption baskets, countries are not very open to trade. In the United States, the openness to trade ratio measured by 1

the sum of exports and imports over gross domestic product (GDP) is only 26% over the period 2000–2007. Given that the US account for about a third of world production, one could expect in a workhorse model of trade with fully tradable differentiated products the US import and export to be about two thirds of their GDP. This would lead to an openness ratio higher than 120%! Not all output produced within an economy can be shipped across borders but even when focusing on manufacturing goods, the consumption home bias remains large. Obstfeld and Rogoff (2000) put forward trade costs as the main explanation for the consumption home bias: indeed, in presence of trade costs, households consume essentially locally produced goods as imported goods are relatively more expensive. In a survey, Anderson and van Wincoop (2004) provides an estimate of the tax equivalent of international trade costs for industrialized countries as large as 74%. This number breaks down as follows: 21% transportation costs (freight costs and tax equivalent of the time value of goods in transit) and 44% border related trade barriers (tariffs and other trade policies, language barrier, currency barrier, information cost, and security barrier).1

Equity and Consumption Biases: Are they Empirically Related? Obstfeld and Rogoff (2000) argue that trade costs in goods markets lead to both consumption and equity home bias. If this is true, then one should observe that both are related in the data: countries which are more open to trade should also have more internationally diversified portfolios. In other words, everything else being equal, countries with higher import (or export) shares should have larger stocks of foreign equities. This is exactly what the data tell us: looking at panel data of a cross-section of countries, there is a positive relationship between trade openness and foreign asset holdings (see van Wincoop and Warnock, 2010 for a discussion of the empirical evidence). Although one should be cautious with such evidence as causality is hard to infer from these regressions, this is an indication that trade costs might affect international portfolios. Another piece of evidence is provided when looking at bilateral data on trade in goods and asset holdings. In a gravity-like framework, Aviat and Coeurdacier (2007) and Lane and Milesi-Feretti (2008) show that country equity portfolios are strongly biased toward trading partners. Using instrumental variables, Aviat and Coeurdacier (2007) show that the causality goes essentially in one direction: reducing trade costs between countries enhances cross-border asset holdings.

This gives a total of 74%: 0.74 ¼ 1.21  1.44  1.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

203

WHY INVESTORS WOULD HOLD DIFFERENT EQUITY PORTFOLIOS?

TABLE 22.1

Home Bias in Equities in 2005

Source country

Domestic market in % of world market capitalization (1)

Share of portfolio in domestic equity in % (2)

Degree of home bias = HBi (3)

Australia

1.9

83.6

0.832

Austria

0.3

58.5

0.583

Belgium

0.7

49.8

0.494

Canada

3.5

76.6

0.757

Denmark

0.4

62.7

0.625

Finland

0.5

63.3

0.631

France

4.2

68.8

0.674

Germany

2.9

57.5

0.562

Greece

0.3

93.4

0.933

Italy

1.9

57.1

0.562

Japan

13.2

91.9

0.906

Netherlands

1.4

32.1

0.311

New Zealand

0.1

59.8

0.597

Norway

0.5

52

0.517

Portugal

0.2

77.8

0.777

Spain

2.3

86.3

0.859

Sweden

1

59.4

0.589

Switzerland

2.2

59.9

0.589

United Kingdom

7.3

65

0.622

82.2

0.700

70.58

0.70

United States Average

40.5 3.67

source: CPIS

These empirical observations are challenging for economic theory. We now turn to the potential theories linking trade costs in goods markets and portfolio decisions.

WHY INVESTORS WOULD HOLD DIFFERENT EQUITY PORTFOLIOS? While the purpose of this chapter is not to review extensively the various explanations proposed to explain the equity home bias puzzle, it is useful to understand the main reason why investors from different countries hold different portfolios before looking at the role of trade costs specifically. As stated above, in a world with perfect financial markets, homogenous investors should hold the same portfolio, the world market portfolio, thus diversifying optimally idiosyncratic national risks. Thus, the home bias in equities is seen as a failure of the standard diversification motive in portfolio theory.

However, one should be cautious with such a statement: investors across the world would hold the same portfolio, the market portfolio, only if they were homogenous. In reality, heterogeneity across investors from different countries leads to departure from the world market portfolio and potentially a bias toward national assets. Various sources of heterogeneity leading to equity home bias have been explored in the literature and one can distinguish among the following set of candidates: (i) hedging domestic risk, (ii) implicit and explicit costs of foreign investments (such as transaction costs, differences in tax treatments between national and foreign assets. . .), (iii) information asymmetries, (iv) corporate governance, transparency and expropriation risk, and (v) behavioral biases (such as ‘familiarity biases’). We will start with the first explanation since this is the main channel through which trade costs in goods markets interact with portfolio decisions. Later in the chapter, we will briefly discuss how information flows,

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

204

22. TRADE COSTS AND HOME BIAS

familiarity biases or expropriation risks can affect portfolio decisions in the presence of trade costs. We mean by hedging domestic risk, choosing the appropriate financial assets that insulate at best investors from local sources of risk that affect their income streams. The sources of domestic risk extensively developed in the literature are mostly of two kinds: • The presence of real exchange rate risk: prices of investors’ consumption goods are fluctuating which affects the purchasing power of their income. • The presence of nontradable income risk: investors have a part of their income (in the form of wages in particular) that cannot be traded in financial markets. In other words, because investors in different countries have different exposure both to real exchange rate risk and/or to nontradable income risk, they will hold different portfolios. We now turn to the interaction between trade costs in goods markets and the hedging of these domestic risks.

HOME BIAS IN EQUITIES AND THE HEDGING OF REAL EXCHANGE RATE RISK From Partial Equilibrium . . . As explained in trade costs generate some ‘home bias in consumption’ by increasing the price of imported goods compared to local goods. Hence, fluctuations in the relative price of local goods (over foreign goods), or equivalently fluctuations in the real exchange rate affect the purchasing power of investors across countries differently. This source of heterogeneity among investors can lead them to choose different portfolio of assets. In other words, households have the desire to build the appropriate portfolio to hedge against fluctuations in the real exchange rate. The optimal hedging of real exchange rate risk depends on two forces going in opposite direction: when local goods are more expensive, consumers need to generate more income in order to stabilize their purchasing power. On the other hand, since local goods are more expensive, households could be better off consuming when goods are cheaper. The dominating effect depends on how much households want to smooth their consumption (across states of nature), in other words how risk averse they are. For consumers sufficiently risk averse (with a relative risk aversion above unity as usually assumed), the former effect dominates and households want to increase their income when their consumption goods are more expensive. Thus, they build their portfolio by choosing assets with a high pay-off when local goods are expensive. This is at the heart of the potential

divergence of portfolios across investors in the partial equilibrium portfolio choice models with real exchange rate risk (see Adler and Dumas, 1983). The key issue is whether local equities are a good hedge against relative price (real exchange rate) fluctuations, that is, whether local equities have higher returns when local goods are (relatively) more expensive. If this is the case, then local investors should favor local equities.

. . . To General Equilibrium In general equilibrium, the exact main mechanism is at work but relative prices are determined endogenously in the model (rather than taken as given in partial equilibrium models). As recently emphasized by Coeurdacier (2009), standard general equilibrium models cast doubt on the ability of local equities to provide a good hedge against real exchange rate risk. Coeurdacier (2009) solves for equity portfolios in a two-country/two-good stochastic equilibrium model with trade costs in goods markets. Uncertainty in the model is driven by supply (endowment) shocks in both countries. Due to the presence of trade costs, the real exchange rate fluctuates following supply shocks, which in turn affect portfolio decisions. Unfortunately, trade costs do not turn to be helpful to solve for the equity home bias. The main intuition goes as follows: when local output is low (relative to foreign output), the price of local goods increases due to their scarcity and investors require a high return on their equity portfolio to stabilize their purchasing power. They would rather bias their portfolio toward foreign equities as foreign equities have a higher pay-off than local equities when local output is lower. Contrary to conventional wisdom, trade costs generate equity foreign bias and make the equity home bias puzzle worse than we think! Such a result relies on three important assumptions: First, as explained above investors must be sufficiently risk averse (they must have a relative risk aversion above unity as usually assumed). Second, a fall in the supply of local goods must lower the return of local equity. This is not always true as it depends on the response of relative prices. If consumers cannot easily substitute local and foreign goods (technically speaking if the elasticity of substitution between local and foreign goods is below unity), the fall in supply triggers a large increase in local goods prices, which can increase local equity returns. In that case, local equity returns are precisely high when prices of local goods are high. Investors would rather hold local equities. The response of relative prices depends on the elasticity of substitution between local and foreign products. While time series macro data estimating the response of trade to exchange rate changes suggests a low elasticity of substitution, between 0.5

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

HOME BIAS IN EQUITIES AND THE HEDGING OF NONTRADABLE RISK

205

and 1.5, bilateral trade data suggests a large elasticity, above 5 for most sectors (see Coeurdacier, 2009 for a more detailed discussion). The parameter uncertainty makes it hard to get a conclusive answer from this class of models. Third, output fluctuations in this class of models are driven by supply shocks. In the presence of demand shocks, equilibrium portfolios could turn out to be different: when local demand is high, both prices of local goods and pay-off of local firms increase. Hence, demand shocks can generate positive co-movements between local equity returns and the price of local goods. In order to be able to consume when demand is high, local investors would prefer local equities.

increases output and this additional output is shared in constant proportion between capital and labor (assuming a Cobb–Douglas production function using capital and labor). Hence, labor and capital incomes are perfectly correlated in this class of models. As investors are already strongly exposed to domestic risk due to their labor income, they should not hold local capital whose returns move in lock-steps with returns on nontradable wealth. Investors should short-sell local equities and diversify their nontradable risk by holding foreign equities. The equity home bias puzzle is worse than we think!

Are Equities Empirically a Good Hedge against Real Exchange Rate Fluctuations?

Heathcote and Perri (2008) show that Baxter and Jermann’s (1997) result relies on a very strong assumption: one unique and perfectly tradable good. Relaxing this hypothesis (as in the bench-mark two-country/ two-good RBC model) and introducing differentiated product across countries together with trade costs (or equivalently ad hoc consumption home bias) drastically changes the picture and help to solve the equity home bias puzzle. Their result relies on two key elements: endogenous investment and a strong adjustment of relative prices. The main intuition goes as follows: suppose a positive (persistent) productivity shock hits the Home economy. This leads to:

While relaxing these assumptions could theoretically help to solve the equity home bias puzzle, such an explanation would still face a major empirical issue. Both in the partial and general equilibrium literature, the hedging of real exchange rate risk would lead to equity home bias if local equities have higher returns (than abroad) when local prices are higher (than abroad). In other words, equity home bias appears if excess local equity returns (over foreign) increase when the real exchange rate appreciates. As shown by van Wincoop and Warnock (2010), the empirical correlation between excess equity returns and the real exchange rate is very low, too low to explain observed equity home bias. Furthermore, most of the fluctuations in the real exchange rate represent fluctuations in the nominal exchange rate: this could be easily hedged using positions in the forward currency markets or the currency bond market. In other words, equities do not seem empirically to be an appropriate asset to insure investors against real exchange rate fluctuations (see van Wincoop and Warnock, 2010 among others). Hence, there is now quite a consensus that the hedging of real exchange rate risk cannot account for the equity home bias. We now investigate another source of domestic risk, namely the presence of nontradable income risk.

HOME BIAS IN EQUITIES AND THE HEDGING OF NONTRADABLE RISK The International Diversification Puzzle is Worse than we Think . . . In an influential paper, Baxter and Jermann (1997) argues that the presence of nontradable income risk worsen the equity home bias puzzle. Their argument goes as follows: in a standard two-country real business cycle model, productivity shocks in one country

Or Better Than We Think . . .

(i) a fall in the relative price of Home goods (Foreign goods are scarcer). (ii) an increase in Home investment (more than abroad) as Home investment uses more intensively cheaper Home goods (due to trade costs). (iii) an increase in Home wage incomes (more than abroad) and in the return on nontradable wealth (thus despite the fall in the price Home goods). (iv) A decrease in the returns on Home capital (relative to Foreign) if the (relative) price response of Home goods is strong enough. The main difference with Baxter and Jermann (1997) is the last point (iv): excess returns on Home capital fall following a positive productivity shock if the adjustment of (relative) prices is large enough. Indeed, if the market price of Home goods falls sufficiently and Home investment is increasing, dividends distributed by Home firms (which are of net investment) are lower than abroad so are Home returns to capital following the shock. Hence, the model is able to generate negative co-movements between Home (excess) return on nontradable wealth and Home excess returns to capital. Contrary to Baxter and Jermann’s (1997) argument, if the response of relative prices is strong enough, hedging nontradable income risk can lead to home bias in equities. In other words,

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

206

22. TRADE COSTS AND HOME BIAS

when productivity is high at Home, Home households have a wage increase and can finance the increase in investment at Home without cutting their consumption, the reason why ex-ante they hold more shares of Home firms. Trade costs (or consumption home bias) are important as they trigger a stronger response of investment at Home (see (iii)). This is why in such a set-up trade costs do help to solve the home equity bias. Importantly, the model generates a positive link between consumption home bias and equity home bias as found in the data and argued by Obstfeld and Rogoff (2000). Finally, one could argue that the hedging of real exchange rate risk, still operating in their framework due to consumption home bias, should play in the opposite direction as explained in section ‘. . .To General Equilibrium.’2 In related work, Coeurdacier et al (2009) show that Heathcote and Perri’s (2008) findings are robust when investors want to hedge real exchange rate risk (and are more risk averse than the log-investor) as long as real exchange rate risk can be hedged by using the appropriate position in the currency bond market. This theoretical result echoes the empirical findings of van Wincoop and Warnock (2010) described in section ‘Are Equities Empirically a Good Hedge Against Real Exchange Rate Fluctuations?’: equities are not the appropriate asset to hedge real exchange rate risk once bond trading is allowed.

TRADE COSTS AND PORTFOLIO HOME BIAS: ALTERNATIVE STORIES The Role of Expropriation/Sovereign Risk Rose and Spiegel (2004) proposes a model of international lending where cross-border capital flows are sustainable because of trade in goods markets. The paper is in the vein of the literature on sovereign risk. Because debt contracts cannot be enforced internationally and foreign creditors face a risk of expropriation, they are willing to lend capital to foreign countries only if they can threaten debtor countries with a credible sanction in case of default. In their model, penalties go through trade: creditors exclude their defaulting partners from trade relationships. The intuition is similar to the one developed by Guibaud (2008) in a set-up with endogenous borrowing constraints a` la Kehoe and Perri and trade costs. Countries cannot commit to repay their debt and are excluded from trade relationships in case of default. When trade costs are lower, gains from trade increase and so does the cost of being in autarky. This endogenously relaxes the borrowing constraint and fosters 2

international risk-sharing. In other words, trade acts as a collateral which relaxes credit constraints in international markets. As a consequence, falling trade costs increase cross-border capital flows and international risk-sharing.

The Role of Information and Behavioral Biases An alternative story could be based on information asymmetries: because trading partners share information, the information flows through trade will enhance trade in assets (and vice versa). In other words, because information flows (or social networks) positively affect both cross-border finance and trade, trade in goods and trade in assets are mutually reinforcing: firm managers learn about each other by trading goods and/or securities. Therefore, lowering trade costs reduces informational asymmetries in the financial markets. Empirically, Portes and Rey (2005) among others show how information flows are a strong determinant of crossborder equity trade. In a similar vein, recent studies put forward a behavioral explanation for the equity home bias: they suggest that familiarity might be the main determinant of portfolio choice. One could argue that foreign firms which sell domestically become more familiar to investors and are favored in their investment decisions. Economies more open in goods markets will be more likely to diversify internationally in asset markets.

CONCLUSION The consumption and equity home biases are two of the main puzzles in international finance (Obstfeld and Rogoff, 2000). They seem to be empirically related: countries more open to international trade have larger stocks of foreign equities. This would suggest a common explanation and Obstfeld and Rogoff (2000) claim that international trade costs can solve both puzzles simultaneously. While there is quite a consensus that international trade costs can explain the consumption home bias, the trade costs explanation for the equity home bias remains an open question. In this chapter, we have reviewed the different theoretical channels through which trade costs might affect equity portfolio decisions. The most natural one related to the hedging of real exchange rate risk turned out to be challenged by the data: equities are empirically a poor hedge for real exchange rate risk. Heathcote and Perri (2008) provides a new channel based on the hedging of nontradable risk (driven by fluctuations in labor income): in their model,

Note that Heathcote and Perri (2008) focuses on log-utility which cancels out any real exchange rate risk hedging.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

CONCLUSION

trade costs are key to generate a negative covariance between returns on nontradable wealth and returns on local capital, making local equities attractive to insure against fluctuations in labor income. Finally, we explored how a decrease in trade costs can relax some financial market imperfections such as expropriation/ sovereign risk, information asymmetries, and foster cross-border equity flows and international risk-sharing. This is unfortunately hard to disentangle empirically between the various potential channels and there is certainly scope to test one theory against the other.

Acknowledgments The author wish to thank Helene Rey for comments on an earlier version and Phil Lane for giving me the opportunity to write this chapter.

References Adler, M., Dumas, B., 1983. International portfolio choice and corporation finance: a synthesis. Journal of Finance 38, 925–984. Anderson, J., Van Wincoop, E., 2004. Trade costs. Journal of Economic Literature 42, 691–751. Aviat, A., Coeurdacier, N., 2007. The geography of trade in goods and asset holdings. Journal of International Economics 71, 22–51.

207

Baxter, M., Jermann, U., 1997. The international portfolio diversification puzzle is worse than you think. American Economic Review 87, 170–180. Coeurdacier, N., 2009. Do trade costs in goods market lead to home bias in equities? Journal of International Economics 77, 86–100. Coeurdacier, N., Kollmann, R., Martin, P., 2009. International portfolios, capital accumulation and foreign assets dynamics. Journal of International Economics. French, K., Poterba, J., 1991. Investor diversification and international equity markets. American Economic Review 81, 222–226. Guibaud, S., 2008. A Tale of Two Frictions: Endogenous Borrowing Constraints with Trade Costs. Mimeo, London School of Economics: London. Heathcote, J., Perri, F., 2008. The International Diversification Puzzle is Not as Bad as You Think. Unpublished Working Paper, Federal Reserve Bank of Minneapolis. Lane, P., Milesi-Feretti, G.M., 2008. International investment patterns. The Review of Economics and Statistics 90, 538–549. Obstfeld, M., Rogoff, K., 2000. The six major puzzles in international macroeconomics: is there a common cause? NBER Macroeconomics Annual. Portes, R., Rey, H., 2005. The determinants of cross-border equity flows. Journal of International Economics 65, 269–296. Rose, A.K., Spiegel, M.M., 2004. A gravity model of sovereign lending: trade, default, and credit. IMF Staff Papers 51, 50–63. van Wincoop, E., Warnock, F.E., 2010. Is home bias in assets related to home bias in goods? Journal of International Money and Finance. Working Paper 12728, National Bureau of Economic Research.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

Intentionally left as blank

C H A P T E R

23 Explaining Deviations from Uncovered Interest Rate Parity P. Bacchetta University of Lausanne, Lausanne, Switzerland Centre for Economic Policy Research (CEPR), London, UK O U T L I N E Introduction

209

Deviations from Rational Expectations

211

Risk Premium with Representative Investors

210

Limited Participation

210

Conclusion References

212 212

INTRODUCTION A key relationship in international finance is uncovered interest rate parity (UIP). An approximated version of this relationship is Et stþ1  st ¼ it  it

ð23:1Þ

where st is the log nominal exchange rate (domestic per foreign currency), ii and it* are domestic and foreign one-period nominal interest rates, and Et is market expectation based on information at time t. This relationship, described in numerous textbooks, means that an expected depreciation should be compensated by an interest rate differential. It also implies that there should be no expected excess return from arbitraging across currencies. Define the linearized excess return from a foreign currency investment as xtþ1  st þ 1  st þ it*  it. UIP means that Etxt þ 1 ¼ 0. However, the empirical evidence shows that expected excess returns are nonzero. Moreover, excess returns can be systematically predicted by interest rate differentials. This can be seen from the famous Fama (1984) regression that implies that xt þ 1 is predictable by the interest differential it  it*. More precisely, consider the regression: xtþ1 ¼ a þ bðit  it Þ þ etþ1

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00017-9

ð23:2Þ

The coefficient b is typically significant, which indicates predictability of excess returns. Moreover, b is systematically negative and is often below 1. This means, for example, that if the foreign interest rate, it*, increases, the foreign currency appreciates so that the excess return in foreign currency increases more than the interest rate. Furthermore, the interest differential (or the forward discount) can predict excess return at future dates. Consider a regression of a future 3-month excess return qt þ k, from t þ k  1 to t þ k, on the current interest rate differential it  it*. Bacchetta and van Wincoop (2010) show that there is significant predictability with a negative sign for five to ten quarters. Over longer horizons, however, the slope coefficient becomes insignificant or even positive. This is consistent with the findings that UIP holds better at longer horizons. The persistence in the predictability of excess returns is related to the phenomenon of the so-called delayed overshooting. The presence of expected excess returns naturally makes the foreign exchange market more interesting for investors, as the recent waves of carry trade illustrate. The basic question, however, is why we observe these predictable excess returns. Deviations from UIP, traditionally called the forward premium puzzle, have received extensive attention in the literature. While there is no consensus on a single explanation, the recent literature has made useful progress. The objective of this chapter

209

# 2013 Elsevier Inc. All rights reserved.

210

23. EXPLAINING DEVIATIONS FROM UNCOVERED INTEREST RATE PARITY

is to review some of the most recent theoretical papers providing explanations for the puzzle. Previous literature reviews include Froot and Thaler (1990) and Engel (1996). Theoretical explanations of the forward premium puzzle can be grouped into three main categories: (i) risk premium, (ii) limited market participation, and (iii) deviations from rational expectations.1

RISK PREMIUM WITH REPRESENTATIVE INVESTORS In the presence of risk aversion, we have Etxtþ 1 ¼ t, where t is a risk premium. For example, the risk premium for a small open economy would typically be positively related to the covariance between the excess return and the stochastic discount factor, Mt, that is, with cov(xt, Mt). This means that investors demand a premium because of the covariance between the expected excess return and the discount factor. It is useful to write the excess return as xtþ1 ¼ t þ xtþ1  Et xtþ1 |fflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflffl}

ð23:3Þ

prediction error

Theories relying on the risk premium consider models with rational expectations where the risk premium t is negatively related to the interest differential and where there is no systematic prediction error. This means that, in a small open economy, cov(xt, Mt) should decrease with the interest rate differential. The earlier literature had clearly rejected the explanations based on risk premium (e.g., see Engel, 1996) as they could not match the empirical evidence. In recent years, however, several papers had more success in matching the data. For example, Verdelhan (2010) proposes a two-country consumption-based model with habit formation that generates a negative relationship between the excess return and the interest differential. In his framework, a low domestic consumption implies higher risk aversion and therefore a higher risk premium. Lower consumption also leads to a lower domestic real interest rate. Therefore, a lower real interest rate differential coincides with a higher expected excess return. Another way to introduce time-varying risk premia is to assume disaster risk. Farhi and Gabaix (2008) consider a multicountry model where a disaster takes the form of a drop in world consumption in tradeables and in countries’ productivities. However, productivities are affected differently across countries so that some countries face larger disasters. In this context, countries with

higher disaster risk face both a higher risk premium, and thus a higher expected excess return, and a lower interest rate. This explains the forward premium puzzle. Other papers that introduce disaster risk to explain the forward premium puzzle are Guo (2007) and Gourio et al. (2010). While these recent papers generate a negative relationship between excess returns and interest rate differentials, they face some serious challenges with other features of the data. For example, it is well known that exchange rates are much more volatile than macroeconomic fundamentals (e.g., aggregate consumption). See Burnside et al. (2011a) for a more detailed discussion.

LIMITED PARTICIPATION As models with representative individuals and rational expectations have difficulties matching certain aspects of the data, some recent papers have considered limited participation in the FX market, in the sense that only a subset of potential investors is active in a given period. Alvarez et al. (2009) consider a model with market segmentation, where an endogenous fraction of households is active in asset markets. The source of uncertainty comes from monetary shocks that generate a risk premium and expected excess returns. In this context, an increase in domestic money growth may lead to a negative correlation between excess returns and the interest differential. If money growth is moderately persistent, an increase in domestic money growth leads more households to be active in assets markets since higher inflation increases the cost of not participating. This reduces the risk premium. At the same time, it increases the domestic nominal interest rate due to an expected inflation effect. Hence, the expected excess return and the interest rate differential are negatively related. While the endogeneity of the fraction of active investors is likely to be an important feature to explain the FX market, the robustness of the specific mechanism proposed by Alvarez et al. remains to be determined. Another form of limited participation is due to the fact that investors only change infrequently their international portfolio positions. Froot and Thaler (1990) have informally argued that models where some agents are slow in responding to new information lead to predictability in the right direction. The argument is simple. An increase in the interest rate of a particular currency will lead to an increase in demand for that currency and therefore an appreciation of the currency. But when investors make infrequent portfolio decisions, they will gradually buy the currency as time goes on. This can

1

Some explanations of the forward premium puzzle do not fit in these categories. For example, one strand of the literature argues that there are econometric problems with the Fama regression. It is not the purpose of this chapter to give an exhaustive review of the literature and most of the relevant references can be found in the papers mentioned in this chapter.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

DEVIATIONS FROM RATIONAL EXPECTATIONS

cause a continued appreciation of the currency, implying that a higher interest rate raises the expected excess return of the currency.2 In reality, only a small fraction of foreign currency holdings is actively managed. Outside a small industry that actively manages foreign exchange positions of investors, there is little active currency management over horizons relevant to medium-term excess return predictability. For example, banks conduct extensive intraday trade but hold virtually no overnight positions. Mutual funds do not actively exploit excess returns on foreign investment since they only trade within a certain asset class and cannot freely reallocate between domestic and foreign assets. Finally, most large financial institutions do not even devote their own proprietary capital to currency strategies based on the forward discount bias. Therefore, most of the international portfolios held over the medium run belong to institutions (or rich individuals) that are not active in the foreign exchange market. Bacchetta and van Wincoop (2010) (henceforth BW) formalize this idea and examine the impact of infrequent portfolio decisions in a simple two-country general equilibrium model that is calibrated to data. Agents have the choice between actively managing their foreign exchange positions, at a cost, and making infrequent portfolio decisions. The cost of active currency management is measured by the actual fees charged by the active currency management industry. BW find that all or most investors do not find it in their interest to actively manage their foreign exchange positions as the resulting welfare gain does not outweigh the cost. Infrequent portfolio decisions mean that investors will look at excess returns over several periods. If we assume that investors change their portfolio every T periods, the relevant excess return is the cumulative excess return from t to t þ T: xt,tþT ¼ xtþ1þ  þxtþT. This implies the following approximated constant relative risk aversion optimal portfolio rule: bIt ¼ bI þ

Et xt;tþT gs2I

ð23:4Þ

where bI is a constant and sI2 is a measure of the portfolio variance. It is interesting to notice that there is also a welldefined risk premium for investors making infrequent portfolio decisions. For those investors, a T-period Euler equation applies: Et ðctþT Þg xt;tþT ¼ 0

ð23:5Þ

where ctþT is consumption at t þ T. The risk premium for infrequent investors applies over T periods and is equal to

211

the rate of risk aversion times the covariance of the excess return over T periods and consumption in T periods. In their model, BW find that infrequent portfolio decisions lead to a delayed impact of interest rate shocks on exchange rates. This can explain the phenomenon of delayed overshooting, whereby the exchange rate continues to appreciate over time after a rise in the interest rate. This delayed overshooting leads to excess return predictability in the direction seen in the data. Even future excess returns continue to be predictable by the current forward discount, with the magnitude of the predictability declining as time goes on. While the model with infrequent decision making can explain the forward premium puzzle, it also matches other aspects of the data, in particular, various univariate properties of exchange rates and interest rates (volatility and persistence). BW show that excess return predictability resulting from infrequent portfolio decisions is even stronger when agents condition exchange rate expectations on a limited set of variables. In reality, the most common active currency management strategy is carry trade, which is mostly based on current interest rate differentials. When exchange rate expectations are based on either current interest rate differentials alone or random walk expectations, excess return predictability is larger than in the case where expectations are conditioned on the entire information set. Bacchetta and van Wincoop (2007) show that if investors have random walk expectations but trade frequently, high interest rate currencies depreciate much more than what UIP would predict. However, when agents make infrequent FX portfolio decisions, random walk expectations can explain the forward premium puzzle.

DEVIATIONS FROM RATIONAL EXPECTATIONS The other strand of the literature focuses on prediction errors in Eq. (23.3). On average, investors make mistakes in predicting excess returns. Moreover, these errors must be negatively related to the interest rate differential so that the coefficient b in Eq. (23.2) is negative. This is consistent with empirical evidence. Expectational errors derived from survey data are clearly negatively correlated with the interest differential (e.g., see Bacchetta et al., 2009).

2

Notice that models with imperfect information and learning can also produce this type of response. However, rational learning models usually converge to the full information case. Moreover, learning models do not necessarily generate excess return predictability with the right sign.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

212

23. EXPLAINING DEVIATIONS FROM UNCOVERED INTEREST RATE PARITY

A simple formal model was presented by Gourinchas and Tornell (2004). They assume that investors think that shocks to the interest rate differential are more temporary than they actually are. To understand how this affects the dynamics of exchange rate expectations, it is useful to remember that an increase in it  it* implies an initial domestic currency appreciation, followed by a gradual depreciation. The gradual depreciation is required to compensate for the interest differential, as Eq. (23.1) indicates. The smaller the future interest rate differential, the smaller the gradual depreciation and thus, the smaller the required initial appreciation. Therefore, when investors think incorrectly that the interest differential will decline more quickly than it actually does, the initial appreciation is smaller than it should be. But when investors realize that the interest differential declines more slowly than they expected, they revise their expectations and the domestic currency appreciates. Hence, in this learning period, we are in a situation of gradual appreciation, accompanied by a higher interest differential, which means a negative excess return. This clearly implies a negative relationship between the excess return and the interest differential. When this learning period is over, the currency starts to depreciate. This dynamic gradual appreciation followed by depreciation leads to the delayed overshooting described earlier. Deviations from standard rational expectations in Gourinchas and Tornell are modeled in an ad hoc way. Recent research has modeled more carefully the behavior of investors. For example, Ilut (2010) introduces ambiguity aversion in a context where investors have imperfect knowledge of the underlying model. The author considers an example where investors do not know the true variance of temporary shocks to the interest rate differential. Ambiguity aversion leads investors to give more weight to the possibility of high volatility of temporary shocks. This leads to an effect which is similar to an overestimation of the variance of temporary shocks. Therefore, the mechanism in Ilut leads to results similar to the Gourinchas and Tornell model and can also explain the forward premium puzzle. Burnside et al. (2011b) consider overconfident investors in the sense that they overestimate the precision (or underestimate the variance) of their private signals. In their model, signals are about future monetary policy. A signal about a future domestic inflation is taken ‘too seriously’ on average so that the interest differential increases excessively and the currency depreciates excessively, both due to an increase in expected inflation. This implies that in subsequent periods, when investors realize that inflation is not so high, the currency has to appreciate while the interest differential has to decline.

Again, this implies a negative relationship between excess return and the interest differential.

CONCLUSION It is relatively easy to design models that generate deviations from UIP as observed in the Fama regression. What is more challenging is to find mechanisms that are robust to various types of shocks and that can match the various aspects of the data, such as high exchange rate volatility. It is unlikely that a simple explanation can be sufficient to solve the puzzle and a combination of factors is therefore needed. For example, Bacchetta and van Wincoop (2010) show that infrequent decision making is likely to be a key ingredient. But to this feature they need to assume limited arbitrage, due to risk aversion, by active traders. Moreover, their results get closer to the data when some limited information processing is assumed.

