university of groningen foreign bank entry and …...remco van der molen, thanks for being a great...

University of Groningen

Foreign bank entry and performance with a focus on Central and Eastern EuropeNaaborg, I.

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2007

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Naaborg, I. (2007). Foreign bank entry and performance with a focus on Central and Eastern Europe. [S.l.]:[s.n.].

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 08-08-2020

https://www.rug.nl/research/portal/en/publications/foreign-bank-entry-and-performance-with-a-focus-on-central-and-eastern-europe(2e0bd200-8881-4c8d-8f9b-9e7c0fd4eb24).html

https://www.rug.nl/research/portal/en/publications/foreign-bank-entry-and-performance-with-a-focus-on-central-and-eastern-europe(2e0bd200-8881-4c8d-8f9b-9e7c0fd4eb24).html

F o r e i g n B a n k E n t ry a n d P e r f o r m a n c e

with a focus on Central and Eastern Europe

Rijksuniversiteit Groningen

Foreign Bank Entry and Performancewith a focus on Central and Eastern Europe

(met een samenvatting in het Nederlands)

Proefschrift ter verkrijging van de graad van doctor

aan de Rijksuniversiteit Groningen op gezag van de Rector Magnificus

Prof. dr. F. Zwartsingevolge het besluit van het College voor Promotiesin het openbaar te verdedigen op 22 februari 2007

des donderdags te 14.45 uur

door

Ilko Joost NaaborgGeboren op 7 september 1974 te Delft

�

Promotores Prof. dr. J. De HaanProf. dr. B.W. LensinkProf. dr. L.J.R. Scholtens

Committee Prof. dr. J.H. Garretsen (Utrecht University)Prof. dr. K.H.W. Knot (University of Groningen/ DNB) Prof. dr. R. Vander Vennet (Ghent University)

© 2007 Ilko NaaborgUitgeverij EburonPrinted on Biotop paper: made entirely from wood thinnings and chippings, totally chlorine free.Bookdesign by Donkigotte.nlCoverphoto Karluv most, Prague. ©iStockphoto.com/Tim Starkey, 2005

ISBN 978-90-5972-170-8

�

Contents

Acknowledgements viiList of Tables, Figures and Appendices ixAbbreviations xiIntroduction 1

I Foreign Bank Behaviour in Central and Eastern Europe 11

1 Foreign Bank Entry and Credit Performance 131.1 Introduction 13

1.2 Theory 141.3 The importance of foreign banks in CEE 171.4 The origin of foreign direct investment in the banking sector 221.5 Intermediation and the role of foreign owned banks 271.6 Financial performance of foreign banks 321.7 Summary and conclusion 35

2 Size and Operations 392.1 Introduction 392.2 Methodology and data 402.3 Bank size 402.4 Assets 442.5 Liabilities 48 2.6 Foreign banks’ balance sheets 502.7 Conclusion 53

3 Entry Strategies 573.1 Introduction 573.2 Hypothesis development 573.3 Methodology and data 633.4 Results 673.5 Summary and Conclusion 76

�i �ii�ii

II Foreign Ownership and Performance 79

4 Financial Performance: Foreign Banks in European transition Countries 81 4.1 Introduction 81 4.2 Data and methodology 84 4.3 Empirical results 87 4.4 Conclusion 93

5 Financial Performance: Foreign Banks Worldwide 95 5.1 Introduction 95 5.2 The data 95 5.3 Model and estimation technique 96 5.4 Empirical results 101 5.5 Conclusion 101

6 Efficiency. Do Institutions Matter? 107 6.1 Introduction 107 6.2 Econometric framework 110 6.3 Data 106 6.4 Empirical results 118 6.5 Conclusion 119

7 Summary and conclusion 127 7.1 Introduction 127 7.2 Summary of the main findings 127

7.3 Policy implications, suggestions for further research and conclusion 130

Bibliography 133Appendices 145

Samenvatting en conclusie (Summary in Dutch) 173

�i �ii�ii

Acknowledgements

While bearing my name as sole author, the chapters in this book are largely a collective undertaking and would not exist without the support and assistance of a large number of people. Foremost I thank Jakob de Haan, who brought this PhD project under my attention and invited me to apply. His straightforward way of supervising has been most helpful, his advice invaluable for completing this book and the many informal Groningen dinners most pleasant.

I am most thankful to Robert Lensink for the opportunity of learning from him. In particular, I owe gratitude to him for his guidance with the empirical work in this book. His dedication to both theory and empirical research has inspired me in the various papers we worked on together. On a more personal level, I thank him for the fruitful research meetings at his home in Annen and the pleasant hours we spend at the SUERF conference in Tallinn and the Tor Vergata Conference in Rome.

I wish to thank Bert Scholtens for his assistance during my PhD-project and his proposal to write a joint-paper for the International Finance Review. I also thank Bert for his many valuable remarks on the several other studies, which make up this dissertation.

My appreciation goes out to Aljar Meesters for being a great research partner and colleague. What started out as two people having interest in the same data set, turned into a fruitful teamwork, which resulted in a rather adventurous and successful trip to the 2006 Prague EIASM conference.

I am grateful to Lammertjan Dam for the exchange of ideas and experiences when we both started out as PhD students. I was lucky to share my office at Zernike with Remco van der Molen, thanks for being a great colleague and office mate and for the experience of being paranimph. I would like to thank Suwarni and Barka for their valuable help with the information technology, I thank Grietje Pol for her secretarial assistance and I thank Ellen and other members of the SOM Research School for being most accessible and helpful. I also thank the research school for providing me with the financial means to collect data from foreign banks throughout Europe.

With great pleasure I take the opportunity to express my gratitude to Martine, Rutger, Martien, Hellen, Petr, Joop, Stamatis, Roan, Meindert and Niek for sharing ideas and for being such great individuals to know. Also, I want to express my deep appreciation for the guidance of my mother for the assistance of my father, and for the invaluable friendship and understanding with Menno and Hijko. Directly or indirectly you all contributed to the realisation of this book.

The chapters in this book benefited substantially from the input of researchers and anonymous referees, who helped ameliorating its content. I am grateful for comments and suggestions by Allen Berger, Abe de Jong, Luca Papi, Linda Toolsema, Katharina Steiner, Elmer Sterken, Semi Yildirim, Laurent Weill, participants of the Lodz 2003

�iii ixix

Open Minds Conference, participants of the Groningen SOM 2003 EU Enlargement Conference, participants of the Rome 2004 XIII Tor Vergata Conference, participants of the Århus 2004 PhD Workshop on Banking and Financial Markets by Mark Flannery, participants of the Prague 2006 EIASM Workshop on Financial Market Development in Central and Eastern European countries, and participants of several seminars and PhD conferences organized by the SOM Research School Groningen.

Part of the research was conducted abroad. I owe many thanks to bankers from ABN AMRO Bank, ING Bank, KBC, HVB Group/BACA, ERSTE Bank, Raiffeisen International, Swedbank, SEB and SAMPO Bank, their subsidiaries in Estonia, the Czech Republic, Hungary and Poland, and staff of Eesti Pank, Ceská Národní Banka, Rahoitustarkastus, Finansinspektionen, Narodowy Bank Polski, Magyar Nemzeti Bank and the Oesterreichische Nationalbank for their time and effort to cooperate in this research project. The project was undertaken in cooperation with the Dutch central bank. I thank Ralph de Haas for his help in coordinating the meetings and his assistance in structuring the data.

Another part of the research was conducted with bank specific data from BankScope, which combines data from Fitch Ratings with other sources. I thank Bob Vaanhold and Mark Wessels at Bureau van Dijk Electronic Publishing in Amsterdam for providing me with additional information and data when I needed it.

The chapters in this book are based on several publications. Chapter 1 is based on publications in the Journal of Emerging Market Finance (Naaborg et al., 2004) and in Masciandaro, D. (ed.), “Financial Intermediation in the New Europe: Banks, Markets and Regulation in EU Accession Countries”, Edward Elgar (De Haan and Naaborg, 2004). Chapter 2 was published in the International Finance Review (Naaborg and Scholtens, 2004). Parts of Chapter 4 will be forthcoming in the European Journal of Finance (Naaborg and Lensink, 2006), Chapter 5 in Applied Financial Economics (Lensink and Naaborg, 2006). At the time of printing, Chapters 6 was under review with the Journal of Banking and Finance. I would like to thank Elsevier Publishing, Edward Elgar Publishing Limited, SAGE Publications and Taylor and Francis Group for allowing me to reprint (parts of) these articles in this thesis.

Ilko NaaborgGroningen, September 2006

�iii ixix

List of Tables, Figures and AppendicesTables Table 1.1 Determinants of real growth in bank loans to customers 17Table 1.2 Number of foreign banks (% total number of banks), 1995-2004 19Table 1.3 Foreign banks’ assets (% total bank assets), 1995-2004 21Table 1.4 Origin of financial FDI in CEE banks, 2002 23Table 1.5 Origin of financial FDI in CEE’s biggest banks, 2005 24Table 1.6 Domestic credit to the private sector (in % GDP), 1995-2004 28Table 1.7 Distance of domestic credit to the benchmark (%-points), 1995-2003 29Table 2.1 Average bank size (€mln), 1995-2002 41Table 2.2 Decomposition of banks assets (% total assets), 1995-2002 44Table 2.3 Decomposition total customer loans (% customer loans), 1995-2002 46Table 2.4 Average equity and liabilities (%) assets), 1995-2002 48Table 2.5 Decomposition of liabilities (% liabilities), 1995-2002 49Table 2.6 Balance sheet ratios of foreign and domestic banks, 1995- 2002 52Table 3.1 Literature summary of foreign bank entry determinants 59Table 3.2 Selected parent banks 65Table 3.3 Selected foreign owned local banks 66Table 3.4 Profile of selected foreign owned local banks, per end 2004 68Table 3.5 Determinants of foreign bank entry, summary of results 70Table 4.1 Descriptive statistics 85Table 4.2 Levels of foreign ownership in CEE banks, 2001 86Table 4.3a Foreign ownership and performance 88Table 4.3b Foreign majority ownership and performance 89Table 4.4 Test 1: does economic development matter? 90Table 4.5 Test 2: does concentration matter? 91Table 4.6 Test 3: does the relative number of foreign banks matter? 92Table 4.7 Test 4: does the relative amount of foreign banks’ assets matter? 93Table 5.1 Foreign ownership worldwide 97Table 5.2 Descriptive statistics 98Table 5.3 Foreign ownership and net interest revenues 102Table 5.4 Foreign ownership and profit 104Table 6.1 Summary of the findings on the efficiency of foreign banks 108Table 6.2 Institutional indicators 116Table 6.3 Correlation matrix of efficiency correlates 118Table 6.4a Bank inefficiency: ownership and host country institutions (all countries) 120Table 6.4b Bank inefficiency: ownership and host country institutions (low income) 121Table 6.5 Foreign bank inefficiency: institutional similarity and home institutions 123Table 6.6 Foreign bank inefficiency: inst. similarity and country host institutions 124 FiguresFigure 1.1 Number of foreign and domestic banks in CEE, 1995-2004 18Figure 1.2 Average foreign ownership, 1995-2004 20Figure 1.3 Bank assets and credit to the private sector (% GDP), 1993-2000 26

x xixi

Figure 1.4 Average credit to the private sector (% total bank credit), 1993-2000 26Figure 1.5 Average credit by foreign and domestic banks (% total credit),1993-2000 30Figure 1.6 Foreign banks’ relative share in credit to the private sector, 1995-2000 31Figure 1.7 Foreign banks’ relative share in credit to the private sector, 2000 32Figure 1.8 Average ROE of foreign and domestic banks, 1995-2000 33Figure 1.9 Average ROE and net interest revenues, 1995-2000 33Figure 1.10 Average overhead costs (% assets), 1995-2000 34Figure 2.1 Average bank size in Central and Eastern Europe (€mln), 2002 41Figure 2.2 Average growth in bank size (1995 = 100), 1995-2002 42Figure 2.3a Average foreign bank size, EU15 (in €bn), 1995-2002 43Figure 2.3b Average foreign bank size, CEE countries (in €bn), 1995-2002 43Figure 2.3c Average foreign bank size, SEE/CIS countries (in €mln), 1995-2002 43Figure 2.4a Maturity of total customer loans, EU15, 1995-2002 45Figure 2.4b Maturity of total customer loans, CEE countries, 1995-2002 45Figure 2.4c Maturity of total customer loans, SEE/ CIS countries, 1995-2002 46Figure 2.5a Maturity of customer deposits (% customer deposits), EU15, 1995-2002 49Figure 2.5b Maturity of customer deposits (% customer deposits), CEE, 1995-2002 49 Figure 2.5c Maturity of customer deposits (% customer dep.), SEE/CIS, 1995-2003 50Figure 3.1 Determinants of the mode of entry, stylised results 74

AppendicesAppendix 1 Description of the central bank survey 145Appendix 2 Total number of banks in CEE countries, 1994-2004 146Appendix 3 Number of foreign banks in CEE countries, 1994-2004 146Appendix 4 Asset share of state-owned CEE banks, 1995-2004 147Appendix 5 Number and origin of banks in sample chapters 1 and 4 148Appendix 6 Number and origin of banks in sample chapter 2 148 Appendix 7 Banking assets per country (% assets per region), 1995-2002 150Appendix 8 Number of foreign and domestic banks in sample chapter 2 151Appendix 9 Average bank size, per country (in €mln), 1995-2002 152Appendix 10 Banks, supervisory institutions, the representatives and their position 154Appendix 11 Initial entry strategies and the sequence of events 156Appendix 12 Number and origin of banks in sample chapter 5 160Appendix 13 Macro-economic descriptives of chapter 5 bank sample 161Appendix 14 Descriptive statistics of chapter 6 bank sample 163Appendix 15 Correlation matrix of inefficiency correlates (foreign banks only) 167Appendix 16a Estimation of the cost frontier (model specification Table 6.4a) 168Appendix 16b Estimation of the cost frontier (model specification Table 6.4b) 169Appendix 17 Estimation of the cost frontier (model specification Table 6.5) 170Appendix 18 Estimation of the cost frontier (model specification Table 6.6) 171Appendix 19 Generalized likelihood ratio tests 172

x xixi

Abbreviations

CEE Central and Eastern EuropeCEO Chief Executive OfficerCFO Chief Financial OfficerCIS Commonwealth of Independent StatesCOR Control of CorruptionCONC ConcentrationCSTF Customer and Short Term FundingDEA Data-Envelopment AnalysisDCPS Domestic Credit to the Private SectorDF Degrees of FreedomDFA Distribution Free ApproachDNB De Nederlandsche BankEBRD European Bank for Reconstruction and DevelopmentECB European Central BankEQ Ratio of Equity over Assets EU European UnionFB Foreign BankFDI Foreign Direct InvestmentFOR Foreign OwnershipGDP(PC) Gross Domestic Product (Per Capita)GMM Generalized Methods of MomentsGNP Gross National ProductGOV Government EffectivenessIFC International Finance CorporationIFS International Financial StatisticsIT Information TechnologyIMF International Monetary FundINST InstitutionalLAW Rule of LawLLP Loan Loss ProvisionsLR Likelihood RatioM&A Mergers and AcquisitionsN.A. Non-available NEA Non-Earning AssetsNIR Net Interest RevenuesNII Non Interest IncomeNPV Net Present ValueNR NumberGMM System Generalized Methods of Moments

xii xii

OLS Ordinary Least Squares OHC Overhead CostsPBT Profit before TaxPF Price of FundsPL Price of LabourPIV Political Instability and ViolenceREG Regulatory BurdenREL RelativeREP RepublicROA Return on AssetsROE Return on EquitySFA Stochastic Frontier AnalysisSEE Southern and Eastern EuropeSME Small and Medium-sized Enterprises ST Short TermTA Total AssetsTC Total CostsLT Long TermUK United KingdomUS(A) United States of AmericaVAC Voice and Accountability

�

Introduction

One of the biggest changes brought about by international deregulation and harmonization of the financial sector has been the rise in foreign banking. Foreign ownership of banks is particularly important in the Caribbean, Central and Eastern Europe, Central Asia, Latin America and Sub-Saharan Africa. This book aims to extend the evidence on foreign banking by examining three research questions: (i) What do foreign banks do in their role as intermediaries? (ii) Why does a bank start foreign operations and does its entry motive affect its entry mode? (iii) Are profitability and efficiency of a bank affected by foreign ownership? In addressing the first two questions, I specifically look at foreign bank behaviour in Central and Eastern Europe. Since 1989, many of the former centrally planned economies in Central and Eastern Europe are returning to a market based economy. During this transition, foreign banks have gradually received the opportunity to establish branches and subsidiaries in Central and Eastern Europe. Now, the importance of foreign banks in Central and Eastern Europe exceeds that in most countries worldwide. In 2006, foreign banks in the new EU member states own 82 percent of bank assets, compared with 5 percent in 1995. The remainder of this introduction has a dual purpose: it explains the books’ approach and the organization of the chapters; and it identifies some important topics that receive little attention in the book.

Foreign banks’ lending behaviourThe study of foreign owned banks has attracted wide attention in recent years from the academic world. Increasingly, the role of foreign banks in (the stability) of credit supply to the private sector is being studied as foreign banks may have more troubles in assessing the emerging private sector, notably small and medium-sized enterprises than domestic owned banks. In a recent study, Detragiache et al. (2006) find that in poor countries, a stronger foreign bank presence is robustly associated with less credit to the private sector. In addition, in countries with more foreign bank penetration, credit growth is slower and there is less access to credit. Similarly, both Berger et al. (2001) and Clarke et al. (2005) find that foreign banks in Argentina, Chile, Colombia and Peru have problems supplying funds to small businesses. Once established abroad, foreign banks focus their activities mainly on large enterprises. Stanley et al. (1993a) find that in the US, foreign–owned banks commit a significantly larger portion of their loanable funds to business lending than domestic banks. Regarding the distribution of business loans between domestic and foreign borrowers in the US, large foreign banks tend to shift over time in favour of domestic business loans (Stanley et al. 1993b). Seth et al. (1998) find that foreign-owned banks first and foremost lend to borrowers other than customers from the home nation. This is in line with a theoretical paper by Du (2003) who argues that multinational enterprises prefer their foreign subsidiaries

� �

being co-financed with local participation as it hardens the budget constraint because local banks have more strength in seizing firm assets in liquidation.

A related topic is the impact of foreign bank entry on the stability of credit supply. According to a 2000 IMF report, it is still unclear whether a greater foreign bank presence contributes to a less volatile supply of credit in emerging economies, especially during adverse economic times. Indeed, Jeon et al. (2006) find that foreign banks supplied more foreign currency loans than domestic banks in pre-crisis Korea for a given home-country growth rate, than in post-crisis Korea. Peek and Rosengren (2000a) find that the collapse of the Japanese equity and real estate markets and the subsequent banking crisis in Japan had an impact in the US, as Japanese bank subsidiaries in the US reduced their lending in response to the problems in Japan. Similarly, Martinez Peria et al. (2002) find evidence of foreign banks transmitting home country shocks to Latin America and Peek and Rosengren (2000b) show that cross-border lending did in some cases retrench during economic slowdowns in Latin America. However, Dages et al. (2000) and Crystal et al. (2002) show that foreign banks that have been present in Latin America for a relatively long time exhibit stronger and less volatile loan growth than domestic banks. Likewise, Peek and Rosengren (2000b) and Martinez Peria et al. (2002) find that foreign bank subsidiaries did not reduce their credit supply during adverse economic times in the host country.

Why do banks go abroad?A second branch of foreign banking examines the relevance of foreign banks’ reasons for entry to their nature of financial intermediation. There are various motives for banks to go abroad. One of the main reasons is that foreign banks follow their customers (Goldberg and Saunders, 1981; Brealey and Kaplanis, 1996; Konopielko, 1999; Buch, 2000; Moshirian, 2001; Green et al. 2004; Williams 2002)1. A second reason why a bank might go abroad is the attractiveness of the host market. The market of the host country may offer new opportunities to make money, depending on the characteristics of the market (like size, stability of the country and features of the local banking sector). A number of recent studies offer support for the view that profitable opportunities in host countries drive foreign bank entry. For example, Claessens et al. (2000) model foreign bank presence in 80 countries, including some transition countries, from 1988-95 and find that foreign banks are attracted to markets with low taxes and high per capita income. In addition to control for the degree of economic integration between countries (non-financial FDI, bilateral trade, and geographical distance) and regulatory restrictions, they include variables that measure the prospects for economic growth and the competitiveness of the banking sector of the host countries. Likewise, Papi and Revoltella (1999) find in their study

1 However, often it is unclear to what extent there is also a causal relationship, as causation may run from FDI of banks to FDI in the non-financial sectors. It is also possible that some omitted factor(s) is (are) driving FDI in both sectors.

� �

Introduction

on 9 transition countries that various variables that proxy the attractiveness of the new market are positively and significantly correlated with FDI initiatives by foreign banks. Similar results are reported by Mathieson and Roldes (2001) in their study on 15 emerging countries, including the Czech Republic, Hungary and Poland. Variables that are significant in this study include the rate of return on equity, non-performing loans and banking crises. Again, for some of these variables causality may go in both directions. For instance, a foreign bank may be attracted to a country with an efficient banking system, but the entry of foreign banks may enhance the overall efficiency of the banking sector. A third consideration for foreign bank entry is host country regulation, which generally limits competition and protects inefficient domestic banks (Clarke et al. 2003). Foccarelli and Pozzolo (2001) find that foreign banks prefer to invest in countries with fewer regulatory restrictions. However, this study does not deal with transition countries. Lensink and De Haan (2002) provide strong evidence that foreign bank entry in transition countries positively responds to economic reform measures. Moreover, they find that reforms significantly affect foreign bank entry via the efficiency of the financial sector and by stimulating domestic investment. Papi and Revoltella (2000) find that the attitude of the authorities in the host country concerning entry of foreign banks is a significant determinant of FDI initiatives of banks in contrast to bank taxes.

The activities of foreign banks are also dependent on the mode of entry. A take-over goes along with the acquisition of the existing client base, including its retail banking activities. Along with the greenfield investment goes the ability to build up the activities from scratch. The bank does not have to cope with the potential of a bad loans portfolio from past activities. Regulation may also affect the activities of foreign banks. In Slovenia, for example, liberalization of foreign borrowing by residents and the abolition of interest rate ceilings on deposits have created a more competitive environment, but only since 1999 (EBRD, 2001). Likewise, Hungary initially did not permit banks to provide financial and insurance services. Changed legislation in 1999 resulted in a movement towards a model of universal banking (ECB, 2001).

Ownership and performance A third branch of foreign banking literature is concerned with the effect of foreign ownership on bank performance. According to Claessens et al. (2001), foreign banks are more profitable and efficient than domestic banks in developing countries, while in developed countries domestic banks are more profitable and efficient than foreign banks. These differences can reflect a differential impact of informational (dis)advantages, customer bases, bank procedures as well as different relevant regulatory and tax regimes. In contrast, DeYoung and Nolle (1996) and Berger et al. (2000) find that foreign banks are less efficient than host nation banks in developed nations. Berger, Dai, Ongena and Smith (2003) find that foreign affiliates of multinational firms use host nation banks for cash management services. This choice appears to affect the geographic scope and size of the chosen bank, the so-called bank reach.

� �

Furthermore, they find that legal and financial development of the host nation affects both bank nationality and bank reach. Focarelli and Pozzolo (2001) and Buch and DeLong (2004) analyse cross-border M&A in the banking industry. These, in general, appear to be relatively unprofitable (see also Berger et al., 2000; DeLong, 2001).

Few studies have been published that focus on transition countries only. Kraft and Tirtiroglu (1998) have studied X-efficiency and scale-efficiencies for both old and new (state and private) banks in Croatia. New banks (i.e. foreign banks) are shown to be more X-inefficient and more scale-inefficient than either old privatised banks or old state banks. However, new private banks are highly profitable. According to these authors, this abnormal situation has been the result of free-riding opportunities created by distressed borrowers, limited competition and start-up difficulties of the new banks. Grigorian and Manole (2002) have analysed 17 transition economies over the 1995-1998 period and find that foreign ownership with controlling power and enterprise restructuring enhances commercial bank efficiency. Weill (2003) has studied banks in the Czech Republic and Poland using the stochastic frontier approach to compute cost efficiency scores. Also this study reports evidence that foreign banks are more efficient than domestic banks. Weill (2003) also concludes that the advantage of foreign banks does not result from differences in the scale of operations or the structure of activities. Yildirim and Philippatos (2003) study twelve transition economies of Central and Eastern Europe over the period 1993-2000 and find that foreign banks are more cost efficient but less profit efficient relative to domestically owned private banks and state-owned banks. Bonin et al. (2003) have used data from 1996 to 2000 for eleven transition countries to investigate the effect of foreign ownership on the banking sectors in general and bank efficiency in particular. Using stochastic frontier estimation procedures, they compute profit and cost efficiency scores. Their results indicate that banking sectors in these countries became more efficient and more competitive toward the end of the 1990s. They also find that government banks are less efficient than their private counterparts and that majority foreign ownership generates higher efficiency scores. Moreover, they report that banks with participation by international institutional investors, about 10 per cent of their observations, earn higher returns on assets and are more efficient by the profit measure; however, these banks are not significantly more cost efficient. A recent paper by Green et al. (2004) focuses on economies of economies of scale and scope in transition countries, distinguishing between domestic and foreign banks. Their conclusion is somewhat out of line with those of most previous studies on foreign bank entry. Using a panel of 273 foreign and domestic banks located in nine European transition economies (Bulgaria, Croatia, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland and Romania) during 1995 – 1999, they reject the hypothesis that foreign banks are more efficient than domestic banks in these economies. They also find that foreign ownership is hardly an important factor in reducing the banks total costs. Havrylchyk (2006) has investigated the efficiency of the Polish banking industry between 1998-2000, using data envelopment analysis (DEA), which allows distinguishing between cost, allocative, technical, and scale efficiency.

� �

Introduction

This author concludes that bank efficiency has not improved during the analysed years. Foreign banks were found to be more efficient than their domestically owned counterparts. Loan portfolio quality, higher productivity of labour, and market power have the largest impact on efficiency.

Externalities of foreign bank presenceA final strand in the literature highlights that the entrance of foreign bank ownership may also have various positive externalities on the development of the domestic financial system. First, foreign banks will help improve the quality, pricing, and availability of financial services, both directly as providers of these services and indirectly through increased competition. The ECB (2002) notes that foreign investors bring capital and know-how and foster the implementation of best practices of the domestic banking system. In addition, foreign banks are often seen as improving the allocation of credit since they have more sophisticated systems for evaluating and pricing of risks. They are also more experienced in the use of derivative products (Mathieson and Roldos, 2001). Also the likely improvement of human capital due to foreign bank presence will be beneficial, because the skills required for the banking business were scarce during the first years of transition. Second, entry of foreign banks may reduce the market power of domestic banks, which was important as at the beginning of transition the creation of a two-tier banking system produced an oligopolistic market structure (Papi and Revoltella, 2000). It is widely believed that allowing foreign bank entry as part of a liberalization process will enhance the efficiency of the banking system (Litan, Masson and Pomerleano, 2001). Third, foreign bank presence may also lead to improvements in bank regulation and supervision, since these banks may demand improved systems of regulation and supervision from the regulatory authorities in the recipient countries. This may contribute to improving the quality of banking operations of domestic banks (Hermes and Lensink, 2004). All these spillover effects may contribute to more efficient domestic banking practices, which, in turn, may enhance economic growth in transition countries (Koivu, 2002).

Coverage of the book and methodological approach This book is divided in two parts. Part I of this book focuses on Central and Eastern Europe and examines the presence, financial intermediation and entry strategies of foreign banks1. Part II of this book focuses on tests for the relevance of foreign ownership on bank performance. There are several reasons why studies on foreign bank entry, intermediation in Central and Eastern Europe and foreign bank performance in general are justified. First, (central) banks and policy makers

1 Throughout this book, the Central and Eastern European countries include Bulgaria, Croatia, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, the Slovak Repu-blic and Slovenia.

� �

in developing countries may benefit from studies into the nature and performance of foreign banks in transition economies. In the meantime, an EBRD/World Bank Business Environment and Enterprise performance survey found that lack of access to finance is frequently cited by enterprises throughout the transition region as one of the main obstacles to starting, operating or expanding a business (EBRD, 2005)2. Given the dominant presence of foreign banks, the question is whether the very fact that these banks are foreign owned correlates with the survey findings.

Also, contradictory evidence on foreign bank performance, measured by financial indicators as profitability and efficiency, gives rise to the need for further examination of the circumstances under which foreign banks are able to show desirable performance3. Otherwise, policy makers have to stick with the conclusion of Berger et al. (2005) that “a (…) policy-related observation can be made about the appeal to developing countries of encouraging entry by large multinational banks. Having foreign banking giants set up shop in a developing economy no doubt has a number of significant benefits. For example, they are probably more likely to be stable and financially sound. They might also be less likely to engage in the sort of corrupt related-lending practices documented by LaPorta et al. (2003). Without denying the importance of these factors, our analysis points to a potential trade-off. If large foreign banks substantially crowd out smaller domestic ones, the supply of loans to informationally opaque small businesses could be negatively affected.”

Next, many of the countries in the CEE region have entered the European Union or will, within the coming years. EU membership facilitates banks to establish branches in any EU country. This could support spreading the activities of EU licensed banks or changing the format of subsidiaries in branches4. As branch structures are regulated and supervised by the foreign authorities in the country where the foreign owner is located, CEE banking supervisors have signalled concern over the potential loss of supervision over their foreign owned local banks, despite the existence of Memoranda of Understanding5. Studies on foreign banking could further support the insight into

2 According to the survey, enterprises in the transition countries finance 75 per cent of new invest-ments with internal funds, and loans from family and friends, including remittances, compared with 65 per cent in mature market economies.

3 A 2002 European Central Bank (ECB) report notes that in many of the foreign-owned banks, trading and other key activities were shifted to the headquarters, so that the subsidiaries in CEE lost some of their important functions.

4 Illustrative of this trend are recent branch openings of Dutch Fortis Bank and Bank of Tok-yo Mitshubisi UFJ in the Czech Republic and the transformation of Crédit Lyonnais former Czech subsidiary Calyon into a branch.

5 In the 2001 Report on Financial Crisis Management the Economic and Financial Committee of the ECB concluded that ‘Regarding the institutional framework on crisis management, the EFC is of the opinion that closer co-operation among the competent authorities is required in the context of increasing financial integration (EFC, 2001). The banking supervisory authorities and the central banks of the new Member States of the European Union (EU) have agreed to adhere to the Memorandum of Understanding on high-level principles of co-operation in crisis management situations, in force since 1 March 2003. The MoU, which is not a public document, consists of a set of principles and procedures for cross-border co-operation between

� �

Introduction

the dynamics and characteristics. Fourth, the role of foreign capital in CIS banking systems is expected to increase.

In 2001 the limit in Russia of foreign ownership in the banking industry was raised from 12% to 25%. In Kazakhstan and Ukraine, the role of foreign banks has been gradually growing since 1999. About half of Kazakhstan’s banks have some foreign capital in them and total assets of foreign banks in Ukraine currently make about 16% of total Ukrainian banking assets (Golodniuk, 2005). The experience of the transition countries in CEE may be of some help in formulating proper policies.

A special feature of this book is that it makes use of different methodological approaches: Chapter 1 consists of an analysis based on a survey among central banks in the region. The analysis is based on macro data, and data on individual banks from BankScope. Chapter 2 performs a comparative analysis based on micro data. In chapter 3 conclusions are based on structured interviews with bankers and supervisors and in Part II empirical results are based on various econometric techniques like the generalized method of moments (GMM) and ordinary least squares (OLS) (Chapter 4) and stochastic frontier analysis (SFA) (Chapter 5).

Chapter 1 attempts to answer the question: Does foreign bank entry and financial intermediation show similarly trends in Central and Eastern Europe? The chapter examines the growth in the number of foreign banks and their share in total banking assets in CEE. The chapter then looks at growth in financial intermediation in the region. The chapter examines the trend in foreign bank growth versus growth in financial intermediation. Finally, the chapter specifically focuses on the differences between credit extension by foreign owned banks and locally owned banks. The chapter is based on both macro-data and micro data. Due to scarce data for the early transition period, the University of Groningen and the Dutch central bank (DNB) initiated a survey among central banks in Central and Eastern Europe. To present a comprehensive analysis of foreign bank entry the data was supplemented with more recent data from the European Bank for Reconstruction and Development (EBRD). The origin of bank ownership in the region was tracked with Bureau van Dijks’ BankScope database.

Chapter 2 uses bank balance sheets to answer the question: have the nature and size of foreign bank intermediation in Central and Eastern Europe changed over time? The chapter analyzes the size and composition of (foreign) banks’ balance sheets in CEE. The results are compared to two benchmark regions. The first consists of the old EU15 countries as the CEE countries have recently joined the EU (or will soon). The second consists of a set of transition countries in South Eastern Europe

banking supervisors and central banks in crisis situations. These principles and procedures deal specifically with the identification of the authorities responsible for crisis management, the re-quired flows of information between all the involved authorities and the practical conditions for sharing information at the cross-border level. The MoU also provides for the setting-up of a logistical infrastructure to support the enhanced cross-border co-operation between authorities. (ECB, 2004).

� �

(SEE) and the Commonwealth of Independent States (CIS), as these countries have similar features as the CEE countries but do not have the prospect of joining the EU. The chapter first compares trends in foreign bank size in the EU15 and in transition countries Assuming that foreign banks over time learn to deal with information asymmetry and/ or other comparable disadvantages vis-à-vis domestic banks, the chapter hypothesizes increasing customer loans over time and presents the dynamics in the composition of the asset side of foreign banks. Third, it shows how foreign banks have organized the liability side. We show how the maturity of customer loans compares to that in banks in the EU15 and other transition countries. The chapter is based on data from Bureau van Dijks’ BankScope database.

Chapter 3 attempts to reveal potential structural features in foreign bank entry strategies and specifically attempts to answer the question: did banks follow their customers in Central and Eastern Europe? First, it analyzes why banks have entered the region. The chapter then focuses on the choice between greenfield and acquisition and reveals the main determinants influencing this choice. Specifically, the chapter intends to reveal the interdependence of the reasons for foreign banks to start operations in another country, the mode of entry i.e. by acquiring an existing bank or by establishing a new bank, and, in case of establishing a new bank, the choice for i.e. a branch or a subsidiary. For this study a joint interview project was set up in collaboration with the Dutch central bank. For this project, representatives of the most important foreign banks in the region were interviewed. To control for the effects of customer focus, the sample was supplemented with interviews with representatives of global banks present in Central and Eastern Europe. To validate the data additional interviews were conducted with representatives of the local foreign owned banks and local financial supervisory authorities.

Part II shifts the focus from foreign bank entry and intermediation to the financial performance of foreign banks and aims to extend the scarce literature on the relationship between foreign ownership and bank performance. Chapter 4 examines foreign bank performance in Central and Eastern Europe. Performance indicators are profit, overhead costs and net interest revenues. A special feature of the research in Part II is that we use a handmade continuous foreign ownership variable. The implication is that we can conclude on the effect a rise in foreign ownership has on performance. This approach slightly differs from using a dummy variable. The dummy variable approach explicitly identifies banks with banks with foreign majority control, i.e. foreign banks, versus banks with minority foreign majority control, i.e. domestic banks. In chapter 4, ownership of a sample of 211 banks in 22 countries in 2001 is analysed in order to construct a continuous foreign ownership dummy. In line with most other studies in this strand we start with ordinary least squares (OLS) regressions to investigate how foreign ownership is related to bank profitability, costs and interest revenues. Several robustness checks are performed controlling for (i) economic development of the transition country, (ii) concentration of the banking sector, (iii) the presence of the relative number of foreign banks, and (iv) the presence

� �

Introduction

of relative foreign banks’ assets. Chapter 5 extends the analysis of chapter 4 in several ways. A first extension

consists of the use of data of banks worldwide. Second, instead of a cross section we use a panel including data from the period 1998-2001. Similar to chapter 4, for each bank is the exact level of foreign ownership is determined in order to construct a continuous foreign ownership variable. A third extension is that in estimating the impact of foreign ownership on bank performance, chapter 5 uses the generalized method of moments (GMM) technique in addition to the OLS approach. In the model specifications we again control for the influence of economic development.

In Chapter 6 we shift from foreign bank profitability research to foreign bank efficiency research. Literature on the relationship between foreign ownership and foreign banks is scarce and most foreign bank efficiency literature focuses on transition economies. A first contribution is that, using a stochastic frontier approach, we estimate cost efficiency correlates of 1,636 banks in 97 countries worldwide. In addition we test for a sub set of banks in 43 lower income countries. The most innovative feature of this study is that we investigate the role of the institutional context. Specifically, we aim to answer the following research questions: i) Does bank efficiency depend on foreign ownership? ii) Does foreign bank efficiency benefit from improvements in the institutional framework in the host country? iii) Does foreign bank efficiency depend on the institutional framework in the home country? and iv) Does foreign bank efficiency depend on the distance between /similarity of the institutional framework in the host country and the institutional framework in the home country?

Chapter 7 summarizes the main conclusions of this study.

��

Part IForeign Bank Behaviour in Central and Eastern Europe

��

��

Chapter 1 Foreign Bank Entry and Credit Performance*

“Nursing small and medium-sized enterprises in Central Europe to create a much needed Mittelstand is still too labour-intensive to be profitable: it needs technical assistance and some subsidy.” The Economist, September 14th, 2002

1.1 IntroductionAs mentioned in the introduction of this book, an important strand of literature in the field of foreign banking examines the impact of foreign banking on the financial system and the economic development of emerging and transition countries (Claessens, Demirgüç-Kunt and Huizinga, 2001; Iakova and Wagner 2001; Mathieson and Roldos 2001). The aim of this chapter is to analyse the development of the banking sector during the transition process in the Central and Eastern European countries. The chapter presents the main developments and intra-regional differences in (i) financial foreign direct investment in the banking sector and its origin, and (ii) growth in credit to the private sector. In examining banks’ credit supply in Central and Eastern Europe, this chapter particularly focuses on the role of foreign-owned banks as it turns out that they play an important role in the development of the financial system of European transition countries. A bank is defined as foreign owned if more than 50% of the shares are owned by non-domestic investor(s).

A problem in analysing the banking sector transition process is the lack of reliable data for the early years of the transition period. For the analysis in this chapter, new data was gathered about several aspects of the development, structure, conduct and performance of the banking sector in the region. Co-operation of central banks in the region was sought in a joint-project with the Dutch central bank. For a description of the survey, see appendix 1. Several central banks helped to gather aggregated data required to analyse the development of the banking sector and the role of foreign banks.1 As such, this chapter is one of the first to analyse the complete region on the basis

* Based on De Haan, J. and I. Naaborg (2004). Foreign intermediation in Accession countries: The role of foreign banks, in Masciandaro, D. (Ed.) Financial Intermedation in the New Europe: Banks, Markets and Regulation in EU Accession Countries, 181-207, © (2004), with permission of Edward El-gar Publishing Limited. Also based on Naaborg et al. (2004). How important are Foreign Banks in the Financial Development of European Transition Countries? Journal of Emerging Market Finance, Vol. 3, No.2. Copyrights © ICICI Bank Limited, Mumbai, Delhi. All rights reserved. Reproduced with the permission of the copyright holders and the publishers, Sage Publications India Pvt. Ltd, New Delhi.

1 Appendix 1 presents the data requested in the central bank survey.

��

Chapter 1

��

of an identical methodology for all countries.2 This information was supplemented with data from the IMFs International Financial Statistics and data from the EBRD. However, for an understanding of the origin of financial foreign direct investment in the banking sector in the transition economies, adequate knowledge of individual banks is essential. Data on individual banks were therefore gathered using the BankScope database.

The structure of the chapter is as follows. Section 1.2 discusses some theoretical concepts regarding the effect of foreign ownership on foreign owned banks’ credit supply. Section 1.3 reveals the main trends in foreign bank entry in the CEE banking sector. The section shows the trends in both the number and assets of foreign banks, and presents the diversity within the region. In section 1.4 the origin of financial FDI in Central and Eastern Europe is analysed. Section 1.5 aims to answer the question: what is the role of foreign banks as intermediates? First, the section compares financial development in the CEE region with that of the EU15 countries. Next, it examines the role of foreign banks as suppliers of credit to the private sector. Finally, section 1.6 presents some simple performance ratios of foreign versus domestic banks, which were also gathered in the central bank survey.3 Section 1.7 concludes.

1.2 TheoryDo foreign banks in transition economies lend differently than domestic banks and if so, why? Previous studies on emerging markets in Latin America and poor countries have found support for the idea that foreign banks (i) cause less access to credit, (ii) have problems in supplying funds to small business, and (iii) focus their activities mainly on large enterprises. A straightforward, clear-cut hypothesis on the effect of foreign bank entry on financial intermediation in transition economies is troublesome due to conflicting theories. The question boils down to which of the inherent attributes of foreign banks prevails.

According to the global advantage theory (Berger et al., 2001), foreign owned banks have comparative advantages relative to domestic owned banks leading to a better performance. One of the main arguments is that foreign owned banks use more advanced technologies related to risk pricing, screening and monitoring. Expressed in the well-known terms of the finance literature, the theory assumes that foreign banks have state of the art practices to deal with adverse selection and moral hazard. With respect to adverse selection, banks, both foreign and domestic, are concerned by the prospect of lending to a firm with poor prospects, the so-called ‘bad firms’. The theory assumes the foreign owned bank to export best practices regarding screening at low costs to the local owned bank in the host country. Moral

2 Up to now, most studies focused on a limited number of countries within the region. For example, Dobosiewicz (1995), Mervart (1995), Sabi (1996), Buch (1997), Steinherr (1997), Bonin, Mizsei, Székely and Wachtel (1998), Scholtens (2000), Galac and Kraft (2000), Bonin and Abel (2001), Barisitz (2001), Schardax and Reiniger (2001), ECB (2002a), and Hasan and Marton (2003).

3 Chapters 4-6 will assess the effect of foreign ownership on bank performance empirically.

��

Foreign Bank Entry and Credit Performance

hazard occurs when the borrower does not act in the interest of the lender, e.g. due to managerial misbehaviour. Managerial misbehaviour comes with (i) insufficient effort of the borrower, such as insufficient oversight of subordinates or overcommitence with other activities, (ii) extravagant investments, e.g. empire building, (iii) entrenchment strategies, by investments in activities that make the borrower indispensable in its firm, and (iv) self dealing in which the borrower increases her private benefits by consuming perks. In the global advantage theory, foreign banks are assumed to be able to export best-practice monitoring to the subsidiary in the host country and lower the chance of loan default due to moral hazard. Thus, ceteris paribus, we would expect foreign bank entry in transition economies to increase credit availability due to a superior ability to allocate and monitor credit.

Although the EBRD (2006) argues that the entry of foreign financial institutions is generally associated with improved access to credit for small and medium-sized enterprises, some features of the credit market in transition economies gives rise to caution. For example, banks in transition economies in Central and Eastern Europe are troubled in their screening by the lack of historical company data, such as income statements and balance sheets. Especially in the 1990s, track records of company performance were virtually non-available. If banks lack hard data, the principal source of information that remains is soft information. If soft information is the principal source of information, foreign banks usually refrain from extending credit. In this respect, Mian (2006) mentions “[…] if a foreign bank was to take much risk in a developing country and causes a fear of bank failure, it would reap large negative consequences, through reputation, on its operations worldwide.” Additionally, due to cultural differences domestic banks might be better at collecting and interpreting soft information. As foreign banks have less access to soft information and have more troubles interpreting this information, they are at a disadvantage when it comes to screening. For example, Mian (2006) finds that in Pakistan, foreign banks systematically shy away form lending to ‘soft-information’ firms such as small firms, firms in smaller cities, firms not affiliated with a major business group, firms seeking first-time loans and firms seeking long-term relational financing.

Even with the assumption that firms with good prospects, would have a preference for engaging in a long term credit relationship with a foreign owned bank, the adverse selection problem for the foreign bank can not be solved4. Assuming that the foreign bank has no track record of the ‘good firm’, the ‘good firm’ would have to signal that it’s a ‘good firm’ by showing a rating, obtained with a rating-agency, or by using collateral. However, especially in the early transition period, foreign banks faced several problems assessing the value and seizing of collateral. First, banks lacked a sophisticated collateral platform and standardized catalogues. Second, collateral prices were subject to uncertainty related to inflation and the predictability of the collateral value was troubled by the absence of a liquid market for collateral. Creditors’ rights in

4 Signalling is assumed to be costly. In addition, it is assumed that ‘bad firms’ cannot mimic ‘good firms’.

��

Chapter 1

��

seizing the collateral were subject to uncertainty as well, as legislation in this area was only slowly developing in the Central and Eastern European transition economies. Going to court in a transitory economy is a costly and time-consuming business, troubled by corruption. Finally, a second way a firm can signal it is a ‘good firm’ is with a rating. However, rating agencies were not available in the early transition period and would face the same information asymmetry regarding past performance as (foreign) banks.

Next to the adverse selection problem, foreign banks might also have more troubles than domestic banks in solving the moral hazard problem in transition economies, even if the former are endowed with best-practice monitoring skills. An active monitor collects information that some policy proposed or followed by management is value decreasing and intervenes to prevent or correct this policy. The mechanism by which the change is implemented depends on the identity of the active monitor (Tirole, 2006). Due to cultural similarities, domestic banks might have better skills to collect soft information of borrowing companies and to prevent or correct policies of these.

So far, this section has focused on the consequences of frictions in the relationship between loan officer of the foreign bank and the borrower, both residing in the host country. Mian (2006) focuses on the causes of the distance between the loan officer in the host country and the controlling shareholder at the parent bank in the home country. This type of distance comes in different metrics: geographical, cultural, institutional or hierarchical, but all lead to a restricted credit extension of the foreign banks. The author argues that distance implies that working in an environment with a different corporate culture, legal environment, or regulatory framework might increase the information asymmetry and makes it more difficult for the home country CEO of a foreign bank to design policies that are specifically tailored for the developing country.

Fries and Taci (2002) have estimated a reduced form model that relates the real growth of bank loans to observable demand and supply factors. Their data covers 515 banks from transition countries over the years 1994-99. They differentiate between high and low reform countries and years. The accession countries generally belong to the high reform countries and years. Table 1.1 is reproduced from this study. The results of Fries and Taci (2002) show that loan growth is significantly associated with lagged real GDP growth. However, the estimated coefficients are in the range of 0.9-1.0, which implies that lending is expanding at about the same rate as output. The estimates also show that budget deficits crowd out real growth in customer lending. From the point of view of this chapter it is interesting that foreign ownership does not influence loan growth. Foreign banks on average expand at the same rate as privatised or ab initio banks. Still, the share of foreign banks in the total assets of the banking sector has a positive effect on the expansion of customer loans. According to Fries and Taci (2002), this may reflect the positive effect of the presence of foreign banks in the loan market by increasing the competitive pressure in the sector as well as spillover effects

��


from the introduction of better banking technologies, skills and lending products.5

Table 1.1 Determinants of real growth in bank loans to customers

Model 1 Model 2

Lagged real GDP growth 0.87 (2.57)*** 1.02 (3.10)***

Lagged real loan growth 0.23 (1.93)** 0.27 (2.06)**

General government balance (per cent GDP) 0.69 (2.06)** 0.84 (2.50)**

Inflation (per cent) -0.21 (-2.81)*** -0.21 (-2.98)***

Asset share of foreign banks (per cent) 0.12 (1.66)* 0.12 (1.66)*

Total real assets 0.09 (0.61)

Deposits market share (per cent) -0.05 (-0.95)

State-owned banks (dummy) -1.56 (-0.97) -2.26 (-1.51)

Ab initio private banks (dummy) -2.11 (-1.03) -2.77 (-1.34)

Foreign-owned banks (dummy) 1.00 (0.54) 0.53 (0.27)

Lagged Equity/Total assets (per cent) 0.32 (4.31)*** 0.23 (2.90)***

Constant -1.75 (-0.50) -0.23 (-0.07)

Nr. Obs. 672 661

R2 0.23 0.22

*, **, *** indicate statistical significance confidence levels at 10, 5 and 1 percent respectively. Source: Fries and Taci (2002)

1.3 The importance of foreign banks in CEEIn order to examine the importance of foreign banks in the financial system in Central and Eastern Europe, this section uses two measures of foreign bank presence. One way to measure the importance of foreign banks in the banking sector is simply to calculate the share of the number of banks that are foreign-owned. Alternatively, the proportion of total bank assets held by foreign banks is often employed. The relative number of foreign banks is an appropriate measure if the number of domestic and foreign banks determines competitive conditions. This might be the case if domestic banks adjust their activities as soon as foreign entry occurs to prevent the new entrants from capturing a significant market share. The share presence measure can be appropriate if foreign banks have an impact on domestic banks only after gaining market share (Claessens, Demirgüç-Kunt, and Huizinga, 2001).

5 The estimation results for the sub-sample of low-reform countries and years stand in sharp con-trast to those reported in table 1.1, as none of the estimated coefficients is significantly different from zero, with the exception of the inflation variable.

��

Chapter 1

��

1.3.1ForeignownedbanksFigure 1.1 reflects the number of foreign owned banks and domestic owned banks in the whole region for the period 1995-2004; the data were mainly gathered via a survey among central banks.6 Appendix 2 lists the total number of foreign banks for each CEE country. Figure 1.1 combines two striking developments in the banking system in the transition economies. First, between 1995 and 2004, 28 per cent of total banks in 1995, i.e. 125 banks, disappeared. Second, up to the year 2000 the number of foreign banks increased and domestic owned banks became a minority within the banking sector. Foreign bank presence in all transition countries is considerably higher than in the European Union countries, with the exception of Luxembourg (Claessens et al., 2001; Noyer, 2001). However, the importance of foreign banks varies a lot within the CEE region.

Figure 1.1 Number of foreign and domestic banks in CEE, 1995-2004

050

100150200250300350400450500

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Foreign banks Domestic banks

Data include banks from Bulgaria, Croatia, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania and the Slovak Republic. Source: Central bank survey and EBRD.

Table 1.2 shows the relative number of foreign banks per country and Appendix 3 lists the absolute number of foreign banks per country. In 1995, 114 foreign banks were present in the countries in our sample, accounting for 25 per cent of total banks. In that year, the Czech Republic, Hungary and the Slovak Republic already showed relatively high levels of foreign bank presence. In the second part of the 1990s, the relative number of foreign banks grew strongly especially in Bulgaria, Croatia, Hungary, Lithuania, Poland, and Romania. In 2000, the number of foreign banks

6 The main reason for the survey was to gather financial sector information from the early transi-tion period. Survey data on foreign bank entry and the financial performance of foreign banks allows for an analysis from 1995 onwards (sections 1.3 and 1.6). Survey data for a smaller set of countries allow for an analysis of the role of foreign banks as intermediators from 1993 onwards (section 1.5).

��


reached a peak of 214 and that number has slightly decreased since, especially due to foreign banks leaving Croatia and Hungary.

Table 1.2 Number of foreign banks (% total number of banks), 1995-2004

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Bulgaria 7 7 25 50 65 68 74 76 71 69

Croatia 2 9 11 17 25 49 56 50 45 41

Czech Rep. 42 43 48 56 64 65 68 70 74 74

Estonia 26 27 31 50 43 57 57 57 57 67

Hungary 49 57 67 64 67 79 76 71 76 71

Latvia 26 40 47 56 52 57 43 39 43 39

Lithuania 0 25 33 42 36 46 46 50 54 50

Poland 22 31 35 37 51 63 67 76 79 77

Romania Slovak Rep.

3355

3248

3945

4441

5640

6457

7357

7775

7076

7276

Slovenia 15 11 12 10 16 21 21 27 27 32

Average 25 30 36 42 47 57 58 61 61 61

Relative number of foreign banks is measured as the ratio of the number of foreign banks and total number of banks. Source: Central bank survey and EBRD (Appendices 2 and 3).

1.3.2Foreignownedbanks’assetsA second way of measuring foreign ownership in the banking sector is to calculate the ratio of the sum of foreign banks’ assets and the sum of total banking assets, including those of domestic owned banks. Figure 1.2 reveals the relative asset shares of foreign owned banks state-owned banks, and domestic owned banks in the period 1995-2004. The share of state-owned banks evaporated from 51 per cent in 1995 to 3 percent in 2004. After several banking crises hit most transition countries in the mid-1990s (see Caprio and Klingebiel (2003) for an overview of the different crises), bank privatisation furthered foreign participation. Appendix 4 shows the share of state-owned banks in each CEE country during 1995-2004. Only in Poland and Slovenia governments are still important shareholders of banks. Similarly to state-owned banks, domestic banks lost importance with a lowest level of relative assets of 9 per cent in 2000.

Table 1.3 shows that foreign banks’ assets reached a peak of 84 per cent in 2002 and remained relatively stable at that level. The table also shows the difference between countries regarding the timing of foreign bank entry. Hungary and Latvia were among the first countries were foreign banks’ assets dominated domestic bank assets. In 1996/97, foreign banks’ assets in Hungary rose from 19 per cent up to 62 per cent of total banking asset as Dutch ABN AMRO bought 80 per cent in privatisation of Magyar Hitel Bank in 1996, while German LB increased its stake to 50 per cent in Hungarian foreign Trade Bank MKB. In 1997, Austrian ERSTE

�0

Chapter 1

��

Bank acquired 64 per cent in privatisation of Mezobank, while Belgian KBC and Irish Life bought 56 per cent in the privatisation of K&H. In the same period, foreign banking assets in Latvia rise to 72 per cent as the EBRD and other foreign investors buy Unibanka in 1996 and Estonian Hansapank buys Deutsch Lettische Bank, establishing Hansabankas. A year later, Estonian Eesti Ühispank became owner of Latvijas Zemes Banka.

In 1998, the share of foreign banks’ assets in Estonia rose to 90 per cent reflecting that its largest banks, Hansapank and Eesti Ühispank, were sold to two Swedish banks: Swedbank and Skandinaviska Enskilda Banken. In 1999, foreign banks’ assets in Poland increased from 17 per cent to 49 percent as Allied Irish Banks bought 80 per cent in Bank Zachodni, Italian Unicredito acquired 52 per cent in Pekao in a second stage privatisation and Bank Austria raised its stake with 20 per cent in PKB up to 44 per cent in 1999.

Figure 1.2 Average foreign ownership, 1995-2004

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Foreign banks' assets State-owned banks' assets Domestic banks' assets

Foreign bank (state-owned banks, domestic banks) presence by assets is measured as the ratio of total assets of foreign banks (state-owned banks, domestic banks) and total banking assets. Data presented are means from Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithua-nia, Poland, Romania, the Slovak Republic, Slovenia and Bulgaria. Relative shares are means. Source: central bank survey and EBRD.

In the years 2000-2001 sharp rises in foreign banks’ assets took place region wide, except in Hungary, Latvia, Romania and Slovenia7. Due to mergers and acquisitions, relative foreign banks’ assets spur from an average of 40 per cent up to a 70 per cent level. In Bulgaria for example, foreign banks’ assets rose with 30 per cent-points up to

7 Foreign bank entry in Hungary and Latvia had taken place in 1996/97. In Romania foreign owned bank assets are growing by a steady few percentages a year, rather than in shocks. In Slovenia foreign banks are of minor importance in the banking sector.

�0 ��


72 per cent of total banking assets, as Bulgarian largest bank Bulbank was privatised and sold to Unicredito. In the meanwhile, foreign banks’ assets in Croatia also rise. The increase amounts to 45 per cent points up to 89 per cent of total banking assets, as the third and fourth largest banks Splitska Banka and Rijecka Banka were acquired in privatisation by Unicredito8 and German Bayerische Landesbank9.

Table 1.3 Foreign banks’ assets (% total bank assets), 1995-2004

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Bulgaria 1 2 18 25 42 72 71 72 82 82

Croatia 0 1 4 8 39 84 89 90 91 91

Czech Rep. 17 20 24 27 40 66 89 86 86 85

Estonia 2 2 90 90 97 98 98 98 98

Hungary 19 46 62 63 62 67 67 85 84 63

Latvia 53 72 81 74 74 65 43 53 49

Lithuania 0 28 41 52 37 55 78 96 96 91

Poland 4 14 15 17 49 73 72 71 72 71

Romania 36 44 47 51 53 55 59

Slovak Rep. 19 23 30 33 24 43 78 84 96 97

Slovenia 5 5 5 5 5 15 15 17 19 20

Median 4 17 21 33 42 67 72 84 82 82

Share of foreign banks’ assets is measured as the ratio of assets of foreign banks and total banking assets. Source: Central bank survey and EBRD.

In addition, Croatia’s second largest bank, Privredna Banka Zagreb, was sold in privatisation to Banca Commerciala Italiana. In the Czech Republic foreign banks’ assets grew to 89 per cent of total banking assets, as Austrian Erste Bank acquired 52 per cent in privatisation of savings bank Česká Spořitelna, Belgian KBC’s Czech subsidiary CSOB took over assets and liabilities of IPB, and German Bankgesellschaft Berlin raised its stake from 47 per cent to 85 per cent in Zivnostenska Banka. A year later, French Société Générale bought Komercní Banka in privatisation and Erste Bank acquired 71 per cent of the preferential shares in Česká Spořitelna. In Lithuania, 78 per cent of banking assets were foreign owned in 2001 as Swedish Skandinaviska

8 The subsequent changes in ownership of Splitska Banka are exemplary for the consolidation that took place in CEE. In 2002, Unicredito sold the bank to Austrian Bank Austria Creditan-stalt (BACA) due to anti-trust reasons. Following German HVBs acquisition of BACA, the bank merged with HVB Bank Croatia in 2003. Following the Italian Unicredito acquisition of HVB in 2005, French Société Générale bought Splitska Banka in 2006, again as a result of anti-trust reasons.

9 Due to large losses on the foreign exchange market and subsequent solvency problems, the Cro-atian State Agency for Banks bought Rijecka Banka for €1, recapitalized, and sold the bank to Austrian Erste and Steiermarkische Bank.

��

Chapter 1

��

Enskilda Banken raised its stake in Lithuanian Vilniaus Bankas to 100 per cent after the latter had merged with Bankas Hermis in 2000, while Finnish Sampo bought majority stake in privatisation of Lithuanian Development bank. A year later, Estonian Hansapank, owned by Swedish Swedbank, bought 90 per cent in savings bank LTB, Lithuanians second largest bank and merged it with Hansabankas. Finally, In Poland relative foreign banks’ assets increased up to 72 per cent in 2001 as Citibank bought 88 per cent in Bank Handlowy w Warzawie. In the Slovak Republic, Hungarian OTP Bank bought a majority stake in Investicna a Rozvojá Bank establishing OTP Bank Slovensko and Erste Bank acquired 87 per cent of savings bank Slovensko Sporitelna in privatisation. In 2001, foreign banks’ assets in the Slovak Republic have increased up to 78 per cent of total banking assets.

1.4 The origin of foreign direct investment in the banking sectorIn 2002, The Economist noted that ‘Central Europe tends to be the stamping ground of second-tier banks. Where are the Citibanks, J.P. Morgan’s and Deutsche’s? And where are the British, especially HSBC, that emerging-market specialist? […] Perhaps they are planning to wait for a few years and then buy one of the Austrians or Italians along with their Central European network. (The Economist, Sept 14th, 2002)

So far, our analysis has been based on aggregated data mainly provided by national central banks. To shed some light on the nature of foreign ownership of CEE banks, this section aims to reveal the origin of financial FDI. The analysis is based on a sample of 244 banks in CEE and SEE/CIS countries, available from the November 2002 BankScope database issue provided by Bureau van Dijk. Banks included in BankScope roughly account for 90 per cent of the assets of all banks. For a bank to be included in the sample, it was required that ownership details of at least 60% of the shares were available.

Using a micro-approach, the origin of ownership of each bank was unravelled. First, the origin of each owner in each CEE bank was established. Next, each owner was assigned the weighed amount of assets according to her level of ownership. Third, to determine total financial FDI for each origin, ownership of residents with the same origin are summed as follows, where banki,j denotes bank i in CEE country j:

Total financial FDI per origin z = S (shares i,j owned by residentz * assets banki,j)

10 (1.1)

For some banks, ownership was not completely available in BankScope. In the process, 78% of the bank assets in CEE could be traced to their owners. Table 1.4 presents the origin of financial FDI in CEE banks. For the aggregate, total foreign

10 For example, if a Belgian bank owns 50% of the shares of a €6bn bank in Hungary, a Dutch bank owns 25% and the Hungarian state ownes the remaining 25%, financial FDI with origin ‘Belgium’ is assigned €3bn and financial FDI with origin ‘The Netherlands’ is assigned €1.5bn.

��


ownership amounts to 63.3 per cent. Non-CEE foreign investors own 62.0% and only 1.3% is owned by foreign investors coming from CEE countries. Investors from Austria, Belgium, Germany, Italy, the US, France and the Netherlands together own more than 51% of CEE banking assets. Austria’s number one position as foreign investor of banking assets is mainly due to the presence of three large Austrian banks: HVB-GROUP/ Bank Austria Creditanstalt11, Erste Bank12 and Raiffeisen. Proximity and/or the cultural ties of Austria with the region originating from the Austrian-Habsburger Empire probably explain the Austrian dominance. Chapter 3 examines the reasons for foreign entry in detail. Belgium’s second position is due to the large amounts of banking assets of Belgian KBC in the region through its stakes in the Czech CSOB, the Polish Kredytbank, the Hungarian K&H Bank and the Slovenian Nova Ljubljanska Banka. The interregional financial FDI comes mainly from Estonia and Hungary. Estonian Hansapank owns Latvia’s second largest bank, Hansabankas and the largest bank in Lithuania, Bankas Hansabankas. Hungary’s largest bank, OTP bought Slovakian IRB, now OTP Banka Slovensko.13

Table 1.4 Origin of financial FDI in CEE banks, 2002

Non-CEE/CIS origin Assets (%) CEE/CIS origin Assets (%)

Austria 14.1 Estonia 0.3

Belgium 8.6 Hungary 0.3

Germany 6.5 Czech Republic 0.2

Italy 6.4 Poland 0.2

USA 6.3 Bulgaria 0.1

France 4.6 Romania 0.1

The Netherlands 4.1 Russia 0.1

Luxembourg 3.2 Slovak Republic 0.1

Sweden 2.2

EBRD and IFC 2.1

Other 3.9

Total 62.0 Total 1.3 %

The table presents the origin of foreign owners coming from non-CEE countries and the origin of foreign owners coming from CEE countries. Ownership of assets (Assets) by origin is scaled to total banking assets in the sample. The sample includes banks from Albania, Ar-menia, Azerbaijan, Belarus, Bosnia-Herzegovina, Bulgaria, Croatia, Czech Republic, Estonia,

11 In the summer of 2005, Italian UniCredito took over HVB Bank.12 Erste Bank is the number one largest acquirer in the region according to deal value during

1990-2004 (ECB, 2005). 13 Recently, OPT increased its foreign by acquiring DSK in Bulgaria, Robank in Romania, Nova

Banka in Croatia and Niska Banka in Serbia and Montenegro.

��

Chapter 1

��

Hungary, Kazakhstan, Latvia, Lithuania, Macedonia, Poland, Romania, Russia, Slovak Repu-blic, Slovenia, Ukraine, Uzbekistan, and Yugoslavia. Source: November 2002 issue of Bureau van Dijk’s BankScope database. For sample details see Appendix 5.

Table 1.5 provides in CEE foreign bank ownership 2005. The sample consists of the three biggest banks in each of the eleven CEE countries. The majority shareholder, if any, is traced together with her origin. The considerable presence of Austrian investors is clearly visible in Croatia, the Czech Republic, Romania and the Slovak Republic. Belgium is represented in big banks in the Czech Republic and Hungary, and Italy invests in the most important banks in neighbour Croatia. In Romania and Poland, the largest bank is still government owned, and in Slovenia foreign investors own none of the largest banks. Table 1.5 suggests that distance between foreign investor and ownership of large banks plays a role in financial FDI decisions. In Bulgaria, Greek and Hungarian investors provide for financial FDI owning two of the three largest banks. In Croatia, neighbour Italy owns the two largest banks. Swedish and Finnish banks own the largest banks in Estonia and a similar picture arises in Latvia and Lithuania. In contrast to Austrian and Italian banks, German and French banks rarely show up as owners of CEE largest banks. Finally, investors from the USA and the Netherlands own none of the biggest banks in Central and Eastern Europe. Chapter 3 will focus in more detail on foreign banks’ entry decisions.

Table 1.5 Origin of financial FDI in CEE’s biggest banks, 2005

Country Bank Foreign owner (>50%) Origin

Bulgaria BULBANK UniCredito Italiano Italy

DSK BANK OTP Bank Hungary

UNITED BULGARIAN BANK National Bank of Greece Greece

Croatia ZAGREBACKA BANKA UniCredito Italiano Italy

PRIVREDNA BANKA Banca Intesa Italy

ERSTE & STEIERMARKISCHE BANK

ErsteAustria

Czech Rep. CSOB KBC Belgium

ČESKÁ SPOŘITELNA Erste Bank Austria

KOMERCNI BANKA Société Générale France

Estonia HANSAPANK Swedbank Sweden

SEB EESTI ÜHISPANK SEB Sweden

SAMPO BANK Sampo PLC Finland

��


Country Bank Foreign owner (>50%) Origin

Hungary OTP BANK *

KERESKEDELNI ES HITELBANK KBC Belgium

MKB BANK Bayerische Landesbank Germany

Latvia PAREKSS BANKA *

HANSABANKA Hansapank Estonia

SEB LATVIJAS UNIBANKA SEB Sweden

Lithuania SEB VILNIAUS BANKAS SEB Sweden

BANKAS HANSABANKAS Hansapank Estonia

BANKAS NORD/LB LIETUVA NORD/LB Denmark

Poland PKO BP Polish government Poland

BANK PEKAO Unicredito Italiano Italy

BANK BPH HVB Group/ BA-CA Germany

RomaniaROMANIAN COMMERCIAL BANK *

BRD Société Générale France

RAIFFEISEN BANK Raifeissen Austria

Slovak Rep. SLOVENSKA SPORITELNA ERSTE Austria

VSEOBECNA UVEROVA BANKA Banca Intesa Italy

TATRA BANKA Raiffeisen Austria

Slovenia NOVA LJUBLJANSKA BANKA *

NOVAK KREDITNA BANKA Government of Slovenia Slovenia

ABANKA VIPA *

The table shows the majority owner of each of the top 3 biggest banks, by assets, in every CEE country. Some banks have no majority owner and are marked with * and their ownership is as follows. OTP BANK is listed. Two private individuals own PAREKKS BANKA. Shareholders of the ROMANIAN COMMERCIAL BANK are the Agency for Privatisation and Manage-ment of State ownership APAPS (37%), 5 regional private investment funds (30%), and the EBRD and the IFC (both 12.5%). The Slovenian state (35%) and Belgian KBC (34%) own NOVA LJUBLJANSKA BANKA. Largest shareholder of ABANKA VIPA is insurance com-pany Triglav (33%). Source: June 2006 edition of Bureau van Dijks’ BankScope.

��

Chapter 1

��

Figure 1.3 Bank assets and credit to the private sector (% GDP), 1993-2000

0%

20%

40%

60%

80%

100%

120%

140%

160%

1993 1994 1995 1996 1997 1998 1999 2000

Credit to the private sector (CEE) Credit to the private sector (€-area)

Deposit money bank assets (CEE) Deposit money bank assets (€-area)

Credit to the private sector and deposit money bank assets are averages and presented as a ratio of GDP. The CEE area covers Croatia, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, the Slovak Republic and Slovenia. The euro-area covers Austria, Belgium, Finland, France, Germany, Greece, Ireland, Italy, Portugal, Spain and the Nether-lands. Source: Central bank survey and IMF’s IFS (for euro-area deposit money bank assets, credit to the private sector and GDP). Luxembourg was excluded in determining euro-area indicators.

Figure 1.4 Average credit to the private sector (% total bank credit), 1993-2000

0%

10%

20%

30%

40%

50%

60%

1993 1994 1995 1996 1997 1998 1999 2000

Average Standard deviation Credit to the private sector is the sum of credit to the private sector/ GDP by foreign banks and credit to the private sector/ GDP by domestic banks. The sum of private sector credit/ GDP is presented as a ratio of domestic banks assets/ GDP. Data include banks from Croa-tia, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland and Slovenia. Source: Central bank survey.

��


1.5 Intermediation and the role of foreign owned banksWith the elimination of the mono-bank system, most transition countries experienced a rapid expansion of the banking sector. Big increases in the amount of non-performing loans led to banking crises in many countries during the 1990s.14 Many of these crises arose out of insolvencies of state-owned or formerly state-owned banks, caused by bad loans inherited from the communist era. Moreover, the transition reduced profitability in certain sectors, thereby reducing the abilities of firms to service their debts (Koivu, 2002). Despite the rapid developments, the banking sectors in most transition countries are still underdeveloped.15 Still, they are the most important channels for domestic financing as capital markets in these countries are small, volatile and illiquid. As pointed out by Koivu (2002), stock markets in some transition countries even face a danger of becoming irrelevant, as many companies have been delisted in recent years. Thus, enterprises often have to finance capital investment out of their own cash flows. Berglöf and Bolton (2002) argue that banks in transition countries have not yet begun extending significant long-term finance16.