References Alvarez, F., Atkeson, A., Kehoe, P.J., 2009. Time-varying risk, interest rates, and exchange rates in general equilibrium. Review of Economic Studies 76, 851–878. Bacchetta, P., Mertens, E., van Wincoop, E., 2009. Predictability in financial markets: what does survey expectations tell us? Journal of International Money and Finance 28, 406–426. Bacchetta, P., van Wincoop, E., 2007. Random walk expectations and the forward discount puzzle. American Economic Review 97, 346–350. Bacchetta, P., van Wincoop, E., 2010. Infrequent portfolio decisions: a solution to the forward discount puzzle. American Economic Review 100, 837–869. Burnside, C., Eichenbaum, M., Kleshchelski, I., Rebelo, S., 2011a. Do peso problems explain the returns to the carry trade? Review of Financial Studies 24 (3), 853–891. Burnside, C., Han, B., Hirshleifer, D., Wang, T.Y., 2011b. Investor overconfidence and the forward premium puzzle. Review of Economic Studies 78 (2), 523–558. Engel, C., 1996. The forward discount anomaly and the risk premium: a survey of recent evidence. Journal of Empirical Finance 3, 123–192. Fama, E.F., 1984. Forward and spot exchange rates. Journal of Monetary Economics 14, 319–338. Farhi, E., Gabaix, X., 2008. Rare disasters and exchange rates. NBER, Working Paper No. 13805. Froot, K.A., Thaler, R.H., 1990. Anomalies: foreign exchange. Journal of Economic Perspectives 4, 179–192. Gourinchas, P.-O., Tornell, A., 2004. Exchange rate puzzles and distorted beliefs. Journal of International Economics 64, 303–333. Gourio, F., Siemery, M., Verdelhan, A., 2010. International risk cycles. Mimeo. Guo, K., 2007. Exchange rates and asset prices in an open economy with rare disasters. Mimeo. Ilut, C., 2010. Ambiguity aversion: implications for the uncovered interest rate parity puzzle. Mimeo. Verdelhan, A., 2010. A habit-based explanation of the exchange rate risk premium. Journal of Finance 65, 123–145.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

C H A P T E R

24 Valuation Effects, Capital Flows and International Adjustment C. Tille Graduate Institute of International and Development Studies and CEPR, Gene`ve, Switzerland O U T L I N E Introduction

213

Financial Globalization and Valuation Effects Stock and Flows in External Accounts The Drivers of Net Foreign Assets Yield and Return Differentials

214 214 214 216

International Portfolio Choice and Adjustment in Theory

217

INTRODUCTION Financial globalization has been one of the major developments in the world economy since the mid-1990s. While capital had been flowing across borders well before then, the last decade and a half saw an important shift in the pattern of international investment. The initial situation could be broadly characterized as a ‘one-way’ financial integration where some countries held assets abroad, and other countries were indebted to foreign investors. The more recent period by contrast is characterized by a ‘two-way’ integration where all countries hold substantial assets abroad, while at the same time having substantial liabilities to foreign investors. The current configuration is thus one with large holdings of gross international assets and liabilities that dwarf the value of net foreign assets.1

Modeling Financial Integration Portfolio Choice and Capital Flows Valuation Gains and External Adjustment

217 218 218

Interpretation of the External Accounts

219

Concluding Remarks Glossary Further Reading

220 221 221

This ‘leveraged’ pattern where a country has sizable holdings of both gross assets and liabilities2 implies that moderate movements in asset prices can lead to large fluctuations in the country’s net foreign assets as they apply to large gross holdings. As shown below, such valuation effects have indeed become quite prominent. This raises the question of whether a country indebted to the rest of the world could reduce its debt burden through valuation effects instead of having to repay its debt. This chapter first illustrates how valuation effects have affected the dynamics of international assets and liabilities by focusing on the case of the United States. The rising prominence of these effects in the real world has led to an active theoretical literature aiming at bringing them into the workhorse open economy macroeconomic models. It then reviews the main features of this literature, with an emphasis on the dynamics of external

1

Gross assets are defined as the value of claims by a country’s residents on the rest of the world. Gross liabilities are claims by foreign investors on the country. Net foreign assets are the difference between gross assets and gross liabilities. A country with negative net foreign assets is a net debtor to the rest of the world.

2

It is important to bear in mind that a country is not leveraged in the same sense as an individual investor is. Specifically, an investor chooses his leverage by borrowing to purchase assets. By contrast, a country’s assets and liabilities reflect the decision of a large array of different agents instead of a centralized investor.

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00019-2

213

# 2013 Elsevier Inc. All rights reserved.

214

24. VALUATION EFFECTS, CAPITAL FLOWS AND INTERNATIONAL ADJUSTMENT

assets and liabilities. The final section reviews how financial globalization can impact the current account, both in terms of its sustainable level and in terms of its volatility.

FINANCIAL GLOBALIZATION AND VALUATION EFFECTS Stock and Flows in External Accounts The discussion of a country’s external accounts often focuses on the flows of money across borders. For instance, the current account denotes the difference between what the country earns abroad, through exports or earnings on its assets, and what it pays to the rest of the world, in the form of imports and interest payments to foreign creditors. However, the stock of international assets and liabilities is a more meaningful measure when one wants to assess whether a country is in a worrisome situation or not. The logic is the same as for a person. That someone spends more than his/her income in a given period says little about whether this is troublesome. To make such an inference, one needs to know the person’s wealth: the excess spending is an issue if the person is already in debt, but is of little concern if he/she has a substantial net worth. A country’s net worth relative to the rest of the world is measured by its net foreign asset position.3 In parallel with the above example, one should be more concerned about a country’s running of a large current account deficit (i.e., borrowing additional funds from the rest of the

world) if it is already a net debtor than if it has a comfortable positive net foreign asset position. Measures of stocks of international holdings are thus central to the analysis of the global economy. As indicated in the introduction, most countries experienced a surge in the value of their assets abroad, as well as their liabilities to foreigners. This was most marked among developed economies, where these values tripled as a share of gross domestic product (GDP), but was also observed in emerging markets. This pattern is illustrated by considering the case of the United States (Figure 24.1).4 The value of assets held by US residents abroad (thin solid line) rose from 54% of GDP in 1995 to a peak of 124% in 2007, before falling during the global financial crisis. The gross liabilities to foreign investors (dashed line) followed a similar path. While the US moved from being a net creditor to the rest of the world to becoming a net debtor, the magnitude of its net asset position (thick solid line) is quite small compared to the gross assets and liabilities. While the focus has been on the United States, given its prominent role in the world economy, a similar pattern can be observed in most countries.

The Drivers of Net Foreign Assets Even though stock measures are the relevant concept for this analysis, one could argue that looking at flows is also valid. After all, stocks are the sum of all past flows, so if a country runs persistent current account deficits, these will ultimately add up to a sizable stock of net liabilities to the rest of the world, in the same way as an FIGURE 24.1 US international assets and liabilities.

150% 130% 110% 90% 70% 50% 30% 10% −10%1982

1986

1990

1994

1998

2002

2006

−30% Gross assets

Gross liabilities

Net assets

3

This is only a partial measure of the country’s wealth, as it excludes the value of domestic assets.

4

The data are taken from the US Bureau of Economic Analysis.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

FINANCIAL GLOBALIZATION AND VALUATION EFFECTS

215

FIGURE 24.2 Net position, cumulated flows, and cumulated valuation gains (USA).

40% 30% 20% 10% 0% 1982 −10%

1986

1990

1994

1998

2002

2006

−20% −30% −40% −50% Net position

Cumulated financial flows

Cumulated valuation gains

individual living beyond his/her means will face a large balance on her credit card. Financial flows to and from the rest of the world (i.e., the current account) are however not the only drivers of a country’s external position. Specifically, the change in the net foreign asset position between two periods is Net positiont ¼ Net positiont1 þ Net tradet þ Earnings on assetst  Earnings on liabilitiest þ Valuation effect on assetst  Valuation effect on liabilitiest ¼ Net positiont1 þ Net tradet þ Net earningst |fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl} Financial flowst

þ Net valuation effectt ð24:1Þ In addition to financial flows, the so-called valuation effects matter. These represent changes in the value of existing assets and liabilities that are not associated with any flows. Going back to the example of a person, consider that he/she has 1000 dollars invested in stocks and owes 1000 on a credit card loan, so that his/her net worth is zero. If stock prices increase by 10%, the value of his/her stock holdings is boosted by 100 dollars, and so is his/her net worth, without any flow of income or spending. The impact of this valuation effect on the net worth is larger the bigger the value of gross assets and liabilities. If the stock holdings and bank loan amount to 10 000 dollars each (so the net worth is still zero), the higher stock price translates into a gain of 1000 dollars. The impact of financial flows and valuation gains for the United States is illustrated in Figure 24.2. The thick solid line represents the actual net foreign asset position,

with the United States being a net debtor to the tune of 22% of its GDP at the end of 2009. The role of financial flows is shown by the dashed line, which adds up the current account deficit starting from the observed net foreign asset position in 1982. This line trends downward, reflecting the persistent current account deficits. Financial flows alone, however, would imply a net debt of 48% of GDP at the end of 2009. The large difference reflects cumulated valuation gains on US gross assets and liabilities, shown by the dotted line. These gains have been positive, and have taken a prominent role since 2000, despite some payback in the current crisis. As a result, the United States has managed to limit the size of its external indebtedness despite running large and persistent current account deficits. Two main aspects emerge from Figure 24.2. First, valuation gains are much more volatile than financial flows. This can be seen from the fact that the ‘financial flows’ line is much smoother than the ‘valuation gains’ line, which shows large movements in both directions. For instance, the United States experienced valuation losses in 2008 that amounted to 17% of GDP. Second, while valuation gains drive substantial gaps between the cumulated financial flows and the actual net foreign asset position, these gaps tend to be relatively short lived. Financial flows remain the main driver of the net foreign asset position in the long run. These aspects are also observed in other countries. The valuation effects reflect three main components: Overall valuation effect ¼ Valuation effect from exchange rates þ Valuation effect from asset prices þ Other valuation effect

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

ð24:2Þ

216

24. VALUATION EFFECTS, CAPITAL FLOWS AND INTERNATIONAL ADJUSTMENT

FIGURE 24.3

40%

Composition of cumulated valuation

gains (USA).

35% 30% 25% 20% 15% 10% 5% 0% 1982 −5%

1986

1990

1994

1998

2002

Total cumulated effect

From exchange rates

From asset prices

From other sources

2006

These components are shown in Figure 24.3.5 The solid line shows the impact of exchange rate movements. As a large fraction of US gross assets are in foreign currencies, while the quasi-totality of gross liabilities is in dollars, a depreciation of the dollar boosts the dollar value of assets denominated in foreign currencies. A weakening of the dollar, such as in the early 2000s, generates valuation gains (the line goes up). The second term in eqn (24.2) is the impact of movements in asset prices in local currencies (dashed line). It played only a moderate role until the early 2000s, as asset prices moved broadly in steps in the United States and abroad, leading to similar valuation effects on gross assets and liabilities that are canceled out in net terms. While the United States has benefited from the stronger performance of foreign asset markets since 2004, the gains have been lost during the ongoing crisis. Finally there is an ‘other’ source of valuation gains (dotted line) which primarily reflects statistical revisions that cannot be clearly linked to financial flows or specific valuation gains. This last source highlights the challenges faced by statisticians in building the data. As a result, the data have to be taken with care, and researchers have shown that considering these other valuation changes to be true capital gains reflecting exchange rate or asset prices can lead to misleading inferences.

Yield and Return Differentials In addition to the valuation gains discussed earlier, the rise in financial integration has an impact on the current account itself. While for most countries the bulk of the current account reflects the trade balance 5

between exports and imports, eqn (24.1) shows that the current account also includes the net stream of dividends and interest payments between a country and the rest of the world. Earning streams are financial flows and are thus distinct from the valuation effects, in the same way as the dividend on a stock differs from the capital gain on that stock. The yield on an asset is defined as the earning streams in percentage of the asset value, while the overall rate of return also includes the valuation gain. A striking feature of the United States is that while one would expect it to have a net asset income deficit reflecting its position as a net debtor vis-a`-vis the rest of the world, it instead earns a net asset income surplus. In other words, the United States gets a better yield on its gross external assets than it pays on its gross liabilities. Such gaps in yields have a much larger effect on the current account when financial integration is high. Consider for instance that a country earns 1% point more on its assets than it pays on it liabilities. If its holdings are balanced between 100 worth of assets and 100 of liabilities, the spread translates into a net asset income surplus of one. If, however, the value of assets and liabilities is 1000, the same yield gap translates into a surplus of ten. Thus, financial integration magnifies the impact of yield differentials in the same way as it magnifies that of differences between rates of valuation gains across assets and liabilities. Gaps in yields and rates of return can allow a country to enjoy an ‘exorbitant’ privilege, that is, a situation in which the country earns a higher rate of return on its assets and thus can keep its net foreign asset position stable despite running persistent current account deficits. Some researchers argue that this situation is observed

The data breaking down the overall effect into the three components are available only since 1989.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

INTERNATIONAL PORTFOLIO CHOICE AND ADJUSTMENT IN THEORY

for the United States not only in terms of yields, but also in terms of rates of return, allowing it to systematically offset its current account deficit through valuation gains. While the pattern is more contrasted when one considers a broader range of countries, there are many cases of systematic gaps between the rates of returns on assets and liabilities. It is important, however, to bear in mind that computing rates of return is a difficult task for two reasons. First, they include the rates of valuation gains which are driven by changes in exchange rates and asset prices. As these are very volatile, the confidence intervals around the estimated rates of return are fairly large. This is less of an issue when estimating yields that only reflect relatively steady earning streams. Second, the computation of rates of valuation gains from aggregate data can be substantially distorted by statistical issues, such as the nature of the ‘other’ valuation gain discussed above. Researchers who focus on disaggregated data and correct for statistical problems find much smaller (if any) gaps between the rates of return on assets and liabilities.

INTERNATIONAL PORTFOLIO CHOICE AND ADJUSTMENT IN THEORY Modeling Financial Integration The surge of financial integration since the mid-1990s has led researchers to include this into theoretical open economy models. While international asset holdings were hardly new in the literature, until the late 1990s the existing frameworks fell into two broad categories, each with limited appeal. One modeling strategy was to consider that only one asset was traded across borders, the case of a risk-free bond being the most standard one. While this approach is perfectly valid when one is interested in modeling intertemporal borrowing and lending decisions in a ‘one-way’ financial integration, it has nothing to say about a ‘two-way’ integration. The second modeling strategy was to consider that asset markets were complete, that is, that one could insure for all possible future contingencies by holding an appropriate portfolio. In addition to being of limited realism, this approach completely splits the real economy from portfolio decisions. All that matters for the dynamics of real variables (such as output or consumption) in response to shocks is to know that financial markets are complete. How this is achieved is an entirely separate question. In addition, many ways of generating complete markets, such as assuming that there exists one asset for each possible state of nature is unrealistic as it generates no capital flows. There has thus been a need to build models that fall between the two extremes. These models have more than

217

one internationally traded asset, but do not have too many assets either in order to keep financial markets incomplete. In addition to being the most empirically relevant pattern, such a setting combines both the presence of an endogenous portfolio decision and an interaction between this decision and the properties of the real economy. The inclusion of portfolio choice in open economy models has faced two challenges. First, portfolio choice is about the hedging of risk. Investors prefer assets that deliver a high payoff when their own utility of income is high, due for instance to low wages. The optimality conditions that reflect an agent trading off alternative assets must capture this hedging. The standard linear approximation techniques are of no use in that respect as they by construction ignore terms beyond the linear dimension of the model, including the variances and covariances at the core of the hedging motive. This hurdle has been overcome by extending the standard approximation techniques to go beyond the linear terms. While this entails additional technicalities, the approach is fundamentally in the line of standard solution techniques. An additional aspect of the first challenge is that researchers are interested not only in computing the portfolio that agents choose before shocks are realized, but also in assessing how the occurrence of a particular shock leads agents to reallocate their portfolio across different assets. Understanding this aspect requires one not only to include the variances and covariances in the optimality conditions for portfolio choice, but also to determine how these moments shift in response to a shock, that is, to model the dynamic response of variances and covariances. This dynamic response implies that the hedging properties of various assets can be timevarying, leading agents to reshuffle their portfolio through time. Modeling this dynamic dimension requires an additional extension of the approximation techniques. While this involves nontrivial technical aspects, it remains in the line of the standard approximation method. The second challenge faced by researchers is that the analysis cannot be parsimonious. In models that abstract from portfolio choice, one can easily analyze the response of the economy to a particular shock, say a fiscal expansion, without having to worry about all the other potential shocks that can affect the economy. This parsimony is lost once portfolio choice is included because agents allocate their holdings across assets before the shocks are realized, and do so by taking account of all possible shocks. A researcher interested solely in assessing the impact of a fiscal expansion cannot abstract from technology, demand, and other shocks, as they affect the portfolio chosen by the investors, and this portfolio in turn plays a role in determining how the fiscal expansion is transmitted across the economy.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

218

24. VALUATION EFFECTS, CAPITAL FLOWS AND INTERNATIONAL ADJUSTMENT

Portfolio Choice and Capital Flows The readjustment of portfolios in response to shocks has a direct link to international capital flows. Consider a shock that increases the flow of savings by a country’s agents, with some of these savings invested abroad. Cross-border capital flows can be split into two components. The portfolio growth component reflects the allocation of the additional savings along the line of the portfolio held by the agents prior to the shock. The second portfolio reallocation component reflects any change in the allocation of assets that agents undertake in response to changing hedging properties. In a situation where cross-border asset holdings are sizable, even a moderate reallocation of portfolio can lead to substantial capital flows between countries as it applies to a large existing stock of assets. The literature shows that the link between the dynamics of portfolio allocation and capital flows can be subtle. This reflects the fact that agents already hold a diversified portfolio before shocks occur. Consider for an investor who, following a shock, decides to increase the share of foreign assets in his portfolio. While this seems to necessarily imply that he purchases foreign assets, this need not be the case. If the shock also boosts the price of the foreign asset, this automatically raises the value of the investor’s existing holdings of foreign assets without any capital flow taking place. The question is then how this higher ‘passive’ portfolio share (i.e., the share that merely reflects the asset price increase) compares to the investor’s desired share. If the passive and optimal shares coincide, no capital flows take place. If the passive share is smaller than the optimal one, the investor purchases additional foreign assets, leading to a capital flow toward the foreign country. By contrast, if the passive share exceeds the optimal one, the investor liquidates foreign assets and repatriates the money, leading to a capital flow away from the foreign country. Models with portfolio choice have demonstrated the ability to match two empirical features of international capital flows. First, flows are more volatile at short horizons than at long ones. Second, flows in and out of a country tend to move in opposite directions at short horizons, with both domestic and foreign investors pulling money back to their own country for instance. This ability reflects a different mix of the portfolio growth and portfolio allocation components at different horizons. When a shock hits, the reallocation component leads to an immediate reshuffling of the large existing stock of assets, with large capital flows. By contrast, the additional savings are relatively moderate and the portfolio growth component plays a small role. In subsequent periods, the reallocation component is limited as portfolios have by then been rearranged, while the flow of additional savings remains at a similar level. One thus gets a pattern

of large movements in capital flows on impact reflecting portfolio reallocation, followed by smaller effects in subsequent periods driven by portfolio growth. The initial portfolio reallocation can also lead investors to move in opposite directions. For instance, a persistent increase in productivity in a country leads all investors to want to increase the share of that country’s equity in their portfolio. In equilibrium, domestic and foreign investors cannot both purchase this equity, but must buy and sell it between themselves. Consider the situation where domestic investors want to reallocate into the local equity more than what foreign investors want. In equilibrium, the equity price increases, reflecting the higher productivity, leading to an automatic increase in the passive share of local equity in all portfolios. That passive share falls in between the optimal shares of local and foreign agents, and as a result, foreign agents sell some local equity to domestic agents. The domestic agents finance this by selling foreign equity, while the foreign agents repatriate the proceeds of their sales of local equity to buy foreign equity. The model then generates a situation where all investors pull money back into their own country. Information asymmetries can also lead to investors from different countries readjusting their portfolios in different ways. Consider a situation where investors learn that asset payoffs will be favorable in the future, but do not know exactly in which country. If investors are more confident in the prospects of their own country, they liquidate assets abroad and repatriate the funds to invest them domestically.

Valuation Gains and External Adjustment The theoretical literature sheds light on the mechanisms through which a country can address an imbalanced external position, and especially on the role of valuation gains in the process. It is important to distinguish between unexpected and expected valuation gains. Consider a country with liabilities in its own currency and assets in foreign currencies. An unforeseen depreciation of its currency boosts the value, in its currency, of foreign assets, that is, generates an unexpected valuation gain. If the exchange rate subsequently reverts to its initial value in a predictable way, the country faces expected valuation losses. The literature shows that while unexpected valuation effects play a large role in the change in country’s net foreign asset positions at the time of shocks, expected valuation effects play a more moderate role along the subsequent adjustment path. A country’s intertemporal budget constraint shows that an external debt can be reduced in two ways: Debtt ¼ Trade surplusest;tþ1;tþ2;... þ Return on assetst;tþ1;tþ2;...  Return on liabilitiest;tþ1;tþ2;...

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

ð24:3Þ

INTERPRETATION OF THE EXTERNAL ACCOUNTS

The first way to pay off the debt is to run trade surpluses in the future. The second way is for the country to earn more on foreign assets than it pays on liabilities, for instance by having a higher rate of return (including valuation gains) on its external assets. As eqn (24.3) holds in expected value, the gap in the rates of return between assets and liabilities must be predictable. At this point, it is important to distinguish between expected valuation gains and expected rates of return, which include both valuation gains and expected yields. Consider for instance a country which gets a steady 5% rate of valuation gains on its assets with no valuation losses on its liabilities. This gap in the rates of capital gains does not translate into a gap in the rates of returns if the country gets a 2% yield on its assets, but pays a 7% yield on its liabilities. The presence of predictable valuation effects is thus no guarantee that there is a predictable gap in the overall rates of returns, which are the relevant measures for the adjustment process. Theoretical contributions cast doubt on the ability of a country to narrow its debt through return differentials. This reflects the fact that if assets deliver different returns in a predictable way, agents will be unwilling to hold the low-return asset. This is not to say that the expected rates of returns must be exactly the same across assets, but that any gap must remain on the same order of magnitude as hedging considerations, a magnitude that is too small to make much difference in the adjustment process. Consider for instance an investor who can purchase two assets, A and B. Asset A pays an expected return of 10%, with the realized return falling close to this value, say between 9 and 11%. If asset B provides a better hedge to the investor, for instance by offering a better offset to his/her labor income risk, the investor will be willing to hold asset B even if it offers a lower return. That return must, however, remain fairly close to that of asset A. For instance, the return on asset B can fluctuate between 8.5 and 10.5%, averaging 9.5%. Such a gap is consistent with the investor holding both the higher return asset A and the better hedge asset B. If the return on asset B, however, moves between 5 and 7% (averaging 6%), the hedging benefits of that asset are too small to offset that large return gap compared to asset A, and the investor only holds asset A. Substantial predictable gaps in the rates of return across different assets are thus not compatible with investors holding all assets in their portfolios, even to a different extent. The absence of predictable gaps in rates of returns is not incompatible with predictable differentials in valuation effects on a country’s assets and liabilities. These effects are, however, offset by predictable gaps in yields. Consider for instance the case of a temporary (albeit persistent) increase in a country’s productivity. The productivity shock leads to an immediate increase in the price of the country’s equity, followed by a gradual

219

decrease back to the original value. The initial jump in the asset price generates large unexpected valuation gains for foreign investors who hold some of the equity, thereby pushing the country into a net debt position. In subsequent periods, foreign investors face predictable capital losses as the equity price decreases to its initial value. The high productivity of the country, however, implies that its equity pays high dividends that offset the valuation losses. Overall, domestic and foreign equity earn the same return along the adjustment path, albeit with different combinations of yields and valuation gains. The only way to generate a large gap between the expected rates of return is through a segmentation of markets. For instance, one can assume that foreign investors are willing to accept a lower return on their holdings in a country than they get in their own domestic market. This could, for instance, reflect a strategy of accumulating reserves to fix the value of the exchange rate. Assuming such an exogenous investment motive is, however, not enough, as one also need to restrict optimizing investors from purchasing assets in the foreign country. If not, these investors would arbitrage as much as possible between the high-return foreign asset and the low-return domestic one. Without additional restrictions, this arbitrage would be infinite, which is not consistent with equilibrium. Generating large predictable gaps in rates of returns in a theoretical setting thus requires the imposition of very strong restrictions that are clearly unappealing.

INTERPRETATION OF THE EXTERNAL ACCOUNTS While from a theoretical perspective, financial integration does not offer a powerful adjustment channel to address imbalances, in addition to the standard trade one, it still impacts the interpretation of the current account. First, financial integration affects the current account that a country can run along a balanced growth path. It has long been understood that a country with a growing GDP can run a current account deficit and still maintain a constant ratio of net foreign debt to GDP, as long as the deficit does not add additional external liabilities faster than the economy grows. Consider for instance a country that grows by 5% a year, and has net external liabilities of 50% of GDP on which it pays a 10% interest rate. With a balanced current account, growth would reduce the net liabilities to 47.5% of GDP after a year. Keeping the net external debt at 50% of GDP then allows the country to run a deficit of 2.5% of GDP that exactly offsets the dilution of the debt through growth. In this perspective, the sustainable current account deficit is a

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

220

24. VALUATION EFFECTS, CAPITAL FLOWS AND INTERNATIONAL ADJUSTMENT

direct function of the level of the country’s net foreign position. While the country can run a current account deficit, it still needs to run a trade surplus of 2.5% of GDP to cover the interest payments of 5% of GDP.6 The presence of valuation gains implies that the composition of external assets and liabilities matters in addition to the level of the net foreign position. Consider that the country’s gross assets amount to 100% of GDP. They consist entirely of equity holdings abroad that pay a dividend yield of 2%, with equity prices increasing by 8% a year, for a total rate of return of 10%. The country also has gross liabilities of 150% of GDP which pay an interest rate of 10% with no valuation gain. Gross assets and liabilities then have the same rate of return. The net asset income of the country (dividend minus interest) now amounts to a deficit of 13% of GDP. Keeping the net foreign position constant at 50% of GDP is still consistent with a trade surplus of 2.5% of GDP. While this implies a much larger current account deficit than in the previous example (10.5% instead of 2.5%), this is now offset by valuation gains on the equity assets equal to 8% of GDP, so the change in the net position remains at 2.5% of GDP, which exactly offsets the impact of GDP growth. The increase in financial globalization since the mid1990s then implies that countries whose asset composition is more tilted toward assets that earn a return through capital gains instead of yields (such as equity) that its liabilities can maintain a larger current account deficit along a balanced growth path. By contrast, the sustainable trade balance is not affected by the composition of assets and liabilities, and is thus a better gauge of long-term sustainability. Another consequence of financial globalization is that projections of countries’ net external positions have become more uncertain. When valuation gains played a minor role, one could project a baseline path for the net foreign asset position relying only on a forecast for the current account. Given the limited volatility of the current account, the interval of confidence around the baseline projection for the net foreign asset position was then fairly narrow. Things are quite different in the presence of large potential valuation gains. First, the baseline projection is also affected by what one assumes for the average gap in the rate of return between asset and liabilities. Second, and more importantly, these rates of returns include highly volatile capital gains, and are thus estimated with a large interval of confidence. The range of confidence around any projection of the net foreign asset position has thus become much broader. While a country that is a net creditor used to be able to count on its net assets

6

for many years, its position can now change quickly because of adverse developments in financial markets. A similar pattern also emerges in the current account, even though it is not directly affected by valuation effects. This is because the earning streams on assets are now playing a larger role, next to trade flows, than they used to. This development is a direct consequence of the large increase in the value of external assets and liabilities. While yields are far less volatile than the rates of capital gains, they are still more volatile than real variables, and as a consequence earnings streams are more volatile than trade flows. The rising role of earnings in the current account could then lead to current accounts that are more volatile and harder to forecast with confidence. While one could see the heightened volatility of the net foreign asset position and the current account with concern, this inference is not necessarily valid. In fact, valuation gains and losses can simply reflect risk sharing.

CONCLUDING REMARKS Financial globalization has been one of the main developments in the global economy since the mid-1990s. It has led to a renewed interest in bringing endogenous portfolio choice into the workhorse macroeconomic models, and the literature has made substantial advances in recent years. Looking forward, a promising avenue of research is to enrich the various dimensions along which investors differ across countries. The literature so far has focused on different exposures to shocks, for instance through wage income on which workers cannot issue claims to be sold to investors. Investors, however, differ in additional ways. One particularly relevant aspect is to include asymmetries in information and knowledge, with investors being for instance better informed of the prospects in their own country than abroad. Researchers have shown that heterogeneous information can substantially enrich the theoretical understanding of international portfolios and their connection to the real economy. An additional line of research is to focus on the prominent role that some categories of investors play in financial markets. While mutual and pension funds account for the vast majority of asset holdings, the short-term functioning of markets could be more dependent on investors such as hedge funds that are smaller, but trade more actively. Shocks hitting these investors could then have a disproportionately large impact on markets and international capital flows.

The current account (2.5%) is the sum of the trade balance (þ2.5%) and the net asset income (5%).

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

CONCLUDING REMARKS

SEE ALSO Theoretical Perspectives on Financial Globalization: Intertemporal Approach to the Current Account; Endogenous Portfolios in International Macro Models; Financial Development and Global Imbalances; International Trade and International Capital Flows.

Glossary Capital flows The accumulation of new claims by residents on the rest of the world (gross outflows), or of new claims by foreign investors on the country (gross inflows). Net capital flows (outflows minus inflows) correspond to the current account balance. Complete markets The situation where financial markets include enough different assets to cover all possible contingencies. Earnings The streams of payments to the owners of an asset, in the form of dividends or interest payments. Exorbitant privilege The ability of a country to earn a higher return on its foreign assets than it pays on its liabilities. Expected valuation effects The effects in a period that can be foreseen prior to knowing the value of shocks in the period. The effects that cannot be foreseen are the unexpected valuation effects. Financial globalization The increase in the value of countries’ claims on the rest of the world, as well as of the value of their liabilities to foreign investors. Gross assets The value of claims by a country’s residents on the rest of the world. Gross liabilities The value of claims by foreign investors on a country. Hedging The ability of a portfolio to generate returns that offset the returns from other sources, such as wages. Net foreign assets The difference between gross assets and gross liabilities. Rate of return The ratio of earnings and valuation effects in a period to the value of the corresponding holdings at the beginning of the period. Valuation effect The change in the value of assets (or liabilities) driven by changes in asset prices or the exchange rate. It is also referred to as capital gains (or losses). Yield The ratio of earnings in a period to the value of the corresponding holdings at the beginning of the period.

Further Reading Bosworth, B., Collins, S.M., Chodorow-Reich, G., 2007. Returns on FDI: does the U.S. really do better? Brookings Trade Forum 2007: FDI. Brookings Institution Press, Washington, DC, pp. 201–206. Caballero, R.J., Farhi, E., Gourinchas, P., 2008. An equilibrium model of ‘Global imbalances’ and low interest rates. American Economic Review 98, 358–393. Coeurdacier, N., Gourinchas, P.O., 2008. When bonds matter: home bias in goods and assets. Federal Reserve Bank of San Francisco Working Paper 2008–2025. Curcuru, S., Dvorak, T., Warnock, F.E., 2007. The stability of large external imbalances: the role of returns differentials. International Finance Discussion Paper 894. Curcuru, S., Dvorak, T., Warnock, F.E., 2008. Cross-border returns differentials. Quarterly Journal of Economics 123, 1495–1530.