Figures 1.3 and 1.4 reveal the poor development in the banking industry in CEE countries in the period between 1993 and 2000. The lines in Figure 1.3 show the development of the average assets of deposit money banks as a ratio to GDP in CEE countries and the euro area. The ratio between banks assets and GDP illustrates the importance of the financial services provided by banks relative to the size of the economy (Beck et al., 2000). In general, the average of deposit money bank assets to GDP in the euro area is at least twice as high as in the transition countries17. The columns in Figure 1.3 present the development of the level of bank claims on the private sector as share of GDP in CEE countries and the euro area. This ratio shows the importance of one of the main functions of financial intermediaries, i.e. channelling funds to investors. Private credit captures the loans to the private non-financial sector (Beck et al. 2000). The difference with total deposit money bank assets mainly consists of claims on the government and on (semi-) public enterprises. Figure 1.3 also shows that, in contrast to the euro-area, credit to the private sector on average hardly increased in CEE countries in the period 1993-2000. Not only did

14 Three types of strategies were used to deal with the crises. The CEE countries generally pur-sued extensive restructuring and recapitalizing of banks; most CIS countries pursued large-scale liquidation; and the Baltic states generally pursued a combination of liquidation and re-structuring. See Zoli, Tang, and Klytchnikova (2000) for further details.

15 At the same time, transition countries may be over-banked. Jaffee and Levonian (2001) cal-culated a benchmark value for the number of banks in transition countries, using data on the number of banks in OECD countries. They find that the number of banks is lower than the benchmark value in the Czech Republic, Hungary and Poland; in all the other countries it is higher.

16 In Chapter 2 details of the maturity of customer loans in Central and Eastern Europe will be reviewed.

17 As argued by Koivu (2002), in transition countries the size of the financial sector does not tell much about the quality of the sector, as soft budget constraints are still prevalent in various countries.

��

Chapter 1

��

private credit/ GDP remain constant over this period, Figure 1.4 shows that credit to the private sector as share of total credit, did not change either. The conclusion is that banks in CEE mostly lend to the public sector throughout 1993-2000. From the point of view of the private sector these observations are confirmed by Lízal and Svejnar (2002). The authors use data from most of the industrial firms located in the Czech Republic between 1992 and 1995 and find that domestic firms cannot easily borrow investment funds externally. Their study lends support to the argument that the financial sector in the Czech Republic was not yet playing the necessary role of intermediary.

Table 1.6 Domestic credit to the private sector (in % GDP), 1995-2004

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Bulgaria 21 35 12 12 14 12 15 18 26 23

Croatia 23 21 25 27 22 28 34 44 49 52

Czech Rep. 47 47 54 47 40 36 27 24 25 27

Estonia 14 18 26 26 24 24 26 28 33 43

Hungary 23 22 24 24 26 30 31 35 43 46

Latvia 8 7 11 15 15 18 25 31 39 50

Lithuania 12 10 9 10 11 10 11 14 20 25

Poland 13 16 17 18 19 21 22 22 23 23

Romania 8 12 8 12 8 7 8 8 10 10

Slovak Rep. 26 30 42 44 39 34 28 25 25 26

Slovenia 27 28 28 33 36 39 40 40 43 48

Average 20 22 23 24 23 23 24 26 30 34

Domestic credit to the private sector is defined as the ratio of total outstanding bank credit to private sector at end-of-year, including households and enterprises, to GDP. Source: EBRD

Though intermediation remained stunted until 2000, recent data draw a more optimistic picture. Table 1.6 shows that from 2002 onwards, average domestic credit to the private sector seemed to pick up. It also reveals the big differences in growth and level between countries. Credit growth takes place mainly in the Baltic States, Croatia, Hungary, and, moderately, in Slovenia. In the Czech Republic, Poland, Romania and the Slovak Republic, private sector credit to GDP hardly picks up after 2002. This growth has only exacerbated the intra-regional differences regarding credit to the private sector. In Croatia, Estonia, Hungary, Latvia, and Slovenia credit to the private sector as a ratio of GDP is 40 per cent or higher. In contrast, in Bulgaria, the Czech Republic, Lithuania, Poland, and the Slovak Republic this ratio is between 20 and 30%. Romania is at the lower end of the scale with domestic credit to the private sector being only 10 per cent of GDP.

��


Table 1.7 Domestic credit and distance to the benchmark (%-points), 1995-2003

1995 1996 1997 1998 1999 2000 2001 2002 2003 Av.

Bulgaria -26 -10 -33 -34 -33 -38 -37 -35 -29 -30

Croatia -28 -32 -30 -30 -34 -31 -26 -18 -15 -30

Czech Republic -19 -20 -13 -20 -28 -34 -44 -48 -49 -25

Estonia -37 -35 -31 -32 -34 -37 -38 -39 -35 -35

Hungary -36 -37 -37 -38 -38 -36 -36 -34 -28 -37

Latvia -38 -40 -39 -36 -38 -37 -33 -29 -23 -37

Lithuania -38 -42 -44 -46 -45 -48 -49 -48 -45 -45

Poland -40 -39 -40 -41 -41 -40 -41 -41 -41 -40

Romania -40 -37 -39 -35 -39 -41 -42 -43 -44 -39

Slovak Republic -32 -30 -19 -19 -25 -31 -38 -43 -45 -28

Slovenia -40 -41 -42 -39 -38 -37 -37 -38 -36 -39

Average -34 -33 -33 -34 -36 -37 -38 -38 -35 -35

The table shows own calculations based on the Fries and Taci (2002) benchmark model. The specified model reads as: y = 34.7 – 20.1ln(x) + 2.5(lnx)2 , where x is GNP per capita in international US dollars at purchasing power parity and y is domestic credit to the private sector/GDP. Av. denotes the 1995-2001 average. Sources: EBRD and World Bank WDI.

Fries and Taci (2002) have estimated a benchmark model to examine to what extent transition countries deviate from countries with a similar income per capita when it comes to the domestic credit to GDP ratio. Table 1.7 is an update from this study.

Several conclusions follow from Table 1.7. First, the average gap between actual credit to the private sector and the benchmark has increased since 1995 and seems to narrow only since 2003. Second, differences between countries seem to be persistent over time. On average, Bulgaria and Croatia had the smallest gap with its benchmark and from 2001 onwards this gap is narrowing steadily. In Lithuania, Poland and Romania the average distance over time between actual credit to the private sector and the benchmark is relatively largest and the situation is improving slightly only in Lithuania while even deteriorating in Romania. In the group of intermediate countries, the credit to the private sector gap is narrowing in Estonia, Hungary, Latvia and Slovenia. Exceptions are the Czech Republic and the Slovak Republic where the average gap is low but steadily increasing from 1999 onwards. The EBRD (2006) argues that the declining credit/GDP ratios are due to bank restructuring in these countries. The overall picture leads us to conclude that for more than a decade of transition, bank intermediation in Central and Eastern European countries remained stunted. Only from 2002 onwards credit to the private sector as a ratio of GDP has picked up. Results from the benchmark model confirm these findings.

�0

Chapter 1

��

Does foreign bank entry enhance the level credit to the private sector? A quick glance at Tables 1.3 (foreign bank entry) and 1.6 (credit to the private sector) learns that average foreign presence increased steadily over time and average credit to the private sector remained virtually unchanged. However, the relationship between foreign bank entry and growth of private credit to GDP differs per country. For example, countries that received many foreign banks already by 1996, such as Estonia, Hungary and Latvia, showed the highest credit to GDP level in 2004. However, by 2004, Croatia had reached a similar level of private credit to GDP, while foreign banks started entering the country only in 1999. Second, in Slovenia credit to GDP picked up in 1998, but the foreign bank presence remained unchanged in that period. Finally, in the Czech and the Slovak Republic foreign bank entry and credit to GDP show an inverse relationship after 1997, respectively 1998. So, if there is no direct one-on-one positive or negative relationship between foreign bank entry and credit to the private sector/GDP, what has been on average the role of foreign banks as intermediaries in Central and Eastern Europe?

Figure 1.5 Average credit by foreign and domestic banks (% total credit), 1993-2000

0%

10%

20%

30%

40%

50%

60%

1993 1994 1995 1996 1997 1998 1999 2000

Private credit by foreign banks Public credit by foreign banksPrivate credit by domestic banks Public credit by domestic banks

Sources: Central bank survey for data on: a) deposit money bank assets, b) foreign banks’ as-sets/ GDP, c) credit to the private sector by deposit money banks/ GDP, and d) foreign bank credit to the private sector/ GDP. Foreign bank credit to the public sector was calculated by subtracting d) from b). Domestic bank assets were calculated by subtracting b) from a). Credit to the private sector by domestic banks was calculated by subtracting d) from c). Credit the public sector by domestic banks was calculated as a) minus b) minus credit to the private sector by domestic banks. Data include from Croatia, Estonia, Hungary, Latvia, Lithuania, Poland and Slovenia. Source: Central bank survey.

�0 ��


Figures 1.5 shows the development of the average volume of credit to the private and public sector as share of total bank credit over the period both by domestic and foreign banks. First, it is clear that domestic banks have been replaced by foreign banks as creditors. In 1993, domestic banks were the primary source of credit for the public and the private sector, while in 2000 foreign banks dominate both markets. Second, Figure 1.5 shows that only from 1998 onwards, foreign banks started to lend more to the private sector than to the public sector. We conclude that foreign banks behave no differently than domestic banks in this respect. Only in 2000 foreign banks started to extend more credit to the private sector than domestic banks, while domestic banks focused predominantly on the public sector. As the huge amount of credit to the public sector in the Czech Republic influences the two averages for the foreign banks to a great extent, the figure is presented without the Czech Republic. With the Czech Republic taken into account, foreign banks extend more credit to the public sector than to the private sector for the entire period. Similarly, domestic banks’ credit to the public sector would exceed credit to the private sector throughout the 1990s.

Figure 1.6 Foreign banks’ relative share in credit to the private sector, 1995-2000

Data include banks from Croatia, Estonia, Hungary, Latvia, Lithuania, Poland and Slovenia. Presence is measured as the average of foreign banks’ assets / total banks assets and number of foreign banks/ total number of banks. Credit is measured as the share of foreign banks’ credit to the private sector in terms of total supply of credit to the private sector. Source: Central bank survey and EBRD.

The next question is: was the share of foreign banks’ supply of credit to the private sector in CEE in line with their presence? Figure 1.6 shows that up to 1997 foreign banks extended less private credit than one would expect based on their presence in the region. It was only from 1998 that foreign banks lend relatively more to the private sector than domestic banks; in previous years domestic banks were more important in serving the private sector. As can be observed from Figure 1.7, the averages in Figure 1.6 mask the substantial differences among the countries with respect to the relative importance of private lending by foreign banks. In countries marked left of the 45-degrees line, foreign banks share of total banking assets would justify a higher share in credit to the private sector. In countries marked on the right of the 45-degrees foreign banks take up a larger share of credit to the private sector than the scale of

��

Chapter 1

��

their presence would suggest. By 2000, after a decade of transition, foreign banks in Croatia, the Czech Republic, Lithuania and Poland still underservice the private sector more than the scale of their presence would justify. In contrast, foreign banks in Latvia, Hungary and Slovenia lend more to the private sector than domestic banks based on their mutual market shares measured in banking assets.

Figure 1.7 Foreign banks’ relative share in credit to the private sector, 2000

0%

25%

50%

75%

100%

0% 25% 50% 75% 100%Credit

Presence

Croatia

CzechRepublicEstonia

Hungary

Latvia

Lithuania

Poland

Slovenia

Presence of foreign banks is determined by relative assets of foreign banks as a ratio of total bank assets. Credit is defined as foreign banks’ credit tot the private sector (% total credit tot the private sector).

1.6 Financial performance of foreign banksA final topic regarding the role of foreign banks in the banking system in Central and Eastern Europe concerns their performance vis-à-vis domestic banks. Part II of this book will assess the relationship between foreign ownership and bank performance empirically. However, the data obtained from the national central bank will be presented now. We have asked national central banks to inform on the financial performance of foreign and domestic banks. Financial performance indicators consist of equity (ROE), return on assets (ROA), overhead costs and net interest revenues18. Figure 1.8 shows that average ROE of foreign banks exceeded ROE of domestic banks throughout the period 1995-2000. The findings are in line with ROA data in Figure 1.9. The columns in Figure 1.9 show the average profitability of foreign and domestic banks, measured by the average after-tax profit over total assets.19

18 In the literature net interest revenues are sometimes referred to as net interest margin. In this book, the net interest margin refers to the difference between the lending rate and the borrow-ing rate.

19 Unfortunately, we do not have access to reliable information in order to adjust these data for risk.

��


Figure 1.8 Average ROE of foreign and domestic banks, 1995-2000

-70,0%

-60,0%

-50,0%

-40,0%

-30,0%

-20,0%

-10,0%

0,0%

10,0%

20,0%

30,0%

1995 1996 1997 1998 1999 2000

ROE Foreign banks ROE domestic banks

ROE is determined as after tax profit over total equity. Data include banks from Croatia, the Czech Republic, Estonia, Hungary, Lithuania, and the Slovak Republic. Source: Central bank survey.

Figure 1.9 Average ROA and net interest revenues (NIR), 1995-2000

-6,0%

-4,0%

-2,0%

0,0%

2,0%

4,0%

6,0%

8,0%

1995 1996 1997 1998 1999 2000

ROA foreign banks ROA domestic Banks NIM foreign banks NIM domestic banks

In the figure bank performance is presented by the ratio of net income over assets (ROA) and net interest revenues over assets (NIR). Data include banks from Croatia, the Czech Republic, Estonia, Hungary, Lithuania, Romania, the Slovak Republic and Slovenia. Source: Central bank survey.

��

Chapter 1

��

Figure 1.10 Average overhead costs (% assets), 1995-2000

0,0%

1,0%

2,0%

3,0%

4,0%

5,0%

1995 1996 1997 1998 1999 2000year

Foreign banks Domestic banks

Data include banks from Croatia, the Czech Republic, Estonia, Hungary, Lithuania, the Slovak Republic and Slovenia. Source: Central bank survey

The distance between the ROA of foreign banks and the ROA of domestic banks seems to behave concavely in time, as relative profitability of foreign banks increases up to 1998 and decreases from 1999 onwards. It is important to note that the ROA values of are severely affected by bad results of domestic banks in 1998 and 1999. In all other years, the ROA of domestic banks did not diverge much from that of foreign banks. The ROA of foreign banks in Croatia sharply increased since its opening up to foreign banks. In Romania and Hungary, the ROA of foreign banks fell in the period under consideration. Estonian domestic banks had a particular bad year in 1998 when their combined ROA was –24 per cent. Without Estonian banks, domestic banks ROA in transition countries would have been –1.5 per cent instead of –4.3 per cent. Figures 1.8 and 1.9 suggest that foreign banks are more profitable than domestic banks in terms of ROE and ROA. This is in line with Claessens et al. (2001) and Kraft and Tirtiroglu (1998) who conclude that foreign banks are more profitable than domestic banks in transition economies. However, one must not disregard that in this period domestic banks proved to be more sensitive to the economic and financial crisis in 1998, caused by the Russian debt crisis, than foreign banks.

Figure 1.9 suggests that the net interest revenues of domestic banks are higher than that of foreign banks.20 Net interest revenues are defined as banks net interest income as a share of its total assets. With respect to this performance indicator, domestic and foreign banks appear to move closely together. Domestic banks in the Czech Republic

20 Koivu (2002) finds in his unbalanced panel data set from 25 transition countries that the margin between lending and deposit rates is significantly and negatively related to economic growth. In contrast, the amount of outstanding credit allocated to the private sector is not related to economic growth. Koivu argues that this could reflect the numerous banking crises that the transition countries have experienced and the soft budget constraints that were still prevalent in many transition countries at the time.

��


witnessed the lowest net interest revenues. Domestic and foreign banks in Croatia and Estonia earned, especially during the late 1990s, high net interest revenues.

A stylised fact from the economic literature as referred to above is that the costs of foreign banks are lower than those of domestic banks. However, for the transition countries, figure 1.10 reveals that the differences between domestic and foreign banks in the overhead costs as percentage of the total assets are rather small. In addition, one can observe the downward trend for both domestic and foreign banks. In the Czech Republic, the costs of domestic banks are even lower than the costs of most foreign banks in other countries. However, the foreign banks in the Czech Republic also have a lower cost level than the domestic ones. Estonia and Lithuania show a strong decrease in costs of domestic banks during the second part of the nineties. According to Lensink and Hermes (2004), the costs of domestic banks in developing countries rise with foreign bank entry. However, the data in Figure 1.10 do not confirm these findings for our set of transition economies.

1.7 Summary and conclusionChapter 1 examines the role of foreign banks in the banking sector in Central and Eastern Europe. The relationship between the presence of foreign banks and the level of credit to the private sector in European transition economies has received much attention in the literature. Theory assumes that especially in emerging markets and transition economies foreign banks, usually based in developed economies, have superior techniques to their disposal regarding screening, monitoring and risk pricing compared to domestic banks. This theory is best described by Berger (2001) in his global advantage theory. However, many factors related to banking in European transition economies sabotage the comparative advantages of foreign banks. For example, the lack of track records of firms makes it more difficult for foreign banks to solve the adverse selection problem. Adverse selection occurs when banks lend to companies with poor potential of repaying the loan. I assume that domestically owned banks have superior skills to collect and interpret the soft information of potential borrowers. The result is that domestic banks’ decision to whether or not extend credit qualitatively exceeds that of foreign banks. Possibly, foreign banks abstain from extending credit to companies lacking a track record.

The first section of this chapter sketches foreign bank entry in Central and Eastern Europe. It turns out that despite the upward sloping trend, major intraregional differences can be discerned. The increase in the number of foreign banks has continued up to 2000. After 2000 the number of foreign banks has stabilized and from 2002 onwards the relative number of foreign banks has stabilized on a 61 per cent level. In 1995 the relative number of foreign banks was only 25 per cent.

The amount of relative foreign assets exceeds the importance of the relative number of banks. From 2002 onwards, foreign banks own 84 per cent of the banking assets in Central and Eastern Europe. In 1995 this ratio was only 4 per cent. The regional differences regarding foreign bank entry also stem from differences in time.

��

Chapter 1

��

Already in 1996, foreign banks in Hungary and Latvia administered a large part of the assets. In 1997 foreign banks also entered Lithuania, in 1998 they entered Estonia, and in 1999 they entered Poland. The largest increase on average of foreign banks took place in 2000, especially in Bulgaria, Croatia, the Czech Republic, the Slovak Republic and Slovenia. An exception is Latvia in which relative foreign banking assets declined from 81 percent in 1998 to 29 percent in 2004.

The chapter continues with an analysis into the nature of foreign ownership. In 2002 The Economist noted that in Central and Eastern Europe the big US and UK banks are not present. It turns out that in 2005 the most important investors in the financial sector are the Austrian, Italian, Belgian and German banks. The biggest intra-regional investors come from Estonia, Hungary, the Czech Republic and Poland. However, the fraction of investments from CEE is only 1.3 percent and sharply contrasts with that of banks from Western Europe (62 percent).

Regarding the size of the banking sector in relation to gross national product, the chapters find that, despite a slightly increasing trend, the former is still half the size of what is prevailing in the countries that have introduced the euro. The level of credit to the private sector has even remained at an unchanged low level in the period 1993-2000. Measured as a percentage of gross domestic product, the level of credit to the private sector is a fourth of that in the euro area. Also in terms of total supply of credit, credit to the private sector has remained unchanged in this period in Central and Eastern Europe.

Only since 2002, credit to the private sector seems to be on a rise. However, the increasing average masks large intra-regional differences. Credit growth turns up mainly in the Baltic States, Croatia, Hungary and the Slovenia. In the Czech and the Slovak Republic credit to the private sector diminishes constantly since 1997. Based on a benchmark model, that measures to what extent credit growth in the economies in Central and Eastern Europe compares to what is common in western economies with comparable gross domestic product per capita, it turns out that the countries in which credit growths from 2002 onwards, are the same countries where the average distance to the credit-benchmark was smallest between 1995 and 2001. Exceptions are the Czech and the Slovak Republic in which the distance to the benchmark has unceasingly increased since 1999.

The analysis shows that there is no one-to-one relationship, positive nor negative between the entry of foreign banks and growth of credit to the private sector. The chapter finds two answers to the question on the role of foreign banks as intermediaries in the banking sector in Central and Eastern Europe. First, foreign banks have taken over the role from domestic banks as suppliers of credit from 2000 onwards. The reason is that the number of foreign banks exceeds that of domestic banks from that point in time, also measured in relative assets. However, that does not answer the question how foreign banks act as suppliers of credit. The second finding is that foreign banks extend credit to an extent that meets their relative presence only since 1997. Before 1997, domestic banks extended more credit to the private sector than one would expect based on their relative share in the banking sector. Even in 2000,

��


foreign banks in Croatia, Lithuania, Poland and the Czech Republic extended less credit to the private sector than would be justified based on relative assets in the banking sector.

Chapter 1 concludes with some indications regarding foreign bank financial performance in relation to that of domestic banks. It turns out that profitability of foreign banks, measured in terms of return on equity and return on assets, exceeds domestic bank profitability in the period 1995-2000. The higher profitability does not stem from lower costs or from higher net-interest revenues. After 2000 bank profitability of foreign and domestic banks seems to converge. A caveat of the analysis is that it does not conclude on the statistical significance of the difference in performance between foreign and domestic banks. However, the estimation results in chapters 4-6 will examine the relevance of foreign ownership on bank performance in more detail.

The following conclusions emerge from this chapter. First, we conclude that foreign banks have not entered the individual Central and Eastern European transition countries at the same time or in the same degree. Second, foreign banks in Central and Eastern Europe usually come from countries of the former EU15. This finding suggests a relationship between foreign banking and distance between host country and home country. The hypothesis would then be that foreign banks have followed their clients to Central and Eastern Europe. This picture is confirmed by the initial hesitation of foreign banks to take part in extending credit to the private sector in line with their relative presence. Chapter 3 will examine this hypothesis.

Another cause of the relative low credit extension of foreign banks could be traced to the specific characteristics of banking in transition countries. One of these characteristics is the difference between the institutional context of host country and home country. One can think of time-consuming legal procedures for foreign banks in seizing collateral of a borrower in default. These procedures can be troubled by corruption of those who confiscate and the judiciary. This theory is supported by the findings of Hainz (2003), who mentions that a low quality of the institutional context induces banks to use collateral to solve the problem of moral hazard. For example, she finds that in Estonia and Romania, the demand for collateral diminishes when the institutional environment improves. Bratkowski et al. (2000) also find that the probability of obtaining bank credit increases with the supply of collateral. More specifically, the authors argue that, to solve the adverse selection problem, banks apply credit rationing when information asymmetry regarding the assets of a company. It is likely that foreign banks in Central and Eastern Europe were unknown with the ways to seize collateral, if any, and to sell it in a liquid market. A 2002 study of the European Investment Bank (EIB) confirms that even in 2002, there was a need to improve and strengthen the legal system and the judiciary regarding seizing collateral in the Czech Republic, Hungary, Slovenia and the Slovak Republic. The cause of the recent credit growth in the Central and Eastern European countries must be attributed to increased competition in the banking sector and favourable macro

��

Chapter 1

economic developments (EBRD, 2006). The conclusions of this chapter are in line with Mian (2006) who argues that

important restrictions accompany the potential of capital mobility to solve financial underdevelopment. The analysis from chapter 1 raises new questions. First, to what extent are foreign banks in Central and Eastern Europe different from those in the euro-area or from those in European transition economies that do not have the prospect of joining the EU? Why have foreign banks in Central and Eastern Europe committed themselves only from the end of the 1990s to extending credit to the private sector? What can we conclude regarding the influence of foreign ownership of banks on their financial performance? And finally, to what extent does the institutional environment affect foreign banking? Chapter 2 aims to find an answer to the first question, using a microanalysis. Chapter 3 aims to answer the second question by asking banks themselves what has been the strategy behind their entry to Central and Eastern Europe. The other questions will be examined empirically in Part II of this book.

��

Chapter 2Size and Operations *

2.1 IntroductionThe previous chapter showed how foreign banks gradually took hold of most of the banking assets in Central and Eastern Europe, and discussed the importance of foreign banks as suppliers of credit to the private sector. Overall, it took up to 1998 before foreign banks started lending to the private sector in a way proportionate to the relative scale of their assets in the banking sector. This chapter extends the analysis of the role foreign banks play in the CEE banking sector examining the following topics: (i) how does foreign bank size relate to the size of a domestically owned bank? (ii) what are the main components of foreign banks’ assets? In detail I will examine (iii) the content and the maturity structure of the loan portfolio and (iv) the main items of the liability side and the maturity of customer deposits. This chapter presents the main trends in foreign banks’ balance sheets over a period of 8 years. Our starting point is the year 1995, when state-owned banks still owned the majority of assets in Central and Eastern Europe, and foreign banks hardly played a role in the financial system. The analysis goes up to 2002, when credit to the private sector in Central and Eastern Europe started to pick up after more than a decade of stunted intermediation as shown in chapter 1.

This chapter is comparative in nature. Whereas chapter 1 predominantly focused on the role of foreign banks in Central and Eastern Europe, this chapter introduces two benchmarks, both related to the (future) accession of Central and Eastern European countries as new Member States of the European Union. The first benchmark consists of banks in the fifteen countries that used to be the corpus of the European Union up to May 2004: hereafter referred to as the EU15. Ex ante I hypothesize that, due to relative geo-political closeness and the prospect of EU accession, size and operations of CEE (foreign) banks converge to EU15 levels. The second benchmark consists of banks in European transition economies which are geographically, culturally and institutionally more distant from the EU15. These countries do not have the prospect of EU accession and their economies are less developed than those of the CEE countries. More specifically, the second benchmark contains 16 transition countries in South Eastern Europe (SEE) and the Commonwealth of Independent States (CIS). SEE countries include Albania, Bosnia-Herzegovina, Macedonia and Serbia and Montenegro, and the CIS countries include Armenia, Azerbaijan, Belarus, Georgia, Kazakhstan, Kyrgystan, Moldova, Russia, Tajikistan, Turkmenistan, Ukraine and Uzbekistan. Recently, the EBRD 2006 Transition Report has stated that the banking

* Based on Naaborg, I.J. and L.J.R. Scholtens, A comparative study of banks’ balance sheets in the European Union and European transition countries, 1995-2003, International Finance Review, 6, pp 157-178, © 2006, with permission from Elsevier.

�0

Chapter 2

��

reform in the CIS countries is much less advanced than in the CEE countries. Ex ante I hypothesize that a microanalysis of bank balance sheets will confirm that the scale and scope of CEE (foreign) banks exceed those of SEE/CIS banks. The structure of this chapter is as follows. Section 2.2 introduces the methodology and the data. In section 2.3, trends in foreign bank size are discussed. Sections 2.4 and 2.5 present analyses on assets and liabilities. Section 2.6 summarizes and concludes.

2.2 Methodology and data The methodology and variables in this chapter differ from those in chapter 1. Instead of using macro-data, the chapter takes a micro approach and the analysis is based only on information contained in banks’ balance sheets. The balance sheets are examined in three ways. First of all, they are informative on the size of the banks. Within the analysis of bank size, the section will distinguish between foreign banks and domestic banks in order to compare bank size in the three regions. Second, assets and liabilities are analysed separately on (i) the main components and (ii) maturity of loans and deposits. Within the set of loans the chapter aims at assessing the main loan products.

The analysis in this chapter is based on 31,319 observations from the balance sheets of commercial banks, savings banks and cooperative banks in the period 1995-2002, provided by Bureau van Dijk’s BankScope database. Appendix 6 shows the number of banks for each country in the sample.1 Appendix 7 illustrates the structure of the sample by showing relative bank assets of each country as a share of total bank assets in each of the three regions. In Central and Eastern Europe, banks in Poland, the Czech Republic and Hungary together account for almost 70 per cent of banking assets, while in the SEE/CIS group Russia clearly dominates, with approximately three quarters of the region’s assets.

Using bank ownership information from several issues of the BankScope database made it possible to identify domestic owned banks and foreign owned banks2. Appendix 8 shows the number of foreign and domestic banks in each of the three regions.

2.3 Bank sizeTable 2.1 shows the average bank size in each country for the period 1995-2002. Banks in the three regions differ considerably in size. In 1995, the average bank in the EU15 was 8 times larger than an average bank in Central and Eastern Europe;

1 For the purpose of sample homogeneity, only the commercial banks, savings banks and coo-perative banks in BankScope were selected out of a total set of nine specializations defined by BankScope.

2 One of the drawbacks of BankScope is that bank ownership information is given only for the most recent update. I thank Mark Wessels and Bob Vaanhold at Bureau van Dijk Electronic Publishing for making previous versions available, and Aljar Meesters for his assistance with the preparation of the dataset.

�0 ��

Size and Operations

by 2002 that multiple had increased to 11. However, the average CEE bank grew considerably relative to a SEE/CIS bank. In 1995 the average CEE bank was only twice as large as a SEE/CIS bank, while in 2002 this multiple had risen to 4.

Table 2.1 Average bank size (€mln), 1995-2002

1995 1996 1997 1998 1999 2000 2001 2002

EU15 4,626 8,055 10,493 11,425 12,470 13,806 14,056 15,591

(4,457) (6,374) (9,929) (9,459) (10,046) (11,081) (12,504) (13,704)

CEE countries 560 590 562 761 775 915 1,171 1,328

(521) (604) (564) (726) (615) (803) (966) (1,050)

SEE/CIS countries 291 262 360 261 372 356 328 309

(488) (467) (597) (384) (520) (627) (515) (389)

Bank size is determined as the average of country averages. The standard deviation of the country averages is in parentheses beneath the averages. Source: BankScope

Figure 2.1 Average bank size in Central and Eastern Europe (€mln), 2002

Source: BankScope, EBRD (2006) and national central banks. SR and SV refer to the Slovak Republic and Slovenia.

Appendix 9 shows the average bank size in each country of the three regions. In 2002, average size of banks in the Czech Republic and Poland started to be comparable to that of banks in Italy, Austria and Germany. Figure 2.1 shows the negative correlation, with a value of -0.34, of average bank size with the number of banks relative to the size of the population. Romania, Bulgaria and Lithuania have small banks. Latvia, Croatia and Slovenia have most banks. A ratio of four banks per one million inhabitants is common in the region.

Figure 2.2 shows growth in bank size in the three regions. In CEE, no bank

Czech Republic

Poland

Hungary

EstoniaSR

LithuaniaBulgaria

RomaniaLatvia

Croatia

SV

€ 0

€ 500

€ 1.000

€ 1.500

€ 2.000

€ 2.500

€ 3.000

€ 3.500

€ 4.000

0 2 4 6 8 10 12

Number of banks/ Population (mln)

Average bank size (€mln)

��

Chapter 2

��

growth occurred up to 1998. Next, bank growth took off, especially in 2000/01. This development coincides with the privatisations, mergers and foreign bank entry in those years. See chapter 1 for details. Despite the fact that CEE bank growth takes off only from 1997, the trend is similar to that of former EU banks. In contrast, bank growth in the SEE/ CIS countries is virtually absent.

Figure 2.2 Average growth in bank size (1995 =100), 1995- 2002

Bank size is determined as the average of country averages. Source: BankScope. Bank size is based on nominal assets.

Figure 2.3a-c relates previous findings to the development in size of foreign and domestic banks in the three regions. The dotted columns in each of the figures show the average size of all banks, in line with Table 2.1. For the CEE area, figure 2.3a shows that the average size of all banks and the average size of foreign banks exhibit similar trends. The figure also shows that domestic banks decrease in size. Specifically, foreign bank size has been steadily growing from €300mln in 1995 to over €2bn in 2002. In the same period, domestic bank size shrunk from an average of over €2bn in 1995 to only €750 mln in 2002. The increasing foreign bank size might be explained by the acquisitions by foreign banks of large privatised state-owned banks in the region and by mergers between foreign owned local banks.3 The diminishing role of domestic owned banks is in line with the results of chapter 1.

In contrast, in the former EU average bank size is predominantly determined by growth in domestic owned banks, as both groups exhibit a similar trend. The average size of a domestic bank grew from €6bn in 1995 to nearly €20bn. Foreign bank size reached a little more than €5bn over the period. Data for foreign and domestic banks in CIS/SEE countries is more difficult to interpret due to scarcity of ownership data. Figure 2.3c suggests that in the less developed transition economies, domestic bank

3 For example, in 1998, Austrian Creditanstalt absorbed Bank Austria. As a result, the respective subsidiaries of both parent banks in Central and Eastern Europe are merged. In 2001, German HVB Bank acquired the Bank Austria Creditantalt combination. Again, subsidiaries of both parent banks in the CEE region were merged.

0

50

100

150

200

250

300

350

400

1995 1996 1997 1998 1999 2000 2001 2002

Former EU-15 CEE countries SEE and CIS countries

��

Size and Operations

size generally exceeds the size of foreign banks and that the size of the latter hardly increased in the 1990s. Second, one may argue that in 2002, average foreign bank size in the CEE/CIS region was similar to the size of foreign banks in CEE countries five years earlier.

Figure 2.3a Average foreign bank size, EU15 (€bn), 1995-2002Source Figures 2.3a-c: BankScope. For a summary of the sample, see Appendix 8.

Figure 2.3b Average foreign bank size, CEE countries (€bn), 1995-2002

Figure 2.3c Average foreign bank size, SEE/ CIS countries (€mln), 1995-2002

0

5.000

10.000

15.000

20.000

25.000

1995 1996 1997 1998 1999 2000 2001 2002

Domestic Banks All banks Foreign banks

0

500

1.000

1.500

2.000

2.500

3.000

1995 1996 1997 1998 1999 2000 2001 2002


0

100

200

300

400

500

600

700

1995 1996 1997 1998 1999 2000 2001 2002


��

Chapter 2

��

2.4 AssetsTable 2.2 shows the main banking assets in the three regions classified by its main components: (i) total loans - net, (ii) total other earning assets, (iii) total non earning assets, and (iv) fixed assets.4 The four components are presented per region as a ratio of total banking assets. Total loans –net are lowest in the CEE region. Only from 2002 the loan ratio seems to pick up significantly in terms of total bank assets. This result is in line with the findings in chapter 1 that credit to the private sector/GDP picked up in 2002. In the same period total other earning assets remained at a stale level around 42-45 percent of total assets, while in the EU other earning assets exhibited a clear declining trend. What did drop in terms in the average CEE bank were the non-earning assets: from 9.5 percent in 1995 to 6.5 percent in 2002. On this asset item CEE banks converged to the average EU15 banks. The average CEE bank also converged to a EU15 bank in terms of fixed assets, although its level is still higher. Combining branches and the liquidation of branches in rural areas may have added to the lower fixed assets in CEE. In the average CIS/SEE bank total non-earning assets and fixed assets in 2002 were on a level, which was normal in a CEE bank in 1995.