221

Devereux, M.B., Sutherland, A., 2010. Valuation effects and the dynamics of net external assets. Journal of International Economics 80, 129–143. Devereux, M.B., Sutherland, A., 2011. Country portfolios in open economy macro models. Journal of the European Economic Association 9 (2). Engel, C., Matsumoto, A., 2009. The international diversification puzzle when goods prices are sticky: it’s really about exchange-rate hedging, not equity portfolios. American Economic Journal: Macroeconomics 12, 155–188. Evans, M., Hnatkovska, V., 2007. Financial integration, macroeconomic volatility, and welfare. Journal of the European Economic Association 5, 500–508. Evans, M., Hnatkovska, V., 2008. A method for solving general equilibrium models with incomplete markets and many assets. National Bureau of Economic Research Technical Working Paper 318. Gourinchas, P., Rey, H., 2005. From world banker to world venture capitalist: U.S. external adjustment and the exorbitant privilege. In: Clarida, R. (Ed.), G7 Current Account Imbalances: Sustainability and Adjustment. The University of Chicago Press, Chicago. Gourinchas, P., Rey, H., 2007. International financial adjustment. Journal of Political Economy 115, 665–703. Habib, M., 2010. Excess returns on net foreign assets: the exorbitant privilege from a global perspective. European Central Bank Working Paper 1158. Hau, H., Rey, H., 2008. Global portfolio rebalancing under the microscope. CEPR Discussion Paper 6901. Hellerstein, R., Tille, C., 2009. The Changing Nature of the U.S. Balance of Payments. Current Issues in Economics and Finance, vol. 14. Federal Reserve Bank of New York, New York. Higgins, M., Klitgaard, T., Tille, C., 2007. Borrowing without debt? Interpreting the U.S. international investment position. Business Economics 42, 17–27. Kraay, A., Ventura, J., 2000. Current accounts in debtor and creditor countries. Quarterly Journal of Economics XCV, 1137–1166. Lane, P.R., Milesi-Ferretti, G.M., 2003. International financial integration. International Monetary Fund Staff Papers 50. Lane, P.R., Milesi-Ferretti, G.M., 2007. A global perspective on external positions. In: Clarida, R. (Ed.), G7 Current Account Imbalances: Sustainability and Adjustment. The University of Chicago Press, Chicago. Lane, P.R., Shambaugh, J.C., 2010. Financial exchange rates and international currency exposures. American Economic Review 100, 518–540. Meissner, C.M., Taylor, A., 2006. Losing our marbles in the new century? The great rebalancing in historical perspective. National Bureau of Economic Research Working Paper 12580. Obstfeld, M., 2004. External adjustment. Review of World Economics 140, 541–568. Pavlova, A., Rigobon, R., 2010. An asset pricing view of the international adjustment process. Journal of International Economics 80, 144–156. Tille, C., 2008a. Composition of international assets and the long run current account. Economic Notes by Banca Monte dei Paschi di Siena 37, 283–313. Tille, C., 2008b. Financial integration and the wealth effect of exchange rate fluctuations. Journal of International Economics 75, 283–294. Tille, C.E., van Wincoop, E., 2008. International Capital flows under dispersed information: theory and evidence. NBER Working Paper 14390. Tille, C., van Wincoop, E., 2010. International capital flows. Journal of International Economics 80, 157–175.

II. THEORETICAL PERSPECTIVES ON FINANCIAL GLOBALIZATION

Intentionally left as blank

S E C T I O N I I I

SAFEGUARDING GLOBAL FINANCIAL STABILITY

Intentionally left as blank

C H A P T E R

25 Safeguarding Global Financial Stability, Overview J.R. Barth*,†,{, D.G. Mayes}, M.W. Taylork *Auburn University, Auburn, AL, USA Milken Institute, Santa Monica, CA, USA { Wharton Financial Institutions Center, Philadelphia, PA, USA } University of Auckland, Auckland, New Zealand k Bank of International Settlements †

O U T L I N E Financial Stability

225

Global Approaches

229

Establishing and Maintaining Financial Stability

227

Crisis Management and Avoidance

228

Other Issues References

229 230

FINANCIAL STABILITY Tommaso Padoa-Schioppa once remarked that a concern with financial stability was part of the ‘genetic code’ of central banks (Padoa-Schioppa, 2004). Similarly, Charles Goodhart’s monograph on the Evolution of Central Banks (Goodhart, 1988) emphasized that the function of safeguarding financial stability had historically been the primary purpose of central banks. Yet, for several decades prior to the financial crisis of 2007–08, the financial stability function of central banks remained in abeyance, as their core focus was increasingly on the manipulation of short-term interest rates in pursuit of the goal of price stability. The pursuit of inflation targeting and the adoption of quantifiable goals for price stability reinforced this trend. By contrast, financial stability was seen as at best a subsidiary aim of policy that would be attained as an inevitable consequence of achieving a central bank’s price stability objective. This is not to suggest that financial stability was completely neglected during this entire period. From the mid-1990s onward, central banks did establish specific departments labeled ‘financial stability’ that had responsibility for producing detailed assessments of

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00047-7

financial sector vulnerabilities, often published in the form of Financial Stability Reports. However, these departments – and the central banks in which they were located – often lacked specific tools with which to back their analyses with policy actions. This situation did not prevent the emergence of a fruitful academic and policy debate on how financial stability was to be defined and measured, but it did result in the relative neglect of the stability of the financial sector as an explicit objective of policy. A further factor reinforcing this state of affairs was the fashion of removing responsibility for banking supervision from central banks and handing it over to a single financial services regulator. These unified regulatory agencies tended to focus on microprudential regulation at the level of the individual firm. With central banks increasingly focused on the price stability objective, financial stability was too often left to fall between the institutional gaps – as was arguably the case in the United Kingdom in the run-up to the global financial crisis that began to emerge fully in 2007. Since the financial crisis, however, the pursuit of financial stability has been restored to its place as a core central-banking function. There has been a renewed

225

# 2013 Elsevier Inc. All rights reserved.

226

25. SAFEGUARDING GLOBAL FINANCIAL STABILITY, OVERVIEW

emphasis on macroprudential analysis and tools, while new decision-making mechanisms (such as the Financial Stability Oversight Council in the United States, the United Kingdom’s Financial Policy Committee, and the European Systemic Risk Board) have been established. Nonetheless, important issues remain to be addressed. First, while the banking sector has been traditionally viewed as the primary source of potential financial instability, many believe the chain of collapse that ran through the financial system in 2007–08 following the bursting of the housing bubble in the United States originated in the nonbank sector and then spread to the banks. It can, of course, be argued that the severity of the crisis was a consequence of the fact that the soundness of the banking system was itself brought into question. Nonetheless, many consider that the immediate trigger of the worst part of the global crisis was the failure of an investment bank, Lehman Brothers. In addition, the US government took the position that AIG, an insurance company, needed to be bailed out because of its centrality in the system. Understanding the complex interactions between the different parts of the financial system and how these can create financial instability is a major part of the postcrisis reform agenda. In particular, it involves knowledge of the nature of the interconnections between various institutions and the way shocks might be transmitted through the financial system. Among the issues that are being explored given these complexities are the authorities’ ability to collect the data needed to perform financial stability analysis, including from unregulated or lightly regulated parts of the financial system; the effect of various policy measures in reducing financial system interconnectedness; and last – but by no means least – where to draw the regulatory boundary. A particularly tricky issue concerns the extent and nature of regulatory oversight of the ‘shadow’ (or, less pejoratively, the ‘market-based’) banking system comprising money market mutual funds, some hedge funds, and certain other nonbank financial intermediaries. The common feature of this sector is that it engages in the same type of maturity transformation and has the same or even greater ability to take on leverage as the regulated banking sector. However, lacking the backstop of the financial safety net, it is exposed to the same risk of runs that the formal banking sector experienced before the advent of lender of last resort (LOLR) assistance and deposit insurance. Second, there has been growing recognition that the focus of financial stability analysis and monitoring has to be on the complex interaction of the elements of the system as a whole and not just its individual components. One of the major failures of the supervisory system during the period prior to global financial crisis that fully emerged in September 2008 was to view systemic stability in terms of the aggregation of individual financial firms. If all financial institutions operated

prudently, then shocks would be unlikely to have systemic implications. This incarnation of the fallacy of composition has been dramatically shown to not be the case, particularly as a large portion of the entire sector could be subjected to the same shock, as was the case with collapsing real estate prices in the United States. Third, financial stability policy has shifted to a more overt concern with the ‘plumbing’ of the financial system. Starting in the 1980s, central banks had already moved to reduce payment system risk through the adoption of realtime gross settlement (RTGS) systems. RTGS reduced the systemic risk resulting from the buildup of large overdrafts intraday in traditional systems which settled on an end-ofday net basis. In addition, attention had also been given to improving the robustness of settlement systems used in securities trading, for example, by the implementation of delivery-versus-payment (DvP) technologies. Since the financial crisis, this focus has continued, with increasing emphasis on bringing a large portion of the over-the-counter derivatives markets into centralized clearing systems. The aim of these initiatives, like the earlier ones on RTGS and DvP, is to reduce the risk that the failure of one firm to meet its obligations will set off a chain reaction throughout the financial system. Finally, there has been an increased awareness of the international dimension of financial stability. The interconnectedness of the financial system is not just a fact of life within national jurisdictions, but it also extends beyond national boundaries. As events of 2007–08 demonstrated, the fragility of a specific part of the US financial system – subprime lending for real estate purchase – could be rapidly transmitted through securitization to the global financial system. The transmission mechanisms were many and varied: European banks had been large purchasers of securitized assets based on subprime loans; credit default swaps written or purchased on these same instruments became another instrument of potential contagion. Following the collapse of Lehman Brothers, however, the chief mechanism by which financial instability spread around the global financial system was the traditional one of a loss of confidence leading to the fear that one could not be sure about the soundness of any financial institution. And the worst enemy of well-functioning financial markets is a loss of confidence or uncertainty. These events brought home very clearly that global financial stability is a global public good that may be inadequately supplied unless more formal mechanisms can be established to coordinate policies to assure financial stability better. Among the postcrisis institutional innovations, the formation of the Financial Stability Board, a mechanism of the G20 member countries to coordinate financial stability policy, is designed to meet this need. Given these brief comments, the first task of the contributions to this section is to sort out the fundamental issues, such as addressing the issue of how to define the term ‘financial stability,’ as there is no neat,

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

ESTABLISHING AND MAINTAINING FINANCIAL STABILITY

quantifiable equivalent to an inflation target. And while the risks in the financial system need to be managed, they cannot be eliminated. Indeed, it would not make any sense to try to do so, as managed risk-taking lies at the heart of successful financial intermediation. In a second paper, there is a debate concerning which agency should do what and how the interactions between those responsible for the stability of the system as a whole and those responsible for the stability of individual institutions should be coordinated. A paper by Dirk Schoenmaker (‘The Role of Central Banks in Financial Stability’) specifically examines the role of central banks. It also discusses how accountability should take place for financial stability. Central banks have become increasingly independent in order to conduct monetary policy focusing on the medium term, which is beyond many governments’ electoral horizon. But at the same time, their performance in this regard is readily measurable and they can be held accountable for their actions. The same accountability is not possible in the case of financial stability. Supervisory failures are easy to identify but not successes. It is quite difficult to decide whether financial stability has been achieved because of or despite supervisory actions. Furthermore, it is difficult to assign any grades for relative successes and failures. This issue is therefore the topic of a specific paper (Michael Taylor, ‘The Independence and Accountability of Regulatory Agencies’). Still another paper considers what regulation and supervision can hope to achieve and where the boundaries of its oversight should be drawn (David Llewellyn, ‘Safeguarding Global Financial Stability: Role and Scope of Regulation and Supervision’).

ESTABLISHING AND MAINTAINING FINANCIAL STABILITY It is difficult to take these discussions much further without understanding what establishing and maintaining the stability of the financial system entails. Normally, this involves a combination of having an appropriate structure on the one hand and trying to work out how various specific risks might be handled on the other. Clearly, this second area of activity also involves an ongoing analysis of the likelihood of such risks over time. An assessment of these risks forms an obvious extension of the activities of central banks in trying to establish the needs of monetary policy, with this area being labeled macroprudential regulation, much in parallel with macroeconomic policy. However, it is one thing to provide risk assessment but quite another to control risk. There is a considerable debate about the appropriate tools for macroprudential policy. On the structural side, clearly there are arguments for ensuring that the system is robust to the failure of any one institution, or where a monopoly is difficult to avoid, such as in payments and

227

settlement, that the system should be extremely robust with backing from the regulatory authorities. The argument is then about who has responsibility for ensuring the appropriate structure. It could well be the supervisory authority but the degree of concentration among financial institutions may properly be considered the preserve of the competition authority. Allowing the macroprudential authority to take and implement such a decision is not currently seen as the way to proceed. The process for handling potential shocks becomes much more difficult. There are few specific tools that are different from those applied by those regulating individual institutions or those conducting monetary policy. In the case of monetary policy, there is a clear link between stability of the overall price level as reflected in consumer prices and measures of asset prices. While the link with stock prices may be limited, the link with property prices is strong. House price booms and slumps are highly correlated with the cycles in economic activity and hence inflation. Thus, a monetary policy aiming to control inflation will react to asset price increases. There will also be feedback in the opposite direction as monetary policy attempts to smooth the shape of the economic cycle and hence asset prices. There is a considerable debate about how explicit the focus on asset prices should be in monetary policy, but they are clearly indicators, albeit noisy ones. On this basis, therefore, there would be little conflict between the two objectives of price and financial stability to be pursued using the single interest rate instrument. However, implementation is obviously not that straightforward or the financial pressure before the global financial crisis would not have been so great. This same debate on the extent to which asset prices should be a target in themselves extends to financial stability. Under the terms of the so-called Greenspan standard, it was suggested that due to the difficulty in deciding whether an asset price boom was based on sustainable increases in future returns or was simply a temporary bubble that a relatively benign approach involving lean against it cautiously was appropriate, as long as it was possible to intervene rapidly if the bubble collapsed to avoid adverse consequences to the real economy. That appeared to be very much what was achieved in the dot. com boom at the beginning of the century. The failure to pursue such an approach prior to the global financial crisis was one of the major contributory causes of the crisis. The second facet of the debate is over adding new instruments purely for financial stability purposes, but which would not conflict with monetary policy objectives. These could include loan-to-value ratios as well as dynamic loan loss provisioning that has been applied in Spain inter alia. The advantage of the dynamic provisioning route is that it is automatic and does not require decision-making by the authorities. Banks have an incentive to provide for extra provisioning for losses in good times because it is tax-efficient to do so and the cushion is

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

228

25. SAFEGUARDING GLOBAL FINANCIAL STABILITY, OVERVIEW

reduced when the losses are taken. To some extent, it would also be possible to make loan-to-value ratios and other measures automatic. There could be automatic rules which prescribe how far loan-to-value ratios need to fall as home prices rise evermore rapidly. Required capital ratios could similarly vary with the state of the economic cycle rather than be based on a risk assessment by the authorities. Nevertheless, monitoring of all such measures would be on a firm-by-firm basis and hence more sensibly acted upon by the microprudential regulator. There are therefore several papers in this section dealing with the relationship between financial stability policy and monetary policy (Ken Kuttner, ‘Financial Stability and Inflation Targeting’), and between macro- and microprudential policy (Avinash Persaud, ‘Micro and Macro Prudential Regulation’). Because of its importance as a real rather than a potential instrument, there is a paper on dynamic provisioning (Alicia GarciaHerrero and Santiago Ferna´ndez de Lis, ‘Spain: Dynamic Provisioning to Reduce Procyclicality’) in addition to a paper on the types of countercyclical tools that could be applied, including loan-to-value and capital ratios (Andrew Haldane, ‘Macroprudential Policy Tools – Countercyclical Capital, LTV Ratios’).

CRISIS MANAGEMENT AND AVOIDANCE Finally, in terms of the scope of topics covered, the tools that can be applied, and the structures that are required, it is necessary to explore how the whole process of crisis management and avoidance can be structured. Much of what has been described so far is a system that operates in good times and in the hope that difficulties will not occur. The authorities also need to be prepared for potential failures or serious threats to the system. One of the major problems in the global financial crisis was countries that found they did not have the powers to resolve problems in the way they would have liked. The simplest example is the case of Lehman Brothers in September 2008. As it was not a depository institution, it fell outside the special resolution regime that provides for early intervention and a means to resolve problems without the need to halt operations in a material way or to draw on taxpayer funding. However, this was not a purely US phenomenon. The United Kingdom found it could not resolve the problems of Northern Rock in late 2007 without nationalization and hence led to a reform of its entire banking resolution regime. The major problem in these situations was simply the inability to handle large organizations, a version of what is called ‘too big to fail’ (TBTF). Smaller banks can be allowed to fail or have their activities transferred to a healthier institution without any serious threat to the stability of the system. But the same is not true for a large

bank. In all cases, nearly every country thinks that financial stability requires that they protect smaller depositors against loss, although there is considerable variation in the definition of ‘small.’ It is therefore necessary to establish a deposit insurance fund of some form to cover the costs of failures. As illustrated in the United States, the authorities require funding to transfer operations to other providers of the same services. However, the European countries feel they will also need explicit resolution funds to handle all eventualities. Discussions over handling the TBTF institutions are continuing, with considerations involving their structure to make it possible to take appropriate action in the time available. Given their complex structures, it has proved too difficult to make an assessment in sufficient time for an alternative to the government providing loans, equity, or some form of nationalization to keep them operating. Now there is an emphasis on having these TBTF institutions provide living wills and funeral plans describing how they can recover in the case of difficulty or be resolved if this proves insufficient. One of the most straightforward areas of debate concerns whether banks should be divided into retail operations that need to be kept operating and investment and other operations that are not so essential. This arrangement is recommended by the Vickers Review of banking in 2011 in the United Kingdom. Only the retail side of a bank would require heavy capitalization. Other ideas for reform include contingent capital, in which debt is converted to equity in conditions of difficulty. In a crisis, not only do asset values fall but the ability and willingness of others to subscribe to new equity can also disappear. Contingent capital has already been subscribed in some instances and hence does not have this desirable feature. The importance of contingent capital and other related measures is that they can take effect before problems mount to the point that they are too large to handle. Most countries do not have mandatory Prompt Corrective Action as part of their arsenal to address problems, but they are increasingly being provided with the ability to act before a financial institution becomes fully insolvent and when it is merely undercapitalized according to the existing rules. Four papers cover these key issues, with the first concentrating on what can be done in designing the system to make crises less likely (David Mayes, ‘Lines of Defense Against Systemic Crises: Prevention’). Simply having a credible means of resolving all banks and the requirement of prompt corrective action is likely to act as a considerable positive incentive for banks to behave prudently, otherwise the shareholders would be wiped out and the management would lose their jobs. The second paper deals with resolution methods, focusing on individual institutions, including those previously

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

229

OTHER ISSUES

thought too big to fail (Eva Hupkes, ‘Challenges in Safeguarding Financial Stability, Lines of Defense Against Systemic Crises: Resolution’), whereas the third paper extends this analysis of crisis resolution to the financial system as a whole (Luc Laeven, ‘Resolution of Banking Crises’). The final paper looks at the improvements that can be made to the regulation of financial markets to improve stability. Some of the difficulties in the global financial crisis, for example, related to the operation of derivatives market. The measures to improve the ability to handle individual institutions therefore need to be accompanied by measures to make markets more secure, more transparent, and less likely to failure. The LOLR function is not dealt with explicitly here although it is mentioned later in the discussion in the context of developments in Europe (Jose´ Manuel Gonza´lez-Pa´ramo, ‘Innovations in the Lender of Last Resort Policy in Europe’). The problem with LOLR has been twofold. The first is that the scale on which it has been required has been enormous, resulting in a major expansion of central bank balance sheets. The second is that when a central bank steps in to provide lending to banks that cannot obtain it through the market but against good collateral and at a premium rate of interest, it acts as a stigma. The original idea was that when a central bank lent in this manner, it would be viewed by other lenders as a source of confidence. Now it is seen as a source of weakness on the part of banks. Various proposals have been made to rectify this problem.

GLOBAL APPROACHES Up until this point, the discussion has been deliberately about safeguarding financial stability at the national level, not only because this is where most of the action has been but because almost all of the relevant authorities are national. However, the problem is clearly global or at least cross-border, not simply because of the existing linkages among international markets but because many of the main institutions operate across borders. This requires cooperation and coordination among the authorities in the countries involved not just in a crisis but all the time while the institution is being supervised and monitored. This group of papers begins with two that set the overall context. The first is historical, looking at how the international financial architecture has developed (Forrest Capie, ‘The Development and Evolution of the International Financial Architecture’), while the second considers the structure as it exists and is likely to develop. This is followed by papers dealing with each of the main institutions: the IMF (Gillian Garcia), the Basel Committee for Banking Supervision, the WTO (Douglas Arner), the Financial Stability Forum/Board (Larry Promisel and Kate Langdon), and the Gn bodies (Andrew Baker), placing considerable

emphasis on the G20, which has taken the lead in the present crisis, taking over from the G10, which had developed the Basel recommendations. This set of papers is then complemented by a series of papers looking at the role and progress of international bodies operating at the regional level. Four of these papers look explicitly at Europe where progress has been considerable as the European Union (EU) has struggled to address the cross-border problems in the face of a strong pressure from the member states to retain control at the national level. Thus, there is no European Deposit Insurance Corporation or Resolution Authority and no European level banking charter – all measures that the United States has found necessary in the handling of countrywide financial stability in the period since the Great Depression. A number of collaborative authorities have been set up: the European Banking Authority, the European Securities Market Authority, and the European Insurance and Occupational Pensions Authority. At the macroprudential level, the European Systemic Risk Board has been created, although it is largely a monitoring institution, supported by the European Central Bank (ECB), with no powers other than revealing the risks to all the competent authorities, particularly at the national level. In addition to a paper summarizing this progress (Kern Alexander, ‘Changing Market Structures and Market Abuse’) and another analyzing supervision in Europe (Dirk Schoenmaker, ‘Financial Supervision in the EU’), there is a third looking at the evolution of the LOLR’s role in Europe (Jose´ Manuel Gonza´lez-Pa´ramo, ‘Innovations in the Lender of Last Resort Policy in Europe’). This paper is extremely topical in light of the euro crisis that has characterized the global financial crisis since the May 2010 rescue package for Greece. Thomas Willett (‘The Variety of European Crises’) provides a longer term view, covering a variety of experience, thereby offering a caution that crises vary and hence measures put in place to handle future crises have to be able to handle something different and not just a repeat of what has been experienced recently.

OTHER ISSUES The final three papers are a little different. The first of these looks at the financial sector assessments made by the IMF and the World Bank (FSAPs) (Susan Marcus, ‘The Financial Sector Assessment Program’). Although FSAPs have been around for a decade or more, it is not clear what they have achieved as the global crisis occurred before many of them could be acted on fully. The notable point, however, is that the problems of having a wellfunctioning and structured financial system were already being put in place at the time of the crisis. Perhaps given

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

230

25. SAFEGUARDING GLOBAL FINANCIAL STABILITY, OVERVIEW

some luck and more time, the crisis might have been largely avoided in the absence of a major shock but in general what we are seeing is ‘too little too late.’ This leaves two more papers on controversial issues to complete the section. The first, on resolving cross-border banks (Penny Angkinand and Clas Wihlborg, ‘Crossborder Banking in Subsidiaries and Branches: Organization, Supervision and Resolution’), strikes at the heart of the difficulties that have yet to be fully resolved. While a home country, where the parent organization is located, is responsible for regulation, supervision, and resolution of the parent, its branches, domestic subsidiaries, and the group as a whole, host countries have responsibility for the subsidiaries in their jurisdiction and, outside the EU, also often for branches. The problem is that the home country may be unwilling to shoulder the whole burden in the event of failure or may choose a solution that will result in financial instability in host countries. Yet, host countries cannot unilaterally maintain the stability of subsidiaries and branches within their jurisdiction unless they make these activities highly separable. A collaborative effort by national regulators requires working across different legal systems, whose entire basis may conflict – where one party applies territoriality in insolvency proceedings and another, universality. Even where countries all applied universality and sought to work collaboratively within the EU, there were major problems, as illustrated by the case of Fortis and the three main Icelandic banks (which were subject to EU rules as Iceland is a member of the European Economic Area). In an attempt to halt the outflow of assets in the last stage of the crisis, the United Kingdom had to invoke legislation primarily designed to combat terrorism and money laundering. New Zealand has a workable scheme by insisting that all foreign-owned systemic banks are not merely domestically incorporated and separately capitalized,

but that they can operate independently from their parent within the transaction day (which would at maximum be over a weekend). However, if this were widely adopted, it could remove much of the cost advantage of operating across borders. While the home country could also solve a problem through recapitalizing the entire group, as was done for AIG, for example, halfway houses are likely to be messy. The final topic covered is compensation practices (Peter Sinclair, ‘The Advantages and Drawbacks of Bonus Payments in the Financial Sector’), which appear to have contributed considerably to financial instability, both by causing an emphasis on short-term returns and by allowing many of those in the financial chain to collect all their fees up front and have no stake in the continuing quality of the assets they have sold, rated, or serviced. These issues can be addressed and the incentives of those assisting aligned more with those of the purchasers of the securities and the shareholders, but much remains to be done and the opposition from the industry is considerable.

SEE ALSO Political Economy of Financial Globalization: Interest Group Politics; Theoretical Perspectives on Financial Globalization: Financial Contagion; Theory of Sovereign Debt and Default.

References Goodhart, C.A.E., 1988. The Evolution of Central Banks. MIT Press, Cambridge, MA. Padoa-Schioppa, T., 2004. Regulating Finance: Balancing Freedom and Risk. Oxford University Press, Oxford.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

C H A P T E R

26 Resolution of Banking Crises L. Laeven, F. Valencia Research Department of the International Monetary Fund, CEPR, Westminster, England, UK Research Department of the International Monetary Fund, Westminster, England, UK O U T L I N E Introduction

231

The 2007–09 Global Crisis: A Synopsis

233

Which Countries Had a Systemic Banking Crisis in 2007–09?

234

Policy Responses in the 2007–09 Crises: What Is New?

235

INTRODUCTION Systemic banking crises are disruptive events not only to financial systems but to the economy as a whole. There is a large literature on the real effects of banking crises providing empirical support to this claim (e.g., Dell’Ariccia et al., 2008a,b; Krozner et al., 2007; Peek and Rosengren, 2000). Systemic crises are not specific to the recent past or specific countries – almost no country has avoided the experience and some have had multiple banking crises. While the banking crises of the past have differed in terms of underlying causes, triggers, and economic impact, they share many commonalities.1 Banking crises are often preceded by prolonged periods of high credit growth and are often associated with large imbalances in the balance sheets of the public or private sector, such as maturity mismatches or exchange rate risk, that ultimately translate into credit risk for the banking sector. Since 2007, the world has experienced a period of severe financial stress, not seen since the time of the Great

How Costly Are the 2007–09 Systemic Banking Crises?

245

Concluding Remarks

248

Appendix Glossary Acknowledgment References

250 257 257 257

Depression. This crisis started with the collapse of the subprime residential mortgage market in the United States and spread to the rest of the world through exposure to US real estate assets (often in the form of complex financial derivatives) and a collapse in global trade. Many countries were significantly affected by these adverse shocks, causing systemic banking crises in a number of countries, despite extraordinary policy interventions. Choosing the optimal path to resolve a financial crisisand accelerate economic recovery is far from unproblematic. There has been little agreement on what constitutes best practice or even good practice. Many approaches have been proposed that have tried to resolve systemic crises more efficiently. Part of these differences may arise from varying policy objectives. Some policies have focused on reducing the fiscal costs of financial crises, others on limiting the economic cost in terms of lost output and on accelerating restructuring, whereas again others have focused on achieving long-term, structural reforms. Trade-offs are likely to arise between these

1

For a review of the literature on macro origins of banking crises, see Lindgren et al. (1996), Dooley and Frankel (2003), and Collyns and Kincaid (2003).

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00043-X

231

# 2013 Elsevier Inc. All rights reserved.

232

26. RESOLUTION OF BANKING CRISES

objectives.2 Governments may, for example, through certain policies consciously incur large fiscal outlays in resolving a banking crisis, with the objective to accelerate recovery. Or structural reforms may only be politically feasible in the context of a severe crisis with large output losses and high fiscal costs. This article reviews the recent systemic banking crises, and contrasts them with past crises to shed light on how the cost of this crisis compares to that of previous banking crisis episodes, and how policy responses differ from those of the past? The article presents new and comprehensive data on the starting dates and characteristics of systemic banking crises over the period 1970–2009, building on earlier work by Caprio et al. (2005), Laeven and Valencia (2008), and Reinhart and Rogoff (2009). In particular, it updates the Laeven and Valencia (2008) database on systemic banking crises to include the recent episodes following the US mortgage crisis of 2007. The update makes several improvements to the earlier database, including an improved definition of systemic banking crisis, the inclusion of crisis ending dates, and a broader coverage of crisis management policies. The result is the most up-to-date banking crisis database available.3 This article contributes to the literature on banking crisis management. Recent contributions to this literature include Borio et al. (2010), who compare the recent crisis with the Nordic crisis of the 1990s to assess the extent to which the recent policies adhered to the principles of bank restructuring and conclude that the measures taken so far remain less comprehensive and deep compared to the intensity of restructuring achieved in the Nordic countries following the Nordic crisis. Boudghene et al. (2010) present a descriptive analysis of institution-specific asset relief measures in the European Union. Panetta et al. (2009) present an analysis of intervention measures and their impact on market sentiment during the recent crises. This article differs from these and similar studies in two important ways. First, the coverage in terms of crises is broader, since all systemic banking crisis events since 1970 are taken into account. Second, it considers a broader set of intervention policies and outcomes, by comparing the impact of conventional and nonconventional intervention and stabilization measures on public debt, output losses, and direct fiscal costs arising from interventions in the financial sector. Furthermore, the comparison here is more homogeneous in the sense that the effects of government interventions are compared only across crisis episodes.

The new data show that there are many commonalities between recent and past crises, both in terms of underlying causes and policy responses, yet there are also some striking differences in the economic and fiscal costs associated with the new crises. The economic cost of the new crises is on average much larger than that of past crises, both in terms of output losses and increases in public debt. The median output loss (computed as deviations of actual output from its trend) is 25% of gross domestic product (GDP) in recent crises, compared to a historical median of 20% of GDP, while the median increase in public debt (over the 3-year period following the start of the crisis) is 24% of GDP in recent crises, compared to a historical median of 16% of GDP. These differences in part reflect an increase in the size of financial systems, the fact that the recent crises are concentrated in highincome countries, and possibly differences in the size of the initial shock to the financial system. At the same time, direct fiscal costs to support the financial sector were smaller this time at 5% of GDP, compared to 10% of GDP for past crises, as a consequence of relatively swift policy action and the significant indirect support the financial system received through expansionary monetary and fiscal policy, the widespread use of guarantees on liabilities, and direct purchases of assets that helped sustain asset prices. Policy responses broadly consisted of the same type of containment and resolution tools as used in previous crisis episodes. As in past crises, policymakers used extensive liquidity support and guarantees. However, recapitalization policies were implemented more quickly this time around. While in previous crises it took policymakers about 1 year from the time that liquidity support became extensive before comprehensive recapitalization measures were implemented, this time recapitalization measures were implemented around the same time that liquidity support became extensive. While all these extraordinary measures have contributed to reduce the real impact of the recent crisis, they also have increased the burden of public debt and the size of fiscal contingent liabilities, raising concerns about fiscal sustainability in a number of countries. The article proceeds as follows. The section ‘The 2007– 09 Global Crisis: A Synopsis’ presents a brief review of the events that led to the 2007–09 global crises. The section ‘Which Countries Had a Systemic Banking Crisis in 2007–09?’ defines a systemic banking crisis and presents a list of countries that meet this definition. The section ‘Policy Responses in the 2007–09 Crises: What Is

2

For an overview of existing literature on how crisis resolution policies have been used and the trade-offs involved, see Claessens et al. (2003), Hoelscher and Quintyn (2003), and Honohan and Laeven (2005).

3

The banking crisis dataset is publicly available at www.luclaeven.com/Data.htm.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

THE 2007–09 GLOBAL CRISIS: A SYNOPSIS

New?’ describes the policy responses and contrasts them with past crises. The section ‘How Costly Are the 2007–09 Systemic Banking Crises?’ presents the cost of recent crises and a comparison with past episodes. The section ‘Concluding Remarks’ concludes the article.

THE 2007–09 GLOBAL CRISIS: A SYNOPSIS Over the decade prior to the crisis, the United States and several other advanced economies experienced an uninterrupted upward trend in real estate prices, which was particularly pronounced in residential property markets. What originated this boom is still a source of debate, though there appears to be broad agreement that financial innovation in the form of asset securitization, government policies to increase homeownership, global imbalances, expansionary monetary policy, and weak regulatory oversight played an important role (e.g., Keys et al., 2010; Obstfeld and Rogoff, 2009; Taylor, 2009). The boom in real estate prices was exacerbated by financial institutions’ ability to exploit loopholes in capital regulation, allowing banks to significantly increase leverage while maintaining capital requirements. They did so by moving assets off balance sheets into specialpurpose vehicles, which were subject to weaker capital standards, and by funding themselves increasingly for short term and in wholesale markets rather than traditional deposits. These special-purpose vehicles were used to invest in risky and illiquid assets (such as mortgages and mortgage derivatives) and were funded in wholesale markets (e.g., through asset-backed commercial paper), without the backing of adequate capital. The growing importance of this shadow banking system highly dependent on short-term funding, combined with lax regulatory oversight, was a key contributor to the asset price bubble (Acharya and Richardson, 2009; Brunnermeier, 2009; Gorton, 2008). Higher asset prices led to a leverage cycle by which increases in home values led to increases in debt. The rise in asset prices decreased measured ‘value at risk’ at financial institutions, creating spare capacity in their balance sheets and leading to an increase in leverage and supply of credit (Adrian and Shin, 2008). A similar mechanism took place in the household sector, as perceived household wealth increased on the back of rising home values. Easy access to the equity accumulated in their homes led households to increase their leverage substantially. Mian and Sufi (2009) estimate that the average homeowner extracted 25–30 cents for every dollar increase in home equity to be used in real outlays. The asset price boom was further fueled by an explosion of subprime mortgage credit in the United States starting

233

in 2002 and reaching a peak around mid-2006 (Dell’Ariccia et al., 2008). The first signs of distress came in early 2007 from losses at US subprime loan originators and institutions holding derivatives of securitized subprime mortgages. However, these first signs were limited to problems in the subprime mortgage market. Later in 2007, these localized signs of distress turned into a global event, with losses spreading to banks in Europe (such as UK mortgage lender Northern Rock), and distress was no longer limited to financial institutions with exposure to the US subprime mortgage market. To alleviate liquidity shortages, the US Federal Reserve reduced the penalty to banks for accessing its discount window, and later that year created the Term Auction Facility. Similarly, a blanket guarantee was issued in the United Kingdom for Northern Rock’s existing deposits. Problems intensified in the United States with the bailout of Bear Stearns, and later in the year with the collapse of investment bank Lehman Brothers, and the government bailouts of insurer AIG and mortgage lenders Freddie Mac and Fannie Mae. By the end of 2007, many economies around the world suffered from a collapse in international trade, reversals in capital flows, and sizable contractions in real output. But as the crisis mounted, so did the policy responses, with many countries announcing bank recapitalization packages and other support for the financial sector in late 2008 and early 2009. While some aspects of this crisis appear new, such as the role of asset securitization in spreading risks across the financial system, it broadly resembles earlier boom–bust episodes, many of which followed a period of financial liberalization. One commonality among these crises is a substantial rise in private sector indebtedness, with the infected sectors besides banks being the household sector (as in the current US crisis and the Nordic crises of the 1990s), the corporate sector (as in the case of the 1997–98 East Asian financial crisis), or both. As in earlier crisis episodes, asset prices rose sharply while banks decreased reliance on deposits in favor of less stable sources of wholesale funding, and while nonbanking institutions (e.g., finance companies in Thailand in the 1990s, and offshore financial institutions in Latin America in the 1980s and 1990s) grew significantly, owing in part to laxer prudential requirements for nonbanks. When such crises erupt, they generally trigger losses that spread rapidly throughout the financial system by way of downward pressures on asset prices and interconnectedness among financial institutions. These broad patterns repeated themselves this time around when losses in the US real estate market triggered general runs on the US shadow banking system, which ultimately hit banks in the United States and elsewhere.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

234

26. RESOLUTION OF BANKING CRISES

WHICH COUNTRIES HAD A SYSTEMIC BANKING CRISIS IN 2007–09? The financial crisis that started in the United States in 2007 spread around the world, affecting banking systems in many other countries. This section defines a systemic banking crisis and identifies which countries experienced one. It also identifies countries that can be considered as cases of borderline banking crises, and countries that escaped a banking crisis (either because they evaded a crisis through successful policies or because they were not hit by the negative shock emanating from the United States). A banking crisis is considered to be systemic if two conditions are met: 1. Significant signs of financial distress in the banking system (as indicated by significant bank runs, losses in the banking system, and bank liquidations) 2. Significant banking policy intervention measures in response to significant losses in the banking system The first year that both criteria are met is considered to be the starting year of the banking crisis, and policy interventions in the banking sector are considered to be significant if at least three of the following six measures have been used4: 1. extensive liquidity support (5% of deposits and liabilities to nonresidents) 2. bank restructuring costs (at least 3% of GDP) 3. significant bank nationalizations 4. significant guarantees put in place 5. significant asset purchases (at least 5% of GDP) 6. deposit freezes and bank holidays In implementing this definition of systemic interventions, liquidity support is considered to be extensive when the ratio of central bank claims on the financial

sector to deposits and foreign liabilities exceeds 5% and more than doubles relative to its precrisis level.5 Direct bank restructuring costs are defined as the component of gross fiscal outlays directed to restructuring the financial sector, such as recapitalization costs. They exclude asset purchases and direct liquidity assistance from the treasury. Significant direct restructuring costs are defined as those exceeding 3% of GDP. Asset purchases from financial institutions include those implemented through the treasury or the central bank. Significant asset purchases are defined as those exceeding 5% of GDP.6 A significant guarantee on bank liabilities indicates that either a full protection of liabilities has been issued or guarantees have been extended to nondeposit liabilities of banks.7 Actions that only raise the level of deposit insurance coverage are not deemed significant. Significant nationalizations are takeovers by the government of systemically important financial institutions and include cases where the government takes a majority stake in the capital of such financial institutions. In the past, some countries intervened in their financial sectors using a combination of less than three of these measures but on a large scale (e.g., by nationalizing all major banks in the country). Therefore, a sufficient condition for a crisis episode to be deemed systemic is considered when either (i) a country’s banking system exhibits significant losses resulting in a share of nonperforming loans above 20% or bank closures of at least 20% of banking system assets or (ii) fiscal restructuring costs of the banking sector are sufficiently high exceeding 5% of GDP.8 For the recent wave of crises, none of these additional criteria are needed to identify systemic events. Quantitative thresholds to implement the authors’ definition of a systemic banking crisis are admittedly arbitrary; therefore, an additional list of ‘borderline cases’ is also maintained that almost met the definition of a systemic crisis. At the same time, the more quantitative

4

When possible, the magnitude of policy interventions and associated fiscal costs have been expressed in terms of GDP rather than banking system size to control for the ability of a country’s economy to support its banking system; this naturally results in higher measured fiscal costs for economies with larger banking systems.

5

Domestic nondeposit liabilities are excluded from the denominator of this ratio because information on such liabilities is not readily available on a gross basis. For Euro-area countries, liquidity support is also considered to be extensive if in a given semester the increase in this ratio is at least 5% points. The reason is that data on Euro-area central bank claims are confounded by large volumes of settlements and cross-border claims between banks in the Eurosystem. As a result, the central banks of some Euro-area countries (notably Germany and Luxembourg) had already large precrisis levels of claims on the financial sector.

6 Asset purchases also provide liquidity to the system. Therefore, an estimate of total liquidity injected would include schemes such as the Special Liquidity Scheme (185 billion pounds sterling) in the United Kingdom and Norway’s Bond Exchange Scheme (230 billion kronas), as well as liquidity provided directly by the Treasury. 7

Although a quantitative threshold is not considered for this criterion, in all cases guarantees involved significant financial sector commitments relative to the size of the corresponding economies.

8

One concrete historical example is Latvia’s 1995 crisis, when banks totaling 40% of financial system’s assets were closed, depositors experienced losses, but few of the interventions listed above were implemented.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

POLICY RESPONSES IN THE 2007–09 CRISES: WHAT IS NEW?

approach is a major improvement over earlier efforts to date banking crises (such as Caprio et al., 2005; Laeven and Valencia, 2008; Reinhart and Rogoff, 2009) that relied on a qualitative approach to determine banking crises as situations in which ‘a large fraction of banking system capital has been depleted’. Table 26.1 provides the list of countries that meet the definition during the recent episode. For each case, the exact criteria that are met are also indicated. A separate column on deposit freezes and bank holidays is not included because no episode during this recent wave of banking crises made use of banking holidays, while deposit freezes were used only for Parex bank in Latvia. In total, 13 systemic banking crises and 10 borderline cases were identified since the year 2007.9 Table A.1 in the Appendix presents more detailed information about the policy interventions in these cases. As in the previous crisis database release (Laeven and Valencia, 2008), the starting year of the banking crises in the United Kingdom and the United States is 2007, while for all other cases the starting date is 2008.10 Most policy packages announced in countries that do not meet the definition can be seen as preemptive interventions. In a large subset of G-20 countries, direct policies to support the financial sector were quite modest. For instance, Argentina, Brazil, China, India, Indonesia, Mexico, Saudi Arabia, South Africa, and Turkey all did not announce direct financial sector support measures that involved fiscal outlays. Some did issue or increase guarantees on bank liabilities, creating contingent liabilities for the government. For example, Mexico announced a guarantee on commercial paper up to a limit of 50 billion pesos. Other G-20 countries were more seriously affected by the financial turmoil and reacted more strongly, but ultimately did not intervene at a large enough scale to be deemed a systemic crisis. Table A.2 in the Appendix provides more detail about these cases, including the actual usage of policy measures. The differences between announced and actually used amounts are quite striking in a number of cases, with the actual usage of announced packages on average being small

235

(see also Cheasty and Das (2009) for a comparison between announcements and used amounts). In Laeven and Valencia (2008), only the first year of the crisis was included but no end date for the crisis episode was reported. The previous release is now expanded by dating the end of each episode. The end of a crisis is defined as the year before two conditions hold: real GDP growth and real credit growth are positive for at least 2 consecutive years. In case the first 2 years record growth in real GDP and real credit, the crisis is dated to end the same year it starts.11 Admittedly, this is an oversimplification given the many factors that come into play in a crisis, and the different nature of crises and recoveries across the sample. In a number of cases this methodology results in long crisis durations, which sometimes is the consequence of additional shocks affecting the country’s economic performance. In order to keep the rule simple, the duration is truncated to 5 years, beginning with the crisis year. As of end-2009, none of the recent crises had ended according to the definition used in this article. The median duration of a crisis for the old episodes is 2 years. Start and end dates for all episodes are reported in Table 26.2, in which output losses (to be defined in the next section) are also reported.

POLICY RESPONSES IN THE 2007–09 CRISES: WHAT IS NEW? Crisis management starts with the containment of liquidity pressures through liquidity support, guarantees on bank liabilities, deposit freezes, or bank holidays. This containment phase is followed by a resolution phase during which typically a broad range of measures (such as capital injections, asset purchases, and guarantees) are taken to restructure banks and reignite economic growth. It is intrinsically difficult to compare the success of crisis resolution policies given differences across countries and time in the size of the initial shock to the financial system, the size of the financial system, the

9

The new definition of a systemic banking crisis is somewhat more specific than the one used in Laeven and Valencia (2008), where systemic crises were considered to include events with ‘significant policy interventions’. As a consequence, a few cases listed as systemic banking crisis in the previous release, would under this definition be considered borderline cases: Brazil 1990, Argentina 1995, Czech Republic 1996, Philippines 1997, and United States 1988. 10

While undoubtedly the most salient events of the UK and US financial crises took place in 2008 (such as the bailout of Bear Stearns, the collapse of Lehman Brothers, the takeover of the GSEs, and the TARP programs in the United States; and the nationalization of the Royal Bank of Scotland in the United Kingdom), significant signs of financial sector distress and policy actions directed to the financial sector were in both cases already observed in 2007.

11

In computing end dates, bank credit to the private sector (in national currency) from International Fund Services (IFS) is used (line 22d). Bank credit series are deflated using CPI from World Economic Outlook (WEO). GDP in constant prices (in national currency) also comes from the WEO. When credit data are not available, the end date is determined as the first year before GDP growth is positive for at least 2 years. In all cases, the duration of a crisis is truncated at 5 years, including the first crisis year.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

236 TABLE 26.1

26. RESOLUTION OF BANKING CRISES

Systemic Banking Crises (2007–09)

Country

Extensive liquidity support

Significant restructuring costs

Significant asset purchases

Significant guarantees on liabilities

Significant nationalizations

✓

✓

✓

✓

Systemic cases Austria

✓

Belgium

✓

Denmark

✓

✓

✓

Germany

✓

✓

✓

Iceland

✓

✓

✓

✓

Ireland

✓

✓

✓

✓

Latvia

✓

✓

✓

Luxembourg

✓

✓

✓

✓

Mongolia

✓

✓

✓

✓

Netherlands

✓

✓

✓

✓

Ukraine

✓

✓

United Kingdom

✓

✓

✓

✓

✓

United States

✓

✓

✓

✓

✓

✓

✓

Borderline cases France

✓

✓

Greece

✓

✓

Hungary

✓

✓

Kazakhstan

✓

Portugal

✓

✓

Russia

✓

✓

Slovenia

✓

✓

Spain

✓

✓

Sweden

✓

✓

Switzerland

✓

✓

✓

Note: Systemic banking crises are defined as cases where at least three of the listed interventions took place, whereas borderline cases are those that almost met the definition of a systemic crisis. Extensive liquidity support is defined as a situation where central bank claims on the financial sector exceed 5% of deposits and foreign liabilities and is at least twice as large as precrisis levels; direct bank restructuring costs are considered significant when they exceed 3% of GDP and exclude liquidity and asset purchase outlays; guarantees on liabilities are considered significant when they include actions that guarantee liabilities of financial institutions other than just increasing deposit insurance coverage limits; nationalizations are significant when they affect systemic financial institutions. Source: Authors’ calculations.

quality of institutions, and the intensity and scope of policy interventions. With this caveat, policy responses during the recent crisis episode can now be compared with those of the past. The policy responses during the 2007–09 crises episodes were broadly similar to those used in the past. First, liquidity pressures were contained through liquidity support and guarantees on bank liabilities. Like the crises of the past, during which bank holidays and deposit freezes have rarely been used as containment policies, there are no records of the use of bank holidays during the recent wave of crises, while a deposit freeze

was used only in the case of Latvia for deposits in Parex Bank. On the resolution side, a wide array of instruments was used this time, including asset purchases, asset guarantees, and equity injections. All these measures have been used in the past, but this time around they seem to have been put in place quicker (for detailed information about the frequency of policy interventions in past crisis episodes, see Laeven and Valencia, 2008). One commonality among the recent crises is that they mostly affected advanced economies with large, internationally integrated financial institutions that were deemed too large and/or interconnected to fail.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

TABLE 26.2

Banking Crisis Start and End Dates Output loss (%)

Country

Start

End

Albaniad

1994

1994 b

Output loss (%)

Country

Start

End

Costa Rica

1987

1991

0

Output loss (%)

Country

Start

End

Kenya

1985

1985

24

Country

Start

End

Russiac

2008

–

Output loss (%) 0 a

Algeria

1990

1994

41

Costa Rica

1994

1995

0

Kenya

1992

1994

50

Sa˜o Tome´ & Prı´ncipe

1992

1992

2

Argentina

1980

1982a

58

Cote d’Ivoire

1988

1992b

45

Korea

1997

1998

58

Senegal

1988

1991

6

Argentina c

1989

1991

13

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

Argentina

1995

1995

0

Argentina

2001

2003

71

d

Armenia

1994

a

1994

Austria

2008

–

Azerbaijand

1995

1995a

Bangladesh

1987

1987

0

Belarus

1995

1995

Belgium

2008

–

Benin

1988

Bolivia

1986

Bolivia

1994

–

36 b

1995

Dominican Rep

2003

2004

Ecuador

1982

1986b

Ecuador

1998

2002

43

98 25

Latviad Latvia

2002

1995

1996

Sloveniad

1992

1992

c

2008

–

–

116

1981

59

1995b

Spainc

2008

–

39

1996

Sri Lanka

1989

1991

Liberia

1991 1995

1989

1990

0

Macedonia, FYRd

1993

1995

1992b

15

Equatorial Guinea

1983

1983b

0

Madagascar

1988

1986

49

Eritrea

1993

1993a

Malaysia

1997

0

62

Brazil

1994

1998

0

Bulgaria

1996

1997

60

1994

1992

1994

Finland

1991

1995

Francec

2008

–

Estonia

d

1991

1995

Germany

2008

–

Ghana b

1998

b

1991

121 106

1982 c

Greece

Guinea

2008 1985

1985

Swaziland

1995

1999

46

0

Sweden

1991

1995

33

1988

0

Swedenc

2008

–

31

1999

31

Switzerlandc

2008

–

0

b

0

Tanzania

1987

1988

0

70

Mauritania

1984

1984

8

Thailand

1983

1983

25

21

Mexico

1981

1985a

27

Thailand

1997

2000

109

Mexico

1994

1996

14

Togo

1993

1994

39

Mongolia

2008

–

29 a

20

1991

45

–

47

b

1987

19

1983

102

Mali

b

Georgia

37 b

1977

1993

El Salvador

b

Slovenia Spain

1990

23

1994

0

1998

–

d

34

b

1994 d

Slovak Rep

Lebanon

Lithuania

1990

1999

2008

d

Sierra Leone

b

1995

2008

1994b

1987

2008

143

b

Luxembourg

1990

Cameroon

Denmark

Kyrgyz Rep

1985

1

Brazilc

1994

2000

1982 d

1980

1996

Burundi

1996

1991

Kuwait

b

1980

1992

1990

1999

Egypt

Bosnia and Herzegovinad

Burkina Faso

1998

Czech Republicc,d

Djibouti 17

d

Croatia

d

0

Cameroon

1995

1997

8

Guinea

1993

1993

0

Cape Verde

1993

1993

0

GuineaBissau

1995

1998

30

Morocco

1980

0

Tunisia

1991

1991

1

b

22

Turkey

1982

1984

35

b

2000

2001

37

1994

1994

0

1984

Mozambique

1987

1991

0

Turkey

Nepal

1988

1988

0

Uganda d

Netherlands

2008

–

25

Ukraine

1998

1999

0

Nicaragua

1990

1993

11

Ukraine

2008

–

5 Continued

TABLE 26.2

Banking Crisis Start and End Dates—cont’d Output loss (%)

Country

Start

End

Central African Rep

1976

1976

Central African Rep

1995

Chad

1983

Start

End

0

Guyana

1993

1993

0

1996

9

Haiti

1994

1998

38

1983

0

Hungaryd

1991

1995b

0

c

b

Chad

1992

1996

0

Chile

1976

1976

20

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

Chile

1981

b

1985

9

Congo, Dem Rep Congo, Rep

1994 1992

1998

1994

79 47

1983

1985

97

United Statesc

1988

1988

Nigeria

1991

1995b

0

United States

2007

–

0 25

b

2008

–

42

Panama

1988

1989

85

Uruguay

2002

2005

27

India

1993

1993

–

130

Niger

24

Iceland

2008

b

–

38

Ireland

1994

2007

1985

47

1991

United Kingdom

1981

1982

Congo, Dem Rep

0

Uruguay

1982

1

2001

5

Colombia

b

2000

1993

2001

1983

End

1991

1997

1983

Start

Norway

Indonesia

Congo, Dem Rep

Nicaragua

Output loss (%)

Country

42

19

43

End

–

1998

2000

Start

2008

1998

1998

Output loss (%)

Country

Hungary

China, Mainland

Colombia

Output loss (%)

Country

Israel

1977

Jamaica

1996

Japan

1997

Jordan

1989 c

Kazakhstan

2008

b

69 110

1977 1998 b

2001

1991 –

0

76 38 45 106 0

Paraguay

1995

Peru Philippines Philippines d

Poland

Portugal

c d

Romania d

Russia

c

1995 a

b

15

Venezuela

1994

1998

1

1983

1983

55

Vietnam

1997

1997

0

1983

1986

92

Yemen

1996

1996

16

0

Zambia

1995

1998

31

1997

b

2001

1992

1994

2008

–

1990 1998

0

Zimbabwe

1995

b

1999

10

37 a

1992

0

a

1998

Output losses computed as the cumulative difference between actual and trend real GDP, expressed as a percentage of trend real GDP for the period [T, T þ3] where T is the starting year of the crisis. Trend real GDP is computed by applying an HP filter (l ¼ 100) to the GDP series over [T-20, T-1]. a Credit data missing. For these countries, end dates are based on GDP growth only. b The duration of crises is truncated at 5 years, starting with the first crisis year. c Borderline cases. d No output losses are reported for crises in transition economies that took place during the period of transition to market economies. Source: World Economic Outlook (WEO), Laeven, L., Valencia, F. 2008. Systemic Banking Crises: A New Database. IMF Working Paper No. 08/224, Washington, DC, and authors’ calculation.

POLICY RESPONSES IN THE 2007–09 CRISES: WHAT IS NEW?

The large international networks and cross-border exposures of these financial institutions helped propagate the crisis to other countries. Failure of any of these large financial institutions could have resulted in the failure of other systemically important institutions, either directly by imposing large losses through counterparty exposures or indirectly by causing a panic that could generate bank runs. This prompted large-scale government interventions in the financial sector (including preemptive measures in some countries). Given that the crisis started in US subprime mortgage markets, financial exposure to the United States was like the key propagation mechanism of the crisis (see Claessens et al., 2010). Figure 26.1 shows foreign claims by nationality of reporting banks, from the Bank for International Settlements (BIS) consolidated banking statistics, expressed as percentage of GDP, as of end-2006. Cross border banking exposure to the United States varied a great deal across countries, ranging from less than 1% for Mexico to 300% for Switzerland. Eight out of 10 of the most exposed economies meet the authors’ systemic banking crisis definition or are categorized as borderline cases. Liquidity support was used intensively as a first line of response to this shock emanating from the United States. Not only was liquidity provision large, as illustrated in Figure 26.2, but it was also made available more broadly through a larger set of instruments and institutions (including nonbank institutions), and under weaker collateral requirements. Examples of unconventional liquidity measures include the Federal Reserve’s decision to grant primary broker–dealers access to the discount window, the widening of collateral accepted by the Federal Reserve and many other central banks, and the purchase of asset-backed securities by the Federal Reserve. These actions were also accompanied in some cases by the introduction of nonconventional facilities to fund nonfinancial companies directly, such as the Federal Reserve’s Commercial Paper Facility and the Bank of England’s Asset Purchase Facility. This significant liquidity provision is reflected in a large increase in central bank claims against the financial sector. The median change from the precrisis level to its peak in the ratio of central bank claims against the

239

financial sector to deposits and foreign liabilities amounts to 5.5%.12 This is about half its median in past crisis episodes. For comparison purposes, Figure 26.2 also reports the historical median of liquidity support among high-income countries only, since most recent crises occurred in high-income economies (all crisis countries except Mongolia, Latvia, and Ukraine).13 In some cases, liquidity was also provided directly by the treasury, as indicated in Figure 26.2. Slovenia shows the largest increase in liquidity funded by the treasury, amounting to close to 5% of deposits and foreign liabilities. Similarly, government deposits at Parex Bank in Latvia constituted an important source of liquidity assistance for this bank.14 Liquidity injected in countries labeled as borderline has also been significant, in particular for Greece, Russia, and Sweden. For Greece, liquidity support increased steadily starting in September 2009. In Russia, liquidity support subsided quickly after reaching a peak of 20% of deposits and foreign liabilities in 2009. Guarantees on bank liabilities have also been widely used during recent crisis episodes to restore confidence of bank liability holders. All crisis countries except Ukraine (and Kazakhstan, Sweden, and Switzerland among borderline cases) extended guarantees on bank liabilities other than raising deposit insurance limits. Coverage extended varied widely, though (Appendix Table A.1). While guarantees on bank liabilities have not been uncommon in past crises, asset guarantees have been used less frequently in the past. This time around, asset guarantees were used in some cases, including Belgium and the United Kingdom. For instance, the Bad Bank Act in Germany, passed in July 2009, provided private banks relief on holdings of illiquid assets by allowing them to transfer assets to a special entity and receive government-guaranteed bonds issued by this special entity in exchange. While direct fiscal costs for Germany amount to just above 1% of GDP, total guarantees (including those associated with the Bad Bank and financial institutions debt) reached about 6%.15 One measure of the length of a crisis is the time it takes central banks to withdraw liquidity support. As a measure of the time it took to withdraw liquidity

12

For Germany and Luxembourg, while at the peak this variable reached 9.4% and 14.7% respectively, the increments look small because banks in these countries already maintained high balances prior to the crisis due to cross-border settlements. Liquidity support is computed as the ratio of Central Bank Claims on deposit money banks (line 12 in IFS) to total deposits and liabilities to nonresidents. Total deposits are computed as the sum of demand deposits (line 24), other deposits (line 25), and liabilities to nonresidents (line 26).

13

It is worth clarifying that there are only five historical (pre-2007) crisis episodes among high-income countries in this historical sample.

14

In the case of Latvia, the threshold used in the definition of extensive liquidity support is satisfied once government deposits at Parex are counted as liquidity support. 15

Because Germany’s Bad Bank implies asset transfers, it could also be treated as asset purchases. Yet, it is treated here as guarantees, and therefore Germany is not listed as also having met the significant asset purchases threshold.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

240

26. RESOLUTION OF BANKING CRISES

FIGURE 26.1 Foreign claims by nationality of reporting banks. Note: Dark solid bars denote systemic banking crises and light solid bars denote borderline cases. Figures denote foreign claims by nationality on the United States at end-2006 as percentage of home country GDP. Source: BIS.

300 80 50 45 40 35 30 25 20 15 10 5

support, the number of months between the peak of liquidity support and the month when liquidity support declined to its precrisis level are computed. In earlier crises, emergency liquidity support was withdrawn within 14 months (median). However, this time around,

Mexico

Turkey

Brazil

Greece

Chile

Italy

Australia

Portugal

Finland

Denmark

Austria

Spain

Sweden

Japan

Ireland

Germany

France

Canada

Belgium

UK

Switzerland

Netherlands

0

as of end-2009, only Denmark, Germany, Hungary, Luxembourg, the Netherlands, and Switzerland saw their liquidity support returned to precrisis levels, suggesting that liquidity remained an issue for a prolonged time in recent crises.

25

20

15

10

5

Dliq

All (old)

Switzerland

Sweden

Spain

Slovenia

Russia

Portugal

Kazakhstan

Hungary

Greece

France

USA

UK

Ukraine

Netherlands

Mongolia

Luxembourg

Latvia

Ireland

Iceland

Germany

Denmark

Belgium

Austria

0

High-income (old)

FIGURE 26.2 Emergency central bank liquidity support. Note: The shaded figures represent the change in the ratio of central bank claims on the financial sector over total deposits and foreign liabilities between the peak of this ratio and the average for the year before the crisis. The nonshaded figures represent the total amount of liquidity support funded directly by the Treasury (between 2007 and 2009) over total deposits and foreign liabilities. Dark-shaded bars denote systemic crisis cases, while light-shaded bars denote nonsystemic crises. For Iceland, liquidity data were available only up to March 2008. Horizontal lines denote the medians classified by countries’ income level for historical data. All (old): all previous countries; High income (old): high-income previous episodes. Reproduced from Laeven, L., Valencia, F. 2008. Systemic Banking Crises: A New Database. IMF Working Paper No. 08/224, Washington, DC, IFS, and authors’ calculations.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

241

POLICY RESPONSES IN THE 2007–09 CRISES: WHAT IS NEW?

The overall size of monetary expansion is also considered, by computing the change in the monetary base, and finding that monetary expansion is significantly higher in recent crises compared to past crises. Figure 26.3 shows the change in the monetary base between its peak during the crisis and its level 1 year before the crisis, expressed in percentage points of GDP.16 One finds that the median monetary expansion of about 6% this time significantly exceeded its historical median of about 1%, though it is not that different from its historical median among high-income countries. Relatively larger financial systems and credibility of monetary policy in high-income economies may explain this difference. About 70% of fiscal outlays correspond to public sector recapitalizations of financial institutions. Bank recapitalizations, while not more common than in earlier crisis episodes, have been implemented much faster than in

the past. The median difference between the time it took to implement public recapitalization programs and the time that liquidity support became extensive (i.e., when liquidity support exceeded 5%) is no months for the recent crises compared to 12 months for past crises (Figure 26.4).17 Addressing solvency problems with public money is generally a complex and lengthy process because it requires political consensus and legislation. Policymakers therefore often prolong the use of liquidity support and guarantees in the hope that problems in the banking sector subside. This time around, though, policymakers acted with relative speed, at least in some countries.18 Governments typically acquire stakes in the banking sector as part of government recapitalization programs, and such ownership stakes often end up being held by the government for a prolonged period of time.

14% 12% 10% 8% 6% 4% 2%

All (old)

Switzerland

Sweden

Spain

Slovenia

Russia

Portugal

Kazakhstan

Hungary

Greece

France

USA

Uk

Ukraine

Netherlands

Mongolia

Luxembourg

Latvia

Ireland

Iceland

Germany

Denmark

−4%

Belgium

−2%

Austria

0%

High-income (old)

FIGURE 26.3 Monetary expansion. Note: Change in monetary base in percentage points of GDP between the peak and its level 1 year before the crisis. Horizontal lines denote the medians by country groups. All (old): all countries; High income (old): high-income countries. Reproduced from IFS, Laeven, L., Valencia, F. 2008. Systemic Banking Crises: A New Database. IMF Working Paper No. 08/224, Washington, DC and authors’ calculations. 16

Data on reserve money come from IFS. For Euro-area countries, reserve money corresponds to the aggregation of currency issued and liabilities to depository corporations, divided by Euro-area GDP. 17 For bank recapitalizations, only ‘comprehensive’ recapitalization packages are considered in which public funds were used, thereby excluding ad hoc interventions and biasing upward the estimate of the response time. In the new crises, three recapitalization programs targeted specific banks: Iceland (the three largest banks), Luxembourg (Fortis and Dexia), and Latvia (Parex). The last two are included in the calculation because of the size of the affected institutions. However, the median does not change if they are excluded. Iceland is not included because of limited data on liquidity support as to compute the date when it became extensive. 18

In many cases, banks were able to raise capital in private markets or from parent banks, and generally this took place before public money was used. In addition, many banks have temporarily been allowed to avoid the recognition of market losses and thereby overstate regulatory capital.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

242

26. RESOLUTION OF BANKING CRISES

In months

FIGURE 26.4 Timing of recapitalization

14

policies. Note: Time in months between moment when liquidity support became extensive and implementation of recapitalization plans. Horizontal line denotes the median across all and high-income past crises in which recapitalization plans were put in place. Reproduced from Laeven, L., Valencia, F. 2008. Systemic Banking Crises: A New Database. IMF Working Paper No. 08/224, Washington, DC and authors’ calculations.