Table 2.2 Decomposition of bank assets (% Total Assets), 1995-2002

Year Loans - net Other Earning Assets Non Earning Assets Fixed Assets

EU CEECIS/SEE EU CEE

CIS/SEE EU CEE

CIS/SEE EU CEE

CIS/SEE

1995 46.3 44.0 31.6 46.8 41.7 43.6 5.5 9.3 17.3 1.5 5.0 7.5

1996 47.1 40.7 40.3 45.9 44.5 37.2 5.6 9.0 13.7 1.4 5.8 8.8

1997 48.1 40.0 42.5 44.4 45.4 35.6 6.1 9.0 13.6 1.4 5.6 8.3

1998 48.9 42.6 44.1 43.1 43.1 36.1 6.7 8.8 12.3 1.3 5.6 7.6

1999 50.2 43.1 45.3 41.4 43.0 36.2 7.2 8.2 11.1 1.2 5.6 7.5

2000 51.1 43.4 43.7 40.4 44.3 38.4 7.4 7.5 11.6 1.1 4.8 6.3

2001 51.4 42.0 46.4 39.4 45.7 35.6 8.0 8.1 12.6 1.1 4.1 5.4

2002 52.3 46.9 48.0 38.7 42.8 36.5 8.0 6.6 10.5 1.0 3.7 5.0Yearly data are averages of country averages. EU refers to the member states of the EU15. Source: BankScope. Total loans – net (loans –net) is the sum of total customer loans and problem loans, minus loan loss reserves. Total other earning assets (other earning assets) include the sum of total securities and bonds. Total securities include deposits with banks, due from central banks, due from other banks, government securities, investment securities, and trading securities.

Figures 2.4a-c decompose total loans by maturity. Maturity refers to the time to maturity, i.e. the time until the loan contract ends, if not renegotiated. First, up to 2002, the maturity structure of loans extended by CEE banks hardly changed. The average

4 The four main components in Table 2.2 of the assets side consist of several sub-items. The remainder of this section will further examine the components and maturity of total loans and discards the components of the rest of the three main items of the asset side.

��

Size and Operations

maturity even dropped in 1997 when the relative amount of loans with maturities up to 6 months replaced loans with longer maturities.

Figure 2.4a Maturity of total customer loans, EU15, 1995-2002

Data refer to Denmark, Finland, Ireland, Italy, Luxembourg and Sweden. Source: BankScope.

Figure 2.4b Maturity of total customer loans, CEE countries, 1995-2002

Data refer to the Czech Republic, Hungary, Poland, Romania and Slovenia. Source: BankScope.

0%20%40%60%80%

100%

1995 1996 1997 1998 1999 2000 2001 2002

Sub 3 months 6 months- 1 year 1-5 years 5 years + No split available

0%

20%

40%

60%

80%

100%

1995 1996 1997 1998 1999 2000 2001 2002

3-6 months 6 months- 1 year 1-5 years 5 years + No split available

��

Chapter 2

��

Figure 2.4c Maturity of total customer loans, SEE/CIS countries, 1995-2002

Data refer to Armenia, Azerbaijan, Belarus, Bosnia-Herzegovina, Kazakhstan, Moldova, Rus-sia and Uzbekistan. Source: BankScope

However, in 2002 the maturity of credit in Central and Eastern Europe seemed to rise as loans with 1-5 year maturities replace loan contracts ending within the year. Second, in addition to the increasing maturities of customer loans, the diversification in the maturity structure of the loan portfolio has increased.

Both features hint at the convergence of the loan portfolio of the average CEE bank to that of a bank in Western Europe. In the EU15 long-term lending clearly dominates (Figure 2.4a). For example, loans maturing 5 years from now make up more than 30 percent of loan portfolio. In the CEE region the share of long-term credit does not exceed 10 percent of the portfolio. EU15 banks also lend a substantial amount short term: 30 per cent of total loans have a maturity of less than 3 months. Short term lending by a CEE bank is not so common. In the CIS and SEE countries, at least two thirds of all loans mature within 1 year, and this percentage hardly changed over the period.

Table 2.3 Decomposition total customer loans (% total customer loans), 1995-2002

Public sector Mortgages Lease Other Loans Other Corporate

EU15 CEE EU15 CEE EU15 CEE EU15 CEE EU15 CEE

1995 9.0 10.9 6.5 0.0 1.3 0.0 80.9 59.6 2.1 26.9

1996 3.7 6.2 13.3 0.0 1.4 0.3 79.2 46.9 2.1 41.1

1997 3.4 4.9 10.2 0.0 1.6 1.9 82.3 26.6 2.1 58.9

1998 3.2 3.9 11.3 0.0 2.0 4.3 80.2 21.4 3.0 61.4

1999 3.1 2.3 12.2 0.0 2.0 3.9 79.1 22.1 3.3 65.3

2000 3.0 3.3 13.2 0.0 2.3 4.3 78.8 20.4 2.6 63.0

2001 2.7 4.0 13.4 0.0 2.5 8.2 78.8 15.1 2.6 58.0

2002 2.7 1.9 15.9 0.0 2.7 9.1 75.9 25.4 2.8 49.6

0%

20%

40%

60%

80%

100%

1996 1997 1998 1999 2000 2001 2002

3-6 months 6 months- 1 year 1-5 years 5 years + No split available

��

Size and Operations

Notes: Loans to the public sector include loans to municipalities and government, loans to state enterprises and advances to or guaranteed by public authorities. Lease includes leasing assets and lease receivables. Loans to other corporate include loans to domestic and private companies and loans to financial (non-bank) companies. Other loans include to trade receivables, overdrafts, se-curities held under repurchase agreements, accrued interest and factoring, overdue instalments, financing for import/export and foreign currency loans, doubtful loans, and bills. Data refer to France, Italy, Luxembourg, the Netherlands, Spain, Latvia and Lithuania. Source: BankScope.

Table 2.3 decomposes total loans in: (i) loans to the public sector (ii) mortgages, (iii) lease products, (iv) other loans, and (v) loans to other corporate.5 For the CIS/SEE region no information on loan products is available. The conclusion for the development in loan products in CEE is twofold. First, loans to municipalities and government decreased in the period 1995-2002. Second, the amount of hire-purchase (HP) and lease products of CEE banks exceeds that of EU15 banks.6 Although Table 2.3 gives loan product information in the CEE region only based on observations in Latvia and Lithuania, the conclusions seem to hold for the whole CEE region. For example, the first finding is in line with the statement of the EBRD that ‘[…] governments in transition countries have scaled back their borrowing from the banking system, thereby freeing up credit to the private sector’ (EBRD 2005). Thus, the observed shift from credit to the public sector to credit to the private sector noted in chapter 1 is consistent with the information from CEE banks’ balance sheets in Table 2.3. The growth in HP and lease products seems to hold for Central and Eastern Europe in general. Although leasing in the EU is often tax-driven, De Haas and Naaborg (2006, p.173) find other explanations for the leasing boom in Central and Eastern Europe. Based on bank interviews, the authors find that ‘when leasing, the bank retains the ownership of the object leased and can simply repossess if the debtor goes bankrupt, thus avoiding any lengthy bankruptcy procedures to take possession of collateral. Especially in the Baltic States, the rapid spread of leasing at the beginning of the transition process was related to the limited protection of collateral. Banks also used leasing to get around the insufficient track record of firms, which made credit screening difficult. […] Although the track records of firms have improved in recent years, leasing continues to be an important way of financing in Central Europe and the Baltic States.’

5 Unfortunately, Bureau van Dijk Electronic Publishing was not able to provide the definitions of the loan products as available in the BankScope database. CEE banks did not provide informa-tion regarding the mortgages.

6 Hire Purchase (HP): a method of buying goods in which the purchaser takes possession of them as soon as he has paid an initial instalment of the price (a deposit) and obtains ownership of the goods when he has paid all the agreed number of subsequent instalments. A hire-purchase agreement differs from a credit-sale agreement and sale by instalments (or a deferred payment agreement) because in these transactions ownership passes when the contract is signed. It also differs from a contract of hire, because in this case ownership never passes. A lease is a contract by which the owner of property allows another to use it for a specified time, usually in return for payment. (Financial Glossary, UK)

��

Chapter 2

��

2.5 LiabilitiesFigure 2.4 presents the ratio of equity over total assets. This ratio has remained about 5 per cent over the period for banks in the EU15, while the equity ratio for CEE banks has risen from about 8 per cent of total assets to over 10 per cent. The equity ratio is highest for banks in the SEE/CIS countries at around 14 per cent. This may reflect these banks’ high levels of risk, as equity acts as a buffer against unexpected losses. A more appropriate measure of strength is the BIS capital ratio, but unfortunately this was not available for most of the banks outside the EU15.

Table 2.5 decomposes banks’ liabilities into: (i) deposits, (ii) money market funding, (iii) other funding, (iv) loan loss and other reserves, and (v) other liabilities. First, regarding the set of liabilities CEE banks are funded less diversified than EU15 banks. In this respect CEE banks hardly differ from SEE/CIS banks as both are for almost 90% funded by deposits. EU15 banks are funded for less than 70% with deposits. Deposit funding is decreasing in both the EU15 and the CEE countries. However, where money market funding seems to be the substitute for deposit funding in the EU15, money market funding seems to be less popular in the CEE, another feature that CEE banks have in common with SEE/CIS banks. The use of this type of funding is even diminishing from 1999 onwards. This is probably due to the less developed money markets in both regions. From Table 2.4, it is difficult to identify a single alternative CEE banks prefer over deposit funding.

Table 2.4 Average equity and liabilities (% assets), 1995-2002

Year EU15 CEE SEE and CIS

Liabilities Equity Liabilities Equity Liabilities Equity

1995 95.1 4.9 91.8 8.2 79.9 20.1

1996 95.2 4.8 92.7 7.3 87.7 12.3

1997 95.4 4.6 90.0 10.0 84.6 15.4

1998 95.2 4.8 90.0 10.0 85.4 14.6

1999 94.8 5.2 89.2 10.8 85.6 14.4

2000 94.8 5.2 89.5 10.5 85.6 14.4

2001 94.9 5.1 89.8 10.2 86.2 13.8

2002 95.1 4.9 89.7 10.3 86.4 13.6

Source: BankScope.

Figures 2.5a-c decompose total deposits in the three regions by maturity. Deposits in the EU15 are highly liquid - about 50 percent are in the form of demand and savings deposits, which can be with drawn at call, another 20 per cent of deposits have a maturity within 3 months. The proportion of liquid deposits is slightly lower in CEE countries, with 60 percent having a maturity of less than 3 months, and it is much lower in the SEE/CIS countries where only 35 percent of deposits are available on an immediate basis. Deposits with a maturity of 6-12 months are dominant in the

��

Size and Operations

SEE/CIS, constituting about two-thirds of all deposits. In the EU15, the maturity of the deposit portfolio changed little over the period 1995-2003, whereas in CEE there was a gradual increase in the relative size of long-term maturing deposits (1-5 years) from 1 per cent in 1995 to around 11 per cent in 2003. In the SEE/CIS countries long-term deposits seem to be non-existent.

Table 2.5 Decomposition of liabilities (% of liabilities), 1995-2002

Year Deposits Money Market Other funding LLR Other

EU CEE S/C EU CEE S/C EU CEE S/C EU CEE S/C EU CEE S/C

1995 78.8 87.2 72.4 7.0 1.3 4.0 8.1 1.9 5.4 0.2 0.0 1.6 5.8 6.3 16.6

1996 76,7 85.3 78.3 6.4 1.9 3.9 9.9 3.7 7.9 0.3 0.0 0.1 6.7 5.3 9.8

1997 74.9 88.4 82.5 6.8 0.8 3.7 10.6 3.5 6.3 0.3 0.0 0.1 7.4 4.8 7.4

1998 74.1 90.4 75.9 7.0 1.7 4.6 10.4 2.0 7.8 0.4 0.1 0.0 8.1 4.7 11.7

1999 72.1 91.9 80.7 8.1 2.0 3.5 11.1 3.9 10.2 0.6 0.1 0.1 8.0 4.4 5.4

2000 69.8 90.6 83.4 8.7 1.6 2.1 11.9 3.3 10.0 1.1 0.1 0.1 8.5 5.1 4.5

2001 69.3 90.8 87.9 8.8 0.8 1.5 12.6 2.6 6.9 1.1 0.2 0.0 8.3 4.7 3.6

2002 68.0 89.6 88.8 8.7 0.9 1.7 13.1 3.0 6.7 1.1 0.3 0.1 9.1 3.9 2.8

Total Deposits (deposits) is the sum of customer deposits and banks deposits. Total money market funding (money market) includes certificates of deposit, debt securities, and other ne-gationable instruments. Total other funding (other) is the sum of other bonds, subordinated debt, and other funding. Loan loss reserves (LLR) include other reserves. The column ‘Other’ refers to other liabilities. EU refers to the Member States of the EU15; S/C refers to the tran-sition economies in South Eastern Europe and the Commonwealth of Independent States. Source: BankScope

Figure 2.5a Maturity of customer deposits (% customer deposits), EU15, 1995-2002

0%20%

40%

60%80%

100%

1995 1996 1997 1998 1999 2000 2001 2002

No split available 5 years + 1-5 years 6 months-1yearSub 3 months Savings Demand

Source: BankScope. Data refer to Denmark, Finland, Ireland, Italy, Luxembourg, the Netherlands, the UK, and Sweden.

�0

Chapter 2

��

Figure 2.5b Maturity of customer deposits (% customer deposits), CEE, 1995-2002

0%

20%

40%

60%

80%

100%

1995 1996 1997 1998 1999 2000 2001 2002

No split available 5 years + 1-5 years 6 months-1year3-6 monthsSub 3 months Savings Demand

Source: BankScope. Data refer to the Czech Republic, Croatia, Hungary, Lithuania, Latvia, Poland, Romania, and the Slovak Republic.

Figure 2.5c Maturity customer deposits (% customer deposits), SEE/CIS, 1995-2002

0%10%20%30%40%50%60%70%80%90%

100%

1995 1996 1997 1998 1999 2000 2001 2002

No split available 5 years + 6 months-1yearSavings Demand

Source: BankScope. Data refer to the Ukraine, Albania, Belarus, Georgia, and Macedonia.

2.6 Foreign banks’ balance sheet This section aims to answer two questions: (i) how do banking activities of foreign owned banks in Central and Eastern Europe compare to those of domestic banking in the region? and (ii) is foreign banking different in Central and Eastern Europe compared to foreign banking in the EU15 and the SEE/CIS region? Table 2.5 compares the main balance sheet items of foreign and domestic banks in the three regions. In line with Table 2.2 in section 2.4 I start by examining the main components on the asset side. Asset ratios allow for three conclusions. First, given their significant higher total loans - net ratio, foreign banks in Central and Eastern Europe seem to lend relatively more than domestic banks. However, CEE foreign banks have significantly less loan loss reserves than domestic banks. As a result, the levels of total customer loans between

�0 ��

Size and Operations

foreign banks and domestic banks do not differ significantly.7 This finding is in line with the conclusion of chapter 1, that only from 1998, foreign banks’ share in credit to the private sector started to match the share of their presence. However, Table 2.5 shows that foreign banks still have a significantly higher total non-earnings assets ratio compared to domestic banks in CEE. In addition, other earning assets of domestic banks are higher than for foreign banks implicating that domestic banks are focused on niche markets. The table also shows that, similarly to the early transition period in Central and Eastern Europe, foreign banks in the SEE/CIS area less involved in lending activities than domestic banks. This finding supports the hypothesis that foreign banks have more problems in lending in host countries where the cultural and/ or institutional difference with the home country is high. Foreign banks in all three regions suffer significantly less from loan loss reserves and, in the EU15 and SEE/CIS countries, from problem loans. This finding is consistent with the theories predicting a lower risk preference of foreign banks relative to domestic banks (see Section 1.2). Given a significantly higher other total earning assets ratio, foreign banks in the CIS/SEE countries are involved in other banking activities, e.g. securities and bonds.

The loan product section in Table 2.5 supports the hypothesis that foreign banks are relatively more involved in corporate lending than domestic banks.8 The loans to other corporate ratio in both Central and Eastern Europe and the EU15 are significantly higher for foreign banks. In addition, CEE foreign banks are more involved in lease operations. Both findings would confirm theoretical literature that suggests that foreign banks suffer from asymmetric information about the local market and rather focus on the less risky corporate sector. The average higher loans to municipalities/ government ratio of foreign banks seems to be in line with the finding in chapter 1 that for most part of 1995-2002, foreign banks were more involved in public lending than in lending to the private sector. Third, the loan portfolios of foreign banks in transition economies offer a more diversified maturity structure than those of the domestic banks. More specifically, in the CEE and the CIS/SEE region, foreign banks lend more on both the very short and on the very long term than domestic banks do. In the CEE area, foreign banks’ share of short-term loans, with maturities between 3 and 6 months, is significantly higher than that of domestic banks in the region. In the EU15 and the SEE/CIS countries, the loan portfolio of foreign banks contains significantly more loans with 5+ year maturities than the portfolios of domestic owned banks. Domestic banks seem to lend more in the mid-term segment of the maturity spectrum (1-5 years).

7 Total Loans – Net is the sum of total customer loans (a) + problem loans (b) - loan loss reserves (c).8 Table 2.5 presents ratios calculated as averages of total group observations. A caveat of this

approach is that I assume that year effects similarly affect domestic owned and foreign owned banks. However, results are in line with ratios presented by Naaborg and Scholtens (2006), where the latter were determined as averages of yearly averages.

��

Chapter 2

��

Table 2.6 Balance sheet ratios of foreign and domestic banks, 1995-2002

EU15 CEE countries CIS/SEE region

FOR DOM FOR DOM FOR DOM

A. Assets (% total assets)

Total customer loans (a) 46.4 47.9 46.1 42.8 31.6 54.0***

Problem loans (b) 0.1 0.2** 0.2 0.4 0.1 2.8**

Loan loss reserve (c) 0.2 0.4* 2.2 3.1* 2.4 5.4***

Total loans – Net 46.3 47.7 44.0* 40.1 33.6 54.0***

Total other earning assets 47.6 45.0 42.3 48.7** 48.6*** 27.2

Total non-earning assets 4.9 5.9** 8.9*** 6.5 13.5 12.4

Total fixed assets 1.2 1.3 4.8 4.6 6.2 6.4

100.0 100.0 100.0 100.0 100.0 100.0

Products (% total customer loans)

Loans to the public sector 0.9 4.8** 2.6* 0.1 n.a. n.a.

Mortgages 3.6 14.2*** 0.0 0.0 n.a. n.a.

HP/ Lease 3.0 2.1 7.1** 1.3 n.a. n.a.

Other loans 81.9 76.1 6.7 44.8 n.a. n.a.

Loans to other corporate 10.6** 2.8 76.9* 38.8 n.a. n.a.

100.0 100.0 93.3 84.9 n.a. n.a.

Maturity (% total customer loans)

Loans sub 3 months 29.8 31.0 0.0 0.0 0.0 0.0

Loans 3-6 months 0.4 0.3 6.3** 0.0 56.8 41.4

Loans 6-12 months 19.8 21.7 46.9 47.8 29.7 22.6

Loans 1-5 years 20.7 21.0 30.2 38.6 8.6 33.6***

Loans 5+ years 25.8** 19.9 15.0 13.6 3.4* 0.0

Loans (no split available) 1.4 1.5 0.0 0.0 0.0 0.6

98.0 94.0 98.4 100.0 98.5 98.2

B. Equity and liabilities

Liabilities 93.5 94.7*** 88.7 89.7 80.0 83.1

Equity 6.5*** 5.3 11.3 10.3 20.0 16.9

100.0 100.0 100.0 100.0 100.0 100.0

Liabilities (% of total liabilities)

Total customer deposits 81.6*** 74.6 89.4** 86.1 90.1*** 76.9

Total money market funding 4.8 8.6*** 2.0 1.9 2.2 4.8**

Total other funding 6.8 9.7** 2.8 6.2* 2.9 14.2**

Total loan loss & other reserves 0.3 0.4 0.2 0.3 0.0 0.1***

Other liabilities 6.5 6.6 5.6 5.0 4.8 3.9

100.0 100.0 100.0 100.0 100.0 100.0

��

Size and Operations

Ratios for foreign (FOR) and domestic (DOM) banks are averages. Loan product ratios for CEE countries are based on banks in Latvia and Lithuania. Loan product ratios are based on countries in which at least 73 per cent of loans could be categorized. For product defini-tions see Table 2.3. Loan maturity ratios are calculated based only on banks from countries for which at least 78 per cent of loans could be categorized. To identify significance levels of differences in means, test statistic T = [(X1-X2) – (μ1-μ2)]/ [(S1/n1) + (S2/n2)]

1/2 and Student’s t Distribution are used, with ν = min(n1,n2) - 1. *, **, *** represent significance levels at 1%, 5%, and 10%.

Regarding the liability side of banks’ balance sheets, the data in Table 2.5 suggest that foreign banks are relatively equity rich compared to domestic banks. However, means are significantly different only in the EU15. I find convincing evidence that foreign banks in Central and Eastern Europe and the other benchmark regions are relatively more funded with deposits than domestic banks. Instead, domestic banks use a more varied set of funding sources, e.g. money market funding. Data across the three regions suggest that the less developed the host country, the more foreign banks fund themselves with deposits. This finding suggests that CEE foreign banks are relatively more involved in retail banking than foreign banks in the EU15 and those in the SEE/CIS region. The next chapter will examine this hypothesis more in detail

2.7 ConclusionChapter 2 extends the analysis on the role of foreign banks in Central and Eastern Europe in several ways. The chapter aims at comparing developments in foreign banking in Central and Eastern Europe with foreign banking in developed economies and foreign banking in transition countries with low economic development. Specifically, the chapter answers the following questions: (i) how does the size of a foreign bank compare to the size of a domestic bank? (ii) which items make up total assets of a foreign bank? (iii) what is the structure of the credit portfolio, and (iv) how are foreign banks funded? The first benchmark consists of the 15 EU Member States that made up the EU up to May 2004. The second benchmark consists of 16 transition economies in South Eastern Europe and the Commonwealth of Independent States (CIS). The analysis starts in 1995, the year in which local governments in Central and Eastern Europe still owned the majority of banking assets. For a period of eight years bank balance sheets in the three regions are followed until 2002, the year in which credit to the private sector in CEE started to pick up. The analysis is based on 31,319 bank observations of commercial banks, savings banks and cooperative banks in the three regions. For reasons of homogeneity other banks from the BankScope database are disregarded.

The conclusions of this chapter are as follows. First, the difference in size between an average CEE bank and a EU15 bank has increased between 1995 and 2002.

��

Chapter 2

��

During this period, a CEE bank grew from €560mln to over €1.3bn. The biggest banks operate in the Czech Republic, Poland and Hungary; the smallest banks are found in Bulgaria, Latvia, Lithuania and Romania. An average EU15 bank measured €4.6bn in 1995 and in 2002 €15.6. Although the size of CEE banks started to augment from 1997, banks in SEE and CIS countries hardly increased in size. Foreign banks in CEE countries grew, while domestic banks decreased in size. This trend can be explained by the intraregional mergers and acquisitions and the privatisations of former state owned banks to foreign investors. In this respect the CEE banking sector differs from that in the EU15 and SEE/CIS countries, where foreign banks are usually smaller than domestic banks.

Chapter 2 confirms the findings of the first chapter regarding the low degree of intermediation in Central and Eastern Europe in the period 1995-2002. The share of loans in total assets of an average bank in this region is lower than in Western Europe; only in 2002 the loans to assets ratio increases. The same picture arises in banks’ balance sheets in SEE and CIS countries. It must be noted that the share of non-earning assets decreased in banks in the transition regions, just as the share of fixed assets within total assets. Explanations for the former can be the improved macro economic prospects, a decrease of the information asymmetry regarding the borrower and an improved functioning of institutions. The decrease in fixed assets can be explained by the closure, merger and/or restructuring of bank branches. The maturity of loans and the diversity in maturity of the average credit portfolio increases in CEE. The trends point at a convergence of the characteristics of credit portfolios of EU15 and CEE banks. In Western Europe, 40 percent of the loans have a maturity of less than a year, in CEE countries this ratio amounts to 50 percent and in SEE/CIS countries this ratio is more than 60 percent.

From the balance sheets of banks in the Baltic States it turns out that the share of loans to the public sector in total customer loans has decreased from 1995 onwards. This finding is additional evidence to the finding of chapter 1 that in the 1990s the credit market in CEE has been suffering from a crowding out effect that has turned to a halt in 2002. In addition, bank balance sheets confirm that the use of leasing by banks in Central and Eastern Europe is quite popular. Apart from tax reasons, additional explanations are the problems in seizing collateral and the absence of track records of potential borrowers.

The liability side of bank balances in CEE countries and those in SEE and CIS countries are more similar than those of EU15 banks. For example, 90 percent of the liabilities are formed by deposits and only a fraction by money market funding; in EU15 banks 70 percent are deposits and almost 9 percent is funded on the money market.

Next, chapter 2 investigates to what extent the balance sheets of CEE foreign banks and domestic banks differ and whether these differences are characteristic for CEE. The first conclusion is that on average foreign and domestic banks had similar loans to assets on their balance sheets. This finding confirms the picture drawn in

��

Size and Operations

chapter 1 that foreign banks extended relatively more credit after 1998 and relatively little credit before that year. Similarly, in Western Europe there is no difference in the amount of credit extended by foreign and domestic banks. However, in the sample of developing transition countries domestic banks extend more credit than foreign banks.

Furthermore, this chapter finds that foreign banks prefer to focus on corporate banking and leasing activities. Domestic banks in Western Europe predominantly focus on loans to the public sector and mortgages. The credit portfolio of foreign banks in transition countries is characterized by a broader spectrum of maturities than that of domestic banks. This may be interpreted as the introduction by foreign banks of new banking products. A last result is that foreign banks, more than domestic banks, give credit with longer maturities. This finding is additional back up for the hypothesis that foreign banks in Central and Eastern Europe primarily focus on the corporate sector. Chapter 3 will study this hypothesis further.

��

��

Chapter 3Entry Strategies *

3.1 IntroductionInternational mergers and acquisitions in the financial sector in European transition countries reached a peak in 2001. Chapter 1 showed that, since 2000, foreign owned banks in Central and Eastern Europe (CEE) outnumber domestic owned banks. Nowadays, foreign owned banks make up more than 80 per cent of bank assets in the region. This chapter analyses international bank expansion strategies, examining the interdependence of (i) the reasons for foreign entry, (ii) the choice for acquisition or greenfield investment (also known as de novo investment), and (iii) the organizational form, i.e. branch or subsidiary. In contrast to most previous studies on foreign bank activity that generally rely on macro data, our analysis is based on data gathered via structured interviews with heads of the key foreign parent banks active in CEE.1 The method of interviews overcomes the disadvantages of single-case studies and questionnaires with a preset specification of possible answers. The results of the former are often difficult to validate, and the latter method runs the risk that relevant issues are not adequately addressed. We collected primary data on the basis of interviews in which respondents had ample opportunity to raise issues they thought to be relevant. Additional interviews with heads of local foreign owned banks and local financial supervisors were used for validation purposes.

To the best of our knowledge, this is the first study that comprehensively examines foreign bank entry strategies in CEE based on structured interviews. The foreign banks included in our survey cover more than 50% of international banking assets in the area. Our main findings are that that early enterers focus on corporate banking and establish greenfields, where local regulations influence the choice for the branch or subsidiary format. When the environment becomes more competitive, foreign banks acquire domestic banks – which are often merged with previous greenfields – focusing on retail banking. We also find evidence that herd behaviour affects foreign bank activity.

The paper is organised as follows. In the next section, we review related literature to come up with our hypotheses. Section 3.3 outlines our methodology and presents the data, while section 3.4 offers the results. Section 3.5 concludes.

3.2 HypothesesThis section presents the findings of previous studies in order to come up with

* Contents of this chapter are currently under review with a business journal.1 There are exceptions. For instance, Kraft (2004) presents results for foreign bank entry in Croa-

tia based on a survey by the Croatian National Bank.

��

Chapter 3

��

propositions regarding the determinants of foreign bank entry (3.2.1), the choice for greenfield or acquisition (3.2.2), and the choice for branch or subsidiary (3.2.3).

3.2.1DeterminantsofentryThe literature documents several motives for cross border bank expansion. In organising previous findings, we identify three groups of motives. The first entry motive is related to foreign activities of non-financial firms. According to the defensive expansion theory (Grubel 1977), banks follow foreign direct investments (FDI) by the non-financial sector to defend relationships with their clients.2 Banks that follow their customers focus on preventing losses in pre-existing activities, rather than on generating profits in the new location. Apart from FDI also other cross-border activities of non-financial firms, like exporting goods and services, may induce banks to follow their customers abroad.

A second set of entry motives is associated with the potential to increase profitability. Expected economic growth in the foreign country may be high offering profitable business opportunities, which may be attractive especially in case of strong competitive pressure in the home banking market. Investing in the foreign country may also be profitable due to other reasons, like expected exchange rate developments or an attractive tax regime. Furthermore, according to the theory of surplus entrepreneurship (Grubel, 1977) multinational banks may apply ownership-specific factors, such as superior entrepreneurial skills or superior technology and management expertise, to foreign banking markets at low marginal costs.3

The final set of foreign entry determinants is associated with the institutional context of the host market. Institutional parameters include financial regulation, the quality of the financial supervisor, the quality of law enforcement, the openness of the host country authorities towards foreign bank entry, and the role of information costs. Information costs mainly depend on the distance between the home and the host country, and the cultural similarity of both countries.

Table 3.1 summarizes previous studies, indicating whether a particular hypothesis is confirmed or rejected. The majority of foreign entry studies confirm the ‘follow the customer’ view. However, Seth et al. (1998) find that foreign owned banks in the US allocate a majority of their loans to non home-country borrowers. Similarly, Berger et al. (2003) find that nearly 66 per cent of the non-domestic multinationals firms in Europe choose a bank headquartered in the host nation and less than 20 per cent select a bank from their home nation. Finally, a third of the ‘follow the customer’

2 The defensive expansion theory is also referred to as the ‘follow the customer’ argument or the ‘conventional hypothesis’ (Sabi, 1995). See Williams (2002) for further details concerning empirical research on the defensive expansion theory.

3 The theory of surplus entrepreneurship is also referred to as the ‘theory of horizontal integra-tion’. Besides explaining international mergers and acquisitions, this theory also might explain foreign greenfields. It may be criticized for assuming that information technology and manage-ment expertise are generic while such factors may in fact be specific to the environment where they were developed.

��

Entry Strategies

studies that focus on less developed economies do not support the hypothesis.

Table 3.1 Foreign bank entry determinants in the literature

Support No Support

Internalisation theory

Follow FDI of existing client

Brealey and Kaplanis (1996), Buch (2000), Goldberg and Johnson (1990), Goldberg and Saunders (1981), Grosse and Goldberg (1991), Hultman and McGee (1989), Miller and Parkhe (1998), Moshirian (2001), Wezel (2004), Yamori (1998), Berger et al. (2003)*, Dubauskas (2002)*, Farabullini and Ferri (2004)*, Galac and Kraft (2000)*, Konopielko (1999)*, Kowalski et al. (2002)*.

Berger et al. (2003), Kindleberger (1983), Seth, Nolle and Mohany (1998), Clarke et al. (2003)*, Miller and Parkhe (1998)*, Wezel (2004)*.

Follow other foreign activities of existing client

Brealey and Kaplanis (1996), Buch (2000)Goldberg and Johnson (1990), Grosse and Goldberg (1991), Heinkel and Levi (1992), Magri et al. (2005), Moshirian (2001), Kowalski et al. (2002)*.

Focarelli and Pozzolo (2001), Galac and Kraft (2000)*, Wezel (2004)*.

Expected relative profitability

Growth in foreign country

Buch (2000), Claessens (2000), ECB (2005), Focarelli and Pozzolo (2001), Magri et al. (2005), Moshirian (2001), Kraft (2004)* Wezel (2004).

Buch (2000), Buch and DeLong (2004), Wezel (2004)*.

Competitive pressure at home

Dubauskas (2002)*, Florescu (2002)*, Galac and Kraft (2000)*, Konopielko (1999)*, Kowalski et al. (2002)*, Kraft (2004)*. Tschoegl (2004), Mathieson and Roldos (2001)*.

Galac and Kraft (2000)*.

Theory of surplus entrepreneurship

Tschoegl (2002). Vander Vennet and Lanine (2005).

Exchange rate expectations

Goldberg and Saunders (1981), Hultman & McGee (1989), Moshirian (2001).

Host country tax regime

Claessens (2000). Papi and Revoltella (2000)*.

Institutional context and information costs

Institutional context host country

Berger et al. (2004), Buch (2003), Buch and DeLong (2004), Focarelli and Pozzolo (2001), Goldberg and Saunders (1981), Goldberg and Johnson (1990), Hultman & McGee (1989), Magri et al. (2005), Clarke et al. (2003)*, Papi and Revoltella (2000)*, Lensink and de Haan (2002)*.

Information costs of foreign banking

Berger et al. (2004), Buch (2003), Buch and De Long (2004), Magri et al. (2005), Mathieson and Roldos (2001)*, Wezel (2004)*.

Buch (2000), Grosse and Goldberg (1991).

* indicates evidence for developing countries and/or transition countries.

�0

Chapter 3

�0 ��

The importance of profitability in foreign bank entry is less controversial.4 Higher economic growth abroad has caused foreign banks to enter the US, the UK and Germany (Moshirian, 2001), Italy (Magri et al., 2005), the new European Union Member States (ECB, 2005), and Croatia (Kraft, 2004). The role of competitive pressure at home has been confirmed for foreign banks expanding to the US (Tschoegl, 2004) and to Central and Eastern Europe. Evidence is available for profitable banking sectors in the Czech Republic, Hungary and Poland (Konopielko, 1999), Croatia (Kraft, 2004), Estonia (Kowalski, 2002), Lithuania (Dubauskas, 2002), and Romania (Florescu, 2000). Only few studies investigate the theory of surplus entrepreneurship. The results of Vander Vennet and Lanine (2005) do not confirm that European banks acquire badly run banks in Central and Eastern Europe in order to upgrade their performance.5 The literature suggests further that exchange rate expectations and tax regimes may also play a role in foreign banks’ decisions to go abroad.

Other studies emphasise the role of the institutions in the host country. Lowering regulatory barriers facilitates foreign bank entry (Magri et al., 2005; Focarelli and Pozzolo, 2001 and Berger et al., 2004). In addition, the attitude of the authorities towards foreign bank entry significantly determines bank entry initiatives (Papi and Revoltella, 2000). Information costs inherent to foreign banking remain an important impediment to the integration of international financial markets (Buch and DeLong, 2004). Berger et al. (2004) and Buch and DeLong (2004) find that cross border bank merger activity increases when home and host country are geographically close, share a common language and legal system, and have similar sized economies and levels of GDP per capita.6

The conclusions regarding the foreign bank entry literature are twofold. First, most studies confirm the follow the customer theory for a number of developed and less-developed economies. However, the general validity of this conventional theory is controversial due to ample counter evidence. Secondly, home market bank competition and higher host market bank profitability are important determinants to start foreign banking, a fortiori in the new EU Member States. Our first hypothesis is therefore:

Hypothesis 1. Foreign bank entry is largely motivated by the perceived need for support of the existing client base and by the expected profitability of banking abroad relative to banking at home.