12 10 8 5

6

−1

0

−3

−1

−3

USA

0

0

UK

2

Ukraine

1

Netherlands

2

2

5

Luxembourg

4

All (old)

Latvia

Ireland

Germany

Denmark

Belgium

−4

Austria

−2

High-income (old)

Although divestments (or repayments) of government support on average start about 1 year from the start of the crisis, suggesting that the early repayments from the capital support of the Troubled Asset Relief Program (TARP) witnessed in the United States are not uncommon, government participation in banks has in many cases largely exceeded the initially envisioned holding period.19 In many cases, the public sector retained participation for over 10 years (in Japan, for instance, as of end2008 over 30% of capital injected in financial institutions following the crisis in 1997 remained to be sold). In some cases, divestment took place by tender, through sales of entire institutions to foreign investors or large domestic banks; in other cases, it took place more gradually through markets. Bank failures – defined broadly by including institutions that received government assistance – have also been significant during the recent wave of crises. This is striking given that bank failures are rare events in most countries, in part due to regulatory forbearance and ‘too big to fail or close’ problems. Relative to the total assets in the banking system, the bank failures in Iceland are by far the most significant, at about 90% of total banking assets (Figure 26.5). Iceland is followed by Greece and Belgium with banks that failed or received government assistance representing 80% or more of banking system assets in each of these three countries. When using the more conservative definition of failure that excludes government assistance, Iceland is followed by Belgium, Kazakhstan, and the United Kingdom, with banks representing 53%, 28%, and 26%, respectively, of the system failing. In terms of banks receiving government assistance, Greece tops the charts, with banks holding 80% 19

of total banking system assets receiving some form of government assistance. Greece is followed by France and Ireland, with banks holding about 70% and 55%, respectively, of banking system assets receiving government assistance. At least for the United States, for which historical data on bank failures since the 1930s are available, the recent wave of failures including assistance is unprecedented, with banks holding about one-quarter of the deposit market having failed or received some form of government assistance since 2007 (see Box 26.1 for a more detailed analysis of historical US bank failures). Excluding banks that received public assistance, the year 1989 during the US savings and loan crisis is by far the worst year on record, with banks holding over 6% of the deposit market failing. The United States is clearly not an outlier this time around, even when using the broader definition of bank failures that includes government assistance. Of course, the US failure list excludes such large financial institutions as Fannie, Freddie Mac, and AIG because they are not banks, although they meet the authors’ definition of failure, and the magnitude of financial distress in the United States could therefore be underestimated. A consequence of these dramatic bank failures has been a reorganization of the worlds’ financial map, with large players becoming significantly smaller, freeing up space for new players, in particular emerging markets. Bank failures during the recent wave of crises have been particularly dramatic for the United States and some of the countries in Western Europe that before the crisis were top-tier players in global banking. Before the crisis, at end-2006, the top 30 banks worldwide had a total stock market capitalization of about US$ 3.4 trillion, of which

A comprehensive analysis of guidelines for exit strategies from crises can be found in Blanchard et al. (2010).

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

243

POLICY RESPONSES IN THE 2007–09 CRISES: WHAT IS NEW?

100% 80% 60% 40% 20%

United states

United kingdom

Ukraine

Switzerland

Spain

Sweden

Portugal

Netherlands

Luxembourg

Latvia

Kuwait

Korea, Rep. of

Kazakhstan

Japan

Italy

Ireland

Iceland

Greece

France

Germany

Denmark

China

Belgium

Austria

0%

Failed banks, fraction of total banking assets (%) Government-assisted banks, fraction of total banking assets (%)

FIGURE 26.5 Bank failures and interventions in selected countries. Note: Government-assistance implies public capital support resulting in the government holding a minority stake in the bank. Failure implies bank closure; bank taken over by government; nationalization; or public capital support resulting in the government becoming a majority shareholder. Source: Authors’ calculations based on data from IMF, EU, FDIC, and JIDC.

BOX 26.1

US BANK FAILURES: PAST AND PRESENT US bank failures since the 1930s have come in three waves: the Great Depression era of the 1930s, the savings and loans crisis of the 1980s, and the recent mortgage crisis of the late 2000s, with the number of bank failures peaking in the years 1937, 1989, and 2009, respectively. Compared to these earlier bank failure episodes, the current wave of bank failures appears more short-lived and, at least compared to the savings and loans crisis, less dramatic in terms of number of failing banks (Figure 26.6). Note that 2005 and 2006 were the only 2 years since 1934 that reported no bank failures. Owing in part to consolidation following financial deregulation starting in the 1980s, the average US bank has grown substantially in size. After accounting for this development, the recent failures look a lot worse. Failed US banks during the current crisis hold about 26% of the deposit market – that is, when including banks that did not fail but received government assistance, such as Citigroup and Bank of America (Figure 26.7). Using this

definition of failure, 2009 is by far the worst on record. When excluding banks that received public assistance, the year 1989 is the worst year on record. Recent bank failures have generated similar losses compared to the past, with a median loss rate to the deposit insurance fund on assets of failed banking institutions of 19% (Figure 26.8). Median losses are relatively stable over the examined period (data on loss rates are available starting in 1986), and roughly the same during the recent crisis as compared to the savings and loan crisis. The median loss rate peaked in 2008 at 28%. In terms of losses to the deposit insurer, there are no data for the year 2009 yet, but losses were significantly lower in 2008, at 0.12% of US GDP, than the highest loss on record in 1989 of 0.97% of US GDP. Overall, one finds that, in the particular case of the United States, the failure rate of banks and losses incurred by the government in closing failed banks in this crisis is similar compared to the US banking crisis of the 1980s with a median loss rate in bank failures of about 20% of bank assets.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

244

26. RESOLUTION OF BANKING CRISES

4.5% Fraction of failed banking institutions (excluding assistance)

4.0%

Fraction of failed banking institutions (including assistance)

3.5% 3.0% 2.5% 2.0% 1.5% 1.0% 0.5%

2010

2005

2000

1995

1990

1985

1980

1975

1970

1965

1960

1955

1950

1945

1940

1935

0.0%

FIGURE 26.6 US bank failures: fraction of failed banks. Note: The figures include all failures and assistance transactions across 50 US states and Washington DC, as percentage of total number of institutions. 2010 includes data up to April. Source: FDIC.

20% 18%

Failed banking institutions’ share in total deposits (excluding assistance)

16%

Failed banking institutions’ share in total deposits (including assistance)

14% 12% 10% 8% 6% 4% 2% 2010

2005

2000

1995

1990

1985

1980

1975

1970

1965

1960

1955

1950

1945

1940

1935

0%

FIGURE 26.7 US bank failures: market share of failed banks. Note: The figures include the fraction of system deposits held by all bank failures and assistance transactions across 50 US states and Washington DC, as percentage of total deposits. Source: FDIC.

30%

Estimated loss rate (fraction of assets), median Median loss rate (1986-2008)

25% 20%

No failures 2006

5%

No failures

10%

2005

15%

2008

2007

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

0%

FIGURE 26.8 US bank failures: loss rates on assets of failed banks. Note: Includes all failures and assistance transactions across 50 US states and Washington DC. Loss rates are expressed as a percentage of total bank assets. Total assets are for FDIC-insured commercial banks only. Median loss rates are reported by year. The estimated loss is the difference between the amount disbursed from the Deposit Insurance Fund (DIF) to cover obligations to insured depositors and the amount estimated to be ultimately recovered from the liquidation of the receivership estate. Estimated losses reflect unpaid principal amounts deemed unrecoverable and do not reflect interest that may be due on the DIF’s administrative or subrogated claims should its principal be repaid in full. Source: FDIC.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

HOW COSTLY ARE THE 2007–09 SYSTEMIC BANKING CRISES?

TABLE 26.3

Market Capitalization of Top 30 Banks Worldwide, By Nationality (As of End of Year 2006 and 2009) Percentage of market capitalization

Number of banks End-2006 United States

End-2009

End-2006

End-2009

10

5

39.8

20.9

United Kingdom

4

3

12.5

13.9

France

3

3

7.8

8.8

Japan

3

1

11.9

2.5

Netherlands and Belgiuma

3

0

6.9

0.0

Spain

2

2

6.5

9.5

Switzerland

2

2

7.5

4.9

Canada

1

2

1.8

5.0

Italy

1

2

2.9

5.1

Germany

1

1

2.4

2.1

Australia

0

4

0.0

8.8

Brazil

0

2

0.0

4.4

China

0

2

0.0

12.2

Sweden

0

1

0.0

1.9

30

30

100.0

100.0

Total

245

to Citigroup – excluding banks that were acquired by other institutions; however, at the country-level, the Netherlands (including Fortis) experienced the largest average decline, followed by Japan. The latter is surprising given that Japan is not even classified as a borderline systemic banking crisis (because, although announced policy interventions in Japan were significant, the actual usage of these resources was small). How does the list of the world’s top 30 banks look like now? Interestingly, as of end-2009, there are four countries that for the first time entered this list. These include Australia, Brazil, China, and Sweden. On the other hand, Netherlands and Belgium – listed together because of jointly owned Fortis – drop from the top 30 ranking in 2009. The United States has fallen to only five banks in this list, together holding only 21% of the market capitalization of the world’s 30 largest banks in 2009 compared to 40% in 2006. The United States is clearly the country for which the change in market capitalization share is the most dramatic. Other clear losers include the Netherlands and Japan. While in 2006 no emerging market appeared on the list, at end-2009 banks from Brazil and China together were holding 16% of the total market capitalization of the top 30 banks worldwide. Other clear winners are Australia and Canada, whose large banks mostly escaped the US mortgage crisis.

a

Includes two Dutch institutions and Fortis, a Dutch–Belgian financial conglomerate. Source: Bankscope. Banks used in the calculation are listed in Appendix Table A.4.

40% belonged to US banks, 12% to UK banks, and 12% to Japanese banks (see Table 26.3).20 Countries with a systemic banking crisis in 2007–09 dominated the banking arena in 2006 with a share of close to 60% of the total. The crisis changed the map significantly. Twelve banks dropped from the top 30 list of 2006, with three banks being acquired by other institutions. The overall loss in market capitalization of the top 30 banks between 2006 and 2009 was a staggering 52%, a figure that already includes a significant stock market recovery during 2009. The largest decline in market capitalization corresponds 20

HOW COSTLY ARE THE 2007–09 SYSTEMIC BANKING CRISES? The cost of each crisis is estimated using three metrics: direct fiscal costs, output losses, and the increase in public sector debt relative to GDP. Direct fiscal costs include fiscal outlays committed to the financial sector from the start of the crisis up to end-2009 (see Appendix Table A.3 for a list of items included), and capture the direct fiscal implications of intervention in the financial sector.21 Output losses are computed as deviations of actual GDP from its trend, and the increase in public debt is measured as the change in the public debt-to-GDP ratio over the 4-year period beginning with the crisis year.22

A complete list of global top 30 banks in 2006 and 2009 is reported in Appendix Table A.4.

21

It is too early to provide reliable estimates about future recoveries and losses for recent crises, but wherever funds have been recovered, they have been included in Table A.3. Moreover, potential losses arising from contingent liabilities (such as asset guarantees) and schemes funded by the central bank (such as asset purchases) are not included, although it is recognized that losses in those schemes may ultimately have fiscal consequences. Output losses are computed as the cumulative sum of the differences between actual and trend real GDP over the period (T, T þ3), expressed as a percentage of trend real GDP, with T symbolizing the starting year of the crisis. Trend real GDP is computed by applying an HP filter (with l ¼ 100) to the log of real GDP series over (T20, T1) (or shorter if data are not available, though at least four precrisis observations are required). Real GDP is extrapolated using the trend growth rate over the same period. Real GDP data are from WEO. For recent crisis episodes, GDP projections are based on April 2010 WEO. The duration of a crisis is truncated at 5 years, including the first year. Wherever the methodology results in a crisis duration over 5 years, or when data availability impedes the application of this methodology, the end year is set as the fifth year from the start of the crisis year. 22

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

246

26. RESOLUTION OF BANKING CRISES

Output losses and the increase in public debt capture the overall real and fiscal implications of the crisis. The recent crises are overall more costly in terms of output losses and increases in debt, but less so in terms of direct fiscal outlays compared to the average crisis of the past. However, when the comparison is limited to high-income countries – given that they dominate the new crises sample – one finds that output losses are similar compared to the past, increases in public debt somewhat lower, but direct fiscal outlays higher (Table 26.4). The median direct fiscal costs associated with financial sector restructuring for the 2007–09 systemic banking crises amounts to almost 5% of GDP, about half its historical median of 10%.23 Figure 26.9 plots the direct fiscal costs for the recent systemic crises, as well as for the borderline cases. Two horizontal lines indicate the median of fiscal costs in all previous crises and that among previous high-income crisis episodes. Greece, Kazakhstan, Russia, and Slovenia show the highest figures among the borderline cases, although for Slovenia all of it corresponds to liquidity support from the treasury in the form of bank deposits. For Greece and Kazakhstan, at least half of it is also liquidity assistance from the treasury, while only for Russia the entire amount corresponds

TABLE 26.4 Summary of the Cost of Banking Crises (Over the Period 1970–2009) Direct fiscal cost

Increase in public debt

Output losses

Medians (% of GDP) Old crises (1970–2006) Advanced economies

3.7

36.2

32.9

Emerging markets

11.5

12.7

29.4

All

10.0

16.3

19.5

New crises (2007–2009) Advanced economies

5.9

25.1

24.8

Other economies

4.8

23.9

4.7

All

4.9

23.9

24.5

Note: New crises include Austria, Belgium, Denmark, Germany, Iceland, Ireland, Latvia, Luxembourg, Mongolia, Netherlands, Ukraine, United Kingdom, and United States. Source: Laeven, L., Valencia, F. 2008. Systemic Banking Crises: A New Database. IMF Working Paper No. 08/224, Washington, DC and authors’ calculations.

14% 12% 10% 8% 6% 4% 2%

All (old)

Switzerland

Sweden

Spain

Slovenia

Russia

Portugal

Kazakhstan

Hungary

Greece

France

USA

UK

Ukraine

Netherlands

Mongolia

Luxembourg

Latvia

Ireland

Iceland

Germany

Denmark

Belgium

0%

High income (old)

FIGURE 26.9 Direct fiscal costs. Note: Dark-shaded bars denote systemic banking crises episodes, and light-shaded bars borderline cases. The horizontal lines represent the medians across crises prior to 2007. Income groups are based on the World Bank country classification. All (old): all old episodes; High income (old): all old crises in high-income countries. Reproduced from Laeven, L., Valencia, F. 2008. Systemic Banking Crises: A New Database. IMF Working Paper No. 08/224, Washington, DC and Authors’ calculations.

23

These higher fiscal costs in part reflect an increase in average banking system size.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

247

HOW COSTLY ARE THE 2007–09 SYSTEMIC BANKING CRISES?

to recapitalization. As one would expect, on average, direct fiscal costs for borderline cases are lower than those for the systemic crises. Iceland shows up with the highest fiscal outlays, at 13% of GDP.24 The lower direct fiscal outlays associated with highincome countries, relative to all past crises, are regarded as a consequence of the greater flexibility these countries have in supporting their financial system indirectly through expansionary monetary and fiscal policy and direct purchases of assets that help sustain asset prices. Additionally, some high-income countries opted for sizable contingent liabilities to complement direct fiscal outlays (see Table A.3). Given that countries can also indirectly support their financial sector at times of crisis through expansionary fiscal policies that support output and employment, it is useful to also consider the overall increase in public debt as a broader estimate of the fiscal cost of the crisis. The median debt increase among the recent crises is 24% of GDP, about 8% points higher than its historical median of 16%. Thus, public debt burdens have increased significantly as a consequence of policy measures taken during the crisis.

Figure 26.10 shows the increase in the public debt burden for each crisis and also reports the historical median of the increase in public debt at crisis times. The increase in public debt that can be attributed to the crisis by computing the difference between pre- and postcrisis debt projections is approximated. For the 2007–09 crises, the fall WEO debt projections from the year before the crisis year are used as precrisis debt figures (i.e., September 2006 WEO for the United Kingdom and United State and October 2007 WEO for all other recent crises) and the Spring WEO 2010 debt projections are used for the postcrisis debt figures. For past episodes, the actual change in debt is simply reported.25 Among the recent borderline cases, France, Greece, Portugal, and Spain exhibit the largest expected increases in debt. While overall fiscal stimulus packages to counteract the global recession were significant in some of these countries, the direct interventions in the financial sector were not sufficient – as of end-2009 – to qualify as systemic banking crises. Recent developments in Greece, since the cutoff date at end-2009, while significant, are still not sufficient for it to qualify as a systemic banking crisis, at least as of April 2010.

75%

50%

25%

All (old)

Switzerland

Sweden

Spain

Russia

Slovenia

Portugal

Kazakhstan

Greece

Hungary

France

USA

UK

Ukraine

Netherlands

Mongolia

Latvia

Luxembourg

Ireland

Iceland

Denmark

Germany

Belgium

Austria

0%

High income (old)

FIGURE 26.10

Increase in public debt. Note: Dark-shaded bars denote systemic banking crises episodes, and light-shaded bars denote borderline cases. Increase in public debt is the increase in gross general government debt (central government debt if not available) over GDP, estimated over the 3-year period following the start of the crisis using WEO debt forecasts. Horizontal lines denote medians across past crises, classified by income level. All (old): all past crises in emerging and high-income countries; High income (old): all past crises in high-income countries. Reproduced from Laeven, L., Valencia, F. 2008. Systemic Banking Crises: A New Database. IMF Working Paper No. 08/224, Washington, DC, WEO and authors’ calculations.

24

These costs exclude the obligations (mostly to the United Kingdom and the Netherlands) arising from the Icesave crisis, which in net present value terms IMF staff estimates to be around 16% of GDP. The increase in debt measured in percentage of GDP over (T1, T þ3), where T is the starting year of the crisis, is computed. The choice of sources is guided by the availability of general government debt. When it is not available, central government debt is reported instead. The primary data source is WEO. When WEO debt data are not available, the OECD Analytical Database and the IMF’s Government Finance Statistics are used. 25

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

248 Difference between increase in public debt and direct fiscal costs, relative to GDP

26. RESOLUTION OF BANKING CRISES

80% 60% 40% 20% 0% 0 −20%

10

20

30 40 50 Real GDP per capita (in 2000 US$)

60

70 Thousands

−40% Previous crisis episodes

2007-2009 crises

FIGURE 26.11

Increase in public debt and direct fiscal costs. Note: Circles denote the new systemic and borderline episodes, while diamonds denote old episodes. The y-axis corresponds to the difference between the increase in public debt and gross fiscal costs, both in percentage of GDP. The x-axis corresponds to real GDP per capita, measured in 2000 US$. Old episodes exclude countries that experienced a sovereign debt crisis using data from Laeven and Valencia (2008). Reproduced from Laeven, L., Valencia, F. 2008. Systemic Banking Crises: A New Database. IMF Working Paper No. 08/224, Washington, DC, WEO, and authors’ calculations.

The previous two graphs suggest a large difference between increases in fiscal costs arising from direct support to the financial sector and increases in overall public debt. This difference appears to be positively correlated at about 0.4 with an economy’s level of income (Figure 26.11). Given that direct fiscal outlays to support the financial sector generally increase public debt, the difference between the increase in public debt and fiscal costs reflect in part the outcome of measures taken to support the real sector. This difference can in part be explained by discretionary fiscal policy and automatic stabilizers. One possible interpretation of this positive correlation is that high-income economies generally face easier financing opportunities than their low-income counterparts, and therefore may choose to complement financial measures with expansionary fiscal measures to deal with banking crises. Clearly, expansionary fiscal policy indirectly supports the financial sector by stimulating aggregate demand, which in turn props up loan demand and lowers the risk of loan defaults. The fallout from the recent crisis on the real sector was large. Median output losses are estimated for the recent crises of 25% of GDP, which is almost 5% points higher than its historical median of 20%. Output losses are estimated by computing the difference between trend GDP and actual GDP over the 4-year period beginning with the crisis year. Therefore, this methodology does not distinguish between permanent and transitory output losses. For the new crises, spring 2010 WEO

projections are used as actual GDP for the postcrisis years. Figure 26.12 shows the results.26 Output losses differ depending on the size of the initial shock, differences across countries in how the shock was propagated through the financial system, and the intensity of policy interventions. The output losses for Ireland and Latvia stand out at over 100% of potential GDP. Losses among borderline cases are also significant, in particular for Hungary, Portugal, and Spain. On average, countries with larger financial systems, and especially those that experienced rapid expansion prior to the crisis (such as Iceland, Ireland, and Latvia), were hit hardest.

CONCLUDING REMARKS This article updates the Laeven and Valencia (2008) database on systemic banking crises through end-2009 to include the recent wave of financial crises following the US mortgage crisis of 2007. The update results in 13 new systemic banking crises episodes and 10 borderline cases since early 2007. The update makes several improvements to the earlier database, including an improved definition of systemic banking crisis, the inclusion of crisis ending dates, and a broader coverage of crisis management policies. The new data show that there are many commonalities between recent and past crises, both in terms of underlying causes and policy responses. All crises share a

26

The medians reported in the graph are based on output losses that have been recomputed for all crisis episodes using the methodology employed in this article rather than by relying on estimates of output losses in Laeven and Valencia (2008). They computed the real GDP trend using all available data, implying a different horizon for each country. The new output loss estimates are on average similar to those in Laeven and Valencia (2008), though they differ for low-income countries and countries affected by large shocks, such as wars.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

249

CONCLUDING REMARKS

100%

75%

50%

25%

All (old)

Switzerland

Sweden

Spain

Slovenia

Portugal

Kazakhstan

Hungary

Greece

France

USA

UK

Ukraine

Netherlands

Mongolia

Luxembourg

Latvia

Ireland

Iceland

Germany

Denmark

Belgium

Austria

0%

High icome (old)

FIGURE 26.12 Output losses. Note: Dark-shaded bars denote systemic banking crises episodes, and light-shaded bars denote borderline cases. Output losses are computed as cumulative percent difference between actual and trend real GDP over the 4-year period starting with the crisis year. Trend GDP is computed applying an HP filter to the real GDP series over the 20-year period prior to the crisis. Horizontal lines denote the historical medians classified by countries’ income level. All (old): all past crises; High income (old): all past crises in high-income countries. Reproduced from Laeven, L., Valencia, F. 2008. Systemic Banking Crises: A New Database. IMF Working Paper No. 08/224, Washington, DC, WEO, and authors’ calculations.

containment phase during which liquidity pressures are contained through liquidity support and in some cases guarantees on bank liabilities. This phase is followed by a resolution phase during which a broad range of measures is taken to restructure banks and encourage bank lending (including asset purchases, guarantees, and capital injections) to reignite economic growth. These common patterns echo earlier findings summarized by Honohan and Laeven (2005) and Reinhart and Rogoff (2009). However, the recent wave of crises also shows some important differences with previous crisis episodes. • First, the recent crisis was concentrated in advanced economies, in particular those with large and integrated financial systems, unlike many of the boom–bust cycles observed in the past that centered on emerging market economies. Liquidity shortages at systemically important, globally interconnected financial institutions in these advanced economies prompted large-scale government interventions. • Second, while the intensity of policy interventions has been comparable to past crisis episodes, the speed of intervention and implementation of resolution policies was faster this time. This in part reflects that most of the crisis-affected countries are high-income countries with strong legal, political, and economic institutions that create an enabling environment for an effective and speedy crisis resolution.

Recapitalization policies in particular were implemented much sooner than in the past, contributing to lower direct fiscal outlays. • Third, countries used a much broader range of policy measures compared to past episodes, including unconventional monetary policy measures, asset purchases and guarantees, and significant fiscal stimulus packages. These large-scale public interventions were possible in part because most of the crisis-affected countries are high-income countries with relatively greater institutional quality and credibility of policy actions. • Fourth, preliminary estimates indicate that the overall economic costs of the recent crises are higher in terms of output losses and increases in public debt compared to past crises, though fiscal costs associated with financial sector interventions are lower this time. The lower short-term fiscal costs in part reflect the relatively swift action with which governments announced recapitalization measures and took other actions to restore the health of the financial system. However, they are also a consequence of the significant indirect support the financial system received through expansionary monetary and fiscal policy, the widespread use of guarantees on liabilities, and direct purchases of assets that helped sustain asset prices. The significant support deployed through monetary and fiscal policies, including coordinated international efforts to ensure adequate foreign exchange

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

250

26. RESOLUTION OF BANKING CRISES

liquidity, and a timely implementation of measures to address solvency problems in the financial system have significantly contributed to reduce the real impact of the recent crises. Moreover, such indirect support from macroeconomic stabilization policies has also lifted the burden on traditional crisis management policies, ultimately keeping the direct fiscal costs associated with bank recapitalization and other direct interventions into the financial sector lower than they otherwise would have been.

However, over the medium term, these indirect support measures have significantly increased the burden of public debt and the size of government contingent liabilities, raising concerns about fiscal sustainability in a number of countries. Moreover, the crisis is still ongoing in several countries and its ultimate impact will have to be reassessed in the future. It may therefore be premature to hail recent crisis management efforts as being more successful than those of the past.

APPENDIX TABLE A.1

Systemic Banking Crises Policy Responses (During the Years 2007–09)

Country

Liquidity support (percentage points increase in central bank claims on financial institutions over deposits and foreign liabilities)

Gross restructuring costs (recapitalization and other restructuring costs, excluding liquidity and asset purchases, in percentage of GDP)

Asset purchases and guarantees (funded by treasury and central bank, in percentage of GDP)

Guarantees on liabilities (significant guarantees on bank liabilities in addition to increasing deposit insurance ceilings)

Nationalizations (state takes control over institutions; year of nationalization between brackets)

Austria

8.2

2.1

Guarantees: 0.6

Unlimited coverage to depositors Bank and nonbank bond issues

Hypo Group Alpe Adria (2009)

Belgium

14.0

5.0

Guarantees: 7.7

DI raised from €20 000 to €100 000 Deposit-like insurance instruments Interbank loans and short-term debt Specific guarantees on Dexia

Fortis (2008)

Denmark

10.5

2.8

Deposits and unsecured claims of PCA banks

Fionia Bank (2009)

Germany

2.8

1.2

Unlimited coverage of household deposits Interbank loans and bank debt (capped at €400 billion)

Hypo Real Estate (2008)

Iceland

2.4

13.0

Unlimited coverage to domestic deposits

Kaupthing, Landsbanki (2008), Glitnir (2008), Straumur-Burdaras (2008), SPRON, and Sparisjo´dabankinn (2008)

Ireland

13.3

7.6

Unlimited coverage until 9/29/10 to most liabilities of 10 banks DI raised to €50 000

Anglo Irish Bank (2009)

Latvia

3.3

2.5

Guarantee on Parexsyndicated loans

Parex Bank (2008)

Luxembourg

4.3

7.7

DI raised from €20 000 to €100 000

Fortis and Dexia’s subsidiaries (2008)

Purchases: 0.2

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

251

APPENDIX

TABLE A.1

Systemic Banking Crises Policy Responses (During the Years 2007–09)—cont’d

Country

Liquidity support (percentage points increase in central bank claims on financial institutions over deposits and foreign liabilities)

Gross restructuring costs (recapitalization and other restructuring costs, excluding liquidity and asset purchases, in percentage of GDP)

Mongolia

9.4

3.0

Netherlands

3.3

6.5

14.6

4.8

United Kingdom

5.5

5.1

United States

4.6

3.5

Ukraine

Asset purchases and guarantees (funded by treasury and central bank, in percentage of GDP)

Guarantees on liabilities (significant guarantees on bank liabilities in addition to increasing deposit insurance ceilings)

Nationalizations (state takes control over institutions; year of nationalization between brackets)

Unlimited coverage to all deposits

Zoos Bank (2009)

DI raised to €100 000 Interbank loans of solvent banks Fortis bonds (€5 billion) and ING bonds (€10 billion)

ABN AMRO/Fortis (2008)

DI raised from UAH 50 000 to 150 000 until 1/1/11

Prominvest (2008), Nadra (2009), Inprom (2009), Volodimrski (2009), Dialog (2009), Rodovid (2009), Kiev (2009), Ukrgaz (2009)

Purchases: 13.4 Guarantees: 14.5

DI raised from ￡35 000 to 50 000 Guarantee on short- to medium-term debt (capped at ￡250 billion) Blanket guarantee on Northern Rock and Bradford & Bingley wholesale deposits

Northern Rock (2008), RBS (2008)

Purchases: 9.0

DI raised from $100 000 to $250 000 (until end-2009) Money market funds (capped at 50 billion) Full guarantee on transaction deposits Newly issued senior unsecured debt

Fannie Mae (2008), Freddie Mac (2008), AIG (2008)

Guarantees: 3.3

Borderline cases France

6.4

Greece

18.3

1.7

DI raised from €20 000 to €100 000 Funding guarantees up to €15 billion

1.3

0.1

Unlimited protection to depositors of small banks

Hungary

DI already higher than EU new limit €360 billion in guarantees for refinancing credit institutions €55 billions Dexia’s debt

Continued

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

252

26. RESOLUTION OF BANKING CRISES

TABLE A.1

Systemic Banking Crises Policy Responses (During the Years 2007–09)—cont’d

Country

Liquidity support (percentage points increase in central bank claims on financial institutions over deposits and foreign liabilities)

Kazakhstan

4.6

Portugal

5.5

Russia

22.2

Gross restructuring costs (recapitalization and other restructuring costs, excluding liquidity and asset purchases, in percentage of GDP)

Asset purchases and guarantees (funded by treasury and central bank, in percentage of GDP)

2.4

Guarantees on liabilities (significant guarantees on bank liabilities in addition to increasing deposit insurance ceilings) DI raised from T0.7 million to T5 million DI raised from €25 000 to €100 000 Debt securities issued by credit institutions (up to 12% of GDP)

1.0

9.3

Unlimited protection for all deposits by individuals and small enterprises until end-2010 New debt issued by financial institutions until end-2010

Spain

4.1

DI raised from €20 000 to €100 000 Credit Institutions New Debt Issues (capped at €200 billion)

13.1

0.7

2.8

1.1

Switzerland

Banco Portugues de Nego´cios (small bank) (2008)

DI raised from R400 ,000 to R700 000 Interbank lending for qualifying banks

Slovenia

Sweden

Nationalizations (state takes control over institutions; year of nationalization between brackets)

DI raised from SEK 250 000 to SEK 500 000 Medium-term debt of banks and mortgage institutions (up to SEK 1.5 trillion) Purchases: 6.7

DI raised from SFr 30 000 to SFr 100 000 until 12/31/11

Source: IMF Staff Reports, Mayer Brown, Official websites, and IFS.