4 The counter evidence from Buch (2000) is due largely to the homogeneity of the countries in the sample, while the counter evidence of Grosse and Goldberg (1991) is due to a bias towards Japanese banks within the set of foreign banks operating in the US.

5 However, the results in this study might be affected by the fact that most acquired banks in the sample were already restructured by government authorities, in order to raise the return from privatisation, and thus better run before privatisation compared to their domestic peers.

6 In a study on the determinants of international lending and borrowing activities of (foreign) banks Buch (2003) finds that information costs appear to be the main factor segmenting inter-national financial markets.

�0�0 ��

Entry Strategies

Following Buch (2000), we separate the decision on the country to enter from the decision concerning the mode of entry. In section 3.2.2 we focus on the choice between a newly established bank and acquisition and in section 3.2.3 we analyse the factors determining the organisational form of foreign representation, i.e. branch versus subsidiary.

3.2.2ModeofentryStudies on banks’ choices for a greenfield investment, i.e. a newly established bank or a de novo bank, or an acquisition are scarce. Still, this is an important decision as some studies suggest that the mode of entry matters for bank performance. Buch (2000) argues that under perfect information about future business conditions, costs of the two modes of entry should be the same, but that under uncertainty, a greenfield investment is likely to entail higher costs. Dell’ Ariccia and Marquez (2004) argue that the relative cost and information advantages determine whether a foreign bank can profitably enter a market as a de novo entrant. Unfortunately, empirical studies do not yield unanimous conclusions. Fries and Taci (2005) find that privatised banks acquired by foreign banks are more efficient than greenfields. In contrast, Majnoni et al. (2003) find that the profitability of foreign owned banks acquired through mergers and acquisitions is lower than the profitability of greenfields.

Konopielko (1999) describes the modes of entry in Hungary, the Czech Republic, and Poland and finds that minority interest and representative offices are not very popular. The most appealing vehicle of entry into Central Europe is through setting up a subsidiary, rather than through taking over an existing bank. However, this author does not offer an explanation for this finding.

Some studies identify the advantages and disadvantages of acquisitions versus greenfields. For example, greenfields avoid inheriting bad loans from the past (Pomerleano and Vojta, 2001), and post-acquisition integration failures rooted in cross-cultural differences and technological mismatches (Dikova, 2005). However, Pomerleano and Vojta (2001) argue that organic growth of a greenfield investment takes time and involves risks relating to unfamiliarity with local market conditions. In addition, entrants need to build up a reputation whereas they can potentially benefit from existing customer contacts when buying an existing bank (Buch, 2000). Additional advantages from acquisitions include access to valuable practices (Morosini, 1998, and Rugman and Verbeke, 1992), and access to local market knowledge (Wilson, 1980, and Harzing, 2002). With acquisitions, local practices can be difficult to implement into the organisation and communication problems may arise between the acquirer and the acquired firm due to cultural differences. At the same time, Pomerleano and Vojta (2001) argue that acquisition of a minority or majority interest requires internal restructuring and the transfer of staff and operational processes from the foreign parent bank to the acquired bank. VanTassel and Vishwasrao (2005) argue that foreign banks should prefer acquisition over greenfield investment since the former brings in valuable information accumulated from past credit operations. Acquisition should edge out greenfield investment even when the foreign bank faces ex ante difficulties

��

Chapter 3

��

in assessing the value of the portfolio belonging to a local acquisition target. On the basis of this review, our second hypothesis states:

Hypothesis 2. The choice for an acquisition (greenfield) is positively (negatively) related to the level of information asymmetry between the bank and its preferred customer in the host country, and is negatively (positively) related to the (in)direct costs of acquisition.

3.2.3OrganisationalformThe third component of foreign bank entry is the organisational form of representation. Foreign representation can be materialized by a small-scale office, such as a representative office or an agency, or by a large-scale office, such as a branch or a subsidiary.7

A representative office is the most limited but most easily established organizational form. It does not engage in attracting deposits and extending loans, but is generally established to test the possibility of further involvement (Goldberg, 1992). In theory, representatives have little more than a public relations function and are used in situations where regulations are present or profit opportunities are low because of low income per capita (Ter Wengel, 1995).8 An agency is a more costly form of foreign banking operation than a representative office and may be warranted if banks engage in very substantial export servicing and subsequent heavy involvement in the foreign exchange market (Heinkel and Levi, 1992, and Tschoegl, 2002). Representation with an agency also allows a bank to make commercial loans, although business related to consumer loans or deposits is not permitted (Goldberg, 1992).

A foreign branch constitutes a higher level of commitment than a representative office or agency. The crucial difference between a foreign branch and a foreign subsidiary is that, legally, a branch is a unity with its parent and a subsidiary is an independent legal entity. Other differences between branch and subsidiary regard supervision, risk and performance. While home country supervisors supervise branches, local supervisory authorities supervise subsidiaries. Subsidiaries are subject to local lending limits associated with the level of their capital, while for branches no local lending limits are involved as from a consolidated point of view, they rely on the capital of the foreign parent.

Based on a real option approach, Gulamhussen (2004) argues that banks use scale to rationally manage risk in their foreign market entry decision under uncertainty. Banks set up small-scale offices to get a foothold and large-scale offices to seek new customers and compete in foreign markets. Goldberg and Johnson (1990) pose that US banks wanting to expand with large operations should do so in countries with less restrictive regulation, high levels of FDI by US companies, large relative foreign trade, and lower levels of GNP per capita and domestic deposits. If

7 See Heinkel and Levi (1992) for an analysis of the competitive aspects of the different forms of foreign banking operations.

8 However, Konopielko (2000) notes that specialised banks such as investment banks will tend to operate through a number of representative offices.

��

Entry Strategies

it wishes to establish a new branch, the US bank should look for ease of regulation, high relative foreign trade levels, higher population, greater per capita income and lower levels of domestic deposits. Ter Wengel (1995) notes that the size of the host country is an important factor determining the form of representation. Banks from big countries tend to establish branches abroad unless the host country is also large and similarly competitive, or unless the host country has regulations that hinder the establishment of branches. Subsidiaries are employed as an investment alternative in countries with liberal capital regulations and high income per capita. Miller and Parkhe (1998) argue that host countries influence the organizational form that U.S. banks adopt: countries with high tax rates and explicit barriers to subsidiaries have a smaller share of subsidiaries. Similarly, Cerutti et al. (2005) conclude that banks are more likely to operate as branches in countries that have higher corporate taxes and when they face lower regulatory restrictions on bank entry, in general, and on foreign branches, in particular. Subsidiaries are the preferred organizational form by banks that seek to penetrate the local market establishing large and mostly retail operations. Summarizing the evidence on the choice between branch and subsidiary, we formulate the hypothesis that:

Hypothesis 3. If the organisational form of foreign bank representation is not exogenously determined by local regulation, the choice for the branch (subsidiary) format is positively (negatively) related to the focus on corporate banking in the host country.

3.3 Methodology and data3.3.1MethodologyWe consulted heads of foreign parent banks, local foreign owned banks and financial supervisory authorities. In structured interviews the following topics were addressed: (i) What were your main considerations to enter the host country? (ii) Why did you choose to take over an existing bank and not to establish a greenfield establishment (or the question mutatis mutandis)? (iii) After a take-over, which significant changes did you implement? (iv) Which were the main, if any, obstacles you encountered when entering the host country?

The method of structured interviews overcomes the disadvantages of a single case study or a questionnaire with a list of preset possible answers. The results of a single case study are often difficult to validate, while a questionnaire with preset answers runs the risk that relevant issues are not adequately addressed. With structured interviews we are able to identify all factors related to foreign bank entry, the motives underlying the mode of entry, and the choice for the organisational form of foreign representation. Triangulation was accounted for in two ways. For the internal validity of the answers obtained at the foreign parent banks, we interviewed heads of the local foreign owned banks. Second, we checked the information that we received in interviews with local financial authorities. Consultations took approximately 1 – 1½ hour and are recorded and transcribed. Appendix 10 specifies all banks and

��

Chapter 3

��

supervisory institutions consulted, the representatives with whom we conducted the interviews and their position. Banks and financial authorities were asked to review the first draft of the paper.

3.3.2DataOur sample consists of Western foreign parent banks with operations in European transition nations. Table 3.2 presents the selected parent banks and their presence in CEE. The sample includes the four largest international banking groups in the region: ERSTE Bank, KBC, Raiffeisen International and Bank Austria Credit Anstalt (BACA).9 Austrian banks are the leading international banks in CEE. Together with Belgian KBC their market share amounts to 43 per cent of international banking assets in CEE (BACA, 2005). ERSTE focuses on consumers and SME companies. Belgian KBC is the second largest international bank in CEE and combines banking and assuring. The bank focuses largely on private persons and small and medium-sized enterprises in Belgium and in Hungary, Poland, the Slovak Republic, Slovenia, and the Czech Republic. KBC is the second largest investor according to deal value (ECB, 2005). Raiffeisen Group is the direct owner of Raiffeisen–RBHU holding, in which all CEE network banks are assembled.10 Of its assets in transition countries, Raiffeisen International has 52 per cent of Central Europe and 38 per cent in South Eastern Europe and countries of the Commonwealth of Independent States (CIS). BACA is a universal bank and market leader in Austria. Within the HVB Group, BACA is responsible for business development in Austria and in CEE. Through the HVB Group, the bank is also represented in Russia, Ukraine, and the Baltic countries.

We supplemented the data from the four largest international banking groups in the region with those of two banks in the top 15 of largest European banks. ING Bank offers services in over 50 countries and provides life insurance in Central Europe. ABN AMRO is an international bank with European roots and a focus on consumer and commercial banking, supported by an international wholesale business. ABN AMRO has branches in more than 60 countries.

Parent banks in our sample cover 56 per cent of all international bank assets in CEE (RZB, 2005).11

Finally, we added the largest foreign banks in the Baltic States: SEB, Swedbank, and the smaller Sampo. Swedbank is one of the largest banking groups in the Nordic

9 Bank Austria and Creditanstalt separately built up their networks in Eastern Europe, begin-ning in 1990-1991. In 1990 Zentral Sparkass Länderbank took over Bank Austria. In 1997 this combination bought Creditanstalt. BACA is fully owned by Bayerische Hypo- und Vereinsbank (HVB). In November 2005, Italy’s second largest bank, Unicredito Italiano, offered €16bn in stock for German HVB Gruppe.

10 Raiffeisen Group intends to sell 49 per cent of the holding for further expansion in the region.11 RZB defines CEE as the Czech Republic, Hungary, Poland, Slovenia, Slovak Republic, Alba-

nia, Bosnia & Herzegovina, Bulgaria, Romania and Serbia. Our sample covers most of bank assets in the Baltics as well.

��

Entry Strategies

area and in Sweden it is the second-largest real estate agent and lender for housing. The bank’s share of the corporate market is 25 per cent when it comes to deposits, lending, leasing and financing. Its subsidiary, Hansabank, offers a complete array of banking services in the Baltic States. SEB is a North European bank for corporate customers, institutions and private individuals with home markets in the Nordic and Baltic countries, Germany, Poland, and the Ukraine. Sampo has specialised in investment and savings. The Group comprises the insurance company If and Sampo, specialised in long-term savings.

Table 3.2 Selected parent banks

Parent bank Home Country

Assets (€ bn)

CEE Rank

Present in:

ERSTE Bank Austria 139.7 1 Czech Republic, Slovenia, Croatia, Hungary, Serbia.

KBC Belgium 249.2 2 Czech Republic, Slovenia, Slovak Republic, Hungary, Poland.

RaiffeisenInternational

Austria 62.8 3 Albania, Belarus, Bulgaria, Bosnia, Czech Republic, Croatia, Hungary, Lithuania, Poland, Romania, Russia, Serbia, Slovak Republic, Slovenia, Ukraine.

HVB-GROUP/BACA

Austria 146.5 4 Bulgaria, Bosnia, Czech Republic, Croatia, Hungary, Lithuania, Poland, Romania, Russia, Serbia, Slovak Republic, Slovenia.

ING Bank Netherlands 866.1 10 Bulgaria, Czech Republic, Poland, Romania, Russia, Slovak Republic, Ukraine.

ABN AMRO Bank Netherlands 608.6 15 Czech Republic, Poland, Romania, Russia.

SEB Sweden 176.4 >16 Estonia Latvia, Lithuania, Poland, Russia, Ukraine.

Swedbank (FSB) Sweden 113.1 >16 Estonia, Lithuania, Latvia, Russia.

Sampo plc. Finland 19.7 >16 Estonia, Latvia, Lithuania.

Banks are ranked based on CEE Rank. CEE rank is based on total assets in CEE, weighted by ownership share. Source: RZB (2005). Total assets and CEE markets covered are based on 2004 data. Between 2003 and 2006, ABN AMRO withdrew from Bulgaria, Hungary, and the Ukraine.

For triangulation purposes we consulted heads of key local banks in the region. Descriptive statistics of these selected banks are shown in Table 3.3. Belgian KBC owns the largest bank in the Czech Republic, ČSOB, with a market share of 19 per

��

Chapter 3

��

cent. KBC recently bought the remaining shares of ABN AMRO in K&H Bank, Hungary’s second largest bank. Austrian ERSTE Bank owns the second largest Czech bank: Česká Spořitelna with a market share of 18 per cent. In Estonia we consulted heads of the key local banks, with a joint market share of 92 per cent. Since 2005, Hansabank is fully owned by Swedbank and was delisted from the Estonian stock market, while Eesti Ühispank and Sampobank are fully owned by SEB and Sampo, respectively. Representatives of the global parent banks ING and ABN AMRO were consulted in the Czech Republic, Hungary and Poland. Table 3.4 summarizes the profile of the foreign owned local banks. Appendix 11 illustrates some of the entry strategies in the 1990s of foreign parent banks we consulted. The strategies are described by the entry determinant, the entry mode and the organisation form of the initial establishment. The last column in the Appendix describes the sequence of events after the initial entrance with respect to mergers, acquisitions or additional shares that the foreign parent bought in its foreign subsidiary. We see that most of the greenfields in our sample have been merged. The sample also suggests that joint ventures eventually do not survive in their organisational form. Third, the sequence of events in this sample suggests that foreign banks prefer to ultimately extend their ownership in their local subsidiary to 100% and that the delisting of the local bank often accompanies this strategy.

Table 3.3 Selected foreign owned local banks

Foreign owned local bank

Country Assets(€bn)

Rank MarketShare

Foreign Parent

AssetRatio

ČSOB Czech Rep. 20.1 1 19% KBC 8.1%

Česká Spořitelna Czech Rep. 18.7 2 18% ERSTE bank 13.4%

Bank BPH S.A. * Poland 13.2 3 10% BACA 9.0%

ING Bank Śląski S.A.* Poland 8.6 4 7% ING 1.0 %

Hansabank-Hansapank Estonia 8.2 1 59% Swedbank 7.3 %

K&H Bank Hungary 7.0 2 10% KBC 2.8 %

Kredyt Bank S.A. * Poland 5.3 8 4% KBC 2.1 %

HVB Bank CZ** Czech Rep. 4.7 4 5% BACA 3.2 %

Raiffeisen Bank Rt.** Hungary 4.3 6 7% Raiffeisen 6.9 %

HVB Bank Hungary** Hungary 3.8 7 6% BACA 2.7 %

SEB Eesti Ühispank Estonia 2.2 2 26% SEB 1.3 %

ING Bank Rt. ** Hungary 1.2 11 n.a. ING 0.1 %

AAB (Polska)** Poland 1.1 20 n.a. AAB 0.2 %

A.S. Sampo Pank Estonia 0.6 3 7% Sampo Plc 0.0 %

AAB CZ** branch Czech Rep. n.a. n.a. n.a. AAB n.a.

Sources: Annual Reports, BACA (2005) and RZB (2005), per end 2004. Notes: Country rank (Rank) and market share are based on bank size. Asset ratio is defined as the ratio of total assets of the local bank and total assets of the parent bank. AAB refers to ABN AMRO. * Listed at stock market. ** Greenfield investment.

��

Entry Strategies

3.4 Results3.4.1Determinantsofforeignbankentry Table 5 summarizes our results on the determinants of entry using an organizing scheme similar to Table 1, suggesting that the expected profit growth in the host country in CEE has been the primary motive to enter the region. For half of the parent banks, the tight margins due to competition at home were also an incentive to enter this region.

Table 3.5 shows mixed evidence for the follow the customer argument. Dutch banks went abroad in order not to lose their global clients to other banks. Similarly, German HVB Bank serviced the subsidiaries of German clientele in European transition countries. However, the entry of Swedish banks to the Baltic States was not primarily motivated by the follow the customer argument. Swedish financial supervisory authorities confirm that for Swedish banks the attractiveness of the market was the key entry determinant. Similarly, Austrian ERSTE Bank was not primarily motivated by the conventional defensive expansion, although for its (former) domestic competitors Creditanstalt and Raiffeisen mixed evidence is available. Other arguments related to the activities of the non-financial sector that played a role in entering the region are: (i) supporting home country trade companies, and (ii) advising firms how to buy state-owned companies that were privatised.

Information costs related to proximity and cultural factors appear of secondary importance for banks that entered Central and Eastern Europe. However, for Swedish banks going to the Baltic States, the historical ties between Sweden and Estonia and the geographical closeness have been important. In this respect, SEB mentions the Hansa-culture. For some Austrian banks the same argument holds. BACA indicates that the reason to enter the Czech Republic and Hungary first was related to the common history with these neighbouring countries. Being an Austrian bank is thought to be advantageous in getting in touch with the local people in Central Europe. Czech HVB CZ has a similar line of reasoning as the bank claims to understand people’s mentality in these countries better than other foreign banks. Several banks mention the changing institutional context, induced by accession to the European Union (EU), as an important entry determinant. Especially the late enterers argue that the expectation of European transition countries to join the EU was an additional motive to enter. Belgian KBC mentions the improved governance and the adjustment and modernisation of the law as important considerations to enter. Likewise, Austrian ERSTE indicates that EU accession played an important role in its decision to enter the region. Raiffeisen International argues that in the early 1980s, the political situation in Hungary was more relaxed than in other countries, making the country the most attractive one to enter. Similarly, a worsening institutional context has induced foreign banks to leave a country.

��

Chapter 3

��

Table 3.4 Profile of foreign owned local banks, per end 2004

Bank Profile

ČSOB Československa Obchodni Banka keeps the position of the strongest, most stable and most powerful universal bank in the Czech Republic. The bank was sold in privatisation to KBC. Later ČSOB acquired Czech IPB. ČSOB also operates in the Slovak Rep. It is the largest bank domiciled in the CEE, measured by assets.

Česká Spořitelna Česká Spořitelna is geared towards small-scale clients, small and medium-sized companies, towns and municipalities, but also plays an indispensable role in the financing of large companies and corporations

Bank BPH S.A. Bank BPH SA is a Polish universal bank. Bank BPH intends to be the first choice bank for the affluent and middle class customers and small enterprises; “a preferred partner” for medium and large domestic and international companies; a leader in real estate financing and a leader in brokerage and asset management.

ING Bank Śląski The principal objective of ING Bank Śląski S.A. is to provide integrated financial services and to be a customer-oriented bank. The strategy for development of retail banking provides for offering integrated financial services, enhancement of cross-selling and prudent approach to risk management.

Hansabank Hansabank is a universal bank, primarily focused on medium-sized companies and private individuals with above average purchasing power. Home market is the Baltics, but niche segments in Russia are served as well.

Kredyt Bank S.A Universal bank. Established in 1990 as one of the first private banks in Poland. Kredyt Bank is developing the bank assurance concept with insurer Warta.

K&H Bank K&H Bank is one of the largest banks in Hungary, the leader of the corporate banking market and the 2nd strongest bank in the retail segment, offering a full range of financial services through a network of 160 branch offices, various business lines, and subsidiaries.

HVB CZ S.A. HVB Group intends becoming the most attractive European financial institution for private and medium-sized corporate customers. Within the HVB Group, Bank Austria Creditanstalt is responsible for business in Austria and CEE.

Raiffeisen Bank Hungary Rt.

Raiffeisen International acts as holding company for Raiffeisen Bank Hungary. It is majority-owned by Raiffeisen Zentralbank Österreich AG (RZB). In the first years of its operation, the bank offered services primarily to large corporations. In 1994 the bank started to focus on municipalities, in 1996 on private banking customers, in 1999 on retail customers and in 2001 on SME.

��

Entry Strategies

SEB Eesti Ühispank SEB Eesti Ühispank is a universal bank that provides its services to private persons, corporate customers and the public sector in Estonia. The bank offers all financial services available on the Estonian market and leasing services in Russia and supports the customers of the SEB Group in investing assets in CEE.

ING Bank Rt. ING Bank Rt.’s corporate and investment banking business lines primarily cater to subsidiaries of multinational companies, leading large Hungarian companies and dynamically growing developing medium sized companies.

ABN AMRO (Polska) S.A.

ABN AMRO Bank (Polska) S.A. focuses on the corporate market and targets multinationals and major Polish companies, companies doing business with Poland and joint ventures between Polish and international companies.

HVB Bank Hungary HVB Bank Hungary is a universal bank. In 2003 the bank merged with a leading broker house. The market position of the bank was thus far primarily dominated by the corporate and mainly large corporation sector. In the future the bank wishes to open up completely to SME in Hungary and private customers.

Sampo Estonia The strategic goal of Sampo Estonia is to become a leading and the most professional financial group, offering LT savings and investment products.

ABN AMRO CZ ABN AMRO Bank focuses on providing in the fields of corporate and investment banking for selected corporate, institutional and public sector clients. The bank also provides its services through ABN AMRO Asset Management.

Source: Annual reports and web sites. Banks are ranked by total assets.

�0

Chapter 3

�0 ��

Table 3.5 Determinants of foreign bank entry, summary of the results

Support No Support

Defensive expansion

Follow the customer 30 % 40 %

Follow other foreign activities of existing client 20 %

Expected relative profitability

Expected profit growth 70 %

Competitive pressure at home 40 % 10 %

Bank specific advantages 30 %

Information costs and institutional context

Information costs 40 %

Institutional context 30 %

Other determinants

Herd behaviour (reputation, status) 60 %

Wait and see >30 %

Circumstantial factors 20 %

Results are presented as percentages of the foreign parent banks in the sample that relate to the particular entry determinant. Despite their merger in 2001, Bank Austria Credit Anstalt and HypoVereinsbank are included as separate foreign banks.

For instance, Belgian KBC withdrew from the Ukraine due to the corrupt political system. Similarly, Sampo left Latvia because of the problems with the Russian minority and the relatively high level of corruption. A Swedish parent bank adds that in Latvia more than half of the deposits are non-resident deposits and banks are faced with the dilemma of not knowing what kind of money they are taking care of. However, most foreign banks consider that they are able to deal with the relatively poor quality of the institutions in CEE.1

Apart from entry motives as examined in previous studies, we identify other explanations of foreign bank presence in CEE (see Table 3.5). First, we find that

1 However, banks were often critical about the institutional context. For example, in some transi-tion countries banking supervision is extremely strict with respect to provisioning. Regulations in Poland, for example, are so strict that banks have to classify loans as doubtful while they are convinced that the loan will be repaid. Supervisors sometimes lack experience as they offer too low salaries and end up with a too junior staff. Also the inefficiency of the legal system is often mentioned by our respondents. In the mid 1990s, it took banks in the Czech Republic years to realize collateral on real estate. Consequently, cases were dealt with via out-of court settlement. According to ERSTE, new legislation in the year 2000 left the banking sector with the opposite situation: current legislation is said be very creditor friendly and open to abuse, although the Czech National bank strongly disagrees with this statement. In Poland, bank operations were hampered by legal uncertainties and the changing legal environment (including laws about fo-reclosures and bankruptcy). Sometimes bankruptcy administrators demand 30 per cent of the claim for their costs. Corruption of bankruptcy administrators still exists.

�0�0 ��

Entry Strategies

bank reputation and status triggered various banks to go abroad. A Dutch parent bank confirms that Dutch banks inspired each other to enter CEE. Similarly, Finnish authorities mention that NORDEA’s policies motivated a domestic competitor to go abroad. Likewise, the Austrian authorities argue that the positive experience of Austrian Raiffeisen has inspired ERSTE to enter CEE. Early entrant German HVB was driven to the region as this was considered in line with the status of the bank. A Dutch global bank confirms that status played a role since bank management, faced with internal limits to growth, preferred extending operations abroad instead of paying dividend.

Related to the real option approach of Gulamhussen (2004), we find that many banks wanted to test the possibility of further involvement. This motive was especially relevant for early enterers. In Hungary, Austrian Raiffeisen participated in 1986 in the joint venture Unicbank to ‘test the water’ rather than for strategic reasons. Later on, due to the positive experience, Raiffeisen decided to buy all the shares in the joint venture. Czech authorities argue that many foreign banks started operating in the beginning of the nineties, at first only monitoring the market. Real activities only started in the middle of the nineties. For example, ABN AMRO entered the Czech banking market in 1991 mainly to explore the market to investigate whether sufficient business opportunities were available. In Poland, ABN AMRO participated in 1991 with Credit Lyonnais and Dresdner Bank in the joint venture International Bank in Poland with the motive to share the risk and to wait and see how the joint venture would develop. Similarly, ABN AMRO is doubtful that FDI decisions of all foreign banks were fully rational and feels that that many banks just wanted “to go with the flow, to take a chance to see whether business would turn out favourable”.

Finally, foreign presence is sometimes affected by specific circumstances. For example, when Dutch ING bought Barings Bank for £1 in 1995, the assets included a Polish branch network. A second example in which unforeseen factors explain foreign bank presence is the battle for Hansabank in Estonia. In 1996, Swedish Swedbank acquired 33 per cent of Estonian Hoiupank in privatisation. When in 1998 Hoiupank merged with Hansabank, Swedbank suddenly found itself on the same ship with compatriot SEB and a bidding war began. Ultimately, Swedbank gained majority control of Hansabank and now holds 60 per cent of the shares in Hansabank.2

3.4.2 Determinants of greenfield vs. acquisitionThe choice between acquisition and greenfield is influenced by many time-varying factors, like the number and attractiveness of banks available for possible acquisition. Figure 1 is a stylised representation of our evidence on the banks’ main considerations

2 Another motive to enter a foreign country that was occasionally mentioned in the interviews is the so-called transit strategy. Swedish authorities hypothesize that, although never mentioned by the Swedish banks themselves, the latter have entered the small Baltic States to acquire the local knowledge of dealing with Russian customers, to be able to enter the Russian retail market as soon as this becomes possible. Likewise, one CFO mentioned that all banks are looking east-wards.

��

Chapter 3

��

when deciding for a greenfield investment or an acquisition. Panel A relates the probability of the choice for a greenfield to the relative price of

an acquisition. The costs of an acquisition are the sum of the direct purchasing costs and the post-acquisition expenses.3 The former depends on (i) the price quoted and (ii) potential competitive biddings by other banks.4 Post-acquisition expenses are related to (i) reviewing the loan portfolio and costs due to mistakes in estimating the quality of the loan portfolio5 and (ii) the restructuring and integration of the subsidiary in the parent bank.

Purchasing costs played an important for some banks. For instance, Bank Austria/Creditanstalt invested in greenfields because the bank considered prices to be paid during privatisation too high. In the Baltic States, acquisition became more attractive during the Russian crisis of 1998, as listed banks became cheaper. In 1998, Swedish SEB acquired 32 per cent in Estonian Eesti Ühispank, 32 per cent in Latvian Unibanka and an additional 12 per cent later that year, and 42 per cent in Lithuanian Vilniaus Banka. Similarly, HVB entered Central Europe through greenfields but made an acquisition in Poland: “Poland was the big exception; it was a question of price and the uncertainties about the legal nature of all these issues and how far it would be possible to outsource or throw out bad loans. The 1998 disaster made our chairman a little bit scared about a new adventure. The price in Poland was more reasonable than what was quoted for the former state banks in the Czech Republic, Hungary and the Slovak Republic.” Belgian KBC adds that sometimes banks are restricted in their acquisition abilities and that it therefore acquired only a minority acquisition in 1996 of Polish Kredytbank, which was at that time already privatised.

Figure 3.1 Determinants of the mode of entry, stylised results

P(greenfield) denotes the chance of establishing a greenfield, relative to a foreign acquisition

3 Late entrants were often restricted in their means. Only after the merger in 1998, Belgian KBC was able to bid in acquisitions. The fact that Austrian’s ERSTE is not present in Poland is due to the fact that banks were too expensive at the time.

4 In addition, the price demanded may depend on ownership share.5 The price of an unexpected bad loan portfolio component is related to the quality of the local

judiciary as lengthy legal procedures with ex ante unknown outcome raise the costs seizing the assets of debtors that default on their loan.

��

Entry Strategies

Post acquisition expenses of a greenfield are generally lower than those of an acquisition as one can start with a clean loan book, a homemade structure, and an experienced screening staff. The Czech authorities confirm that the credit portfolio problems of the former state-owned bank in the early 1990s were the main reason that only few parties were interested in buying these banks. Austrian Raiffeisen firmly believes that it is always safer to build up what is started from scratch than inheriting skeletons in the cupboard and therefore usually applies the greenfield concept for its foreign banking strategy.6 Post-acquisition costs of an acquisition include the costs of integrating the subsidiary in the structure of the parent bank. Integration implies the implementation of a similar IT-system and the need to implement best practices in the newly acquired bank, like risk management techniques. For example, German HVB (BACA) sent 200 employees from the head office to Polish BPH, who were working for more than a year on the IT-systems. Then 3,500 employees were sent over for nine months to ensure the proper transfer of data.

Implementing best practices can be done at lower costs in greenfields than in acquisitions. Foreign banks usually try to build up new management teams. One Austrian foreign parent bank recalls that this process took a while. Foreign parent banks need to establish trust within the acquired organization and this can take up to two years. Furthermore, as foreign banks initially had difficulties finding skilled employees, they set up exchange programs to transfer knowledge. ERSTE Bank recalls that the transfer program in which employees of the subsidiary were transferred to the head office in Vienna cost the bank a lot of money, but that it did pay off. In the early transition period, the lack of skilled labour in the host countries was evident; it was hardly possible to get anyone who was a skilled accountant or risk manager according to Western standards.7 The situation improved with the first generation students coming from local universities.

Despite cost-advantages of greenfields, foreign banks regard the operational risks of greenfields higher than the operational risks of acquisitions. Except for global banks and ING and ABN AMRO, most foreign banks therefore regard greenfield investments as more costly than acquisitions and argue that a greenfield investment has not been considered, because of the time it takes for such an establishment to develop.8

A second determinant of the choice for a greenfield or an acquisition is the degree

6 According to the Austrian central bank OeNB, a disappointing experience with an acquisition in beginning of the nineties made Raiffeisen confident to continue firmly in their initial ap-proach of greenfields. The exception is the acquisition of Prior Bank, the second largest bank in Belarus, of which Raiffeisen is majority owner.

7 Several banks also mentioned the lack of commercial attitude of bank employees, who used to grant credit on personal rather than on commercial grounds.

8 ING Hungary regards organic growth to be less expensive than buying a bank, especially in light of the battle on the retail market and ABN AMRO considers local acquisitions more costly due to goodwill.

��

Chapter 3

��

of competition in the local banking sector (Figure 1, panel B). In addition, the level of home competition is related to the probability of the choice for a foreign greenfield. For example, the Bank Austria Credit Anstalt Group mentions that in the second part of the 1990s the bank bought Polish PBK as it was lagging behind compatriot Erste Bank in retailing. Local financial authorities in Estonia refer to the high level of competition in the 1990s when local banks were already copying Western banking practices to explain the lack of foreign greenfields in Estonia. Estonian domestic banks had introduced relatively good management, accounting systems and risk management, and had created a banking sector in which greenfields would not have been able to survive. Polish authorities confirm this point of view, stressing that in recent years more and more foreign banks became aware that it is very difficult to expand with a greenfield investment. Hungarian authorities mention the 1996 greenfield of Dutch Rabobank that aggressively tried to get business, but which was not able to gain enough market share and had to cease banking business in 2002. Similarly, Société Générale first participated in a 1988 joint venture CE International Bank (CIB) but sold its shares in CIB in 1997. It showed up in 1998 with a commercial bank, which was then restructured into an investment bank but closed in 2002. Both banks entered at the end of the nineties at a stage where the corporate market was already divided. HVB Hungary summarizes: “They came by far too late. […] there are some latecomers that are successful but not on the corporate side, but in retail.” Similarly, the Austrian central bank OeNB argues that a greenfield is perhaps particularly useful if you are a first mover, referring to Raiffeisen International as this bank started greenfield investments and built up these services very early. Hungarian K&H Bank also relates the moment of entry to success, arguing that first enterers are better off.

Third, we find evidence that the choice between greenfield and acquisition is related to the customer focus of the banks (Figure 1, Panel C). Czech authorities argue that greenfield investments in the form of branches are not suited for retail activities. Belgian CSOB argues that with very demanding clients a Belgian brand name may be needed in order to successfully undertake corporate and private banking with a limited number of offices. Stated differently, corporate and investment banking is usually done from branches with the foreign parent bank name as a brand. HVB Czech Republic mentions that in choosing between greenfield and takeover one has to separate between corporate business and retail business. HVB Hungary confirms this by arguing that the customer base for all greenfields is always focused on corporate customers: “Some banks started servicing large firms and national corporations, others focused on servicing medium sized enterprises. We started with large and gradually went down to medium and even smaller sized companies.” Belgian CSOB argues: “A greenfield retail bank, who does that nowadays? Although Raiffeisen Hungary is the best performing greenfield, so it is possible. In retail banking you need a market share of 15-20 per cent to be able to be cost-effective”. On the question of why not having opted for an acquisition, ABN AMRO argues that it applies the greenfield format in Central Europe since the bank has no retail strategy. Belgian KBC notes that with retail banking an existing infrastructure is needed in order to be cost-effective within

��

Entry Strategies

a reasonable time. The bank claims to have a comparative advantage as a bank-insurance group with the special feature of not only selling life insurance through the banking network, but also indemnity insurance. Bank Austria/Creditanstalt started a greenfield operation with a representative office and built up the corporate banking side, to go with Austrian companies that exported into the CEE regions as well as to support and to service the local firms. As it aims for a universal banking approach, the bank operates different business models enabling it to acquire local banks for retail expansion (e.g. PBK in Poland, Biochim in Bulgaria, and Splitska Banka in Croatia).

3.4.3TheorganizationalformWe find that the choice for the organizational form, i.e. branch or subsidiary is largely influenced by regulations. Since 1992, the general policy of the Polish authorities is to support subsidiaries and not to permit foreign branches. However, ING mentions that differences between central banks exist as other central banks allow for subsidiaries, although some insist that the subsidiary should be capitalized. ABN AMRO circumvents the legal lending limits inherent to a subsidiary by booking most of the loans extended in Poland offshore in Amsterdam. ABN AMRO Poland prefers to be a branch because as it foregoes costs associated with the reporting of board meetings, supervisory board meetings, and the Board being once a year in Poland. In Hungary banks are also obliged to take the form of a subsidiary, so that a local lending limit of 25 per cent of capital is a restriction. Local regulators want to be in charge of oversight of the operations because of the strong competition in the retail sector.

Local supervisory institutions may force foreign banks to keep their acquired subsidiaries listed on the local stock exchange or restrict banks in biddings for banks in privatisation. For example, when buying Bank Slaski, the Polish authorities asked ING to make sure that part of the shares was freely floating. When ING owned more than 75 per cent in ING Bank Slaski, the parent had to sell shares on the Polish stock market. In the Czech Republic, foreign bank KBC could not bid for less than 80 per cent of the shares of CSOB. Although the Czech central bank has no preference in this regard, KBC wants to list CSOB on the local stock exchange to increase the bank’s local exposure and reputation. However, ING doubts whether listing enhances reputation and adds that banks usually want 100 per cent of the shares in order to avoid agency problems. In Slovenia, the authorities forced Belgian KBC in 2002 to buy only up to a maximum of 34 per cent in Nova Ljubljanska Banka.

Finally, local supervisors emphasized their concern regarding the common European passport, which makes it easy for banks to establish a branch in any EU country. This possibility could support the idea of either spreading the activities of EU licensed banks or changing the format of subsidiaries, regulated and supervised by host authorities, into branch structures, which are regulated and supervised by home authorities. For example, in 2006 Dutch Fortis Bank and Bank of Tokyo-Mitsubishi UFJ (Holland) N.V. opened new branches in the Czech Republic while the Credit Lyonnais Calyon subsidiary was changed into a branch.

��

Chapter 3

��

3.5 Summary and conclusion Building on the conventional hypothesis that banks follow their customers abroad, this chapter examines foreign bank entry strategies in European transition countries. The main novelty of this chapter is that it integrates the motives for entering Central and Eastern European transition countries, the moment of entry, the mode of entry, i.e. acquisition versus greenfield investment, and the organizational form, i.e. branch or subsidiary.

Due to data unavailability a joint project with the Dutch central banks has been set up to interview numerous bankers and staff of financial supervisory authorities. The sample of bank head offices consisted of the four most important foreign banks in this region, i.e. (i) the Austrian ERSTE Bank, Raiffeisen International, HVB-Group/BACA and the Belgian KBC, (ii) two big global banks: Dutch ABN AMRO and ING Bank, and (iii) two big Scandinavian banks present in the Baltic States: Swedbank and Skandinaviska Enskilda Banka (SEB). Furthermore, bankers of 15 subsidiaries and staff of seven financial supervisory authorities in the region have been interviewed.

The analysis shows that for the majority of the banks high profit expectations were a key determinant. This finding is in line with the result of the ECB (2005) and with country studies arguing that banks start operations in transition economies to search for new business opportunities (Dubauskas, 2002; Florescu, 2000; Kraft, 2004).

Some banks indicate that they entered Central and Eastern Europe to follow their clients. Usually the banks that followed their clients were the big global banks that were the avant-garde of the foreign banks that entered. Next, in contrast with Galac and Kraft (2000) and Konopielko (1999) this study provides evidence that, in explaining foreign bank presence in Central and Eastern European transition economies, for some banks competition in the home country is more important than following home country clients. Especially, the big regional foreign banks report that the high competition in the home country played an important role in entering the transition countries in Central and Eastern Europe. The reargard of foreign banks indicate that for them the progress in the institutional framework was an important factor in going abroad: due to the prospect of EU accession governance improved and laws were modernized. The fact that foreign banks at first entered Hungary is primarily due to the liberal political situation in that country. Some banks indicate that a worsening institutional context is a reason to leave a foreign country.

In the interviews other, new reasons for foreign bank entry have been indicated that have not yet been mentioned in the literature. The majority of the foreign banks report that reputation and status have been reasons to follow competitors to Central and Eastern Europe. This so-called peer pressure occurs between global banks as well as between the big regional players. Related to this is the finding that for some banks going with the crowd is important: sometimes it is even more important to join the competition than the results of an NPV analysis based on expected revenues, costs and risk. A third newly documented consideration for foreign bank entry mentioned by a financial supervisor is best described as the stepping stone strategy. This strategy

��

Entry Strategies

implies to start foreign banking in country B to get accustomed to a local market that is similar to that in country A. We assume that country B and country A are culturally and possibly geographically closely related and that the institutional environment in country A does not allow for foreign banking yet. The aim is to develop a comparative advantage vis-à-vis other foreign banks, which can be exploited as soon as a change in the institutional environment in country A induces the foreign bank to enter.

Regarding the choice between establishing a de novo bank or to acquire an existing bank, the chapter finds the following. Global banks prefer to set up new offices, the so-called greenfields, usually taking the form of branches or joint ventures in cooperation with other foreign banks. Banks that do not exclusively focus on corporate banking prefer to acquire existing banks. The experiences of banking in Central and Eastern Europe show that banks’ choice for establishing greenfield becomes less attractive as competition increases. Next to customer focus and the level competition a third variable influences the mode of entry. The choice for a greenfield also depends on the costs of an acquisition and with that the costs of integrating the acquired bank within the parent bank. These costs primarily exist of standardizing risk management systems, reporting and the information technology (IT). Another cost of acquisition that specifically applies to acquisitions of former state owned banks is the information asymmetry regarding the amount of bad loans that come as a legacy from former banking practices.

Regarding the organizational form of foreign representation we find local regulation, customer focus and, again, the time of entry to be the main determinants. In an early stage, banks employ representative offices, joint ventures and branches focusing on corporate banking. However, the sequence of events in Central and Eastern Europe shows that joint ventures of foreign banks set up to share risk, rarely survive in their original shape. In line with Cerutti (2005), we find that the focus on retail clients is related to the subsidiary format. In addition, we find that competition in the banking market is positively related to the use of subsidiaries. However, some local regulators do not allow for the branch format due to supervisory issues.

This chapter finds evidence for a new ‘pecking order theory’. First, the avant-garde of foreign banks that entered in the early 1990s focused primarily on existing clients from the home country that went abroad and on the big local companies. Small and medium enterprises were served only later because of the increase in (foreign) banks supplying corporate banking and the diminishing margins in this segment. Also the privatisation in the second part of the 1990s of former state (savings) banks to foreign banks induces the latter to start retail banking. Due to these developments, at the end of the 1990s banks had to diversify their customer focus to survive in the changing banking sector. With these conclusions this chapter finds the explanations for the initial stunted intermediation reported in chapters 1 and 2.

Future research should examine which entry strategy, i.e. which combination of entry motives, moment of entry, mode of entry, customer focus and legal representation optimises bank profitability and efficiency. Chapters 4-6 start by examining empirically the effect of foreign ownership on bank performance.

��

��

Part IIForeign Ownership and Performance

�0�0 ��

�0�0 ��

Chapter 4Financial Performance:

Foreign Banks in European transition countries *

4.1 IntroductionChapters 4-6 in Part II examine the effect of foreign ownership on bank performance. Existing literature on the relationship between foreign ownership and bank performance consists of two strands. One strand connects ownership to financial indicators such as profit, overhead costs, net interest income, return on assets (ROA) or return on equity (ROE). Chapter 4 and 5 will follow this strand and analyse the effect of foreign ownership on three financial indicators (i) net interest revenues, (ii) overhead costs (iii) profit before taxes. Chapter 4 focuses on the performance of foreign banks in Central and Eastern Europe. Chapter 5 extends the analysis by examining the performance of foreign banks worldwide. A second strand in the performance literature researches determinants of bank efficiency. Chapter 6 adds to this literature by examining the effect of foreign ownership on bank efficiency with a specific focus on the role of the institutional context.

Berger et al. (2000) have formulated three hypotheses on the different impact of ownership on bank performance.1 First, the home field advantage hypothesis predicts that domestic owned banks are more profitable due to the absence of the structural agency costs that foreign owned banks are confronted with. Distance between the principal (the parent bank in the home country) and the agent (the subsidiary or branch in the host country), creates a cost for the foreign bank from operating or monitoring the subsidiary or branch from a distance. Other factors leading to a comparative advantage for domestic banks are differences in language, culture, and regulatory and supervisory structures between the home and the host country of the foreign bank. The general global advantage hypothesis predicts foreign owned banks to be more profitable due to some comparative advantages that domestic owned banks lack. These advantages stem from advanced technologies, more efficient organisations due to stiff competition in the home market, a more active market for corporate control and better access to an educated labour force with the ability to adapt to new technologies. The limited form of the global advantage hypothesis states that domestic banks are more efficient than foreign banks in most foreign countries, that domestic banks

* Based on Naaborg, I.J. and B.W. Lensink, Banking in transition economies: a study into foreign ownership and performance, an article forthcoming in European Journal of Finance.

1 Berger et al. (2000) have formulated their hypotheses in terms of the efficiency of banks. In this chapter and the next, I test the hypotheses for alternative performance indicators such as profi-tability.

��

Chapter 4

��

may be equally efficient as foreign banks from some countries, but that domestic banks may be less efficient than foreign banks from one of the foreign countries, i.e. the United States (Berger et al. 2000).

Demirgüç-Kunt and Huizinga (2000) find foreign ownership of banks to have a significant impact on banks’ spreads and profitability. In their sample of banks worldwide, foreign owned banks realise higher interest revenues and higher profitability than domestic banks in developing countries. This finding may reflect the technology of foreign owned banks in developing countries is relatively strong; possibly strong enough to overcome any informational disadvantage. The high net interest revenues also indicate that the banking system in developing countries is relatively inefficient. However, foreign owned banks in developed countries are shown to be less profitable. Micco et al. (2006) also find higher profitability of foreign banks in developing countries but no significant difference in developed countries. In addition, the authors do not find significantly different net interest revenues of foreign and domestic banks.

Evidence on foreign bank performance in transition economies is scarce. Fries, Neven and Seabright (2002) examine the performance of banks in 16 transition economies for the period 1994-99. Their study suggests that bank performance differs significantly depending on the competitive conditions as well as on the reform environment in which banks operate. However, the authors do not focus on differences in ownership. Bonin, Hasan and Wachtel (2005) focus on 1996-2000 in estimating determinants of performance and efficiency of banks in 11 transition economies. They use ROA as the dependent variable measuring bank performance. The authors do not find significant results for foreign banks being more profitable than other banks.2 Mian (2003) finds that for a set of 100 developing countries domestic banks are more profitable than foreign banks on the loan side for the period 1992-1999. However, private domestic banks have higher interest expense on deposits and lower revenue from the sale of banking services. The author concludes that there is no significant difference in the average profitability of private domestic and foreign banks in emerging countries.

A descriptive study of Crystel, Dages and Goldberg (2002) shows lower ROA for foreign banks in Argentina, Chile and Colombia. The authors point at higher provisioning of foreign banks as an explanation.

Two country studies provide evidence for the global advantage theory. Majnoni, Shankar and Várhegi (2003) conclude that during 1994-2000 foreign banks in Hungary were able to achieve a consistently higher profitability than domestic banks. The authors argue that this higher profitability is related to the duration of a foreign banks’ presence in the country and to the nature of the initial investment. In addition,

2 However, they do find that banks in which international institutional investors have a stake are relatively more profitable. The authors conclude that, since international investors do seem to be able to choose banks with higher financial returns and more profit efficiency, the evidence is more supportive of the cherry-picking hypothesis than for the technology-transfer hypothesis.

��

Financial Performance: Foreign Banks in European transition countries

the authors find greenfield investments outperforming other forms of entry of foreign banks. Sabi (1996) also concludes that foreign banks in Hungary are more profitable than domestic banks.Based on data from CEE central banks for the period 1995-2000, chapter 1 suggested that foreign banks are more profitable in terms of return on equity (ROE) and return on assets (ROA) than domestic banks. For the same period, foreign bank operations resulted in lower net interest revenues than those of domestic banks. Overhead costs of foreign and domestic banks were not much different. In this chapter, we extent the empirical literature by testing the relationship between foreign ownership and bank performance in 22 transition countries empirically.3 First, it investigates whether an increase in the level of foreign ownership influences bank performance. This type of research is new. We identify exact foreign ownership of each bank, thereby obtaining a continuous variable. We use the exact shares of foreign ownership, as we assume that foreign investors are able to influence strategic decisions even without owning the majority of the assets. Secondly, we compare performance of domestic and foreign majority owned banks, defining a bank as foreign if foreign investors own more than 50 percent of the shares.4 A dummy variable is used to represent foreign majority owned banks. Both methods are inquiring different issues although they are related.

A second major novelty is that the sample consists of banks from an extensive set of 22 European transition countries. We follow the approach of Demirgüç-Kunt and Huizinga (1998). They specify and estimate determinants of net interest revenues and bank profitability using balance sheet data of banks in 80 countries, of which seven are transition countries, in the period 1988-1995. As Lensink and Hermes (2004) find that economic development influences the impact of foreign bank entry on the performance of domestic banks, we test whether the level of GDP per capita influences the results found. Claessens et al. (2001) suggest that the number of foreign banks rather than their size is associated with competitive conditions in national banking markets. Therefore, we also test whether the two measures of foreign bank penetration influences the result found. Finally, we analyse whether the level of bank concentration affects the relationship between foreign ownership and bank

3 In this study we differentiate between domestic ownership of bank shares by residents and own-ership of bank shares by non-residents. We consider the origin of holders of publicly traded shares of banks in CEE less relevant. First, stock markets in the region are small. Second, al-though in Poland authorities push foreign owners of local banks to have a minimal percentage of total shares traded on the stock exchange, many banks in the region have been delisted. For example, Swedish SEB delisted Estonian Eesti Ühispank in 2001, in the Czech Republic Aus-trian Erste Bank delisted Česka Spořitelna in 2002 and Italian Unicredito delisted Zívnostenska Banka in 2003. Swedbank delisted Estonia’s biggest bank Hansabank in 2005.

4 We assume that if an investor has more than 50 per cent of the shares, he/she has the majority of voting rights and the dummy variable takes a value of 1. In the literature, a dummy variable measuring foreign ownerhsip is usually based on a threshold of 50 per cent ownership of shares. However, in a study of Grigorian and Manole (2002) on bank efficiency, a bank is regarded as foreign on the basis of a minimum of 30 per cent non-domestic shareholders.

��

Chapter 4

��

performance.5 The remainder of this chapter is organised as follows. Section 4.2 presents the methodology and the data. Empirical results will be shown in section 4.3. Conclusions are in section 4.4.

4.2 Data and methodology4.2.1DataIn line with Demirgüç-Kunt and Huizinga (1998), we estimate determinants of bank performance. BankScope provides ownership data only for the last year. As we do not want to assume ownership constant over time, especially in transition economies, we chose to perform a cross-country analysis rather than a panel analysis. We present data on ownership and performance indicators of 244 banks in European transition countries in 2001, as available from the November 2002 BankScope data provided by Bureau Van Dijk.6 Banks included in BankScope roughly account for 90 per cent of the assets of all banks. Appendix 5 shows the origin and the nature of the banks in the sample. Descriptive statistics are in Table 4.1 and involve the entire sample of banks. Regression results are presented for the entire sample as well as for the subset of commercial banks only.

For each bank we identify to what extent bank shares are owned by foreign investors and by domestic investors in order to construct a continuous foreign ownership variable. If more than 40 per cent of a banks ownership data was unavailable, the bank was not included in the sample. If 0-40 per cent of share ownership was unavailable we included the bank and categorized this part as unknown. Table 4.1 shows the levels of foreign ownership for the whole sample. Only a third of all banks are fully domestically owned. Most have banks are, fractionally or fully, foreign owned.

In addition, we investigate whether countries’ GDP per capita, concentration within the banking sector and foreign bank penetration influences the relationship between foreign ownership and banks’ net interest revenues, costs and profitability. Presence of foreign banks is measured in two ways. First, we use the relative number of foreign banks present as a ratio of a country’s total number of banks. For this ratio, a foreign bank is defined as a bank of which more than 50 per cent of the shares are in foreign hands. Secondly, we use the relative amount of the foreign banks’ assets as a ratio of a country’s total bank assets. For the denominator of this ratio, we sum the values of the per bank value of foreign share ownership times bank assets. Bank concentration is defined as the ratio of the total assets of the three largest banks to total bank assets.

5 See Mamatzakis et al. (2005) for details on bank concentration in seven South Eastern European economies. The authors conclude that the banking industry in this region operates under mo-nopolistic competition.

6 Our definition of owner is the direct owner. Sometimes the direct owner is the ultimate owner. In case the direct owner is not the ultimate owner we assume the ultimate owner usually is also a foreign investor.

��


Table 4.2 shows average foreign bank penetration and average concentration of the banking sector in the 11 CEE and in the other transition economies. The two groups differ in four respects. First, foreign bank penetration is more substantial in the CEE group than in the rest of the transition economies. Second, the group of CEE countries is more homogeneous with respect to foreign bank penetration than the group of the other transition economies. Third, foreign banks in the CEE countries appear to be bigger than those in the SEE/CIS countries. Finally, concentration of the banking sector in the other European transition economies is more substantial than in the CEE countries.

Table 4.1 Descriptive statistics

Obs. FOR NFB FBA NIR OHC PROF LOANS NII LLR G INT GDPPC CONC

Albania 3 0.67 0.67 0.21 2.5 2.3 2.3 0.4 0.7 0.0 7.8 7.7 1,389 1.00

Armenia 1 1.00 1.00 1.00 4.6 4.3 2.4 4.7 2.1 0.0 6.0 19.9 796 1.00

Azerbaijan 2 0.13 0.00 0.00 4.7 3.8 2.5 50.0 4.1 2.5 11.1 16.5 789 0.95

Belarus 1 0.00 0.00 0.00 2.5 5.4 2.7 n.a. 6.2 0.5 5.8 61.0 1,242 1.00

Bosnia-H.na 6 0.99 0.33 0.54 3.8 6.7 0.9 41.0 4.6 1.4 5.9 - 1,213 0.58

Bulgaria 18 0.92 0.60 0.72 4.4 5.0 1.1 33.6 2.5 0.5 5.8 4.6 1,835 0.56

Croatia 22 0.74 0.35 0.82 3.6 4.1 1.5 52.8 1.9 0.3 3.7 4.3 5,155 0.58

Czech Rep. 21 0.85 0.59 0.85 2.0 2.2 0.6 31.7 1.1 0.0 2.9 5.1 6,204 0.70

Estonia 3 0.99 0.75 0.99 3.5 3.4 0.9 56.7 2.2 0.9 6.9 4.0 4,292 0.99

Hungary 19 0.98 0.58 0.53 3.9 4.3 1.7 55.4 1.9 0.0 5.2 10.8 5,805 0.49

Kazakhstan 4 0.00 0.08 0.04 5.5 6.3 1.1 64.8 4.8 1.1 9.6 5.3 1,706 0.70

Latvia 9 0.99 0.43 0.49 3.0 3.9 1.0 51.7 2.7 0.1 6.6 5.1 3,683 0.58

Lithuania 5 0.88 0.71 0.96 3.3 5.4 0.6 51.6 2.4 0.3 3.9 3.0 3,683 0.87

Macedonia 4 0.73 0.43 0.55 4.7 3.7 2.0 18.8 4.2 3.2 4.6 11.0 1,970 0.84

Poland 26 0.87 0.70 0.68 2.6 3.6 0.8 45.0 2.7 0.8 4.0 12.0 5,275 0.31

Romania 17 0.85 0.52 0.37 6.0 7.5 2.4 35.2 2.8 0.6 1.8 42.2 1,976 0.41

Russia 25 0.00 0.12 0.12 4.1 6.2 3.6 43.4 2.3 0.2 8.3 12.5 2,422 0.53

Slovak Rep. 15 0.93 0.80 0.95 2.6 3.1 1.3 40.6 1.6 0.1 2.2 7.8 4,158 0.23

Slovenia 13 0.00 0.31 0.21 2.2 3.4 0.9 54.1 2.2 1.0 4.6 10.9 10,673 0.62

Ukraine 5 0.50 0.11 0.11 4.8 6.6 0.8 64.3 2.1 1.2 5.9 17.1 885 0.49

Uzbekistan 1 0.00 0.00 0.00 1.9 1.2 0.6 58.5 0.4 0.6 4.0 34.5 269 1.00

Yugoslavia 4 0.50 0.18 0.07 1.2 2.4 -0.6 n.a. 0.8 0.3 5.0 80.1 1,394 0.71

Macro economic variables include the 2000 GDP growth rate (G), the interest rate (interest) and gross domestic product per capita, in € (GDPPC). Financial sector variables include bank concen-tration (CONC), the ratio of the number of foreign banks over total banks (NFB), and the ratio of foreign banks’ assets over total bank assets (FBA). GDP growth and GDP per capita data are from EBRD. Treasury bill 3-months interest rate data are from IFS and EBRD. Bank concentra-tion and foreign bank presence are from own sample. Bank indicators include net interest revenues (NIR), non-interest income (NII), overhead costs (OHC), loan loss provisions (LLR), profit before tax

��

Chapter 4

��

(PROF), net loans (LOANS) and foreign ownership (FOR). Statistics on equity, non-earning assets and customer and ST funding are available on request. Source: BankScope, November 2002. Data are 2001 median values, all scaled to bank assets.

Table 4.2 Level of foreign ownership in CEE banks, 2001

Foreign ownership (%) No. of banks

Share (%) Cumulativepercent

0 - 20 70 31.25 31.25

20 - 40 11 4.91 36.16

40 - 60 15 6.70 42.86

60 - 80 18 8.04 50.89

80 - 100 61 27.23 78.12

100 49 21.88 100.00

Total 224 100.00 100.00

Level of foreign ownership is the percentage of bank shares owned by foreign investors. The table shows the absolute number of banks and the relative number of banks in the sample for each level of ownership. The sample includes banks from Albania, Armenia, Azerbaijan, Belarus, Bosnia-Herzegovina, Bulgaria, Croatia, Czech Rep., Estonia, Hungary, Kazakhstan, Latvia, Lithuania, Macedonia, Poland, Romania, Russia, Slovak Rep., Slovenia, Ukraine, Uz-bekistan, and Yugoslavia. Source: November 2002 issue of Bureau van Dijk’s BankScope da-tabase (For sample details see Appendix 5).

4.2.2MethodologyThis chapter measures financial performance with three variables from the following accounting rule: profit before taxes is the sum of net interest income and non-interest income minus the sum of overhead costs and loan loss provisions. Profit before tax is the main dependent variable. The other two dependent variables we test for are net interest revenues, the main source of profit, and overhead costs. Non-interest income indicates the non-lending activities of a bank, and loan loss provisioning measures provisioning for bad debts.

To examine whether foreign ownership makes a difference in banks’ performance the three dependent variables are regressed on two separate foreign ownership variables, on balance sheet and income statement data and on macro economic data. We estimate the following basic equation:

Iij = a FORij +b Bij + g Cj + d CDj + e (4.1)

Iij is a vector of performance measures that includes net interest revenues, overhead costs and profit before taxes of bank i in country j. FORij indicates the fraction of shares of bank i in country j in foreign hands. In a second estimation set FORij is a dummy variable taking the value of 1 if foreign investors own more than 50 percent of the shares. Vector B consists of bank specific variables for bank i in country j. These

��


variables include loan loss provisions, equity (a proxy for bank size), non-earning assets, customer and short term funding and loans (a proxy for risk). Table 4.1 shows the descriptive statistics of bank specific variables per country. All bank variables are scaled to total assets. Vector C consists of the control variables GDP growth, the interest rate and GDP per capita. Descriptive statistics of the country specific variables are also shown in Table 4.1 CD is a country dummy variable. Several specifications of (4.1) are estimated. For each specification we present the results of the whole sample as well as for the subset of commercial banks only.

We have estimated the three equations for the net interest revenues, profit before taxes and the overhead costs as a system, using Zellners’ seemingly unrelated regression method. We estimated for the entire sample as well as for the subset of commercial banks. We have used this method to account for the possible cross-equation correlation among the errors. Zellners’ SUR method involves generalized least-squares estimation and achieves an improvement in efficiency by taking into account the fact that cross-equation error correlations may not be zero. SUR gives the same outcome as OLS if the cross-equation covariance equals zero, and if the explanatory variables in the different equations are identical. The possible list of explanatory variables in all equations is the same. However, we drop highly insignificant variables from the different equations, so that the explanatory variables differ per equation, and thus SUR will give other results than OLS. Significant intercept dummies are taken into account as well as a constant term.

4.3 Empirical resultsTable 4.3a shows the results of the estimates using the continuous foreign ownership variable. The level of foreign ownership has a significantly negative impact on all three bank performance indicators. Banks with higher foreign ownership generate lower net interest revenues. Although overhead costs also decrease with a high level of foreign ownership, banks with higher foreign ownership have lower profitability. The level of loan loss provisions has a significantly positive relationship with net interest revenues. This relationship is robust and found in all estimations presented in Tables 4.3-4.7 Loan loss provisions negatively influence profitability. The latter finding follows from the accounting definition. Larger banks measured by equity (EQ) generate more net interest revenues and are more profitable, although the level of significance is higher for the set of all banks than for the sub sample of commercial banks only. These results are robust throughout the estimations presented in Tables 4.3-4.7. The level of non-earning assets (NEA) is positively related to net interest revenues and costs and negatively related to overhead costs. Non-earning assets lowers bank profitability significantly. The findings are consistent throughout the different specifications in Tables 4.3-4.7. The level of customer and short term funding (CSTF) and loans (Loans) do not affect profitability, but Growth in gross domestic product (G) does. The level of interest (INT) does not affect profitability. The results show that gross domestic product per capita (GDPPC) does not influence bank profitability. In more developed countries

��

Chapter 4

��

banks seem to be more efficient as the level of overhead costs diminishes. At the same time, the level of net interest revenues drops. The signs and significance of the macro economic correlates are similar in all estimations presented in Tables 4.3-4.7.

Table 4.3a Foreign ownership and performance

Net interest revenues Overhead costs Profit before taxes

(1) (2) (3) (4) (5) (6)

Sample All Commercial All Commercial All Commercial

FOREIGN (continuous)

-9.37E-05***(-4.14)

-7.88E-05***(-3.31)

-7.63E-05**(-2.13)

-7.20-05*(-1.87)

-0.00012***(-3.34)

-9.94E-05***(-2.71)

LLP 0.248***(3.56)

0.200***(2.84)

-0.776***(-11.17)

-0.804***(-11.45)

-0.157(-1.47)

-0.227**(-2.06)

EQUITY 0.059***(4.75)

0.069***(5.45)

0.030**(2.20)

0.024*(1.72)

0.056***(2.92)

0.061***(3.11)

NEA 0.030**(2.06)

0.029**(2.03)

-0.052***(-2.67)

-0.044**(-2.15)

0.116***(5.25)

0.107***(4.88)

CSTF 0.010(1.24)

0.015*(1.65)

0.029**(2.16)

0.026**(1.82)

LOANS 0.0003***(4.61)

0.0003***(4.51)

0.0003***(3.88)

0.0004***(4.33)

GROWTH 0.0035***(4.61)

0.0035***(4.32)

INTEREST -0.0005**(-2.08)

-0.0004(-1.62)

-0.00097***(-2.59)

-0.0008**(-2.05)

GDPPC -2.20E-06***(-5.02)

-1.94E-06***(-4.17)

-6.15E-07(-0.88)

-3.23E-07(-0.43)

-1.57E-06**(-2.31)

-1.73E-06**(-2.38)

Adj. R2 0.46 0.48 0.38 0.40 0.35 0.39

No. obs. 199 176 216 191 199 176

Note: t-values between parentheses. In the equation for net interest revenues three significant country dummies (Croatia, Hungary and Romania) are added; in the equation for profit sig-nificant country dummies for Macedonia and Poland are added; for overhead costs country dummies for Romania and the Russian Federation are added. *, ** and *** indicate signifi-cance levels of 10, 5 and 1 percent respectively.

Table 4.3b presents the coefficients of similar model specifications, estimated with a foreign ownership dummy variable. The correct interpretation of these results would be that foreign majority owned banks generate lower net interest revenues, overhead costs and profits than domestic majority owned banks. Admittedly, it is not to our surprise that similar results are obtained since Table 4.3b showed that the continuous variable seems to behave to a large extent as a dummy variable. For this very reason,

��


the rest of this section presents the estimation results based on the continuous foreign ownership variable only.

Table 4.3b Foreign majority ownership and performance


(1) (2) (3) 4) (5) (6)


FOREIGN (dummy)

-0.0085***(-4.32)

-0.0069***(-3.30)

-0.0069**(-2.21)

-0.0068**(-2.02)

-0.011***(-3.70)

-0.0085***(-2.58)

LLP 0.243***(3.50)

0.199***(2.81)

-0.779***(-11.23)

-0.806***(-11.50)

-0.164(-1.54)

-0.229**(-2.07)

EQUITY 0.059***(4.77)

0.069***(5.46)

0.029**(2.16)

0.0233(1.64)

0.057***(2.96)

0.062***(3.14)

NEA 0.032**(2.22)

0.031**(2.16)

-0.050**(-2.55)

-0.041**(-2.02)

0.119***(5.40)

0.109***(4.96)

CSTF 0.012(1.35)

0.016*(1.74)

0.030**(2.30)

0.027*(1.92)

LOANS 0.00026***(4.71)

0.00026***(4.56)

0.00034***(4.01)

0.00039***(4.39)

GROWTH 0.0035***(4.61)

0.0034***(4.30)

INTEREST -0.0005**(-2.13)

-0.0004*(-1.65)

-0.001***(-2.71)

-0.0008**(-2.12)

GDPPC -2.17E-06***(-4.98)

-1.92E-06***(-4.11)

-5.87E-07(-0.85)

-3.10E-07(-0.42)

-1.55E-06**(-2.29)

-1.71E –06**(-2.34)

Adj. R2 0.47 0.48 0.38 0.41 0.36 0.39

No. obs. 199 176 216 191 199 176


To examine whether the level of economic development affects the negative relationship between foreign ownership and bank performance, we interacted foreign ownership with GDP per capita. Table 4.4 shows the results. The estimations with the interaction variable suggest that the level of economic development does not alter the negative relationship between foreign ownership and bank performance. Table 4.4 shows the results for the estimations when we control for the competitive

�0

Chapter 4

�0 ��

environment in the banking sector. Concentration in the banking sector seems to have little relevance for bank performance. This is in line with the findings of Demirgüç-Kunt et al. (2003), who find mixed evidence, and the conclusion of the ECB indicating that concentration ratios do not necessarily reflect competitive conditions within the region (ECB, 2005). In addition, the estimates of the interaction variable indicate no significant impact on the negative relationship between foreign ownership and bank performance.

Table 4.4 Test 1: does economic development matter?


(1) (2) (3) (4) (5) (6)


FOREIGN(continuous)

-0.00013***(-2.89)

-0.00012**(-2.45)

-0.00012*(-1.68)

-9.78E-05(-1.32)

-0.0002***(-2.83)

-0.00016**(-2.10)

LLP 0.247***(3.56)

0.202***(2.86)

-0.773***(-11.11)

-0.801***(-11.37)

-0.161(-1.51)

-0.226**(-2.05)

EQUITY 0.059***(4.80)

0.070***(5.52)

0.030**(2.21)

0.024*(1.73)

0.059***(3.04)

0.063***(3.20)

NEA 0.030**(2.09)

0.029**(2.07)

-0.051***(-2.65)

-0.044**(-2.14)

0.117***(5.32)

0.108***(4.94)

CSTF 0.011(1.25)

0.015*(1.70)

0.029**(2.18)

0.027*(1.89)

LOANS 0.00025***(4.56)

0.00025***(4.43)

0.00033***(3.81)

0.00037***(4.23)

GROWTH 0.0035***(4.52)

0.0034***(4.27)

INTEREST -0.00054**(-2.16)

-0.00045*(-1.75)

-0.001***(-2.71)

-0.0008**(-2.18)

GDPPC -2.59E-06***(-4.46)

-2.34E –06***(-3.80)

-1.04E-06(-1.09)

-6.01E –07(-0.59)

-2.38E –06***(-2.62)

-2.30E-06**(-2.38)

GDPPC * FOREIGN

9.15E –09(0.99)

9.71E –09(0.97)

9.29E –09(0.66)

6.29E-09(0.41)

1.91E-08(1.35)

1.38E-08(0.90)

Adj. R2 0.46 0.48 0.38 0.40 0.36 0.39

No. obs. 199 176 216 191 199 176


�0�0 ��


Table 4.5 Test 2: does concentration matter?


(1) (2) (3) (4) (5) (6)


FOREIGN(continuous)

-0.00024***(-3.00)

-0.00022**(-2.47)

-1.84E-05(-0.15)

-2.63E-06(-0.02)

-0.00026**(-2.10)

-0.00029**(-2.08)

LLP 0.2575***(3.69)

0.210***(2.94)

-0.774***(-11.24)

-0.801***(-11.47)

-0.125(-1.17)

-0.1888*(-1.71)

EQUITY 0.055***(4.35)

0.065***(5.01)

0.029**(2.13)

0.024*(1.72)

0.0463**(2.36)

0.050**(2.48)

NEA 0.026*(1.81)

0.026*(1.79)

-0.048**(-2.48)

-0.0418**(-2.05)

0.1195***(5.36)

0.1106***(5.01)

CSTF 0.0079(0.90)

0.012(1.31)

0.0213(1.57)

0.017(1.17)

LOANS 0.00024***(4.21)

0.00024***(4.26)

0.00030***(3.44)

0.00035***(4.00)

GROWTH 0.00414***(5.07)

0.0040***(4.69)

INTEREST -0.00041(-1.56)

-0.00029(-1.08)

-0.0010***(-2.64)

-0.0008**(-1.99)

GDPPC -2.24E-06***(-5.16)

-1.94E-06***(-4.16)

-3.47E-07(-0.49)

-8.84E-08(-0.12)

-1.68E-06***(-2.48)

-1.86E-06**(-2.57)

CONC -0.019*(-1.78)

-0.018(-1.47)

-0.010(-0.63)

-0.0054(-0.30)

-0.030*(-1.79)

-0.036*(-1.85)

CONC * FOREIGN

0.00027*(1.90)

0.00026(1.64)

-0.0001(-0.49)

-0.00012(-0.53)

0.00025(1.15)

0.00033(1.35)

Adj. R2 0.47 0.48 0.38 0.41 0.36 0.40

No. obs. 199 176 216 191 199 176

*, ** and *** indicate significance levels of 10, 5 and 1 percent respectively

Tables 4.6 and 4.7 present estimation results of two specifications of the model that both controls for the competitive conditions of foreign banks. Table 4.6 controls for foreign bank penetration measured by the share of foreign banks’ assets as a ratio of total banking assets. Results show that banks in countries with relatively many foreign assets were less profitable than in countries with less foreign assets. Similarly, Table 4.7 shows that the relative number of foreign banks negatively influences bank profitability. However, foreign bank profitability is higher in countries in which foreign bank presence is more extensive, either measured by foreign banks’ assets or by the

��

Chapter 4

��

number of foreign banks. This contrasts with the findings of Claessens, Demirgüç-Kunt and Huizinga (2001) who argue that the number of foreign banks rather than their size is associated with competitive conditions in national banking markets.