TABLE A.2

Preemptive Crisis Responses in Selected G-20 Countries (During the Years 2007–09)

Country

Liquidity support (Percentage points increase in central bank claims on financial institutions over deposits and foreign liabilities, relative to precrisis level)

Australia

n/a

Gross restructuring costs (recapitalization and other restructuring costs, excluding liquidity and asset purchases, in percentage of GDP)

Asset purchases and guarantees (Funded by treasury and central bank, in percentage of GDP)

Guarantees on liabilities (significant: in addition to increasing deposit insurance (DI) ceilings, guarantees of other liabilities) Unlimited coverage to deposits (if above 1 million, only those with maturity 0 ð28:4Þ j j that is, no individual country is ready to finance the recapitalization itself. Obviously, if this equilibrium is selected, the recapitalization policy is inefficient as banks will almost never be recapitalized. That in most cases the closure equilibrium will occur, can be explained by the fact that a part of the externalities falls outside the home country (although it is safe to assume that in the current setting, the country with the highest social benefits of a recapitalization is the home country).6 The countries are grouped as follows: the home country denoted by H, all other European countries denoted by E, and all other countries in the world denoted by W. The social benefits can then be decomposed into the social benefits in the home country

5

The term ‘improvised cooperation’ has been coined to convey the view of an efficient, although adaptive, exchange of information and decision-making. It relies on the idea that maintaining financial stability is a goal that every individual country is interested in achieving, so there are good grounds for cooperation (Freixas, 2003). It can be argued that improvised cooperation corresponds to the current situation in the EU. 6

This assumption is consistent with the post-BCCI directive that stipulates that banks have to be headquartered in the country where most of their businesses are conducted.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

276

28. CENTRAL BANKS ROLE IN FINANCIAL STABILITY

(h  y ¼ yh), the rest of Europe (e  y ¼ ye), and the rest of the world (w  y ¼ yw): W X j¼1

y j ¼ yh þ

E X j= 2H

ye;j þ

W X

yw;j

ð28:5Þ

j= 2E

In this equation, h, e, and w are indexes for the social benefits (i.e., externalities caused by the possible failure of a financial institution) in the home country, the rest of Europe, and the rest of the world. The sum of h, e, and w is 1. When the total social benefits are close (or equal) to the social benefits of the home country (y is close to yh, so h is close to 1), the home country will be willing to bail out the financial institution. In all other cases (h < 1), the home country will only deal with the social benefits within its territory, while host countries expect the home country to pay for (a part of) the costs in the host country. Current national-based arrangements undervalue externalities related to the cross-border business of financial institutions. As a result, insufficient capital will be contributed and the financial institution will not be bailed out. This model pinpoints the public good dimension of collective bailouts and shows why improvised cooperation will lead to an underprovision of public goods, that is, to an insufficient level of recapitalizations. Countries have an incentive to understate their share of the problem so as to incur a smaller share of the costs. This leaves the largest country, almost always the home country, with the decision whether to shoulder the costs on its own or to let the bank close, and possibly be liquidated. The outcome of this model is consistent with the findings of Schinasi (2007). Applying the theory on ‘economics of alliances,’ he examines decision-making in a group of countries. Schinasi (2007) concludes that the provision of shared financial stability public goods results in an equilibrium that is suboptimal from a European perspective, even though each country views its own decision as optimal and has no incentive to change its resource allocation decision if other countries maintain theirs. A case in point is the rescue operation of Fortis in October 2008. The institutional setting with national authorities was not capable to reach a collective approach for Fortis, a cross-border bank with its main operations in the Benelux countries (Schoenmaker, 2008). National authorities were responsible for crisis management. When Fortis was first recapitalized, the Belgian, Dutch, and Luxembourg governments provided capital injections to the national banking parts (Fortis Bank, Fortis Bank Netherlands, and Fortis Bank Luxembourg, respectively) and not to the Fortis Group as a whole. When the first recapitalization of EUR 11 billion proved to be insufficient, Fortis was torn apart along national lines: the Dutch parts were nationalized by the Dutch government and the solvent Belgian/Luxembourg parts were sold to the French banking group BNP Paribas.

In sum, national financial stability management leads to an underprovision of recapitalization, and therefore, more European-based mechanisms for the management and resolution of cross-border financial crises need to be developed. This is because national authorities (central banks and ministries of Finance) only have a mandate for maintaining national financial stability and may therefore be reluctant to provide liquidity or solvency support to banks in other EU countries. They do not take cross-border externalities caused by financial institutions under their jurisdiction into account. When moving to a European mandate for financial stability (as for monetary stability), these externalities will be internalized, leading to an efficient outcome.

FINANCIAL STABILITY FRAMEWORK In order to maintain financial stability, central banks should have a structure in place that enables them to (i) identify potential vulnerabilities at an early stage, (ii) take precautionary measures, which make it less likely that costly financial disturbances occur, and (iii) undertake actions to reduce the costs of disturbances and restore financial stability after a period of distress. Figure 28.3 shows such a framework (De Haan et al., 2012).

Assessment Central banks need to monitor and analyze all potential sources of risks and vulnerabilities, which requires a systematic monitoring of individual parts of the financial system (financial markets, intermediaries, and infrastructure), the interplay between these individual elements, as well as macroeconomic conditions. To come up with a comprehensive view of the stability of the financial system, different steps have to be taken. First, central banks assess the individual and collective robustness of the intermediaries, markets, and infrastructure that make up the financial system. There is no standard framework to analyze financial stability. In an effort to improve the quality and comparability of basic data, the IMF has developed a set of financial soundness indicators (FSIs) as a key tool for macroprudential surveillance (see IMF, 2004). Central banks need to identify the main sources of risk and vulnerability that could pose challenges for financial system stability in the future and assess the ability of the financial system to cope with a crisis, should these risks materialize. The overall assessment will make it clear whether any (remedial) action is needed. If the assessment does not suggest any immediate dangers, continued supervision, surveillance, and macroeconomic policies are key to preserve the stability of

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

FINANCIAL STABILITY FRAMEWORK

FIGURE 28.3 Framework for maintaining financial stability. Reproduced from Houben, A., Kakes, J., Schinasi, G., 2004. Towards a framework for financial stability. De Nederlandsche Bank, Occasional Study No. 2(1).

Monitoring and analysis Macroeconomic conditions

Financial markets

Financial institutions

277

Financial infrastructure

Assessment

Prevention

Remedial action

Resolution

Financial stability

the financial system. In addition, communication on these issues is important. There are various ways of communicating to the public on financial stability policies. One such method is the publication of a Financial Stability Review (FSR). The purpose of publishing a FSR is to promote awareness in the financial industry and among the public of the issues that are relevant for safeguarding the stability of the financial system. By providing an overview of the possible risks to and vulnerabilities of the financial system, the FSR can also play a role in preventing financial crises. In this respect, Svensson (2003, pp. 26–27) argues that publication of a FSR serves “to assure the general public and economic agents that everything is well in the financial sector when this is the case. They also serve as early warnings for the agents concerned and for the financial regulation authorities when problems show up at the horizon. Early action can then prevent any financial instability to materialize, keeping the probability of future financial stability very low.”

The growing interest of central banks in monitoring and analyzing risks and threats to the stability of the financial system has spurred the publication of FSRs. During the last decade, the number of central banks that publish a FSR has increased rapidly from 1 in 1996 to over 40 in 2005 (see Figure 28.4). The Bank of England was the first to publish a FSR in 1996. The ECB and the Bank of Japan published their first FSR in 2004 and 2005, respectively. Until now, the Federal Reserve Board has refused to publish a FSR. Tools for measuring system-wide risks and calibrating policy tools are far from straightforward (Borio and Drehmann, 2009), and the analyses and recommendations put forward in the FSRs can be improved upon. As for the latter, Ciha´k (2006) argues that this includes clarifying the aims of the FSRs, providing an operational definition of financial sector soundness, clarifying the core analysis that is presented in FSRs consistently across time, making available the underlying data, discussing more openly the

45

FIGURE 28.4 Number of central banks that publish a

40

FSR, 1996–2005. Reproduced from Oosterloo, S., de Haan, J., Jong-A-Pin, R.M., 2007. Financial stability reviews: a first empirical analysis. Journal of Financial Stability 2, 337–355.

35 30 25 20 15 10 5 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Year

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

278

28. CENTRAL BANKS ROLE IN FINANCIAL STABILITY

risks and exposures in the financial system, making greater use of disaggregated data, focusing more on forward-looking measures rather than on backwardlooking description of indicators, and presenting stress tests that are comparable across time, and among other things, include scenarios, liquidity risks, and contagion.

Preventive and Remedial Action The next step in the framework of Figure 28.4 is taking action on the basis of the assessment (something that has clearly been lacking in the buildup of the 2007– 2009 financial crisis). If there are any indications of possible financial distress, it is up to the competent authorities (central banks and supervisors) to react properly. The public authorities can take informal action through correspondence and discussion with the affected institutions(s) to solve these problems. They can also use informal pressure to influence the behavior of financial players. Generally, the public authorities might exert moral suasion in two different situations. First, when they want to influence expectations of the general public through external statements or speeches, and second, when they attempt to persuade financial intermediaries to modify their behavior in the interest of the sound development of markets. If moral suasion fails, other policy instruments, such as surveillance and supervision, need to be intensified in order to correct the situation at hand. If financial conditions nevertheless worsen and a financial crisis occurs, one cannot pinpoint a single set of instruments that should be used. Generally, crises are never exactly alike and options differ as to which particular approach is ‘best’ for resolving them. Although there is no blueprint for crisis resolution, generally four reactive instruments can be considered: (i) (ii) (iii) (iv)

private sector solutions, liquidity support measures, public intervention tools, and winding down.

Crisis management starts with the containment of liquidity pressures through liquidity support, guarantees on bank liabilities, deposit freezes, or bank holidays. This containment phase is followed by a resolution phase during which typically a broad range of measures (such as capital injections, asset purchases, and guarantees) are taken to restructure banks and reignite economic growth (Laeven and Valencia, 2010).

resolution. Two types of private sector solutions can be distinguished: • Ad hoc mechanisms, such as liquidity provision, a merger or acquisition (capital infusion), or other rescue operations, which may be considered in case of an emergency. These solutions can be promoted by the authorities acting as honest broker, especially given the time constraints under which most crises have to be solved and the potential information asymmetries that then exist. • Predetermined mechanisms aimed at preventing the spillover effects of financial crises. An example is the German Liquidity Consortium Bank (LIKO-bank), a semiprivate institution that was founded in 1974 after the failure of the Herstatt Bank in order to bridge possible liquidity shortages of individual banks that are financially sound. However, as a ‘lender of penultimate resort,’ the LIKO-bank may not lend to insolvent institutions. If a private sector solution is not immediately at hand, the public authorities can bridge the gap between failure and resolution by involving a third party (bridge banking). Liquidity support measures According to Frydl and Quintyn (2000), liquidity support from the public authorities to troubled financial institutions starts long before the systemic nature of a banking crisis has been recognized. When a bank, or several banks, start experiencing substantial withdrawals from depositors and creditors, and they cannot borrow directly (or only at high rates) in the interbank market, the public authorities (usually the central bank) can become their lender of last resort (LOLR). In principle, central banks should only support illiquid but still solvent banks. Yet, during the early stages of an unfolding crisis, it is often very difficult to distinguish illiquidity from insolvency. Very often, it turns out that the banks approaching the central bank for liquidity support have been insolvent for a while, without this fact being known by them. In a crisis situation, it is hardly possible to distinguish between illiquidity and insolvency. So, the LOLR interventions by the public authorities mostly involve high-risk loans, which eventually may lead to huge costs to taxpayers. Apart from the fact that liquidity support can be given to individual financial institutions, such support can also be provided to the market as a whole. Emergency assistance to the market is provided temporarily to relieve market pressure following an adverse exogenous shock (e.g., the 9/11 terrorist attacks and the subprime mortgage crisis of 2007/2008).

Private sector solutions

Public intervention tools

If a financial crisis occurs, authorities often try to involve the private sector as much as possible in its

Once the true nature of a crisis has been identified and bank insolvency has been revealed as widespread,

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

FINANCIAL STABILITY FRAMEWORK

facilities like deposit insurance schemes may act as stabilizers to the financial system. There are two rationales for deposit insurance (MacDonald, 1996): • Consumer protection: deposit insurance protects depositors against the consequences of the failure of a bank. It is difficult for (potential) depositors to assess the financial health of banks. Only a small part of the information necessary to make an effective assessment of a bank is publicly available and even then, the general public may have difficulties in interpreting such information. • Reducing the risk of a systemic crisis: without deposit insurance, uninformed depositors might remove their deposits from sound banks in reaction to problems at a single bank (bank run). In order to meet these withdrawals, banks have to liquidate their asset portfolio at a loss, and eventually might fail. If depositors know that their money is safe because of the insurance, they will have no reason to withdraw it. Deposit insurance can thus be seen as a preventative instrument as well. This, however, requires a high coverage level (e.g., 100% deposit guarantee) and rapid payout. Although deposit insurance funds were originally aimed at preventing bank runs, in some countries, these funds may also be used for restructuring failing

279

banks. It is, however, questionable whether this is the purpose of deposit insurance. Quite often, countries have established limited deposit insurance funds. Experience has shown that limited deposit insurance schemes are inadequate to maintain or restore confidence during a (systemic) banking crisis. In order to prevent or stop bank runs, countries can resort to the announcement of full protection for depositors and creditors. However, such a blanket guarantee can come at great costs (as the liability is against assets of uncertain value). When the failure of a financial institution could create systemic problems, the government may decide to recapitalize (or even nationalize) the institution. This option is optimal if the costs of recapitalization are lower than the social benefits of preserving financial stability. Recapitalization may consist of a direct capital injection or the purchase of troubled assets. As the provision of solvency support puts taxpayers’ money at risk, the decision to recapitalize is normally taken by the government, and not by the central bank. Initially, the fiscal costs of nationalization will be relatively high, but the government can try to sell the nationalized institution at a later date. Often, a so-called Banking Restructuring Agency (BRA) is established to restore the health of the banking system (see Box 28.1). In order to protect the BRA from political interference, Enoch et al. (2001) argue

BOX 28.1

Resolving Banking Crises: Experiences of the Nordic Countries and Japan The Nordic countries and Japan experienced severe banking crises in the 1990s. While many comparisons can be made between the Nordic and Japanese banking crises, the approach that was taken to resolve these crises and the actual outcome differed considerably. While the Nordic authorities reacted promptly, the response of the Japanese authorities was slow. As a result, the Nordic banking crises were resolved relatively quickly, while the Japanese banking crisis continued for more than a decade. While the costs of the Nordic banking crises amounted to a fiscal cost of 8% of the GDP, the Japanese authorities spent more than 20% of the GDP on restructuring their banking system. There are a number of substantial differences between the approaches pursued in the Nordic countries and in Japan. First, the Banking Restructuring Agencies formed in the Nordic countries were much more aggressive in disposing of, and restructuring, troubled loans. Klingebiel (2000)

reports that the percentages of assets transferred by the asset management companies (or bank restructuring agency) in Finland and Sweden were 64 and 86%, respectively. In each case, the initial amount of assets transferred was about 8% of the GDP. Both restructuring agencies accomplished their loan disposals within 5 years of their establishment. Second, there was a significant contrast in the willingness to shrink the banking sector. Hoshi and Kashyap (2004) show that in Finland, the total domestic bank assets fell by 33% between 1991 and 1995, while in Sweden, domestic commercial bank assets dropped 11% between 1991 and 1993. In contrast, the total domestic bank assets in Japan fell less than 1% between 1993 and 2003. Third, when downsizing and loan disposal occurred in the Nordic countries, the financial institutions were decisively recapitalized and management was typically. Such a firm line was absent in Japan. There was little public support for banks in Japan. This restricted the ability of the Japanese Ministry of Finance to recapitalize banks (Hoshi and Kashyap, 2004).

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

280

28. CENTRAL BANKS ROLE IN FINANCIAL STABILITY

that the BRA should be functionally independent from the government and publicly accountable. While in previous crises it took policymakers about 1 year from the time that liquidity support became extensive before comprehensive recapitalization measures were implemented, in the 2007–2009 crisis, recapitalization measures were implemented around the same time that liquidity support became extensive (Laeven and Valencia, 2010).

Winding down When systemic risks are negligible, or when the costs of intervention are higher than the potential social benefits, the authorities will opt for the winding down of the troubled institution. However, the closure of a financial institution creates potential for disruption, especially to market functioning and liquidity. Therefore, the authorities should ensure that the winding down is managed in an orderly manner. One way to contain the negative effects is by providing liquidity support to other intermediaries. However, “when financial distress has been broad-based or has involved systemically important institutions, liquidation has rarely been the preferred option” (OECD, 2002, p. 131). The expectation that large financial institutions are ‘too big to fail’ may give rise to moral hazard. Moral hazard refers to the risk that once people know there is some sort of safety net or insurance, they take a greater risk than they would do without this protection. During the 2007–2009 crisis, it appeared impossible to close down the so-called systemically important financial institutions (SIFIs). To address the moral hazard of this ‘too big to fail problem,’ proposals are introduced to ask SIFIs to prepare resolution plans in case they hit the regulatory insolvency trigger point (which must be substantially above zero economic net worth, book value insolvency, or illiquidity). The plan for resolution would be negotiated in advance between the SIFI’s supervisors and its management (Claessens et al., 2010). This process should include the SIFI’s board of directors and its international college of supervisors. The resolution plan should ensure that a SIFI can be dismantled without interrupting the provision of any systemically important services or creating any other major spillovers. The resolution plan will have to be reviewed annually and subjected to stress simulations by the college of supervisors. This process will make it clear to the market that no firm is indispensable and that whatever essential functions a firm performs can continue to be provided. This will help to combat the increase in moral hazard resulting from the bailouts resorted to by countries in the wake of a financial crisis.

FINANCIAL STABILITY FUNCTIONS OF CENTRAL BANKS It is interesting to contrast the financial stability functions of the two main central banks: the Federal Reserve and the ECB. Under the Federal Reserve Act of 1913, 12 Federal Reserve Banks were established under the umbrella of the Federal Reserve Board and designated the Federal Reserve System, the Fed. The functions of the Fed included currency issue, the provision of banking services to the government, the provision of discounting and clearing facilities to member banks, and the regulation and supervision of member banks (Capie et al., 1994). The Fed was also considered as LOLR. The latter function is often labeled as the discount window function of the Fed: discounting bills from banks. The Fed has thus been a broad central bank from its start. It has the full set of financial stability functions, including the LOLR function, and the banking supervision function. During each major reform, the supervisory role of the Fed was intensely discussed. When the Gramm– Leach–Bliley Act of 1999 and the Dodd–Frank Wall Street Reform and Consumer Protection Act of 2010 were introduced, the role of the Fed was strengthened. The Gramm–Leach–Bliley Act enabled financial conglomerates combining banking, securities, and insurance to form separate entities under a (bank) holding. The Fed’s role as a holding supervisor was reinforced. In the Dodd–Frank Act, the Fed’s role overseeing systemically important payment and settlement systems was reinforced. When Economic and Monetary Union (EMU) was designed in the 1990s, Folkerst-Landau and Garber (1992) published a paper titled ‘The European Central Bank: A Bank or a Monetary Policy Rule?’ in which they introduced two concepts of central banking. The narrow concept only includes monetary stability (monetary policy rule), while the broad concept includes monetary as well as financial stability (like the LOLR function and supervision of financial institutions). The ECB is largely modeled after the Bundesbank and follows the narrow central banking concept focusing on monetary stability. The Maastricht Treaty defines maintaining of price stability as the primary objective (article 127, paragraph 1) of the European System of Central Banks (ESCB) and specifies that the ESCB should only contribute to the supervision and financial stability policies of the national authorities (article 127, paragraph 5). Folkerst-Landau and Garber (1992) argue that the narrow mandate for the ECB may hamper the development of the EU financial system. Still, the ECB is gradually moving into the process of becoming a full-fledged central bank by developing its ‘banking’ functions. It has been publishing a FSR since

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

FINANCIAL STABILITY FUNCTIONS OF CENTRAL BANKS

2004.7 Moreover, the ECB acted as LOLR throughout the recent financial crisis during which there were severe problems in the wholesale interbank market. Banks having surplus funds were unwilling to lend to banks that were on deficit because of concerns about solvency due to losses from, and exposures to, subprime mortgages. The ECB was proactive and provided short-term funds to deficit banks and absorbed funds from surplus banks. The ECB provided liquidity through its instrument of open market operations (OMOs) and standing facilities (marginal lending facility and deposit facility). This is the so-called general LOLR function, under which liquidity is available for all banks against collateral in a standardized way. The ECB’s policy was successful in stabilizing the euro-area interbank market.8 By its generous provision of liquidity to all banks (and a broader range of eligible collaterals), the ECB has come close to becoming a LOLR for ailing individual banks. However, the decision to provide emergency liquidity assistance to banks (individual LOLR) is left to the NCBs in the respective countries where banking groups are licensed and operate (Padoa-Schioppa, 1999). This national responsibility can lead to multiple coordination problems (see the section ‘Financial Stability: National or International?’). During the 2007–2009 financial crisis, the ECB had established itself as an effective European crisis manager. But, the ECB has no powers to prevent and manage crises in the financial system, including problems with cross-border banks.

Financial Stability Boards The 2007–2009 financial crisis has raised awareness of the need to ‘manage’ financial stability. While the central bank has a strong role to play in financial stability management, it is dependent on other players such as the ministry of finance for the deep pockets to rescue financial institutions, and the supervisors for obtaining information on individual financial institutions. The United States and Europe take different routes in setting up an FSB. The Treasury is the central player in the new FSOC, established by the Dodd–Frank Act. The ECB is

281

the central player in the new ESRB, proposed by De Larosie`re (2009). Financial Stability Oversight Council The Dodd–Frank Act created two new agencies, the FSOC and the Office of Financial Research (OFR), to monitor systemic risk and research the state of the economy. The two new offices are attached to the Treasury Department, with the Treasury Secretary being the Chair of the Council. The FSOC has to identify threats to the financial stability of the United States, promote market discipline, and respond to emerging risks to the stability of the US financial system. At a minimum, it must meet on a quarterly basis. Specifically, there are three purposes assigned to the council: 1. identify threats to the financial stability of the United States from both financial and nonfinancial organizations; 2. promote market discipline, by eliminating expectations that the government will shield them from losses in the event of failure; and 3. respond to emerging threats to the stability of the US financial system. The council must collate data (received from affiliated agencies and optionally from the companies themselves) to assess risks to the financial system, monitor the financial services marketplace, and make general regulatory recommendations to affiliated agencies reflecting a broader consensus; it may also compel the Federal Reserve to assume an oversight position of certain institutions considered to pose a systemic risk. The council must monitor domestic and international regulatory proposals and developments, and advise Congress in these areas. The council and the associated OFR are entrusted with the responsibility to facilitate information sharing and coordination among the member agencies and other Federal and State agencies regarding domestic financial services policy development, rule-making, examinations, reporting requirements, and enforcement actions. The voting members of the FSOC include the Treasury (chair), the Fed, the Comptroller of the Currency,

7

A particular issue for the ECB was the appropriate level: the EU level (because of the EU Internal Market for financial services) or the euro-area level (because of the monetary responsibility for the euro currency). The ECB decided to examine the stability of the financial system at the euro-area level. Most authors agree that financial stability should be managed at the European level (see the section ‘Financial Stability: National or International?’), but there is no agreement on the precise scope. Some argue that financial stability is primarily a concern for the euro area (Pisani-Ferry et al., 2008), while others consider financial stability as an issue for the EU as a whole (Goodhart and Schoenmaker, 2009; Nieto and Schinasi, 2007).

8

However, at the EU-wide level, the ECB and the Bank of England (BoE) followed different policies and did not coordinate with each other. Initially, the BoE did not provide extra liquidity to banks as it was fearful of banks’ overreliance on central bank funds (moral hazard). Liquidity shortages in the UK interbank market caused severe funding problems for Northern Rock, culminating in a bank run on retail deposits in September 2007. The BoE provided a massive LOLR loan to keep Northern Rock afloat and the UK government subsequently nationalized Northern Rock.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

282

28. CENTRAL BANKS ROLE IN FINANCIAL STABILITY

the Bureau of Consumer Financial Protection (the new financial consumer authority), the SEC, the FDIC, the CFTC, the Federal Housing Finance Agency, and the National Credit Union Administration Board. In addition, there are nonvoting advisory members: the OFR (part of the Treasury), the Federal Insurance Office (a new body at the Treasury), a state insurance commissioner, a state banking supervisor, a state securities commissioner. The membership indicates the dominant role of the Treasury. Moreover, the federal supervisors have the upper hand, with some minor inputs from the state supervisors. The council draws on resources of the Federal government. The council has very broad powers to monitor, investigate, and assess any risks to the US financial system. The council has the authority to collect information from any State or Federal financial regulatory agency and may direct the OFR, which supports the work of the council, ‘to collect information from bank holding companies and nonbank financial companies.’ The council monitors domestic and international regulatory proposals, including insurance and accounting issues, and advises the Congress and the Federal Reserve on ways to enhance the integrity, efficiency, competitiveness, and stability of the US financial markets. On a regular basis, the council is required to make a report to the Congress describing the state of the US financial system. Under specific circumstances, the Chairman of the Council (who is also the Secretary of the Treasury), with the concurrence of 2/3 voting members, may place nonbank financial companies or domestic subsidiaries of international banks under the supervision of the Federal Reserve if it appears that these companies could pose a threat to the financial stability of the United States. Under certain circumstances, the council may provide for more stringent regulation of a financial activity by issuing recommendations to the primary financial regulatory agency, which the primary financial agency is obliged to implement – the council reports to the Congress on the implementation of, or the failure to implement, such recommendations. Finally, the OFR, a department within the Treasury, provides administrative and technical support to the council.

European Systemic Risk Board In October 2008, the European Commission mandated a high-level group, chaired by former managing director of the IMF Jacques De Larosie`re, to give advice on the future of European financial regulation and supervision. The group presented its final report on 25 February 2009 and its recommendation provided the 9

basis for legislative proposals by the commission later that year. According to De Larosie`re (2009) report, a key lesson to be drawn from the crisis is the urgent need to upgrade macroprudential supervision in the EU for all financial activities. In the report of the high-level group, it is stressed that central banks have a key role to play in a sound macroprudential system. However, in order to be able to fully play their role in preserving financial stability, they should receive an explicit formal mandate to assess high-level macrofinancial risks to the system and to issue warnings where required. The high-level group recommends the establishment of a new independent body, the ESRB, responsible for safeguarding financial stability by conducting macroprudential supervision at the European level. The ESRB includes the members of the ECB General Council plus the Chairs of the three European Supervisory Authorities (EBA, ESMA, EIOPA)9 and a member of the European Commission. To ensure appropriate geographical coverage and a well-balanced composition, the De Larosie`re report proposes ECB involvement via the ECB General Council, which includes the president of the ECB, the vice-president of the ECB, and the governors of the NCBs of all 27 EU Member States, rather than that of the Governing Council (which includes only the euro-area members). The main task of the ESRB is to make assessments of stability across the EU financial system in the context of macroeconomic developments and general trends in financial markets. In case of significant stability risks, the ESRB provides early warnings and, where appropriate, issues recommendations for remedial action. The addressees of warnings and recommendations are subsequently expected to act on them unless inaction can be adequately justified.

But Financial Stability Tools are Needed The Dodd–Frank Act and the De Larosie`re Report are silent on the tools for macroprudential supervision (i.e., financial stability). However, the Fed and the ECB need tools to actively manage financial stability. Tinbergen, one of the first winners of the Nobel Prize for economics, already showed that one instrument is needed for each policy goal. Central banks have two goals: monetary and financial stability. They have a clear instrument, setting the interest rate, to serve monetary policy. They also need a clear instrument for financial stability. The Fed and the ECB can then proactively decide about applying the tool.

See the chapter on Financial Supervision in the EU (Schoenmaker, 2012) on the new European Supervisory Authorities.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

283

CONCLUSION

Two different tools have been proposed. The first proposal is to revisit the Basel’s system of capital requirements and make it more cycle neutral (e.g., Brunnermeier et al., 2009; Kremers and Schoenmaker, 2009). The Basel system is geared toward the stability of individual financial institutions and does little to take account of their interaction with their environment and its stability. Capital requirements that ‘breathe with the cycle’ may help avoid banks overly expanding credit when there is ample capital in boom time and, conversely, help avoid them tightening credit after boom time precisely when this is least conducive to financial stability. A simple way to introduce countercyclical capital buffers is to scale the minimum capital requirement multiplicatively. When credit or gross domestic product (GDP) growth is at its neutral level, the multiple is set to 1. If credit/GDP growth is above the trend, the multiple is proportionally set above 1. Vice versa, the multiple is set below 1 if credit/GDP falls below the trend. The challenge is to get a proper indicator for credit and GDP growth and to establish the required adjustment to the minimum capital requirement. A second proposal is to impose liquidity charges. Perotti and Suarez (2010) argue that in all crises that spread beyond the original shock, liquidity runs forcing fire sales are a main cause of propagation. If systemic crises involve liquidity runs, which only liquidity insurance by central banks can absorb, then it is only appropriate that the central bank assumes responsibility to monitor the buildup of risk and to manage the liquidity insurance provision with effective tools. Perotti and Suarez (2010) propose to establish a mandatory liquidity charge, to be paid continuously during good times to the central bank, which in exchange will provide emergency liquidity during systemic crises. The charge would be set according to the principle that future regulation should work like Pigouvian taxes on pollution, discouraging bank strategies that create a systemic risk for everyone. Hence, the charge will increase with the maturity mismatch between assets and liabilities and should be levied on all financial institutions having access to the LOLR. So, if the central bank observes an increase in short-term funding of a bank (while asset maturities remain constant), it will increase its liquidity charge for that bank.

Information Challenge Timely information on the condition of financial institutions and markets is crucial to make an up-to-date assessment of the stability of the financial system and to act swiftly when needed (Schoenmaker, 2010). A key challenge for these new FSBs is the flow of information between the various participants. In particular, will the supervisory agencies, which are closest to the financial

institutions, be prepared to inform, in a timely and comprehensive manner, the other members of the FSB? The rivalry between the US supervisory agencies is well known. An additional challenge for the ESRB is that it has to contend with the full flow of information from NCBs and national supervisors to the ECB. Game theory suggests that the envisaged arrangements are not incentive-compatible. The ECB has a mandate for the stability of the EU-wide financial system (European mandate), while the remit of NCBs is limited to the stability of their respective national financial systems (national mandate). If the interests of the ECB and the NCBs converge, NCBs may provide the necessary information to the ECB. But if there is a conflict of interest between an NCB and the ECB, there is no incentive for the NCB to provide timely information to the ECB (Ciha´k and Decressin, 2007). A case in point is emerging problems with a national bank in one of the EU Member States. While an NCB may have an incentive to help a major player of its national banking system and to wait (and hope) for better times (forbearance), the ECB may want to act swiftly to prevent problems from spreading to the wider EU financial system (prompt corrective action). However, without any information from the NCB that is close to the ailing bank, the ECB cannot act in a timely manner.

CONCLUSION Central banks are currently facing major reforms to foster their financial stability role. Financial stability departments are expanding. In a way, central banks are going back to their roots. Most central banks were established during times of crisis (most notably during wars to finance governments) and took on the role of financial crisis management in their early days. After World War II, the focus had tilted toward monetary policy. This monetary bias has now been reversed. Monetary and financial stability are two sides of the same coin. While the structures for financial stability (the hardware) have been put in place, the analytical models and the macroprudential tools (the software) are still under development. There are major challenges to develop the appropriate tools and to apply them. We should move from issuing warnings, which are often ignored, to applying tools based on a proper legal framework. A key issue is the timely application of these tools: timely action by the central bank may nip an asset price bubble in the bud. Most academic economists favor a rule-based system to ensure that the macroprudential tools are kicking in time. Most central banks favor discretion because each situation is unique. That brings us back to the long-standing central banking debate on rules versus discretion.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

284

28. CENTRAL BANKS ROLE IN FINANCIAL STABILITY

References Borio, C., Drehmann, M., 2009. Towards an operational framework for financial stability: ‘fuzzy’ measurement and its consequences. BIS Working Papers No. 284. Brunnermeier, M., 2010. Macroprudential regulation: optimizing a currency area. Colloquium in Honor of Lucas Papademos, May 20, Frankfurt. Brunnermeier, M., Crockett, A., Goodhart, C., Persaud, A., Shin, H., 2009. The Fundamental Principles of Financial Regulation, Geneva Report on the World Economy, vol. 11. CEPR, London. Capie, F., Goodhart, C., Fisher, S., Schnadt, N., 1994. The Future of Central Banking. Cambridge University Press, Cambridge. Ciha´k, M., 2006. How do central banks write on financial stability? IMF Working Paper No. 06/163. Ciha´k, M., Decressin, J., 2007. The case for a European Banking Charter. IMF Working Paper No. 07/173. Claessens, S., Herring, R., Schoenmaker, D., 2010. A Safer World Financial System: Improving the Resolution of Systemic Institutions, Geneva Report on the World Economy, vol. 12. CEPR, London. De Haan, J., Oosterloo, S., Schoenmaker, D., 2012. Financial Markets and Institutions, A European Perspective, second ed. Cambridge University Press, Cambridge. De Larosie`re, J., 2009. Report of the high-level group on financial supervision in the EU, Brussels. Enoch, C., Garcia, G., Sundararajan, V., 2001. Recapitalising banks with public funds. IMF Staff Papers 48 (1), 58–110. Financial Stability Board (ESB) (2009) Financial Stability Board Charter, Basel. Folkerst-Landau, D., Garber, P., 1992. The ECB: a bank or a monetary policy rule? In: Canzoneri, M., Grilli, V., Masson, P. (Eds.), Establishing a Central Bank: Issues in Europe and Lessons from the US. Cambridge University Press, Cambridge, pp. 86–110. Freixas, X., 2003. Crisis management in Europe. In: Kremers, J., Schoenmaker, D., Wierts, P. (Eds.), Financial Supervision in Europe. Edward Elgar, Cheltenham, pp. 102–119. Frydl, E., Quintyn, M., 2000. The benefits and costs of intervening in banking crises. IMF Working Paper No. 00/147. Goodhart, C., 1988. The Evolution of Central Banks. MIT Press, Cambridge, MA. Goodhart, C., Schoenmaker, D., 2009. Fiscal burden sharing in crossborder banking crises. International Journal of Central Banking 5, 141–165. Goodhart, C., Schoenmaker, D., Dasgupta, P., 2002. The skill profile of central bankers and supervisors. European Finance Review 6, 397–427. Herring, R., Carmassi, J., 2008. The structure of cross-sector financial supervision. Financial Markets, Institutions and Instruments 17 (1), 51–76. Hoshi, T., Kashyap, A.K., 2004. Solutions to the Japanese banking crisis: what might work and what definitely will fail. Hi-Stat Discussion Paper Series D04-35. Institute of Economic Research, Hitotsubashi University. International Monetary Fund (IMF), 2004. Compilation Guide on Financial Soundness Indicators. IMF, Washington, DC. Klingebiel, D., 2000. The use of asset management companies in the resolution of banking crises. World Bank Policy Research Paper No. 2284.