Table 4.6 Test 3: does the relative number of foreign banks matter?


(1) (2) (3) (4) (5) (6)



-0.00015***(-2.99)

-0.000147***(-3.33)

-0.00010(-1.38)

-0.000114(-1.43)

-0.00024***(-3.29)

-0.00019**(-2.53)

LLP 0.240***(3.50)

0.197***(2.87)

-0.165(-1.52)

-0.236**(-2.14)

-0.784***(-11.45)

-0.811***(-11.68)

EQUITY 0.057***(4.65)

0.066***(5.35)

0.056***(2.91)

0.0603***(3.07)

0.0271**(2.00)

0.0212(1.51)

NEA 0.028*(1.95)

0.025*(1.82)

0.116***(5.25)

0.1069***(4.88)

-0.054***(-2.81)

-0.048**(-2.38)

CSTF 0.0096(1.13)

0.013(1.41)

0.028**(2.09)

0.0238*(1.68)

LOANS 0.00025***(4.49)

0.00023***(4.11)

0.00033***(3.83)

0.00037***(4.13)

GROWTH 0.0029***(3.09)

0.0027***(2.85)

INTEREST -0.00080***(-2.73)

-0.0008***(-2.72)

-0.0012***(-2.78)

-0.0011**(-2.53)

GDPPC -2.06E-06***(-4.74)

-1.73E-06***(-3.82)

-1.55E-06**(-2.27)

-1.72E-06**(-2.34)

-5.63E-07(-0.82)

-1.82E-07(-0.25)

REL. FOR.ASSETS

-0.0149***(-2.60)

-0.0220***(-3.55)

-0.0046(-0.49)

-0.0096(-0.92)

-0.024***(-2.64)

-0.025***(-2.59)

REL. FOR. ASSETS * FOREIGN

0.00013(1.63)

0.00021**(2.52)

-1.10E-05(-0.09)

4.40E-05(0.34)

0.00033***(2.76)

0.00028**(2.15)

Adj. R2 0.48 0.51 0.35 0.39 0.40 0.42

No. obs. 199 176 199 176 216 191

*, ** and *** indicate significance levels of 10, 5 and 1 percent respectively.

��


Table 4.7 Test 4: does the relative amount of foreign banks’ assets matter?


(1) (2) (3) (4) (5) (6)



-9.55E-05*(-1.72)

-0.00011*(-1.95)

-0.00025***(-3.16)

-0.00019**(-2.20)

-6.86E-05(-0.81)

-3.84E-05(-0.42)

LLP 0.248***(3.60)

0.203***(2.92)

-0.777***(-11.33)

-0.806***(-11.53)

-0.153(-1.41)

-0.220**(-2.00)

EQUITY 0.058***(4.68)

0.067***(5.39)

0.026*(1.90)

0.0218(1.53)

0.0565***(2.92)

0.0614***(3.11)

NEA 0.028**(1.98)

0.026*(1.87)

-0.056***(-2.92)

-0.049**(-2.39)

0.1166***(5.25)

0.1085***(4.93)

CSTF 0.0112(1.31)

0.015(1.62)

0.0290**(2.17)

0.027*(1.86)

LOANS 0.00025***(4.56)

0.00025***(4.37)

0.00033***(3.83)

0.00039***(4.31)

GROWTH 0.0031***(3.31)

0.0030***(3.06)

INTEREST -0.00061**(-2.25)

-0.00061**(-2.17)

-0.0010***(-2.61)

-0.00079**(-2.01)

GDPPC -2.12E—06***(-4.80)

-1.73E-06***(-3.68)

-5.54E-07(-0.80)

-1.45E-07(-0.19)

-1.57E-06**(-2.29)

-1.77E-06**(-2.40)

REL. NR. OF FOR BANKS

-0.0081(-1.04)

-0.017**(-2.00)

-0.029**(-2.23)

-0.0288*(-1.95)

0.0051(0.40)

0.0074(0.53)

REL. NR. OF FOR BANKS* FOREIGN

2.98E-05(0.27)

0.00012(1.04)

0.00042***(2.60)

0.00033*(1.80)

-0.0001(-0.62)

-0.00014(-0.73)

Adj. R2 0.46 0.49 0.39 0.41 0.35 0.39

No. obs. 199 176 216 191 199 176

*, ** and *** indicate significance levels of 10, 5 and 1 percent respectively.

4.4 ConclusionChapter 4 examines whether the performance of foreign owned banks in Central and Eastern Europe exceeds that of domestic owned banks. The analysis is based on OLS regressions and focuses on bank performance of 244 commercial, savings and cooperative, and investment banks in 20 European transition countries in 2001. The analysis specifically concentrates on whether the level of foreign ownership affects (i) net interest revenues, (ii) overhead costs, and (iii) profit before taxes. Next,

��

Chapter 4

we examine whether the result are robust if we use a dummy foreign ownership variable instead of a continuous foreign ownership variable. Finally, we test whether the relationship between foreign ownership and bank performance is affected by (a) the level of economic development of the host country, (b) concentration within the banking sector, (c) the presence of foreign banks measured by relative assets and (d) the presence of foreign banks measured by the relative number of banks.

Results show that foreign ownership has a negative influence on the net interest revenues of banks in transition countries. This could hint at an aggressive strategy of foreign banks in augmenting their market share as suggested by VanderVennet (1996) who argues that especially when a retail bank with a branch network is acquired by a foreign bank, the acquirer acts within a penetration strategy and often launches an intensive campaign to attract additional business. Another explanation could be that foreign banks face a structural disadvantage relative to their domestically owned competitors. A second conclusion is that in transition countries overhead costs are lower when banks have a higher share foreign ownership. This second result would back up the hypothesis that, at least in transition countries, foreign banks operate more efficiently than their competitors that lack the best-practice way of banking common in more developed banking sectors, or that have not yet been able to implement this knowledge. Overall, the opposite effects of foreign ownership on net interest revenues and overhead costs do not prevent foreign banks to be less profitable. The results are robust: for the subsample of 198 commercial banks the same results are found.

Furthermore, we find the following effects for the above-mentioned factors that may affect the relationship between foreign ownership and bank performance. First, the level of economic development of the host country does not affect the relationship found. However, banks in more developed transition countries (i) have less net interest revenues and (ii) have less overhead costs, than in less developed transition countries. An explanation could be found in the hypothesis that in more developed economies the banking sector is more competitive. Second, we find that the degree of concentration within the banking market in transition economies does not affect the relationship between foreign ownership and bank performance. Moreover, concentration does not seem to have any important effect on bank profitability. Third, a stronger presence of foreign banks, measured in relative assets, is positively related to foreign bank profitability and in countries where foreign banks have a less pronounced role in the financial sector, foreign banks have lower profitability. The cause for this seems to be the higher net interest revenues rather than lower overhead costs. Similarly, foreign banks are more profitably in countries with relatively more foreign banks than in countries with relatively few foreign banks. The cause for this finding is not traceable to higher interest revenues or lower overhead costs.

The results found are the first prove for a so-called home advantage for domestic banks in European transition countries. To meet potential endogeneity problems between foreign ownership and bank performance, chapter 5 improves the estimation technique by applying the generalized method of moments to a sample of banks worldwide.

��

Chapter 5Financial performance:

Foreign Banks Worldwide *

5.1. IntroductionExamining the performance of banks in transition economies in chapter 4, we found evidence that foreign banks are less profitable than domestically owned banks. This chapter aims to extend the analysis on foreign bank performance in two ways. First, we introduce a second estimation technique that adds to the evidence of the OLS approach. The results of the OLS approach must be interpreted with care due to the possible endogeneity problem between the dependent and independent variables. This chapter examines the same research question using the generalized method of moments (GMM) technique. A second contribution of this chapter is that we now use a panel consisting of banks from countries worldwide.

5.2 The dataWe use bank income statement data for the period 1998-2001 of banks listed in the BankScope database of Bureau van Dijk. We include banks for which ownership data was available for at least three subsequent years and of which at least 75 per cent of the ownership information is known. For each bank we determine exact foreign ownership.1

There are 511 banks in 73 countries in the sample with total number of observations of 1.747. Table 5.1 shows the aggregated data of the three groups of banks and makes a distinction in majority owned foreign banks2 and majority owned domestic banks on the basis of the 50%-assets threshold. It is important to note that in Tables 5.1 and 5.2 we use this distinction to give some insight in the dataset, while in the regression analyses we use the continuous ownership variable, i.e. the exact foreign ownership percentages for each bank. Table 5.1 shows that in all four years under consideration foreign banks amount to 1/3 of all banks.

However, if we classify the 73 economies according to GNI per capita in accordance with the World Development Report 2003 of the World Bank we see that the market

* Based on Lensink, B.W. and I.J. Naaborg, ‘Does foreign ownership foster bank performance?’, forthcoming in Applied Financial Economics (www.tandf.co.uk/journals).

1 Foreign ownership measures the percentage of bank shares in foreign hands as a proxy for vot-ing rights by foreigners. In the literature, the aggregated measure is often described as the share of foreign assets in total banking assets.

2 Here, a foreign bank is defined as a bank whose assets are for more than 50% in hands of for-eign investors.

��

Chapter 5

��

share of foreign banks differs between groups of countries (table 5.2). In addition, we see observe contrasts between market share measured by number of foreign banks and market share measured by banking assets in hands of foreign investors.

Market share of foreign owned banks measured by both ratios is highest in upper-middle income countries (UMC). In 2001, 36% of all banks in these countries are a foreign majority owned bank and total foreign owned bank assets amount to 40% of all banking assets. During 1998-2001 foreign majority owned bank presence in high-income countries (HIC) is lower than in UMC countries: foreign investors own only 11% of banking assets. In lower-middle income countries (LMC) foreign majority owned bank presence was on average 33% but only 4% of banking assets was owned by foreign investors and the same picture holds for low-income countries (LIC), where market share of foreign investors in the banking sector is negligible.

The total number of banks in our sample amounts to 511 in 73 countries with a total of observations of 1.747. Appendix 12 shows the number of banks in each country in the sample.3 Macro economic data used in the estimations are taken from the IMF’s IFS and is presented in Appendix 13.

5.3 Model and estimation techniqueWe estimate our model with two estimation techniques. In line with most other papers in the field, we start with ordinary least squares (OLS) regressions, based on the levels equations. There are, however, several problems with this. First, since we base the regressions on levels, we ignore the panel properties of the data, i.e. we do not account for the bank specific effects, represented by hij . The usual way to come around this problem is to difference the equations to eliminate the bank specific effects. For Nir this gives:

Nirijt-Nirijt-1=b(Xijt-Xij-1)+g(Zjt-Zjt-1)+n(Nirijt-1-Nirijt-2)+mijt-mijt-1

where we, for reasons of space, have ignored the time specific effects and the interaction term. The problem with this equation is that the lagged dependent variable is now by construction correlated with the error term. This makes OLS an invalid procedure, and calls for another estimation technique. Even in models without a lagged dependent variable, OLS probably is an unreliable estimation technique because of endogeneity problems with variables in X or Z. Many right-hand side variables, including Foreign, are probably endogenous, and therefore correlated with the error term. Therefore, we also estimate our models using an instrumental variable approach. The instrumental variable estimation technique controls for the fact that the explanatory variables are likely to be correlated with the error term and the firm-specific effect, and deals with possible endogeneity problems. More specifically, we estimate the investment models

3 To correct for outliers, we removed banks with the top five highest and top five lowest values in six bank specific data categories and in three macro-economic control categories.

��

Financial Performance: Foreign Banks Worldwide

with the system-generalized methods of moments (GMM) estimator, using DPD98 for Gauss (Arellano and Bond, 1998).

Table 5.1 Foreign ownership of banks worldwide

YearNo. banks

No. Foreign banks NFB FBA

(1) (2)

By income group

High Income Countries 1998 314 110 35% 12%

1999 368 124 34% 12%

2000 368 122 33% 9%

2001 237 82 35% 11%

Upper Middle income 1998 70 29 41% 20%

Countries 1999 85 36 42% 29%

2000 85 35 41% 26%

2001 28 10 36% 40%

Lower Middle income 1998 39 12 31% 4%

Countries 1999 45 11 24% 3%

2000 45 15 33% 4%

2001 21 9 43% 5%

Low Income Countries 1998 11 1 9% 0%

1999 13 2 15% 1%

2000 13 2 15% 1%

2001 6 1 17% 1%

By year

1998 434 152 35%

1999 511 173 34%

2000 511 174 34%

2001 291 102 35%

NFB is the ratio of the number of foreign banks over total banks. FBA measures foreign owned bank assets over total bank assets. Banks are categorized in income groups according to 2001 GNI per capita, in line with the World Bank World Development Report 2003. Groups are defined as: High Income Countries (> $9,205), Upper Middle Income Countries ($2,976 - $9,205), Lower Middle income Countries ($746 - $2,975) and Low Income Countries (< $745). Source: BankScope 2002.

��

Chapter 5

��

Table 5.2 Descriptive statistics

A Obs. Year Assets Equity STF NIR OOI OI OHC LLP PROF

HIC 110 1998 1,015 8.3 83.4 2.0 1.1 0.0 1.8 0.2 1.0

124 1999 1,037 7.7 84.5 1.5 1.2 0.0 1.8 0.1 0.9

122 2000 1,034 8.2 84.0 1.7 1.1 0.0 1.9 0.1 1.0

82 2001 1,079 8.4 84.1 1.5 1.1 0.0 1.8 0.1 0.8

UMC 29 1998 701 12.7 73.5 3.9 1.4 0.0 3.7 0.7 0.9

36 1999 962 12.1 74.3 3.9 2.0 0.0 3.2 0.6 1.5

35 2000 991 12.1 73.3 3.3 1.5 0.0 3.5 0.5 1.2

10 2001 1,604 10.9 72.7 3.4 2.4 0.0 3.3 0.3 1.8

LMC 12 1998 260 12.6 74.9 4.6 2.2 0.1 6.1 1.2 1.4

11 1999 195 13.5 79.8 4.9 1.0 0.7 5.6 2.2 0.8

15 2000 244 11.4 74.3 4.2 1.9 0.9 5.2 1.1 1.5

9 2001 381 12.4 77.4 4.7 2.2 0.0 3.2 0.4 1.4

LIC 1 1998 42 10.8 82.7 n.a n.a n.a n.a n.a 4.1

2 1999 281 11.1 71.9 3.4 3.1 0.0 2.3 1.0 3.2

2 2000 348 11.6 74.6 4.8 2.0 n.a. 2.4 0.2 4.2

1 2001 560 9.9 73.0 4.1 1.7 -0.1 3.1 0.9 1.6

B Obs. Year Assets Equity STF NIR OOI OI OHC LLP PROF

HIC 203 1998 3,105 7.5 81.0 2.9 1.3 0.0 2.6 0.2 1.5

244 1999 2,823 7.4 80.5 2.8 1.3 0.0 2.6 0.2 1.5

246 2000 2,937 7.9 79.6 2.9 1.3 0.0 2.6 0.2 1.5

155 2001 3,217 8.2 79.9 2.8 1.3 0.0 2.7 0.3 1.4

UMC 41 1998 2,759 9.0 80.9 4.4 1.5 0.1 3.8 0.8 1.3

49 1999 1,002 9.1 80.5 3.9 2.1 0.0 3.8 0.9 1.5

50 2000 1,157 10.0 81.9 4.1 1.5 0.1 3.7 0.7 1.6

18 2001 279 10.0 84.2 3.6 1.5 0.0 3.8 0.7 1.3

LMC 27 1998 913 8.9 81.2 4.0 1.5 0.0 3.8 1.3 1.3

34 1999 1,295 8.6 82.6 3.7 1.6 0.1 3.5 1.0 1.0

30 2000 1,420 8.6 81.9 3.6 1.7 0.1 4.0 0.9 0.6

12 2001 2,699 7.8 84.2 3.1 2.0 0.0 3.3 0.7 0.5

LIC 10 1998 388 8.9 79.6 6.9 3.7 0.0 6.4 1.3 2.8

11 1999 226 8.4 80.7 6.3 3.0 0.2 5.8 0.4 2.7

11 2000 262 11.4 76.7 5.9 3.8 0.1 6.0 0.9 2.1

5 2001 207 11.1 80.9 5.2 2.8 0.0 4.3 0.4 3.2

Foreign banks’ (Table A) and domestic banks’ (Table B) descriptives are medians. Bank assets are in €mln. Equity, Short-term funding (STF), net interest revenues (NIR), other operating income (OOI), other income (OI), overhead costs (OHC), loan loss provisions (LLP) and profit

��


before tax (PROF) are scaled to bank assets. Banks are categorized in income groups accor-ding to 2001 GNI per capita, in line with the World Bank World Development Report 2003. Groups are defined as: High Income Countries (> $9,205), Upper Middle income Countries ($2,976 - $9,205), Lower Middle income Countries ($746 - $2,975) and Low Income Countries (< $745).

The system GMM estimator combines the differenced equation with a levels equation to form a system GMM. Lagged levels are used as instruments for the contemporaneous differences and lagged differences as instruments for the contemporaneous levels. We adopt the system GMM estimation procedure since first difference GMM may suffer from weak instruments problems (Blundell and Bond (1998)).

The system GMM estimator is a two-step estimator. In the first step, homoskedasticity and independent error terms are assumed. In the second step, these assumptions are relaxed by using a consistent variance-covariance matrix that is constructed from the first step residuals. However, the two-step estimator has weak small sample properties: the standard errors are biased downwards. The estimator becomes problematic, especially when there are a small number of cross-section units, in relation to the number of instruments. This might result in biased asymptotic inference. Therefore, we present coefficients and t-values based on the first step GMM estimates. In all estimations we control for time effects by adding time dummies for 1999-2001. These time dummies are used as additional instruments. We initially also added region dummies, but since they were not jointly significant in any of the regressions, we have ignored them in the regressions presented below.

The reliability of the system GMM estimation procedure depends very much on the validity of the instruments. We consider the validity of the instruments by presenting the Sargan test, which is a test on overidentifying restrictions. It is asymptotically distributed as c2 and tests the null hypothesis of validity of the (overidentifying) instruments. p-values report the probability of incorrectly rejecting the null hypothesis. The smaller is the p-value, the less plausible is the null hypothesis. So, a p-value above 0.05 implies that the probability of incorrectly rejecting the null is above 0.05. In this case, a higher p-value makes it more likely that the instruments are valid. The Sargan test is based on the two-step GMM estimator since only the Sargan test for the two-step GM estimator is heteroskedasticy-consistent.

The consistency of the GMM estimates also depends on the absence of serial correlation in the error terms. If the disturbances are not serially correlated there should be evidence of significant first order serial correlation in the differenced residuals and no evidence of second order serial correlation in the differenced residuals1. We present tests for first-order and second-order serial correlation related

1 This can be explained as follows. Let uit be the error term for bank i in period t. First-order serial correlation in the differenced residuals is then given by E(Duit Duit-1)= E(uit-uit-1)(uit-1 -uit-

2)=E(uituit-1-uituit-2-uit-1+uit-1uit-2). Under the assumption of no serial correlation this reduces to E(Duit Duit-1)= -uit-1uit-1, which gives evidence for negative first order serial correlation in the differend residuals. Second-order serial correlation in the differenced residuals is given by E(Duit

�00

Chapter 5

�0�

to the estimated residuals in first differences. The test statistics are asymptotically distributed as standard normal variables. The null hypothesis here relates to no serial correlation, or “insignificance,” so that a low p-value for the test on first-order serial correlation suggests first order serial correlation in the differenced residuals and a high p-value for the test on second-order serial correlation suggests no second order autocorrelation in the differenced residuals. The serial correlation tests (M1 and M2 in the table) refer to the one-step GMM estimates. We also present Wald tests for the joint significance of the time dummies. These test statistics are also asymptotically distributed as c2 variables. The Wald test tests for joint significance of all, or a subset of parameters. The null hypothesis refers to “insignificance”, implying that low p-values suggest joint significance. We consider two proxies for bank performance: profits before taxes (Profbt) and the net interest revenues (Nir). The equations we estimate have the following form:

Nirijt=bXijt+gZjt+fGDPPCjtForeignijt+nNirijt-1+lt+hij+mijt (5.1)

Profbtijt= bXijt+gZjt+fGDPPCjtForeignijt + nProfbtijt-1+lt+hij+mijt (5.2)

where Xijt is a vector of four bank specific explanatory variables. This vector contains:

X = X [Foreignijt,Overhijt,Llossijt,Equityijt] (5.3)

where Zjt is a vector of three country specific explanatory variables. This vector contains:

Z=Z[Infljt, Growthjt, GDPPCjt] (5.4)

The first bank specific variable measures the share of foreign assets in total assets of the banks (Foreign). For the values of the last three variables we take the values as given in the income statement2 and balance sheet of the respective banks. Banks dependent variables, the explanatory variables taken from the income statement, as well as banks equity are all scaled to total assets.

The country specific variables are inflation (Infl), GDP growth (Growth) and GDP per capita (GDPPC) for each country for each year. In line with Demirgüç-Kunt and Huizinga we add the interaction term (GDPPC*Foreign) to measure whether a host country’s level of development affects the relationship between foreign ownership and bank performance. lt are time specific effects. We estimate with time dummies for 1999, 2000 and 2001. hij is an unobserved bank-specific effect. mijt is an error term.

Duit-2)= E(uit-uit-1)(uit-2 -uit-3)=E(uit uit-2-uit uit-3-uit-1uit-2+uit-1uit-3). Under the assumption of no serial correlation, this reduces to E(Duit Duit-2) = 0.

2 The income statement defines profit before tax as the net interest revenues plus other operating income minus overhead costs minus loan loss provisions.

�00 �0�


The subscripts i,j,t refer to bank i in country j in period t. Variables with j,t subscripts are measured at the country level.

In the regressions we estimate different versions of the equations specified above. We estimate equations without lagged dependent variables (i.e. assuming ) and with and without interaction term (i.e. ø=0 ). The interaction term is included to test whether effects of foreign ownership on bank performance depends on the development of a country, as suggested by Demirgüç-Kunt and Huizinga (1998). We estimate our model with the system-generalized methods of moments (GMM) estimator, using DPD98 for Gauss (Arellano and Bond, 1998). For completeness, we also present OLS regressions.

5.4 Empirical results The regression results are presented in tables 5.3 and 5.4. The tables clearly show that a rise in foreign ownership has a negative impact on both the net interest revenues and bank profits. This holds for the OLS regressions and the GMM regressions. Note, however that the GMM estimations give much higher coefficients on the foreign ownership variable in absolute terms than the OLS estimates. Both tables also show that the interaction term is insignificant in all regressions, which suggests that the effect of foreign ownership on bank performance does not depend on the income per capita of the host countries. Thus, our results strongly support the home field advantage theory. Stated simply, banks with a low degree of foreign ownership are more profitable and able to raise more net interest revenues than banks with a high degree of foreign ownership.

5.5 ConclusionElaborating on chapter 4, chapter 5 examines the effect of foreign ownership on the financial performance of 511 banks worldwide. The sample consists of ownership data and income statement data of banks in 73 countries for the period 1998-2001. The average number of foreign banks is one third. However, there are large differences in our sample with respect to foreign bank presence if we categorize the host countries by their level of economic development. Foreign banks are relatively numerous in upper middle income countries with a GNI per capita of $3,000 and $9,200. Countries that belong to this category are Central and Eastern European countries and South- and Middle American countries like Argentina, Brazil, Chile, Mexico and Venezuela. In these countries two out of five banks are foreign, and these foreign banks own one third of banking assets on average. Chapter 1 concluded that these ratios are even higher in Central and Eastern European countries. Foreign banks play a less important role in the high-income countries with a GNI per capita of more than $9,200 than in the upper middle-income countries. Although on average one in three banks is foreign owned, foreign banks own only 10 percent of total banking assets.

�0�

Chapter 5

�0�

Table 5.3 Foreign ownership and net interest revenues

Net interest revenues/ total assets

(1) (2) (3) (4) (5) (6)

Estimation technique

OLS GMM OLS GMM OLS GMM

FOREIGN -0.0096*** (-9.72)

-0.0235** (-2.42)

-0.010*** (-5.74)

-0.0278** (-2.03)

-0.00094 (-1.38)

-0.0109* (-2.32)

OVERHEAD COSTS

0.244***(13.76)

0.479***(3.32)

0.244***(13.74)

0.449***(3.33)

0.068***(5.53)

0.1912**(2.16)

LLP 0.620***(13.94)

0.9913***(2.72)

0.617***(13.77)

1.003***(2.84)

0.2053***(6.64)

0.627***(3.08)

EQ 0.038***(6.62)

0.080(1.68)

0.038***(6.59)

0.075(1.61)

0.0054(1.41)

0.037(1.28)

INFLATION 0.001***(10.68)

0.0012(1.32)

0.0012***(10.57)

0.0008(0.78)

-0.00008(-0.94)

0.0005(0.84)

G -0.00016(-0.716)

-0.00008(-0.10)

-0.00016(-0.74)

-0.000006(-0.008)

0.00015(1.00)

0.00011(0.20)

GDPPC -0.0016***(-5.09)

0.0024(1.33)

-0.0018***(-4.07)

-0.0007(-0.45)

-0.00016(-0.78)

0.00116(1.42)

GDPPC * FOREIGN

0.00031(0.52)

0.0039(1.30)

NIR1 0.8082***(50.38)

0.4884***(4.08)

D1999 0.00017(0.14)

-0.0003(-0.30)

0.00015(0.12)

-0.0006(-0.57)

D2000 -0.00048(-0.40)

-0.0014(-1.47)

-0.0005(-0.41)

-0.0013(-1.39)

0.00017(0.25)

-0.00037(-0.36)

D2001 -0.0025*(-1.72)

-0.0045*(-1.89)

-0.0025*(-1.74)

-0.004*(-1.83)

-0.00007(-0.09)

-0.0018(-1.22)

CONSTANT 0.020***(12.94)

0.0021(0.22)

0.0207***(11.85)

0.0098(1.12)

0.0023**(2.15)

0.0020(0.35)

RSS 0.5626 0.7980 0.5626 0.7343 0.1254 0.1972

TS 1.0181 1.0181 1.0181 1.0181 0.6896 0.6896

Wald time dummy

4.09p=0.252

8.379p=0.039

4.126p=0.248

7.43p=0.059

0.099p=0.952

1.698p=0.428

M1 28.401p=0.00

-3.630p=0.00

28.424p=0.00

-3.703p=0.00

-2.991p=0.003

-4.024p=0.00

M2 19.507p=0.00

-0.074p=0.941

19.471p=0.00

0.167p=0.867

1.833p=0.067

Sargan 19.18p=0.89

22.60p=0.89

39.88p=0.16

�0� �0�


Note: the dependent variable in all equations is the net interest revenues. RSS refers to the residual sum of squares. TSS refers to the total sum of squares. M1 (M2) refers to the test for first-order (second-order) serial correlation, related to the estimated residuals in first dif-ferences, unless the model has been estimated using only the levels equations. Sargan refers to the Sargan test of overidentifying restrictions. The OLS estimates refer to ordinary least squares estimates using only the level equations. The GMM estimates refer to system GMM estimates. Instruments: for the first differenced equations we use the entire history of lagged levels (starting from t-2). For the levels equations we use the one period lagged first differences as instruments. *, ** and *** indicate significance levels of 10, 5 and 1 percent respectively. T-values in parentheses.

The role of foreign banks in lower middle income countries like China, Colombia, Egypt and Iran, and low income countries like Indonesia, Kenya, Nigeria and Pakistan is negligible. In the former foreign banks own on average 4 percent of the assets and in the latter foreign banks own around 1 percent of the assets.

The results of the estimations in this chapter confirm those in chapter 4. Foreign ownership has a negative effect on the level of net interest revenues and on profit before taxes. Due to similar results of the OLS and the GMM method we classify the results as robust. Again, we find that the level of economic development of the host country does not influence the negative relationship between foreign ownership and bank performance. In short, regardless of whether a foreign bank enters a more developed country or a developing country, net interest revenues and profitability of a foreign bank are lower than those of the local domestically owned competitor. Another conclusion is that both domestic and foreign banks generate less net interest revenues in more developed countries than in developing countries. These results confirm those found in chapter 4. In view of the fact that chapter 2 showed that banks in more developed countries have more loans on their balance sheets, the explanation for the negative relationship between economic development and net interest revenues cannot be that banks in more developed countries are less concerned with credit extension. Rather, it seems that in countries with higher GDP per capita the net interest revenues are smaller than in countries with lower GDP per capita.

The conclusions of chapters 4 and 5 are consistent and may be interpreted as evidence for the home advantage theory. Chapter 6 will examine whether foreign ownership has any impact on bank efficiency.

�0�

Chapter 5

�0�

Table 5.4 Foreign ownership and profit

Profit before tax over total assets

(1) (2) (3) (4) (5) (6)

Estimation Technique

OLS GMM OLS GMM OLS GMM

FOREIGN -0.0069***(-7.18)

-0.0097**(-1.94)

-0.0089***(-5.10)

-0.0112(-1.35)

-0.0033**(-3.48)

-0.0086**(-1.68)

OVERHEAD COSTS

0.032*(1.85)

0.0099(0.096)

0.0340**(1.98)

0.0018(0.017)

0.0246(1.48)

-0.0024(-0.03)

LLP -0.470***(-10.94)

-0.590*(-1.80)

-0.4775***(-11.03)

-0.5698*(-1.81)

-0.3602***(-8.54)

-0.4481(-1.26)

EQUITY 0.095***(17.18)

0.085(1.56)

0.095***(17.12)

0.085(1.57)

0.045***(7.79)

0.088*(1.74)

INFLATION 0.00069***(6.62)

0.00078(1.06)

0.0007***(6.43)

0.0006(0.82)

0.00032***(2.63)

0.00058(0.74)

GROWTH 0.00010(0.47)

-0.0008(-0.90)

0.00008(0.39)

-0.0006(-0.83)

-0.00001(-0.05)

-0.00076(-0.99)

GDPPC 0.0014***(4.62)

0.0015(1.34)

0.0010**(2.42)

0.0005(0.34)

0.0003(1.04)

0.0016(1.43)

GDPPC *FOREIGN

0.0008(1.40)

0.0014(0.60)

PROFBT1 0.5433***(24.30)

0.0091(0.072)

D1999 0.0009(0.83)

0.0009(0.93)

0.0009(0.78)

0.0008(0.86)

D2000 -0.0005(-0.43)

0.0002(0.14)

-0.0005(-0.45)

0.00005(0.044)

-0.0013(-1.35)

-0.0007(-0.51)

D2001 -0.0020(-1.43)

-0.0040**(-2.02)

-0.0020(-1.48)

-0.0037**(-2.03)

-0.0036***(-3.19)

-0.0052***(-2.92)

Constant 0.0027**(1.75)

0.0081(0.89)

0.0037**(2.19)

0.0102(1.10)

0.0035**(2.56)

0.0085(1.06)

RSS 0.5260 0.5382 0.5254 0.5342 0.2483 0.3725

TSS 0.6894 0.6894 0.6894 0.6894 0.4921 0.4921

Walt time dummy

5.223p=0.156

8.498p=0.037

5.251p=0.154

8.091p=0.044

10.246p=0.006

9.394p=0.009

M1 21.292p=0.00

-4.173p=0.000

21.318p=0.00

-4.205p=0.000

-5.485p=0.000

-2.073p=0.038

M2 16.253p=0.00

-1.384p=0.166

16.263p=0.00

-1.370p=0.171

-0.491p=0.624

Sargan 27.64p=0.48

35.09p=0.32

36.14p=0.28

�0� �0�


Note: the dependent variable in all equations is profit before tax. See also notes to Table 5.2 *, ** and *** indicate significance levels of 10, 5 and 1 percent respectively. T-values in paren-theses.

�0�

�0�

Chapter 6Efficiency. Do Institutions Matter? *

6.1. Introduction This chapter extends our analysis of the performance of foreign banks. Both chapter 4 and chapter 5 focused on net profit. In this chapter we will turn to efficiency as an indicator of bank performance. Profitability reduces the chance of going bankrupt and allows for paying dividends to the shareholders and for reinvesting in the company. Two ways in which banks can raise profit is by operating more cost efficient and/or more profit efficient. First, banks could improve cost efficiency by reducing the costs per unit of output for a given set of output quantities and input prices. Next, banks could improve profit efficiency by putting together superior combinations of inputs and outputs. This chapter will specifically focus on whether differences in banks’ cost efficiency are associated with foreign vs. domestic ownership. In line with the small amount of papers available on this issue, we examine whether, on average, a rise in foreign ownership has positive or negative effects on bank efficiency. Moreover, and more importantly, we investigate to what extent the foreign ownership-bank efficiency relationship depends on the institutional, i.e. the political, legal and regulatory framework of i) the host country, ii) the home country and iii) the difference between the institutional framework in the home and the host countries. To the best of our knowledge, this is the first study available that provides some empirical evidence on this policy relevant topic.

Berger and Humphrey (1997) survey 130 efficiency studies of financial institutions, of which a few address the impact of foreign ownership. Most importantly, the article clearly shows that a general conclusion regarding the efficiency effect of foreign ownership cannot be drawn based on the available empirical literature. My own literature survey (Table 6.1) suggests that foreign banks in transition and developing markets show higher efficiency than their domestically owned counterparts. Also, foreign banks in developed countries exhibit lower efficiency in comparison with domestic banks. For example, Berger et al. (2000) perform an analysis of cross border banking efficiency in France, Germany, Spain, the United Kingdom, and the United States during the 1990s. On average, they find that domestic banks in these countries have both higher cost efficiency and higher profit efficiency than foreign banks operating in the country. However, the authors also find, after disaggregating their results, that foreign banks may be less efficient than foreign banks from most foreign countries, may be about equally efficient with foreign banks from some foreign countries, but may be less efficient than foreign banks from one (the US) of the foreign countries. Thus, the relative efficiency of foreign vs. domestic ownership appears to depend on host and home country conditions. Does foreign and domestic bank efficiency differ, and to

* This chapter is currently under second review with the Journal of Banking and Finance.

�0� �0�

Chapter 6

�0�

Table 6.1 Summary of the findings on the efficiency of foreign banks

Authors Country Period Technique Empirical Findings

DeYoung and Nolle (1996)

US 1985-1990

DFA Foreign owned U.S. banks are less profit efficient than U.S. owned banks as the former may have placed growth ahead of profitability.

Hasan and Hunter (1996)

US 1984-1989

SFA Japanese multinational banks operating in the US are significantly less cost and profit efficient than their US counterparts.

Bhattacharya, Lovell and Sahay (1997)

India 1986-1991

DEA Foreign owned banks are found to be somewhat more efficient than privately owned domestic banks but government owned banks are more efficient than both.

Chang, Hasan, Hunter (1998)

US 1984-1989

SFA Foreign owned multinational banks are significantly less cost efficient than US owned multinational banks.