Kremers, J., Schoenmaker, D., 2009. The US proposals for regulatory reform. In: Mayes, D., Pringle, R., Taylor, M. (Eds.), Towards a New Framework for Financial Stability. Central Banking Publications, London, pp. 489–497. Kremers, J., Schoenmaker, D., Wierts, P., 2003. Cross-sector supervision: which model? In: Herring, R., Litan, R. (Eds.), BrookingsWharton Papers on Financial Services: 2003. Brookings Institution Press, Washington, DC, pp. 225–243. Laeven, L., Valencia, F., 2010. Resolution of banking crises: the good, the bad, and the ugly. IMF Working Papers No. 10/146. MacDonald, R., 1996. Deposit Insurance. Centre for Central Bank Studies, Bank of England, London. Mishkin, F., 2008. How should we respond to asset price bubbles? Speech at Wharton, Philadelphia, May 15. Nieto, M., Schinasi, G., 2007. EU framework for safeguarding financial stability: towards an analytical benchmark for assessing its effectiveness. IMF Working Paper No. 07/260. OECD, 2002. Experiences with the resolution of weak financial institutions in the OECD area. OECD Financial Market Trends No. 82. Padoa-Schioppa, T., 1999. EMU and banking supervision. International Finance 2, 295–308. Perotti, E., Suarez, J., 2010. Liquidity risk charges as a primary macroprudential tool. Duisenberg School of Finance, Amsterdam DSF Policy Paper No. 1. Pisani-Ferry, J., Aghion, P., Belka, M., et al., 2008. Coming of age: report on the euro area. Bruegel Blueprint No. 4, Brussels. Schinasi, G., 2007. Resolving EU financial-stability challenges: is a decentralized decision-making approach efficient? Mimeograph. Schoenmaker, D., 2008. Resolving the stability trilemma. Financial Regulator 13 (3), 45–53. Schoenmaker, D., 2010. The ECB, financial supervision, and financial stability management. In: de Haan, J., Berger, H. (Eds.), The European Central Bank at Ten. Springer, Heidelberg. Schoenmaker, D., 2012. Financial supervision in the EU. In: Safeguarding Global Financial Stability. Handbook of Financial Globalization. Elsevier, Amsterdam. Schoenmaker, D., Oosterloo, S., 2005. Financial supervision in an integrating Europe: measuring cross-border externalities. International Finance 8, 1–27. Schoenmaker, D., van Laecke, C., 2007. Current state of cross-border banking. In: Evanoff, D., Kaufman, G., LaBrosse, J. (Eds.), International Financial Instability: Global Banking and National Regulation. World Scientific, Singapore, pp. 39–63. Soros, G., 2010. Keynote Address. IIF Spring Membership Meeting, June 10, Vienna. Sullivan, D., 1994. Measuring the degree of internationalization of a firm. Journal of International Business Studies 25, 325–342. Svensson, L., 2003. Monetary policy and real stabilization. NBER Working Paper No. 9486. Taylor, M., 2009. Twin Peaks Revisited. Centre for Study of Financial Innovation, London. Trichet, J.-C., 2007. Towards the review of the Lamfalussy approach: market developments, supervisory challenges and institutional arrangements. Speech at the First CEBS Conference, London, 9 May.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

C H A P T E R

29 Organization, Supervision and Resolution of Cross-border Banking P. Angkinand*, C. Wihlborg† *Milken Institute, Santa Monica, CA, USA † Chapman University, Orange, CA, USA O U T L I N E Introduction

285

The Nordea Case

286

Subsidiary and Branch Organizations in Theory and Practice 287

INTRODUCTION Cross-border banking can take place in subsidiaries or branches. In common terminology, a subsidiary is a host country legal entity with its own capital as a buffer against losses and is subject to host country regulation, supervision, and legislation. A host country branch, by the same terminology, is an integrated part of the home country bank, subject to home country supervision, legislation, and control. The branch does not have its own capital and it is subject to host country legislation and regulation only with respect to its conduct in the host country. The EU has pushed these ‘national’ jurisdictional principles the farthest in the Banking Directive, allowing banks to operate across borders in branches under a ‘Single License’ within the EU, subject to home country control and supervision. Increasing internationalization of banking raises the issue whether the nationally oriented infrastructure for supervision and control, including crisis management, along the lines described above, can be maintained without an increase in systemic risk as a result of contagion from the distress of a large bank or a large part of it. Several countries violate the above jurisdictional principles. Host country regulators, supervisors, and central

Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397874-5.00050-6

Organization of Supervision and Crisis Management: Can National Responsibility Be Effective? 292 Conclusions: Need for Reform of the Architecture for Supervision and Crisis Management 294 References 295

banks may claim some jurisdiction over branches and abstain from jurisdiction over subsidiaries in other respects. For example, branches of foreign banks in the United States are subject to ‘ring-fencing’ of their capital, required participation in the US deposit insurance system, and supervision by the Federal Deposit Insurance Corporation (FDIC). New Zealand does not allow foreign banks to operate as branches. In the EU, there are calls for increased centralization of supervision, control, and crisis management for large international banks in the aftermath of the financial crisis. The Basel rules for capital adequacy assign primary responsibility for supervision of host country subsidiaries to the home country. In banking, the branch versus subsidiary distinction is not as clear as suggested above. Subsidiaries can be operated as more or less closely integrated entities, while branches can be required to have capital set aside for them. The distinction between branches and subsidiaries of nonfinancial corporations is at its sharpest in times of corporate distress. A subsidiary is subject to insolvency law and procedures in the host country, while a branch or other entity, which is not a legal person in the host country, is in distress only if the home country corporation is in distress. Banks in distress are often subject to crisis management without application of insolvency

285

# 2013 Elsevier Inc. All rights reserved.

286

29. ORGANIZATION, SUPERVISION AND RESOLUTION OF CROSS-BORDER BANKING

law, however. Few countries have a special insolvency law for banks, and the countries that do not have such a special law typically do not allow banks to enter bankruptcy procedures under regular corporate insolvency law. Corporate insolvency procedures are too time consuming for financial institutions participating in liquidity creation and the payment system. Fear of contagion compels governments to make sure that banks in distress remain in operation and honor their liabilities to other financial institutions. The strong concern with financial stability and fear of contagion from banks in distress has the consequence that international banks face scrutiny and possible involvement by more than one government in times of distress whether international operations are organized in branches or subsidiaries. Politics of burden sharing in potential losses becomes a major concern, as noted by Goodhart and Schoenmaker (2006). The most effective organization of supervision and crisis management and banks’ organization of crossborder activity in branches and subsidiaries are mutually dependent. Thus, the organization of supervision and crisis management must be designed with awareness of banks’ organizational responses, taking into account consequences for efficiency of the financial system as well as financial stability. An important question for policy makers is whether there is a trade-off between efficiency and financial stability in the allocation of responsibility for supervision and crisis management. The postcrisis calls for an EU supervisory authority indicate that policy makers think so since increased centralization of supervision and crisis management stand in contradiction to the EU Banking Directive that envisions cross-border banking in branches under a Single License under home country control and supervision. We argue in this chapter that lack of predictable and credible distress resolution procedures for banks on the national level contributes to the existence of an efficiency–financial stability trade-off. The implementation of an effective, special insolvency law for banks in host and home countries would enable banks to choose between cross-border banking in branches or subsidiaries based on efficiency considerations without threatening financial stability. The nationally oriented infrastructure for supervision, crisis management, and deposit insurance for domestically incorporated banks, including subsidiaries of foreign banks, could be maintained under certain conditions to be discussed below. To discuss these issues further, we take as a starting point the organization of the Nordic bank, Nordea, and its, so far, unrealized plans for reorganizing with a Single License. The Nordea case illustrates several issues related to cross-border banking in subsidiaries and branches. In the section ‘Subsidiary and Branch

Organizations in Theory and Practice’, we turn to theory and empirics with respect to relative advantages of cross-border subsidiaries and branches from banks as well as financial stability perspectives. In the section ‘Organization of Supervision and Crisis Management: Can National Responsibility Be Effective?’, we turn to supervision and crisis management for cross-border banks and ask under what conditions national responsibility can be maintained. These conditions refer to organization in branches or subsidiaries in host countries and to resolution procedures for banks. The section ‘Conclusions: Need for Reform of the Architecture for Supervision and Crisis Management concludes with a description of the type of reforms required on the national level for cross-border banking to operate in both branches and subsidiaries without being a serious threat to financial stability.

THE NORDEA CASE The Nordea bank was created during the period 1997– 2000 by the mergers of four systemically important banks in four countries, Denmark, Finland, Norway, and Sweden. The strategic objective of the mergers was to provide financial services in the four countries in a unified functional organization, although its legal organization consisted of subsidiaries in the different countries. Formally, subsidiaries remain subject to host country regulation, supervision, and deposit insurance, but the four countries’ supervisors developed new approaches to adapt their supervision to the functionally integrated cross-border bank. A joint Nordea supervisory committee was formed with a number of joint subcommittees to be able to supervise the bank by function across the legally separate country entities. The bank’s market shares range from 15% in Norway to 40% in Finland. Nordea offers a large range of financial services and can be considered a financial conglomerate with hundreds of subsidiaries in the four countries, but the bulk of the activities are conducted through the four Nordea banking subsidiaries incorporated in each of the countries. In 2003, Nordea announced a plan to reorganize in a branch organization by forming an EU company (Societas Europaea) headquartered in Sweden with branches in the other three countries operating with a single license as envisioned in the Banking Directive. The legal organization with branches would become consistent with the functional integration. The Swedish Financial Supervisory Authority would become responsible for supervision, and depositors in all branches would be subject to the Swedish Deposit insurance system. Nordea planned for the reorganization to take effect in the fall of 2005. At mid-2007, the plans seemed to have

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

SUBSIDIARY AND BRANCH ORGANIZATIONS IN THEORY AND PRACTICE

stalled, and in 2011, they are still on ice. After the financial crisis, the vision of banks operating across the EU with a single license issued by the home country seems more distant than ever. The failure of Nordea’s planned reorganization illustrates a number of issues with respect to the organization, regulation, and supervision of cross-border banking. First, the reorganization met resistance from financial supervisors in the host countries because they were unwilling to delegate all supervisory authority of systemically important banking operations to the home country supervisor. Second, the home country supervisor objected because the explicit as well as implicit deposit insurance responsibility of the home country would extend to a bank with assets greater than the whole domestic banking system. The Icelandics were later less reluctant in this respect, and conflicts arose when the Icelandic banks failed in 2008 about the responsibilities of host country authorities in the United Kingdom and the Netherlands, and home country authorities. We return to this issue. A third issue raised by the Nordea reorganization as well as the Icelandic bank failures is the organization of deposit insurance responsibilities between home and host countries. The Nordea bank feared that it would not be able to compete successfully in Norway as a Swedish bank with lower deposit insurance coverage than competing Norwegian banks.1 In contrast to the EU (so far), the United States requires that branches of foreign banks become members of the FDIC and the US deposit insurance system. There are proposals after the crisis to adopt the US system in the EU as well. The complication that arises with the US deposit insurance requirement is that a branch typically is a foreign legal entity without its own capital subject to host country law. Membership in the US deposit insurance system implies that the US has a stake in the bank. For this reason, the US requires ‘ring-fencing’ of branches in the sense that the foreign bank must allocate capital specifically to the branch in the United States. A fourth issue is that the distinction between a subsidiary and a branch is blurred. Nordea’s original strategy implied that subsidiaries were to be operated as if they were branches within a functionally integrated organization. The benefits of the mergers were expected to be maximized with functional integration across borders. One concern of supervisors is that a functionally integrated bank would be able to shift risk among subsidiaries and possibly concentrate relatively risky assets in the subsidiary enjoying the greatest explicit or implicit government support in times of distress, in particular.

287

Nordea’s reorganization would have created one bank with one capital buffer for all risks. The US system of ring-fencing branches of foreign banks is another example of blurring the distinction between a branch and a subsidiary. A fifth issue in cross-border banking is the responsibility for crisis management. Crisis management includes lender of last resort (LOLR) activities and procedures for insolvency as well as various government activities in times of crisis for a bank or the financial system. As Herring (2003) notes, regulators and supervisors have had difficulties in resolving cases of distressed international banks; such cases typically involve subsidiaries in a number of different jurisdictions. The issue of which country’s central bank is responsible for LOLR operations is ambiguous if cross-border banking takes place in systemically important branches.

SUBSIDIARY AND BRANCH ORGANIZATIONS IN THEORY AND PRACTICE Before reviewing the theory of organization of cross-border banking and the interaction between cross-border organization of banks, and the organization of supervision and crisis management, we present data on cross-border banking and its organization in branches and subsidiaries.2 Table 29.1 ranks the top 20 publicly traded global banking groups by total assets at the end of 2010. The balance sheets of the major global banking institutions have grown substantially. These largest 20 banking groups controlled roughly 15% of the world’s total bank assets in 2010, which is more than double of the market share of the top 20 banks in 2001. Half of the largest banking groups have the United States, Japan, and China as home countries. The other half is European. Table 29.1 shows that a small number of countries dominate cross-border banking. Of the top 20 global banking groups, the top European banks are located in France, Germany, Spain, and the United Kingdom, as well as in the Netherlands and Switzerland among smaller European countries. The bank size relative to home country’s GDP and the extent to which each of these banks conducts cross-border banking transactions vary significantly across regions but to a smaller extent within a region. Figure 29.1 summarizes data from Table 29.1 and shows in Figure 29.1(a) the shares of foreign bank subsidiaries controlled by different home

1

This issue was probably not the main reason why Nordea’s enthusiasm for the reorganization waned. The bank originally expected to reduce its value-added tax payments of internal transactions by operating in branches, but this benefit turned out to be unattainable.

2

Allen et al. (2011) reviews costs and benefits of cross-border banking.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

288

29. ORGANIZATION, SUPERVISION AND RESOLUTION OF CROSS-BORDER BANKING

TABLE 29.1

The 20 Largest Publicly Traded Global Banks (Ranked by Total Assets as of 2010)

Rank

Bank name

Home country

Total assets (US$ mil, 2010)

1

BNP Paribas

France

2670

103

92

4028

572

14

80

2

Deutsche Bank AG

Germany

2546

77

55

4820

4467

93

68

3

HSBC Holdings Plc

United Kingdom

2455

109

179

2282

1457

64

58

5

Barclays Plc

United Kingdom

2332

104

51

1510

546

36

49

6

The Royal Bank of Scotland Group Plc

United Kingdom

2275

101

69

272

52

19

18

7

Bank of America Corporation

United States

2265

15

111

3827

825

22

57

8

JP Morgan Chase & Co.

United States

2118

14

163

5429

3195

59

46

4

Mitsubishi UFJ Financial Group Inc

Japan

2095

38

72

2567

944

37

26

9

The Industrial & Commercial Bank of China – ICBC

China

2032

35

245

20

10

50

10

10

Citigroup Inc.

United States

1914

13

123

3545

1551

44

92

12

ING Groep NV

The Netherlands

1666

213

48

3815

3320

87

71

11

Mizuho Financial Group

Japan

1637

30

37

1341

741

55

17

13

China Construction Bank Corporation

China

1632

28

208

37

3

8

3

14

Banco Santander SA

Spain

1627

115

98

888

664

75

41

15

Bank of China Limited

China

1579

27

137

33

16

48

14

16

Agricultural Bank of China Limited

China

1561

27

141

93

7

8

3

17

Lloyds Banking Group Plc

United Kingdom

1552

69

54

139

18

13

4

18

Socie´te´ Ge´ne´rale

France

1513

59

46

1923

1713

89

65

19

UBS AG

Switzerland

1402

268

70

4422

4228

96

62

Total assets (% home GDP)

Market capitalization (US$ bil, as of 5 July 2011)

Total number of subsidiaries

Total number of subsidiaries in host countries

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

Percentage of the number of subsidiaries in host countries

Total number of host countries

289

SUBSIDIARY AND BRANCH ORGANIZATIONS IN THEORY AND PRACTICE

The 20 Largest Publicly Traded Global Banks (Ranked by Total Assets as of 2010)—cont’d

TABLE 29.1

Rank

Bank name

20

Sumitomo Mitsui Financial Group, Inc.

Home country

Total assets (US$ mil, 2010)

Japan

1281

Total assets (% home GDP)

Market capitalization (US$ bil, as of 5 July 2011)

23

Total number of subsidiaries

45

93

Total number of subsidiaries in host countries 12

Percentage of the number of subsidiaries in host countries 13

Total number of host countries 3

Sources: Bank-specific data are from BankScope and Bloomberg. The GDP data are from World Economic Outlook, International Monetary Fund.

countries. Figure 29.1(b) shows the total assets of large banks relative to the GDP of home countries. It can be seen that the countries with the largest shares of foreign subsidiaries are also the countries where the banks’ assets represent a large share of GDP. This means that relatively small countries with large banks have the largest number of foreign subsidiaries. European global banks have by far the largest ratios of bank assets relative to the home country’s GDP. The total assets of ING of the Netherlands and UBS of Switzerland amount to 213 and 268% of the home country GDP in 2010, respectively. The significant cross-border activities of these European banks are shown by a high percentage of the number of subsidiaries abroad (Figure 29.1(a)). The majority of net %

revenues of these European banks also come from foreign countries. For example, 95% of the net revenues of Banco Santander, a Spanish bank, originate abroad. Other European banks such as ING and Deutsche Bank have about 75% of their net revenues originating abroad. In comparison with the top European banks, the total assets of the top US, Japanese, and Chinese banks are substantially smaller relative to the country’s GDP. The global banking groups in Japan and United States have a large number of subsidiaries abroad. Many of these banks including the global European banks have over 1000 subsidiaries in foreign countries. The top Chinese banks still concentrate on the domestic markets both in terms of subsidiaries and net revenues. FIGURE 29.1

The average value of the number of subsidiaries and ratio of bank assets to home country’s GDP of the 20 largest publicly traded global banks by home country. (a) The percent of the number of subsidiaries abroad, by home country (average value, 2010) and (b) total assets (% home GDP), by home country, 2010. Source: See Table 29.1.

The percent of the number of subsidiaries abroad, by home country (average value, 2010)

100 80 60 40 20

D AN

AN Y IT ZE

RL

M

N

SW

ET H

G

ER

ER

LA

SP AI

FR

ED IT N U

(a)

N DS

N

E AN C

U

SA

PA N JA

KI N

C

H

IN

G DO

A

M

0

%

Total assets (% home GDP), by home country, 2010

300 250 200 150 100 50

D AN

DS

RL

N SW

IT ZE

LA ER

N

ET

H

KI D IT E U N

(b)

SP AI N

N

FR A

N

G DO

C

M

E

Y G ER

M

AN

PA N JA

IN H C

U

SA

A

0

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

290

29. ORGANIZATION, SUPERVISION AND RESOLUTION OF CROSS-BORDER BANKING

Figure 29.2 shows the distribution of foreign bank branches and subsidiaries relative to domestic credit institutions in Western and Eastern European countries. The data further distinguish between foreigncontrolled banks from the EU and from the rest of the world. The European Union, especially new Central and Eastern Europe (CEE) members, has seen a significant increase in cross-border banking activity in recent years. The new CEE member states have become dominated by Western European banks. Figure 29.2(a) shows the market shares in terms of total assets of branches and subsidiaries of different origins, while Figure 29.2(b) shows the market shares in terms of number of entities. In Western Europe, the market share of foreign bank subsidiaries relative to the market share of foreign bank branches in terms of total assets is 15.2/5.7%. The corresponding ratio in terms of number of institutions is 9/13.1%. Thus, foreign bank subsidiaries are generally much larger than foreign bank branches in Western

Europe. The number of foreign bank branches is actually larger than the number of foreign bank subsidiaries, but the market share of cross-border branch banking is small. The market share of foreign banks is much larger in Eastern Europe than in Western Europe. In terms of assets, the market share of foreign bank subsidiaries relative to the market share of foreign bank branches is 34.8/4.7%. The corresponding ratio in terms of the number of institutions is 18.8/12.6%. Thus, in Eastern Europe, the market share of foreign banks in terms of total assets is much higher than in Western Europe. There is also a substantial foreign bank branch presence in terms of numbers, but the market share of these branches in terms of assets is relatively small. The market shares of foreign bank branches and subsidiaries in Eastern and Western Europe are dominated by Western EU banks. The relatively small market share of branches even in Western Europe implies that the EU Single License for cross-border banking in the EU has not had a substantial impact on the structure of EU banking. Eastern Europe

Western Europe Foreign EU Foreign RoW branches, branches, 1.1% Foreign RoW 4.6% susbidiaries, 1.6% Foreign EU susbidiaries, 13.6% Domestic CIs, 79.0%

Foreign EU branches, 4.5% Foreign RoW susbidiaries, 1.2%

Foreign EU susbidiaries, 33.6%

Foreign RoW branches, 0.2%

Domestic CIs, 60.5%

(a)

Western Europe Foreign EU branches, 10.6% Foreign RoW susbidiaries, 3.3% Foreign EU susbidiaries, 5.7%

Foreign RoW branches, 2.5%

Domestic CIs, 77.9%

Eastern Europe Foreign EU branches, 11.2% Foreign RoW susbidiaries, 2.8%

Foreign EU susbidiaries, 16.0%

Foreign RoW branches, 1.4%

Domestic CIs, 68.6%

(b)

FIGURE 29.2 (a) Market share in terms of the total assets: domestic credit institutions (CIs), foreign bank branches, and foreign bank subsidiaries, 2009 and (b) market share in terms of the number of institutions: domestic credit institutions (CIs), foreign bank branches, and foreign bank subsidiaries, 2009. Source: ECB. Note: Western Europe: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain, Sweden, United Kingdom. Eastern Europe: Bulgaria, Cyprus, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, Romania, Slovakia, Slovenia.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

SUBSIDIARY AND BRANCH ORGANIZATIONS IN THEORY AND PRACTICE

The branch organization seems limited to relatively small cross-border operations. The Nordea reorganization discussed in the section ‘The Nordea Case’ was expected to become a model for other large banks, but its failure so far indicates that there is resistance to the branch organization in banks, supervisory agencies, or both. Although the EU Banking Directive gives banks the right to operate cross-border in branches under a Single License, the banks are not likely to push such an organization if supervisors or central banks object. The figures presented for Eastern and Western Europe conform to Dermine’s (2006) observations with respect to European banking. Fiechter et al. (2011) present data for a larger number of regions as well as specific countries. They show number of branches and subsidiaries as well as the shares of branches and subsidiaries in the total assets held by foreign banks. Looking at industrialized countries, most foreign-controlled assets are held in subsidiaries in France, Germany, and Switzerland, while branches dominate foreigncontrolled assets in Japan, the United Kingdom, and the United States. In the Middle East, branches dominate, while in Latin America, subsidiaries dominate. In China, Indonesia, and Malaysia, subsidiaries dominate, while in Argentina, India, and Korea, branches dominate. The latter countries place severe restrictions on the branches, however, in terms of separability and independence from the parent. Thus, where branches are common, they are in many ways similar to subsidiaries. These differences among countries raise the issue of what explains the choice of subsidiary versus branch organization in cross-border banking. There is also a choice of cross-border lending. We will disregard this alternative and focus on cross-border banking with a foreign presence. In general, funds lent through local subsidiaries and branches have a longer maturity and can be considered more stable than cross-border lending (Schnabl, 2011). Fiechter et al. (2011) review advantages and disadvantages of branches and subsidiaries for different types of banking activities. Their main conclusion is that ‘one size does not fit all.’ An efficient organizational structure depends on the nature of the business. Banking operations in financial conglomerates are diverse enough to be conducted in different organizational structures. For example, Fiechter et al. argue that most wholesale banking activity is more suitable for branches while subsidiaries seem to be favored for conventional retail activity. Risk diversification, economies of scale and scope, as well as regulatory restrictions can affect the choice of organization as well. A more detailed empirical study of factors affecting the choice of organization of foreign entities has been conducted by Cerutti et al. (2007). Using a sample of the world’s 100 largest banks (by total assets in 2002) with parent banks in 20 countries, they focus on the form

291

of affiliation of these banks in Latin America and Eastern Europe. The banks have a total of 182 subsidiaries and 65 branches in host countries. Home country’s regulatory restriction on opening braches oversea is pointed out as the main reason for banks to operate as subsidiaries. Canada, Germany, India, Italy, and Spain are the 5 out of 20 sampled countries that have additional restrictions on opening a branch relatively to subsidiaries in foreign countries. Germany and Italy, as examples, have together 65 subsidiaries and only 9 branches. The regulatory treatment of branches and subsidiaries by host regulators also explain a relatively high number of subsidiaries over branches in host countries in Latin America and Eastern Europe. Based on the constructed Host-Country Regulation Index which takes into account different regulatory requirements on the organizational form of foreign banks, they find that Croatia, Hungary, and Poland are among countries with the most restrictive legal requirements for opening foreign branches. The restrictions on foreign branches relative to foreign subsidiaries are not very different in Latin America where subsidiaries dominate. Cerutti et al. (2007) perform a Probit model estimation to examine the factors affecting the choice of branch versus subsidiary operations using a sample of 240 foreign banks in Latin America and Eastern Europe. In addition to home and host country regulatory restrictions on foreign banks, they identify several other factors influencing the choice of cross-border organization. Banks prefer to operate as branches in host countries with higher corporate tax rates, plausibly because of advantage of the branch structure in allowing shifting profits across border. Subsidiaries are preferred in countries with a higher degree of foreign penetration as subsidiary operations can receive benefits from establishing local retail operations. Economic and political risks are also found to have a significant effect. Subsidiaries are more common in countries with relatively higher macroeconomic risks as parent banks prefer limited liability provided by subsidiaries, but less common in countries facing higher political risks. Subsidiaries are typically required to hold capital and reserve requirements and the capital and reserve losses can be limited with branch operations in countries with risks of internal conflicts, war, or arbitrary foreign government intervention. There is little theoretical research on the factors influencing the branch subsidiary choice. One exception is Dell’Ariccia and Marquez (2010). Their starting point is a shareholder wealth maximizing bank that can operate foreign entities as branches or subsidiaries. The different entities of the bank are subject to macroeconomic risk as well as political risk such as risk of expropriation unrelated to credit risk. Shareholders can take advantage of limited liability to reduce their exposures to risk by choice of organization. If the bank is exposed

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

292

29. ORGANIZATION, SUPERVISION AND RESOLUTION OF CROSS-BORDER BANKING

to uncorrelated macroeconomic risk (causing credit risk) in countries of operation, the benefit of limited liability is maximized in subsidiaries. The home country capital is protected against macroeconomic risk in the host country. Only if macroeconomic shocks are perfectly correlated across countries are branch and subsidiary organizations equivalent from shareholders point of view. On the other hand, if there is substantial political risk in a country, a branch operation is preferred since capital in the home country is shielded from, for example, expropriation. The above predictions with respect to the organizational responses to different kinds of risk are supported by the mentioned empirical results of Cerutti et al. (2007). Dell’Araccia and Marquez add assumptions about pricing of risk, competitive conditions, and the relative profitability of branches and subsidiaries. These factors influence the results quantitatively, but the relationship between the nature of risk and the relative advantages of branches and subsidiaries remains robust. The model described allows banks to maximize shareholders’ benefits from limited liability by their choice of organization. A richer and more complex setting would consider how incentives and costs of supervisors may depend on risk as well as the organization of banks. Calzolari and Loranth (2010) develop a model wherein banks choose either branch or subsidiary organizations in host countries while supervisors decide on intensity of intervention in risk-taking and monitoring of banks’ risk-taking under imperfect information about banks’ projects. If the bank is operating cross-border in branches, the home country supervisor conducts interventions and monitoring, while the intervention is decentralized to the host countries if the banks operate subsidiaries. Incentives of supervisors to intervene in risk-taking and to become informed by monitoring are based on the objective of minimizing costs to the deposit insurance fund in their jurisdiction. Costs of intervention in host country branches by home country supervisors are assumed to be larger than costs to the host country supervisor. Although highly stylized, the Calzolari and Loranth model allows analysis of efficiency as well as financial stability of branch and subsidiary organizations taking into account differences in supervisory incentives. Financial stability can be linked to intensity of intervention and monitoring, while efficiency takes into account both efficiency of banking operations and endogenous benefits and costs of supervision.

The conclusions of the model with respect to efficiency and stability are not easily summarized. There is no simple trade-off between efficiency and financial stability and no unambiguous result with respect to superiority of either subsidiary or branch organizations. However, there are clear differences in incentives of supervisors to intervene and monitor between the two types of organizations, and these incentives affect the social efficiency of branch and subsidiary organizations. Results are sensitive to parameter values with respect to, for example, the prospects of the different units of the bank. One result is that a branch organization is preferred when the prospects of the entity are good while subsidiary organization is preferred when prospects are bad. The reason is that the host country unit receives more lenient intervention as a branch. The Calzolari and Loranth model does obviously not capture all the complexities of organizational choice and supervisor behavior. There is no agency problem in supervision, no regulatory capture, and organizational choices are well defined. Nevertheless, it represents a promising beginning, and it can be extended, as the authors note, to take into account lobbying incentives, information sharing, negative spillovers from a troubled to a sound unit, and resource transfers between countries. This review does not lead to any unambiguous results with respect to the efficiency of banks’ organizational choices, nor with respect to the organization of regulation and supervision. One conclusion is that from banks’ perspective some activities are more efficiently organized in subsidiaries, others in branches. Thus, strong restrictions on the branch–subsidiary choice are associated with costs. The question is how regulation, supervision, and crisis management can be organized to minimize efficiency losses without contributing to financial instability.

ORGANIZATION OF SUPERVISION AND CRISIS MANAGEMENT: CAN NATIONAL RESPONSIBILITY BE EFFECTIVE?3 Crisis management takes many forms in banking in particular. Formal corporate insolvency procedures are, as noted, rarely used because of the special nature of the liabilities of banks as well as other financial institutions. Only in the United States, where the FDICIA (Federal Deposit Insurance Improvement Act) of 1991 specifies mandatory special procedures for banks, are banks commonly closed down under formal procedures.

3

There is a substantial literature on the issues discussed in this section. The most closely related papers on the issue of insolvency procedures for banks are Eisenbeis and Kaufman (2007) and Lastra and Wihlborg (2007). Closely related papers on the organization of supervision and crisis management are Fiechter et al (2011) and Schinasi and Teixeira (2006). The latter focuses on LOLR responsibility within the EU.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

ORGANIZATION OF SUPERVISION AND CRISIS MANAGEMENT: CAN NATIONAL RESPONSIBILITY BE EFFECTIVE?

However, during the recent financial crisis, large systemically important banks were supported and allowed to continue operations under an escape clause in the FDICIA legislation. In most countries, large banks tend to obtain support, be bailed out, and recapitalized one way or another in such a way that most creditors and sometimes even shareholders become protected. Crisis management in banking involves not only resolution of insolvency. Prior to formal insolvency, supervisors can intervene to limit risk exposures and constrain activities following more or less formalized procedures. Prompt Corrective Action procedures in the United States represent one form of ‘structured early intervention.’ Supervisors and governments can also help negotiate the takeover of a bank or parts of it or they can initiate various forms of ad hoc rescue packages. Liquidity problems must be managed as well. Traditionally, central banks are charged with the task of serving as the LOLR providing emergency liquidity assistance while financial supervisors focus on the solvency of financial institutions. The recent crisis has shown that liquidity and solvency issues are often inseparable. Thus, both central banks and financial supervisors are involved in crisis management with fiscal implications. The complexity of crisis management increases with cross-border banking. Both host and home countries have interests at stake if an international bank faces distress or liquidity problems arise whether the bank has a subsidiary organization or a branch organization. The potential for conflicts of interest among supervisors about burden sharing and potential channels of contagion is likely to be greater for relatively complex international banking organizations. The Nordea case discussed in the section ‘The Nordea Case’ illustrates that neither hostnor home country supervisors have been willing to abandon all control over domestic as well as foreign operations independent of its organization. The failure of the Icelandic banks in 2008 also shows that both home and host country authorities were unwilling to take the consequences of the EU (and EEA) rules of responsibility for subsidiaries and branches. Among the three Icelandic banks that failed, two had subsidiaries in the United Kingdom and one had branches. Formally, Iceland had the responsibility for deposit insurance for the UK branches, while the subsidiaries were within the UK jurisdiction. Iceland was unwilling or unable to protect the depositors in the branches, while the UK government could not accept that any UK depositors lost money. There is now a legal conflict with respect to the funds paid put to depositors when the Icelandic banks failed. Several papers point toward the lack of formal, predictable, and credible crisis management procedures

4

293

as a major obstacle to the implementation of an effective framework for supervision and crisis management.4 The problems of crisis management are compounded when cross-border banks are involved, in particular when there are no formal procedures to rely on. Ad hoc solutions must be found quickly, and those solutions typically involve some support to keep operations going for fear of contagion. A number of issues in cross-border banking must be resolved in order to obtain an effective architecture for supervision and crisis management. By effective, we mean that financial institutions should be able to choose organizational structure in accordance with the needs of their business, that supervision is based on the best possible information with clear allocation of responsibility, and that responsibility for liquidity assistance and crisis management is clearly defined. Furthermore, the architecture should allow insolvent banks to be closed with consequences for creditors and shareholders in accordance with contracts. The following major issues related to cross-border banking in different organizations can be identified: 1. Few countries have insolvency procedures in place that allow insolvent banks to be closed under predictable rules. Thus, various forms of bailouts of large banks are the rule rather than the exception. As a result, market discipline suffers and banks have incentives to shift risk to taxpayers and deposit insurance funds. Insolvency procedures need to be credible, predictable, and nondiscriminatory with respect to origin of a claim on a failing bank. 2. Cross-border banks are often ‘hybrid’ structures between subsidiaries and branches. This contributes to opacity and inability of host country supervisors to obtain a clear picture of the risk of an individual subsidiary, as well as of the systemic risk associated with a branch of a foreign bank. Home country supervisors of banks with foreign subsidiaries may collaborate with a distressed bank to allow shifting of risk to another country, while home country supervisors of branches are unable to assess their exposures. In a crisis, there is a high likelihood of conflicts with respect to burden sharing. Even if a special insolvency law is in place in each country, jurisdictional conflicts may arise. 3. Differences in deposit insurance coverage between domestic and foreign banks operating in the same country could lead to politically unacceptable consequences in case a foreign bank with relatively low coverage fails. In the United States, branches of foreign banks must join the US deposit insurance

For example, Eisenbeis and Kaufman (2007), Lastra and Wihlborg (2007), and Fiechter et al. (2011).

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

294

29. ORGANIZATION, SUPERVISION AND RESOLUTION OF CROSS-BORDER BANKING

system, and, therefore, US regulators also restrict the operations of foreign branches. The benefits of branch banking cannot be realized fully under these conditions, however. 4. The allocation of LOLR responsibility needs to be clarified for cross-border subsidiaries as well as branches. A branch would in principle be part of the banks’ liquidity management and thereby subject to the home country’s LOLR operations. Liquidity is under the domain of central banks, however, and therefore directed at a country’s financial system that may include branches of foreign banks. 5. As the recent crisis shows LOLR operations easily become part of a bailout policy with fiscal implications. Controversies may arise with respect to foreign banks, in branches in particular, enjoying benefits from domestic taxpayers. Incentives to conduct LOLR operations to support banks with insolvency problems could be reduced if effective insolvency procedures were in place. 6. The recent crisis also demonstrated that systemic risk in the financial system arises through nonbank activities as well as through the traditional channels. Thus, parts of the crisis management architecture may appropriately apply to nonbank financial institutions as well. Financial services other than traditional banking are often offered cross-border in branches rather than subsidiaries. Thus, there is a need for effective approaches to deal with branches in host countries. Reforms addressing these issues are necessary to obtain an effective architecture with respect to supervision and crisis management based on national jurisdiction over domestically incorporated entities. Supernational supervision and crisis management are hardly realistic outside the EU and possibly quite distant there as well. We will not focus on EU issues here but refer to Schinasi and Teixeira (2006) for a discussion of potential reforms in the EU. A decentralized, nationally oriented approach to the financial architecture seems necessary. This approach does not preclude joint solutions for supervision and agreements with respect to crisis management as illustrated by the Nordea case.

CONCLUSIONS: NEED FOR REFORM OF THE ARCHITECTURE FOR SUPERVISION AND CRISIS MANAGEMENT We have concluded that the effective organization of cross-border banking is likely to include both subsidiaries and branches. The choice from a bank’s point of view depends on the nature of financial services provided, the nature of the market served, tax systems,

and the types of risk the bank faces in home and host countries. In order to allow banks to choose their most efficient organizations, the architecture for supervision and crisis management needs reforms with the objective to minimize the risk of financial crisis and contagion between countries. These reforms must be implemented on the national level since supervision and crisis management ultimately requires involvement of national fiscal authorities. Incentive compatibility requires that supervisors representing a nation’s taxpayers are responsible for risk-taking in banks that in distress can become a burden for those taxpayers as a result of the coverage of deposit insurance systems. The EU Banking Directive assigning home country control and supervision over host country branches and home country insurance for depositors in these branches would provide appropriate incentives for supervisors charged with the minimization of risk for the deposit insurance fund. The US system of supervisory jurisdiction over branches of foreign banks in combination with US deposit insurance for depositors in these branches would seem to provide appropriate incentives for supervisors as well, but if the branches do not have an independent capital base, the risk associated with the branches’ operations is not under the American supervisor’s control. This problem has been solved in the United States by a requirement that branches have independent capital at risk. This type of ring-fencing implies that the benefits of the branch organization in host countries cannot be fully exploited. Central banks as LOLR have responsibility for emergency liquidity assistance. To the extent LOLR operations are restricted to liquidity assistance and do not include a subsidy component, taxpayers do not bear risk associated with liquidity assistance. The recent crisis has demonstrated that emergency liquidity assistance by central banks often has a strong subsidy component. Thus, if host country operations take place in branches relying on a foreign parent’s capital, LOLR operations with a subsidy component to all banks within a central bank’s jurisdiction would support the foreign bank under foreign supervision. The foreign supervisor representing foreign taxpayers would not take into account potential costs of subsidized host country LOLR operations. Thus, the bank with foreign branches may become undersupervised from the point of view of taxpayers in home and host countries jointly. To avoid undersupervision of banks with branches in host countries, the responsibility for emergency liquidity assistance can be assigned to the home country. Central banks conducting LOLR operations are not concerned with liquidity in other countries, however. Another way to avoid undersupervision of banks with branch operations abroad would be to prevent

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

CONCLUSIONS: NEED FOR REFORM OF THE ARCHITECTURE FOR SUPERVISION AND CRISIS MANAGEMENT

central banks from conducting LOLR operations that include a subsidy component.5 All costs associated with crisis management and insolvency risk would then be borne by taxpayers in the jurisdictions where banks are incorporated. The operation of branches in host countries with home country responsibility for supervision and crisis management requires that host country supervisors and central banks can rely on the home country authorities not to favor domestic interests at the expense of host countries when an international bank is in distress. The operation of subsidiaries in host countries would imply host country responsibility for supervision and crisis management including emergency liquidity assistance. This division of responsibility also requires mutual trust in home and host country supervisors and crisis management authorities. To the extent the subsidiaries are functionally integrated, this mutual trust is particularly hard to achieve. We conclude with a brief description of essential reforms that would be required to assign credible supervisory and crisis management responsibilities for banks incorporated in a particular jurisdiction to authorities in this jurisdiction while allowing cross-border banking in both subsidiaries and branches. 1. Several papers referred to above emphasize the need for a lex specialis for bank insolvency on the national level. The laws need to specify credible, predictable, and nondiscriminatory procedures that enable even a large bank to fail without creating severe contagion effects domestically and abroad. If host country supervisors are not convinced that the country’s interests are considered fairly, they are likely to be uncooperative with respect to the establishment of branches in their countries. 2. Insolvency procedures need to be complemented with ‘structured early intervention’ (SEI) allowing supervisors to intervene and restrict risk-taking well in advance of insolvency. Thus, to make cross-border banking in branches acceptable, home countries must introduce credible SEI as well as insolvency law in order to earn trust as supervisors of large parts of host country banking systems.6 The procedures must assure all countries involved that the intervention will be fair in relation to all branches and creditors independent of country. 3. SEI and insolvency procedures must be strong enough and allow intervention early enough to make

295

subsidized emergency liquidity assistance by central banks essentially unnecessary. Otherwise, the ambiguity about the responsibility for LOLR intervention reduces transparency with respect to crisis management responsibility. 4. For home country taxpayers not to bear excessive risk for deposit insurance for large host country branches, the deposit insurance system must be funded to such an extent that taxpayers face little risk.7 5. The functional ambiguity between branches and subsidiaries needs to be resolved by requirements that legally separate entities are functionally separate as well. New Zealand requires foreign subsidiaries to be ‘operationally separable’ essentially over night in order to avoid contagion from distress in a foreign parent bank to spread to New Zealand. Branches are not allowed in New Zealand, however. Spain also restricts operations of foreign subsidiaries.

References Allen, F., Beck, T., Carletti, E., et al., 2011. Cross-Border Banking in Europe: Implications for Financial Stability and Macroeconomic Policies. Centre for Economic Policy Research, London. Calzolari, G., Loranth, G., 2010. Regulation of multinational banks: a theoretical inquiry. Journal of Financial Intermediation 20 (2), 178–198. Cerutti, E., Dell’Ariccia, G., Soledad Martinez Peria, M., 2007. How banks go abroad: branches or subsidiaries? Journal of Banking and Finance 31, 1669–1692. Dell’Ariccia, G., Marquez, R., 2010. Risk and the corporate structure of banks. Journal of Finance 65 (3), 1075–1096. Dermine, J., 2006. European banking integration and the Societas Europeae: from host country to home country control. In: Caprio, G., Jr. Evanoff, D.D., Kaufman, G.G. (Eds.), Cross-Border Banking, Regulatory Challenges. Studies in International Economics, vol. 1. World Scientific, New Jersey. Eisenbeis, R., Kaufman, G., 2007. Cross-border banking: challenges for deposit insurance and financial stability in the European Union. In: Benink, H., Goodhart, C., Lastra, R.M. (Eds.), Prompt corrective action and cross-border supervisory issues in Europe. London School of Economics, London FMG Special Paper 171. ¨ tker-Robe, I., Ilyina, A., et al., 2011. Subsidiaries and Fiechter, J., O branches: does one size fit all? IMF Staff Discussion Note, March 7, SDN/11/04. Goodhart, C., Schoenmaker, D., 2006. Burden sharing in a banking crisis in Europe. Sveriges Riksbank Economic Review 2, 34–57. Herring, R., 2003. International financial conglomerates: implications for national insolvency regimes. In: Kaufman, G. (Ed.), Market Discipline and Banking: Theory and Evidence. Elsevier, Amsterdam, pp. 99–129.

5

Emergency liquidity assistance by the European Central Bank to banks across the eurozone implies that an authority without accountability to particular taxpayers takes risk on behalf of eurozone taxpayers independent of banks’ incorporation.

6

Home country supervision need not imply that host country expertise is lacking. The home country supervisor can work with and rely on host country expertise to the extent it is needed to obtain appropriate information. 7

Spain is an example of an EU country requiring banks to obtain permission to set up branches abroad.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

296

29. ORGANIZATION, SUPERVISION AND RESOLUTION OF CROSS-BORDER BANKING

Lastra, R., Wihlborg, C., 2007. Law and Economics of Crisis Resolution in Cross-Border Banking. In: Benink, H., Goodhart, C., Lastra, R.M. (Eds.), Prompt Corrective Action and Cross-Border Supervisory Issues in Europe. London School of Economics, London FMG Special Paper 171.

Schinasi, G.J., Teixeira, P.G., 2006. The lender of last resort in the European financial market. IMF Working Paper, WP/06/127. Schnabl, P., 2011. Financial globalization and the transmission of bank liquidity shocks: evidence from an emerging market. Journal of Finance forthcoming.

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

C H A P T E R

30 Dynamic Provisioning to Reduce Procyclicality in Spain S. Ferna´ndez de Lis*, A. Garcia-Herrero† †

*BBVA, Madrid, Spain BBVA, Hong Kong, Hong Kong, SAR, China O U T L I N E

Introduction

297

The Housing Boom and Bust in Spain

298

The Introduction of Dynamic Provisions in Spain How Was the System Expected to Work? How Did the System Work? Measures Introduced During the Crisis

300 301 303 304

INTRODUCTION The procyclicality of the financial system has traditionally been a source of concern in many countries in the past, but this concern has been exacerbated by the recent global financial crisis. Not only did the behavior of financial institutions exaggerate the housing bubble and the ensuing bust but also there is a strong consensus that financial regulation itself fueled this procyclical pattern. Instruments like fair value accounting and the reforms in banking regulation and supervision related to the introduction of Basel II increased the sensitivity of financial institutions to the cyclical situation, thus amplifying the financial accelerator mechanism. The debate on the reform of the international financial regulation has therefore been focused to a large extent on how to avoid this procyclicality. The reports produced by international bodies – like the G-20 or the Financial Stability Board (FSB) – in the 2008–2010 period identified the main deficiencies observed and suggested a series of instruments or policies to correct or at least attenuate this problem. The global financial crisis has therefore heightened the interest on international experiences in the use of macroprudential tools to limit the extent of the boom Handbook of Safeguarding Global Financial Stability http://dx.doi.org/10.1016/B978-0-12-397875-2.00030-1

Comparison with Other Countries: Peru and Colombia

304

Conclusions Glossary Further Reading

307 307 307

and the bust, a number of which concentrated in the area of housing finance, such as the following: credit growth ceilings (overall or sectoral), loan to deposits ratios, restrictions on foreign exchange lending, LTV limits for mortgages, compulsory insurance for loans above a certain LTV, and limits on affordability ratios. Most of these instruments have been used in emerging market economies (EMEs), in particular in Asia and central and eastern Europe. Dynamic provisioning is another policy instrument that has attracted considerable interest, because of its theoretical appeal, universal applicability, and apparent simplicity. The Spanish experience received considerable interest since its inception in 2000. Other countries introduced similar mechanisms in recent years, in particular Latin America (Peru, Colombia, and Uruguay), but their experience is shorter than that of Spain. One important caveat that needs to be taken into account when considering the use of anticyclical tools is that the procyclicality of the financial system cannot (and probably should not) be totally eliminated, to the extent that risk follows a cyclical pattern. It is important to have a clear understanding of what can be expected and what cannot be expected from anticyclical instruments.

297

# 2013 Elsevier Inc. All rights reserved.

298

30. DYNAMIC PROVISIONING TO REDUCE PROCYCLICALITY IN SPAIN

The structure of the chapter is as follows: section ‘The Housing Boom and Bust in Spain’ briefly explains the recent cyclical patterns of the Spanish economy over recent decades and the main traits of the housing bubble and bust in the 2000s; section ‘The Introduction of Dynamic Provisions in Spain’ deals with the experience of dynamic provisions in Spain; and section ‘Comparison with Other Countries: Peru and Colombia’ compares this experience with other Latin American countries and section ‘Conclusions’ presents some conclusions.

THE HOUSING BOOM AND BUST IN SPAIN The Spanish economy has generally been quite volatile compared to the standards of other European Union countries. This is specially the case for bank lending, which has been subject to large swings following the economic cycle. As Chart 30.1 shows, credit is strongly procyclical in Spain. The first credit cycle is related to the Spanish industrialization process and the integration of the economy in world markets. Before the mid-1970s oil crisis, the Spanish economy grew strongly and government-controlled interest rates were set at very low levels. The loose monetary environment, compounded by the absence of correct incentives and management skills among bank managers, contributed to the strong growth of bank lending and to an excessive indebtedness on the part of Spanish nonfinancial firms. Such excessive lending and indebtedness – together with the oil supply shock – fueled the banking crisis that affected half of Spain’s commercial banks between 1977 and 1985. In the context of political instability, Spain underwent a sharp and prolonged slowdown, which led to stagnated bank lending in the mid-1980s.

The subsequent credit boom can be largely explained by the housing market. In fact, house prices more than doubled during the second half of the 1980s. Banks saw the opportunity and shifted their lending to mortgages on households. The sharp increase in interest rates immediately before the ERM (Exchange Rate Mechanism) crisis and the devaluation of the peseta led to a collapse in bank lending between 1992 and 1996. Credit started to recover as expectations that Spain was going to become a member of the European Monetary Union increased. This implied that interest rates went down rapidly to converge with those of Germany. The prolonged period of macroeconomic stability after the euro adoption (in the context of low inflation and negative real interest rates), coupled with high gross domestic product (GDP) growth, led to a rapid growth in credit, which reached double digits in the second half of the 1990s and the first half of the 2000s. Much of this lending has again been fueled to the housing market, given the very sharp increase in prices during the last few years. Growing competition among banks, reflected in declining banks’ intermediation margins, also helped boost credit growth, although profitability, measured by return on equity, has remained high due to cheap funding, growth in business, and also low loss provisions in the 1990s. All in all, the strong procyclicality of credit in Spain and the negative experience of previous cycles explain to a certain extent the Bank of Spain decision to introduce dynamic provisioning in July 2000. On the basis of the pattern of the previous cycles, there was a concern that the expansion of banks’ loan portfolios, together with very low loan-loss provisions, was increasing banking risks that would materialize in the next downturn phase. In other words, provisions did not keep pace with potential credit losses, which were latent in the new lending. But to understand the context of the introduction of dynamic provisions in the early 2000s, it is necessary to

CHART 30.1

30% 25% 20% 15% 10% 5% 0%

Credit

2008

2005

2002

1999

1996

1993

1990

1987

1984

1981

1978

1975

1972

1969

1966

−10%

1963

−5%

GDP

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

GDP and credit growth.

THE HOUSING BOOM AND BUST IN SPAIN

explain first the main features of the housing bubble and bust in the 2000s. As can be seen in Chart 30.2, house prices started to accelerate in the second half of the 1990s, reaching doubledigit annual growth in the early 2000s. As explained in the previous section, this happened in the context of rapid credit growth. The introduction of the euro implied the loss of the monetary instruments to deal with this problem, reducing the array of tools at the disposal of the authorities. The housing bubble in these years can be explained by several reasons: • The main explanation was the euro adoption. The integration in a common currency area implied several effects for the countries in the periphery: a sharp reduction in risk premia and real interest rates that fueled credit demand and house prices, as well as real economic growth and positive price differentials, vis-a`-vis core Eurozone countries (especially Germany), where the effects were the opposite: positive real interest rates, stagnant credit, subpar growth, depressed real estate prices, and below-average inflation. Over time, these patterns explain the accumulation of competitiveness losses (gains) and current account deficits (surpluses) in the periphery (core) of the Eurozone, which are behind the euro crisis in 2010. • Other factors: ○ Demographics: Population increased from 39.8 to 46.7 million from 1998 to 2008, thanks to a great extent to immigration. During the same period, the size of the average Spanish household decreased from 3.7 to 2.7 people. ○ Foreign demand. ○ The increase in banking competition. ○ The tax deductibility of amortization expenses in income taxes.

299

As a consequence of the evolution of real interest rates, disposable income, and general financial conditions, affordability conditions for house purchases improved significantly during the 1990s in Spain (see Chart 30.3). The rise in house prices in Spain took place in the context of dramatic increase in construction activity during the boom years. Housing starts reached more than 700 000 per year at the peak of the bubble in 2006, more than construction activity in the other four big EU countries combined (Germany, France, United Kingdom, and Italy). The case of Spain was not one of inelastic supply that, in the presence of a credit boom, explains house price rises, but rather a case of increase in demand so intense that, despite an unprecedented supply growth, it led to a sizable surge in prices. The rate of increase in house prices peaked in 2004 to above 15% in annual terms and experienced a slowdown thereafter, to attain only slightly positive rate in mid2007, when the US subprime crisis accelerated the then ongoing correction. Price changes have been in negative territory since mid-2008, although as can be seen in Chart 30.2, the rate of decrease has moderated since the second half of 2009. The accumulated correction in house prices since the peak in December 2007 reached 17% in August 2010, comparable to the UK figures for the same period, but significantly smaller than the US data. The crisis had a devastating macroeconomic impact: GDP fell sharply and unemployment reached 20% in 2010 and construction activity contracted sharply. Further consequences have been a credit contraction, a rapid increase in bad loans (especially to the property development sector), a wave of bankruptcies in some of the main construction companies, and funding liquidity problems for Spanish banks, especially the smaller and mostly regional savings banks, whose credit is much more CHART 30.2

Housing prices in Spain (year on year growth in %). Source: Ministerio de Vivienda.

20% 15% 10% 5% 0% −5%

Mar-10

Mar-09

Mar-08

Mar-07

Mar-06

Mar-05

Mar-04

Mar-03

Mar-02

Mar-01

Mar-00

Mar-99

Mar-98

Mar-97

Mar-96

−10%

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

300

30. DYNAMIC PROVISIONING TO REDUCE PROCYCLICALITY IN SPAIN

CHART 30.3 Affordability of housing purchases in Spain (net effort in percent of disposable income).

70% 60% 50%

42,6%

40% 30% 27,0%

20% 10%

concentrated in developers, constructors, and mortgages. The fiscal impact of the crisis, combined with poor growth prospects, accumulated competitiveness losses, and the contingent liabilities related to banking sector losses led to a deterioration of sovereign risk in the aftermath of the Greek crisis. A fiscal austerity package, labor market reforms, and transparency in the publication of the banks’ stress tests in the summer of 2010 led to an improvement in the sovereign debt situation, although the situation of peripheral countries in the euro area remains vulnerable at the time of writing in October 2010.

THE INTRODUCTION OF DYNAMIC PROVISIONS IN SPAIN As mentioned in the previous section, the introduction of dynamic provisioning in Spain in the year 2000 should be seen in the context of the profound impact of the euro adoption in the Spanish economy. The expansionary impact of the reduction in real interest rates (estimated around 4–5% points) on the Spanish economy was very significant. Domestic credit growth, which ranged between 5 and 10% in the mid-1990s, accelerated to rates above 15% in 1998–2000. House prices increased at an annual rate of around 10% in the same period. Inflation accelerated from 1.9% in 1997 to 2.2% in 1999 and 3.5% in 2000. The differential in domestic demand growth between Spain and Germany in the early years of monetary union was around 3.6% points, mostly related to the gains from price stability and policy credibility for Spain (and in general peripheral countries), whereas Germany, where credibility was already high, did not experience a similar effect. The European Central Bank kept interest rates around 4% in the late 1990s, a level which was consistent with the average situation of the Eurozone, but which implied

Mar-10

Mar-09

Mar-08

Mar-07

Mar-06

Mar-05

Mar-04

Mar-03

Mar-02

Mar-01

Mar-00

Mar-99

Mar-98

Mar-97

Mar-96

Mar-95

Mar-94

Mar-93

Mar-92

Mar-91

Mar-90

0%

very lax monetary conditions for the Spanish economy. This expansionary impact was compounded by the depreciation of the euro vis-a`-vis the US dollar in these years. At the turn of the century, therefore, the Spanish authorities saw with increasing anxiety the combination of high credit growth, inflation differentials with the Eurozone average, loss of competitiveness, and widening current account deficits. Monetary policy and the nominal exchange rate were no longer available as policy instruments. In this context, dynamic provisions (or statistical provisions, according to the denomination they received at the time) were seen as an instrument with a double objective: (1) to contain credit growth, by increasing the cost (in terms of provisioning effort) of the granting of new credit, and (2) to protect Spanish banking institutions from future losses as a consequence of the relaxation of lending standards typical of the boom phase. While the first objective was probably more important at the time of adoption of this system, the results were much more satisfactory in terms of the second objective. Dynamic or statistical provisioning was therefore a truly macroprudential tool, in the sense that a prudential instrument (provisions) was used to achieve a systemic or macroeconomic goal (limiting credit growth). As concerns the second objective, it was mostly addressed at ensuring an adequate protection to individual institutions (and therefore could be seen as a microprudential tool), but to the extent that excessive risk assumption was partly a result of herd behavior and collective myopia by credit institutions; it had also a certain macroprudential aim. Before entering into the details of how the dynamic provisioning system worked, it is important to keep in mind that it was one aspect of the strategy of a banking supervisor which relied on other instruments to ensure an adequate risk control. Some of the accompanying mechanisms were the following:

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

301

THE INTRODUCTION OF DYNAMIC PROVISIONS IN SPAIN

• Very intensive supervision, especially in the case of systemic institutions, where on-site inspection played a crucial role. The approach of the Bank of Spain was even considered ‘intrusive’ as compared to the ‘light touch’ philosophy of other countries like the United Kingdom. • Focus on consolidated supervision both in terms of accounting and capital treatment of any conduits and SPVs (Special Purpose Vehicles), so that arbitrage opportunities to save capital adequacy by spurious disintermediation requirements are avoided. • ‘Skin in the game’ in the securitization process. According to market practices, issuers keep the first losses, which generated positive incentives in terms of underwriting and credit risk assessment and followup of borrowers. Furthermore, securitization is used as a liquidity instrument, but not to optimize capital since it does not decrease risk weighted assets (RWA). • A single financial supervisor: All financial intermediaries are supervised with entirely homogeneous regulations by the Bank of Spain. This discouraged regulatory arbitrage.

provisioning effort fuels low risk aversion and credit growth, thus feeding back economic growth. In the downturn, the opposite spiral operates: the difficult economic environment is accompanied by high NPLs, which require a bigger provisioning effort. This in turn decreases risk appetite and feeds credit contraction. Hence the procyclical pattern of normal provisions. The objective of dynamic provisions is to smoothen the provisioning effort along the cycle, as shown in Chart 30.5. How much? This is an open question. While the idea is to avoid the procyclical effect of the normal system, a regulator would hardly aim at an opposite pattern of provisions (i.e., increase in the good times and decrease in the bad times), since risk is cyclical and this reality should be reflected in provisions. A reference would be to try to obtain a flat provisioning effort along the cycle in terms of the ratio of provisions to credit. Chart 30.5 – which should be taken only as a reference – depicts provisions with a smoothed procyclical pattern, which was more or less what was aimed at in Spain. Although it evolved since its introduction in 2000, the formula of the dynamic provisioning system was Generic provisions ¼ aDCredit þ bDCredit  specific provisions

How Was the System Expected to Work? As can be seen in Chart 30.4, under a normal provisioning system provisions are a function of contemporary nonperforming loans (NPLs), although this may be smoothed by the possibility of using ‘generic’ provisions based on the credit stock. In the upturn, when GDP grows above potential, credit growth also accelerates. Since business conditions are favorable, collateral prices are increasing, and optimism is pervasive, debtors have in general no problem in servicing the debt. The low

Normal provisioning

with the following coefficients: Type of risk

a (%)

b (%)

No apparent risk Low risk Low–medium risk Medium risk Medium–high risk High risk

0.0 0.6 1.5 1.8 2.0 2.5

0.00 0.11 0.44 0.65 1.10 1.64

CHART 30.4

Credit

NPL

Provisions

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

Normal provisioning cycle.

302

30. DYNAMIC PROVISIONING TO REDUCE PROCYCLICALITY IN SPAIN

Dynamic provisioning

CHART 30.5

Credit

Dynamic provisioning cycle.

Provisions

NPL

CHART 30.6

Spain: generic provisions and its subcomponents

Spain: Generic provisions and its

subcomponents.

Specific provisions

Cycle (credit)

α * Δ credit

(β * credit) - specific

Generic = α * Δ credit + (β * credit) - specific

Chart 30.6 shows how the different components of the dynamic provisioning formula were expected to operate along the cycle. The interpretation of the coefficients included in the formula is the following: • a is the average estimate of the credit loss in a year neutral from a cyclical perspective for each asset class. It is based on the idea of ‘collective assessment for

impairment,’ which implies that losses have not been identified yet as related to specific loans. • b is the historical average specific provision of each group of loans. When credit growth is high and there is a low risk aversion, specific provisions are low and the component [(b  credit growth)  specific provisions] is positive, adding to the component [a  credit] (see green line in Chart 30.6). In the downturn, when credit growth is low, this component

III. SAFEGUARDING GLOBAL FINANCIAL STABILITY

303

THE INTRODUCTION OF DYNAMIC PROVISIONS IN SPAIN

is negative, thus deducting from the component [a  credit] and eventually causing the generic fund to be drawn down. The fund has an upper (125%) and lower (33%) limit. The former (introduced in the 2004 reform) was designed to avoid excess provisioning, whereas the latter was conceived to cover incurred losses not yet individually identified even in a recession.

How Did the System Work?

Specific provision to credit

Generic provision to credit

GDP (yoy growth, RHS)

Mar-10

Mar-09

Total provision to credit

Sep-09

Mar-08

Sep-08

Sep-07

Mar-07

−6.0% Mar-06

0.0% Sep-06

−4.0%

Mar-05

0.5%

Sep-05

−2.0%

Mar-04

1.0%

Sep-04

0.0%

Mar-03

1.5%

Sep-03

2.0%

Sep-02

2.0%

Mar-02

4.0%

Mar-01

2.5%

Sep-01

6.0%

Mar-00

3.0%

Sep-00

8.0%

Mar-99

3.5%

Sep-99

% to credit

As can be seen in Chart 30.7, credit growth stabilized at around 15% annually after the introduction of dynamic provisioning in 2000 and decreased slightly between 2001 and 2004. It is difficult to assess, however, to what extent this was related to the new provisioning system. Most probably, the impact of the burst of the dotcom bubble was more relevant in this period. After 2004, however – coinciding with a reform of the provisioning system – credit accelerated sharply and reached rates of growth above 25% in 2006. The impact of the global financial crisis since mid-2007 implied a sharp contraction of both credit and GDP. To understand these patterns, it is useful to recall how the system was designed and how it was reformed in 2004. Initially, the system introduced in 2000 was based on three types of provisions: specific, generic (both already existing), and statistical (newly introduced in 2000). Specific provisions depended on current bad loans, generic

Yoy growth

provisions were 1% of the credit stock, and statistical provisions were designed to offset specific provisions and depended on credit growth. This mechanism was criticized (1) by international accounting bodies, which argued that it implied profit smoothing along the cycle which masked the real situation of the banks and (2) by Spanish financial institutions, which complained about being subject to higher provisioning requirements than their competitors, which was considered an important competitive disadvantage in the single European market for financial services. By 2004, there was a sense that these provisions were excessive and perhaps explosive. By that time, they reached a level of more than 2.5% of credit (of which