Berger et al. (2000)

France, Germany, Spain, UK and USA

1993-1998

DFA Foreign banks may be less efficient than foreign banks from most foreign countries, may be about equally efficient with foreign banks from some foreign countries, but may be less efficient than foreign banks from one (the US) of the foreign countries.

Grigorian and Manole (2002)

17 European transition nations

1993-2000

DEA Foreign owned bank are significantly more cost efficient than domestic banks.

Isik and Hassan (2002)

Turkey 1988, 1992, and 1996

Foreign banks seem to be significantly more efficient than their domestic peers.

Jemric and Vujcic (2002)

Croatia 1995-2000

DEA Foreign banks are significantly more efficient than domestic banks.

Miller and Parkhe (2002)

12 EU countries Arg., UK, Switz., Australia, US, Jap Can., Chile, India,

1989-1996

SFA US owned banks are more X-efficient than other foreign owned banks in bank-oriented financial systems, but less X-efficient in capital-market oriented systems.

Nikiel and Opiela (2002)

Poland 1997-2000

DFA Foreign banks are more cost efficient and less profit efficient than other banks.

Yildirim and Philippatos (2003)


1993-2000

SFA and DFA

Foreign banks are more cost efficient but less profit efficient than domestically owned private banks and state-owned banks.

�0� �0��0�

Efficiency. Do Institutions Matter?

Hasan and Marton (2003)

Hungary 1993-1997

SFA Foreign banks and banks with higher foreign ownership involvement are associated with lower inefficiency.

Matousek and Taci (2004)

Czech Republic

1993-1998

DFA Foreign banks are more cost efficient than other banks, although their efficiency has been comparable with the good small banks in early years of operations.

Weill (2003) Czech Republic and Poland

1997 SFA Foreign banks are more cost efficient than domestic banks. This advantage does not result from differences in the scale of operations or the structure of activities.

Green, Murinde Nikolov (2004)


1995-1999

System of equations

Foreign banks are not more efficient than domestic banks. Foreign ownership does not significantly reducing banks costs.

Sturm and Williams (2004

Australia 1988-2001

DEA New foreign banks are more input efficient than domestic banks, mainly due to their superior scale efficiency.

Bonin, Hasan, Wachtel (2005)


1996-2000

SFA Foreign owned banks are more cost efficient than other private banks.

Fries and Taci (2005)


1994-2001 SFA Privatised banks with majority foreign ownership are the most efficient and those with domestic ownership are the least.

Havrylchyk (2006)

Poland 1997-2001 DEA Foreign banks are more efficient than domestic owned banks.

Zajc (2006) 6 CEE nations 1995-2000 SFA Foreign banks are less cost efficient than domestic banks.

what extent is foreign bank efficiency affected by the institutional difference of the host country and similarities between host and home countries’ legal and regulatory frameworks? Theory suggests that there are some inherent attributes of foreign banks that limit their performance vis-à-vis domestic banks. Berger et al. (2000) differentiate between home field advantages and global advantages. The global advantage hypothesis states that foreign banks might benefit from competitive advantages relative to their domestic owned peers. Foreign-owned banks use more advanced technologies due to a stiff home market competition. Foreign banks might also become more competitive relative to domestic banks due to an active market for corporate control in the home country and because they have access to an educated labour force with the ability to adapt to new technologies. Similarly, Havrylchyk (2006) suggests that foreign banks might profit from better risk management and reliance on modern information technologies. The home field advantage hypothesis predicts foreign banks to suffer from disadvantages relative to domestic banks. Foreign-controlled banks are assumed to underperform domestically controlled banks due to higher costs of providing the samefinancial services or lower revenues, from problems in providing the same quality

��0 ��

Chapter 6

��

and variety of services as domestic institutions. Hymer (1976) argues that foreign firms are likely to face competitive disadvantages relative to national firms. The latter have the general advantage of better information about their country’s economy, language, laws and politics. This leads to the hypothesis that foreign banks suffer more from a bad institutional framework in the host country than domestic banks. Foreigners and nationals may receive very different treatment from governments, consumers and suppliers. If this hypothesis holds, the foreign ownership-bank efficiency relationship will e.g. differ for countries that score high or low on a good institutional framework. In countries with a good institutional framework, the impact of foreign ownership on bank efficiency will be less negative or more positive than in countries where the institutional framework is bad. Mian (2006) develops a theoretical model that provides some ideas about the effect of institutional distance on foreign bank behaviour. He assumes e.g. that institutional distance between the home country and the host country will cause higher informational, agency, or enforcement costs for foreign banks operating abroad. Mian states “Working in an environment with a different corporate culture, legal environment, or regulatory frame work might increase the asymmetry in information and make it more difficult for the CEO of a foreign bank to design policies that are specifically tailored for the developing country” (Mian, 2006, p. 1470).

This chapter examines the effect of the institutional framework on the foreign ownership-bank efficiency relationship by using stochastic frontier analysis for a broad sample of 1363 banks in 93 countries. It explores the issue by addressing four related questions: (i) What is the effect of foreign ownership on bank inefficiency? (ii) What is the effect of the institutional framework in the host country on this relationship? (iii) What is the effect of institutional difference between the host and the home country on the inefficiency of foreign banks? (iv) What is the effect of the institutional framework in the home country on the inefficiency of foreign banks? We will show that, on average, foreign ownership has a negative effect on bank efficiency. More importantly, we provide empirical evidence for the hypothesis that in countries with a good regulatory environment and good governmental governance, the bank efficiency reducing effects of a rise in foreign ownership are considerably lower. Third, we find that higher quality of the home countries’ institutional framework reduces foreign bank inefficiency. Fourth, the smaller the institutional distance between the host and the home country, the lower is foreign bank inefficiency.

The chapter is organised as follows. Section 6.2 presents the econometric framework and Section 6.3 the data. In Section 6.4 we present the results of our empirical investigation. Section 6.5 concludes.

6.2 Econometric framework6.2.1 Assumptions of the modelIn line with Berger and Mester (1997) we measure cost efficiency (technical efficiency) as how close a bank’s cost is to what best practice banks cost would be for producing

��0 ��


the same output bundle under the same conditions. As costs functions are not directly observable, inefficiencies are measured relative to an efficient cost frontier. Most studies on the impact of foreign ownership on cost efficiency use data envelopment analysis (DEA) or stochastic frontier analysis (SFA). We use stochastic frontier analysis as it controls for measurement error and other random effects1. More specifically, we use the Battese and Coelli (1995) SFA model, henceforth the BC model. A first advantage of the BC model over the standard two-step SFA approach of Aigner, Lovell and Schmidt (1977) and Meeusen and van den Broeck (1977) is that the former estimates the cost-frontier and the coefficients of the efficiency variables simultaneously2. Wang and Schmidt (2002) show that the two-step approach suffers from the assumption that the efficiency term is independent identical half-normally distributed in the first step, while in the second step the efficiency terms are assumed normally distributed and dependent on the explanatory variables. This method inherently renders biased coefficients. A second advantage of the BC model is that it has the ability of estimating with an unbalanced panel dataset. Using an unbalanced panel leads to an increase of observations.

The general BC model specifies a stochastic cost frontier with the following properties:

lnCi,t=C(yi,t,wi,t,qt;b)+ui,t+ni,t (6.1)

Where Ci,t is the total cost bank i faces at time t and C(yi,t,wi,t;b) is the cost frontier. In this model bank efficiency is measured relative to a global best-practice frontier. The advantage of this approach is that a frontier formed from the complete data set across nations allows for a better comparison across nations (Berger and Humphrey, 1997). However, cross-country comparisons are difficult to interpret because the regulatory and economic environments faced by financial institutions are likely to differ importantly across nations. In our analysis, we therefore specifically control for such cross-country differences in economic and financial environment when estimating the efficiency frontier.

Within the cost frontier, yi,t represents the logarithm of output of bank i at time t, wi,t is a vector of the logarithm of input prices of bank i at time t, q are country specific variables and b is a vector of all parameters to be estimated. The term ui,t

captures cost inefficiency and is independent identical distributed with a truncated normal distribution.3 ni,t captures measurement error and random effects, e.g. good

1 Non-parametric techniques do not allow for measurement error and luck factors, thus attribu-ting any deviation from the best-practice bank to technical inefficiency. For a more extensive review of the non-parametric and the parametric approach we refer to Matousek and Taci (2004).

2 Kumbhakar and Lovell (2000) discuss other SFA models that also solve for exogenous influences on efficiency simultaneously. Coeli (1996) has implemented the BC model into a statistical soft-ware package FRONTIER Version 4.1.

3 Thus, the total costs a bank faces are never lower than the costs of the frontier. For a graphical

��

Chapter 6

��

and bad luck, and is distributed as a standard normal variable. Both ui,t and ni,t are time and bank specific and represented as4:

ui,t~N+(mi,t,su2) and ni,t~iidN(0,s

n2) (6.2)

mi,t = d0+Sdn,i,t zn,i,t (6.3)

n

Equation (6.3) models inefficiency and its explanatory variables. The z in Equation (6.3) represents the vector of n variables that drive the inefficiency (m) of bank i at time t. The deltas represent the coefficients. Equations (6.1) and (6.3) are solved in one step by using maximum likelihood.

6.2.2 Specification of the modelWe select a transcendental logarithmic form for the cost function as developed by Christensen, Jorgenson, and Lau (1973). We use this specification since it results in a better fit of the frontier than e.g. the Cobb-Douglass form (Kumbhakar and Lovell, 2000)5. A time trend is included in the function to allow for changes in technology over time. Since a translog function is a second order approximation, a trend is included with a t and t2 term (Coelli, Rao, and Battese, 1998). In line with the general model we specify the cost function as:6

ln(TC)=b0+b1ln(PF) = b2 ln(loans) + b3 ln (securities) + b3 1(ln(PF))2 (6.4)

PL PL

2 PL

+ b5 1 (ln(loans))2 + b6

1 (ln(securities))2 + b7 ln (PF) ln(loans) 2 2 PL

+ b8

ln (PF) ln(securities) + b9 ln (PF) Year + b10 ln(loans) ln(securities)

PL PL

+ b11 ln(loans) Year + b12 ln (securities)* Year + b13Year +1 b14Year 2

2

+ b15GDPPC + b16r + b17 concentrationrate + b18

privatecredit

GDP

+ b19

LLR + b20

otheroperationgincome+ ui,t + ni,t

grossloans totalassets

representation of the frontier and its dynamics see Berger et al. (1993). The authors show how inefficiency is determined by allocative inefficiency and technical inefficiency.

4 Testing for heteroskedasticity of the inefficiency term (u) is impossible as we use an unbalanced panel. The CB model lacks providing values of the residuals, impeding conclusions on heteros-kedasticity of (v).

5 The Fourier flexible form is even more general than the translog production function. However, the results of both are more or less in line with each other (Berger and Mester, 1997).

6 In contrast to Fries and Taci (2005) we scale total costs and the input price by one price, person-nel expenses, in order to guarantee linear homogeneity of the cost function (see Kuenzle, 2005). Total costs include interest expenses, personnel expenses and other operating expenses.

��


The dependent variable of the cost function is total costs (TC) and includes labour, interest, and other costs. Furthermore the specification assumes two input prices and two output quantities. The first input price is the price of funds (PF). This price is defined as the ratio of a bank’s interest costs, scaled by the sum of deposits and other interest bearing funding. The second input price is the price of labour (PL) and is the ratio of personnel expenses, scaled by total assets. Although scaling over total employees instead of total assets gives a better proxy of price of labour, the latter is chosen since for many observations the former is not available. In order to guarantee linear homogeneity in input prices of the cost function, we scale TC and PF by PL. This scaling implies an estimation of coefficients for PF as well as PL with the restriction that the sum of these coefficients is equal to one (see Kuenzle, 2005). The output quantities used are total loans (Loans) and total securities (Securities).

In estimating one best-practice frontier for all banks we control for country-specific and bank-specific variables. By allowing these variables to influence the position of the efficient cost frontier we control for the fact that technological influence and service qualities can vary systematically across countries. Simply pooling all banks across countries would implicitly assume that efficiency differences across banks are attributed entirely to managerial decisions within banks regarding the scale and mix of inputs. Country specific control variables include the level of economic development, proxied by GPD per capita (GDPPC), nominal interest (r), banking market concentration7 and development of the financial sector, proxied by the ratio of credit to the private sector over GDP (private credit/GDP).8

On bank level, we control for bank risk and the output quality as unmeasured differences in product quality may be incorrectly measured as differences in cost inefficiency. As a measure of risk we used loan loss reserves over gross loans (LLR/gross loans). We control for variation in banking service qualities by including the ratio of banks’ other operating income over total assets (other operating income/total assets). We also control for year effects.

For the entire sample, including domestic and foreign banks, we use two specifications of the (in)efficiency equation. The first specification is as follows:

mit = d0+d1

Equity +

d2 Foreign (6.5)

Totalassets

We use this specification to test whether foreign ownership, measured by Foreign, has

7 Miller and Parkhe (2002) find that foreign bank efficiency is strongly influenced by the competi-tiveness of its home country and the host country in which it operates. Concentration ratios for 1998-2001 are calculated as the sum of assets of the three largest banks over the assets of all banks per country.

8 GDP per capita, nominal interest and credit to the private sector figures for 1998-2003 are from the World Bank World Development Indicators 2005. The intermediation ratio reflects differences among the banking sectors in terms of the extent to which they convert deposits into loans, which may be associated with bank holdings of government securities and crowding out of private borrowing by the public sector (Fries and Taci, 2005).

��

Chapter 6

��

a positive or a negative effect on efficiency. Note that a positive d2 implies that an increase in Foreign is associated with an increase in inefficiency. We include the equity position of a bank, measured by (Equity/Total Assets) since the failure to account for the equity position of a bank seems to yield a strong scale bias making large banks more efficient than small banks by virtue of the equity they have built up over time (Berger and Mester, 1997) 9.

We use a second specification to test whether the efficiency-foreign ownership relationship depends on the institutional framework in the host country (Insthost). This specification reads as follows:

mit = d0+d1

Equity +

d2 Foreign + di=3..8 Foreign*Insthost,j (6.6)

Totalassets

Inst denotes a vector of six institutional indicators, to be explained in the next section. The different institutional indicators are included one by one in the equation, which implies that we estimate six different models.

Next, we continue by only considering a reduced sample: a sample of foreign banks only.10 For this sample, we also specify two types of efficiency equations. First, we estimate the effect of institutional similarity (Similarity) between the home and the host country and the institutional framework in the home country (Insthome),

mit = d0+d1

Equity d2 Similarity + di=3..8 Insthome,j (6.7)

Totalassets

where similarity measures institutional distance between the home and host country. For exact measurement of this variable, see section 6.3. Nest, we estimate the effect of institutional similarity between the home and the host country and the institutional context in the host country:

mit = d0+d1

Equity d2 Similarity + di=3..8 Gvrnncehost,j (6.8)

Totalassets

6.3 Data The sample we use is based on data availability. The entire sample includes 1363 commercial banks in 93 countries. All bank level data are from the BankScope database produced by Bureau van Dijk, which covers 90% of banks worldwide. We use the 2006 BankScope issue for PF, PL, TC, other operating income, equity, LLR, loans and securities. One of the main variables in the analysis is foreign ownership (Foreign).

9 Insolvency risk affects costs and profits via the risk premium the bank has to pay for uninsured debt. Moreover, if some banks are more risk averse than others, the former may hold a higher level of equity than maximizes profits or minimizes costs (Berger and Mester, 1997).

10 A foreign bank is defined as a bank of which more than 50% of the shares are owned by non-domestic resident.

��


Most studies on foreign banking use a foreign ownership dummy, a zero-one dummy, which defines a bank to be foreign if more than 50% of the total stock of shares is held by non-domestic residents. Since we are primarily interested in the effect of an increase in foreign ownership of efficiency, we decided to use a continuous variable foreign ownership, defined as the percentage of shares held by foreigners. Conceptually, this implies that we examine the effects of a rise in foreign ownership on efficiency, rather than the differential effect of foreign vs. domestic banks on efficiency. In alternative tests, for reasons of space not presented, we used a foreign bank dummy in line with the literature. This, however, gave qualitatively the same results, which is not surprising, since the bank ownership data appeared to be skewed, in the sense that most banks either have a very large or a very low percentage of foreign ownership, and hence the variable Foreign behaves like a foreign bank dummy. A problem we were confronted with is that ownership data provided by BankScope are not complete in that for almost all banks part of ownership information is missing. To come around this problem, we include only those banks for which ownership is given for at the least 75% of the shares. Moreover, in the case where part of the ownership information was missing, we rescaled the ownership variables in such a way that foreign and domestic ownership add up to 100%. Another problem we were faced with is that each edition of BankScope only provides information on the last available information on ownership distribution, and hence ownership for a bank does not vary over time. To account for changes in ownership distribution per bank over time, we supplement the data with ownership information of the 1998, 1999, 2000, 2001, 2002 and 2003 editions of BankScope. Descriptive information on all bank level variables is given in Appendix 14. This table suggests that the sample is biased in that there is an overrepresentation of banks from developed countries included in the sample. In order to test to what extent our results are driven by our sample, we re-estimated all equations for a smaller sample only including banks from countries with a GDP per capita lower than $10,000. Qualitatively this gave the same result.

The cost function we estimate also includes macro variables, such as GDP per capita, the nominal interest rate and the ratio of domestic credit to the private sector over GDP. These data are extracted from the World Bank World Development Indicators 2005. The nominal interest rate is calculated using the real interest rate and the GDP deflator. Descriptive statistics for the country variables are also included in Appendix 14.

The main aim of the analysis is to see to what extent the relationship between efficiency and foreign-ownership depends on the institutional framework of the host country, the institutional similarity between the home and the host country and the institutional framework of the home country. We decided to use the six institutional indicators as developed by Kaufmann et al. (2005). These indicators all measure another aspect of the institutional context of a country. The aggregate institutional indicators from Kaufman et al. (2005) cover 209 countries and territories for 1996, 1998, 2000, 2002 and 2004 and are based on 352 individual variables measuring perceptions of the institutional context, taken from 37 separate data sources. For the years that

��

Chapter 6

��

are in between, we take the averages of the year before and after. Perceptions of the institutional context include (i) the process by which governments are selected, monitored and replaced. (ii) the capacity of the government to effectively formulate and implement sound policies, and (iii) the respect of citizens and the state for the institutions that govern economic and social interactions among them. The first cluster includes Voice and Accountability (VAC), and Political instability and Violence (PIV), the second Government Effectiveness (GEF) and Regulatory Quality (REG), and the third cluster includes Rule of Law (LAW) and Control of Corruption (COR). Table 6.2 provides exact definitions of the institutional indicators we use in the analysis.

The gains for foreign banks of a better institutional environment come through various channels of transmission. Regarding the relevance of the voice and accountability indicator, we expect a higher level of media independence to increase the quality of information on local developments. As the owners of foreign banks might have more trouble interpreting local information this would especially benefit foreign banks. An increase in political rights could allow for the establishment of labour unions. If working conditions in foreign banks are better than in domestic banks, due to higher standards of foreign banks, the establishment of unions could diminish the relative cost disadvantage of foreign banks.

Table 6.2 Institutional indicators

Institutional indicator Definition

Voice and accountability Measures the various aspects of the political process, civil liberties, political rights, and independence of the media.

Political stability and violence

Measures of perceptions of the likelihood that the government in power will be destabilized or overthrown by possibly unconstitutional and/or violent means, including domestic violence and terrorism.

Government effectiveness Responses on the quality of public service provision, the quality of the bureaucracy, the competence of civil servants, the independence of the civil service from political pressures, and the credibility of the government’s commitment to policies.

Regulatory quality Measures of incidence of market-unfriendly policies such as price controls or inadequate bank supervision, as well as perceptions of the burdens imposed by excessive regulation in areas such as foreign trade and business development.

Rule of law Measures the extent to which agents have confidence in and abide by the rules of society. These include perceptions of the incidence of crime, the effectiveness and predictability of the judiciary, and the enforceability of contracts.

Control of corruption Measures the exercise of public power for private gain, including both petty and grand corruption and state capture.

Source: Kaufman et al. (2005)

��


Second, better political stability and violence could be beneficial especially for foreign banks if they are more risk-avers and have relatively high loan loss provisions. Mian (2006) reviews the idea on why domestic and foreign banks have different risk preferences. The author assumes that domestic banks may be more willing to take on higher levels of risk because of the moral hazard or option values associated with the limited liability of banks. Foreign banks on the other hand, may not be willing to take such high levels of risk because of their ‘franchise value’ at risk, and the added supervision by their home regulatory authority (Mian, 2006). Higher political stability and less violence could also lower foreign banks’ security costs assuming that foreign institutions run a higher risk of becoming a victim of violence.

Better government effectiveness reduces costs of foreign banks as foreign banks are assumed to face more difficulty in dealing with local bureaucracy. Higher independence of the civil service from political pressure lowers foreign banks’ costs in those countries where political pressure against (entry of) foreign banks is prevailing.

Improvements in the regulatory quality help foreign banks if it is accompanied by more adequate banking supervision. For example, foreign banks in Poland have been faced with very strict regulation: banks were forced to provision for loans even when it was certain that the loans would be repaid –for example by the parent company of the borrowing subsidiary- (section 3.1). In addition, foreign banks could spend less time on excessive regulation to which they might be less familiar than their local domestically owned competitors. Weill and Pruteanu-Podpiera (2006) refer to the lack of prudential regulation in the Czech Republic in the period 1989-1993.

The quality of the rule of law affects cost efficiency through the effectiveness and predictability of the judiciary. When going to court is time consuming, bank costs face downward potential. For example, representatives of foreign banks have argued that the legal system in Central and Eastern Europe does not always work in the most efficient way and that in the mid 1990s it took years to realize collateral on real estate in the Czech Republic (section 3.4.1). Finally, enhanced control of corruption influences costs by lowering the costs of bribing. We assume that foreign institutions in particular are subject to corrupt legal or public officers. Thus, we hypothesize that a better institutional framework, i.e. higher values of the six institutional indicators, lowers foreign bank cost inefficiency. Descriptive statistics for home and host country institutional indicators are given in Appendix 14. Table 6.3 gives a correlation matrix of the explanatory variables of bank efficiency for the complete set. Since the institutional indicators are highly correlated, they are separately included in the different specifications.

To test for the relevance of the (dis)similarity between the institutional environment between the host and the home country for foreign bank efficiency, we introduce the variable institutional similarity. This variable is calculated as the correlation between the six home and the six host country institutional indicators. The home country refers to the origin of bank (or investor) with a physical establishment, the foreign owned bank, in another country: the host country. We expect higher values of institutional similarity to lower bank inefficiency. The correlation matrix of the variables used in

��

Chapter 6

��

estimations with the reduced set, the set of foreign banks only, is in appendix 15.

Table 6.3 Correlation matrix of efficiency correlates

EQ/TA FOR VAC PIV GEF REG LAW COR

EQ/TA 1

FOR 0.168 1

VAC 0.025 -0.040 1

PIV 0.064 0.066 0.775 1

GEF 0.021 -0.013 0.822 0.813 1

REG -0.018 -0.007 0.792 0.774 0.908 1

LAW 0.006 -0.024 0.824 0.851 0.977 0.907 1

COR 0.017 -0.021 0.832 0.825 0.975 0.894 0.982 1

EQ/TA is the ratio of bank equity over total assets. FOR is the variable measuring foreign bank ownership. The variable is determined as the percentage of shares owned by foreign investors. The institutional indicators include voice and accountability (VAC), political insta-bility and violence (PIV), government effectiveness (GEF), rule of law (LAW) and control of corruption (COR). Definitions are in Table 6.2.

6.4 Empirical resultsTable 6.4 shows the estimation results for effect of foreign ownership on bank efficiency and the impact of the institutional context of the host country on this relationship. The estimates of the cost frontier for the different specifications are given in Appendix 16. In the discussion of the results we focus on the inefficiency correlates. Results for the foreign variable in specifications 1-7 (Table 6.4) confirm that an increase in foreign ownership positively affects bank inefficiency. An important finding is that this conclusion holds for the set of developed countries as well as for the set of transition and developing countries. Second, the significant coefficients with negative sign of the interactive variables show that in host countries that score high on the institutional indicators, the negative effect of a rise in foreign ownership on bank efficiency decreases. This holds for all institutional indicators used. We interpret this result as a first empirical justification for the hypothesis that foreign banks may find it difficult to deal with host country’s regulations, banking supervision rules, local judiciary in general and corruption and that especially in countries where the institutional environment is bad, the foreign ownership-bank efficiency relationship is negative. The LR-tests indicate that (in)efficiency has a significant impact on the model. Table 6.5 shows the results for the relevance of foreign banks’ institutional environment in the home country on efficiency. The estimation of the pertaining cost functions is given in appendix 17. The SFA is performed with foreign banks only. A shift in focus of the institutional context is made. Instead of looking to the host country institutional context, the home country institutional context is used for the estimation. The home country is the country in which most of the stocks are held. All

��


institutional indicators used have a negative estimate. This means that the better the institutional context in the home country, the less inefficient a foreign bank performs. As the similarity variable has a negative estimate, foreign bank inefficiency reduces as similarity between the home and the host country’s institutions rises. Also for these models (in) efficiency has a seriously impact, given the LR-tests. Table 6.6 shows the estimates of an extended model that is used for the results in table 4. In this model the institutional context in the host country is also taken into account. The cost functions underlying these models are given in appendix 18. The results are similar to those presented in table 6.4. The added variable of host institutions has also a negative sign, which means that efficiency increases as the institutional context in the host country improves. In conclusion this means that the institutional context in the home as well as the host country is important for bank efficiency.

6.5 ConclusionDespite the extant literature on the determinants of bank efficiency, there is little information on the impact of foreign ownership. Moreover, there are no studies that have examined to what extent the foreign ownership-efficiency relationship depends on the institutional context in the home and/or the host country. Mian (2006) suggests that institutional differences between the host and the home country will cause higher informational, agency or enforcement costs for foreign banks operating abroad. We examine six indicators that capture three areas of the institutional environment: (i) the process by which those in authority are selected and replaced, (ii) the ability of the government to formulate and implement sound policies and (iii) the respect of citizens and the state for the institutions which govern their interaction. The analysis of this chapter leads to four conclusions. First, on average foreign banks are less efficient than domestic banks. This result holds for the estimations with the set of banks world wide as well as for estimations with the set of banks in countries with GDDP per capita below $10,000, i.e. the set of transition economies and lower middle income and low income countries. Thus, we find evidence for the home advantage theory both in developed as in developing countries. This finding contrasts with the findings of previous efficiency studies that foreign banks are more efficient in transition economies. Second, a bad institutional context in the host country affects foreign bank efficiency more than domestic bank efficiency. Third, higher institutional difference between the host and the home country lowers the efficiency of foreign banks. Fourth, enhanced home country institutions positively affect the efficiency of foreign banks. Our main results seem to hold for all institutional indicators introduced. Thus, in accordance with Main (2006) our study finds convincing evidence for the relevance of the institutional environment in explaining differences between foreign and domestic banks. Further research could be directed on whether the institutional environment affects state-owned banks and private domestic banks in the same way.

��0 ��

Chapter 6

��

Table 6.4a Bank inefficiency: ownership and host country institutions (all countries)

(1) (2) (3) (4) (5) (6) (7)

Number of observations

3,850 3,850 3,850 3,850 3,850 3,850 3,850

Correlates with bank inefficiency

Constant -10.92 -8.86 -11.31 -8.91 -9.55 -11.38 -10.63

(-8.3)*** (-23.47)*** (-7.28)*** (-20.29)*** (-22.67)*** (-11.63)*** (-11.21)***

EQ -5.91 -6.02 -5.75 -6.04 -5.82 -5.26 -5.91

(-10.86)*** (-16.11)*** (-8.67)*** (-6.65)*** (-27.5)*** (-13.38)*** (-13.28)***

FOREIGN 1.45 1.57 1.72 1.78 2.36 1.94 2.16

(15.45)*** (19.23)*** (18.91)*** (13.16)*** (26.34)*** (19.73)*** (26.79)***

Voice and accountability Host * Foreign

-0.35

Political instability and violence Host * Foreign

-0.43

Government effectiveness Host * Foreign

-0.45

(-9.04)***

Regulatory quality Host * Foreign

-1.19

(-17.85)***

Rule of law Host * Foreign

-0.58

(-8.62)***

Control of corruption Host * Foreign

-0.82

(-7.23)***

Sigma-squared 4.13 3.30 4.26 3.32 3.51 4.11 4.01

(9.61)*** (27.02)*** (8.71)*** (22.73)*** (26.03)*** (12.97)*** (12.13)***

Gamma 0.99 0.99 0.99 0.99 0.99 0.99 0.99

(1073.2)*** (1022.5)*** (814.4)*** (1814.4)*** (1161.2)*** (904.75)*** (1238.8)***

LR-test1 931.26*** 929.20*** 931.97*** 930.52*** 936.44*** 932.27*** 938.34***

(-6.82)***

(-4.69)***

��0 ��


The sample includes foreign and domestic banks. Estimations for the cost frontier are in Ap-pendix 16. * significant at 0.1 level; ** significant at 0.05 level; *** significant at 0.01 level. Sigma-squared, Gamma, and the LR-test give some information about the fit of the model. An explanation is given in Appendix 19.

Table 6.4b Bank inefficiency: ownership and host institutions (low-income countries)

(1) (2) (3) (4) (5) (6)

Number of observations 1810 1810 1810 1810 1810 1810

Correlates with bank inefficiency

Constant -10.77 -9.99 -8.32 -8.05 -8.22 -8.40

(-5.36)*** (-3.84)*** (-5.57)*** (-4.64)*** (-4.85)*** (-3.84)***

EQ -5.40 -5.69 -6.10 -6.16 -6.14 -6.06

(-8.67)*** (-7.92)*** (-8.07)*** (-6.75)*** (-7.01)*** (-5.61)***

FOREIGN 1.47 1.17 1.29 1.66 1.18 0.98

(14.22)*** (10.8)*** (6.47)*** (5.63)*** (5.62)*** (4.49)***

Voice and accountability Host *Foreign

-1.03

(-5.16)***

Political instability and violence Host *Foreign

-0.001

(-0.02)

Government effectiveness Host *Foreign

-1.00

(-5.81)***

Regulatory quality Host * Foreign

-1.74(-5.66)***

Rule of law Host * Foreign

-0.55(-3.78)***

Control of corruption Host * Foreign

-1.71(-4.48)***

Sigma-squared 3.90 3.67 3.11 3.02 3.08 3.13

(5.97)*** (4.4)*** (6.45)*** (5.4)*** (5.61)*** (4.44)***

Gamma 0.98 0.98 0.98 0.98 0.98 0.98

(303.51)*** (240.43)*** (303.39)*** (257.68)*** (256.3)*** (211.38)***

LR-test 346.12*** 344.71*** 344.17*** 349.21*** 343.78*** 348.95***

��

Chapter 6

��

The sample includes foreign and domestic banks. Estimations for the cost frontier are in Ap-pendix 16. * significant at 0.1 level; ** significant at 0.05 level; *** significant at 0.01 level. Sigma-squared, Gamma, and the LR-test give some information about the fit of the model. An explanation is given in Appendix 19.

��


Table 6.5 Foreign bank inefficiency: institutional similarity and home institutions

(1) (2) (3) (4) (5) (6)

Number of observations 1,385 1,385 1,385 1,385 1,385 1,385

Correlates with bank inefficiencies

Constant -4.46 -5.42 -7.71 -4.38 -7.40 -5.03

(-4.13)*** (-3.67)*** (-3.51)*** (-3.36)*** (-3.86)*** (-3.47)***

Bank specific variables

Equity/ total assets -8.25 -8.86 -7.68 -8.22 -7.79 -8.49

(-6.59)*** (-5.68)*** (-8.08)*** (-5.31)*** (-8.93)*** (-5.27)***

Institutional context home-host country

Institutional similarity -0.50 -0.51 -0.54 -0.44 -0.55 -0.45

(-4.64)*** (-4.14)*** (-4.77)*** (-3.77)*** (-5.03)*** (-3.85)***

Institutional context home country

Voice and accountability Home

-0.72

(-4.84)***

Political instability Home -0.75

(-4.15)***

Government effectiveness Home

-0.58

(-5.21)***

Regulatory quality Home -0.70

(-3.78)***

Rule of law home -0.56

(-5.11)***

Control of corruption Home

-0.43

(-3.93)***

Sigma-squared 2.44 2.80 3.67 2.45 3.48 2.63

(5.39)*** (4.6)*** (4.33)*** (4.33)*** (5.34)*** (4.38)***

Gamma 0.98 0.98 0.98 0.98 0.98 0.98

(212.27)*** (188.03)*** (214.98)*** (167.06)*** (307.08)*** (163.25)***

LR-test 236.87*** 236.58*** 237.14*** 235.97*** 236.78*** 234.79***

The sample consists of foreign banks only. Estimations for the cost frontier are in Appendix 17. * significant at 0.1 level; ** significant at 0.05 level; *** significant at 0.01 level. Sigma-squared, Gamma, and the LR-test give some information about the fit of the model. An ex-planation is given in Appendix 19.

��

Chapter 6

��

Table 6.6 Foreign bank inefficiency: institutional similarity and country institutions

(1) (2) (3) (4) (5) (6)

Number of observations 1,385 1,385 1,385 1,385 1,385 1,385

Correlates with bank inefficiencies

Constant -4.44 -5.53 -3.67 -5.40 -4.51 -4.31

(-3.22)*** (-7.67)*** (-3.69)*** (-4.48)*** (-3.97)*** (-3.19)***

EQ -7.50 -8.21 -7.16 -6.70 -7.63 -7.05

(-4.93)*** (-24.77)*** (-6.42)*** (-8.03)*** (-6.68)*** (-5.53)***

Institutional similarity -0.51 -0.47 -0.19 -0.21 -0.20 -0.08

(-3.86)*** (-6.57)*** (-2.52)** (-1.02) (-2.68)*** (-1.18)

Institutional context home

Voice and accountability Home

-0.72

(-3.41)***

Political instability and violence Home

-0.72

(-8.42)***

Government effectiveness Home

-0.50

(-4.62)***

Regulatory Burden Home

-0.76

(-5.29)***

Rule of law Home -0.50

(-4.95)***

Control of corruption Home

-0.44

(-3.91)***

Institutional context host country

Voice and Accountability Host

-0.52

(-3.6)***

Political stability Host -0.44

(-5.77)***

Government effectiveness Host

-0.68

(-5.28)***

��


(1) (2) (3) (4) (5) (6)

Regulatory quality Host -1.33

(-7.9)***

Rule of law Host -0.68

(-5.7)***

Control of corruption Host

-0.86

(-4.44)***

Sigma-squared 2.53 2.88 2.31 3.19 2.62 2.51481

(4.07)*** (10.28)*** (5.1)*** (6.55)*** (5.29)*** (4.24)***

Gamma 0.98 0.98 0.97 0.982426 0.98 0.98

(163.63)*** (404.16)*** (186.73)*** (383.41)*** (230.3)*** (178.03)***

LR-test 240.27*** 239.31*** 238.26*** 248.06*** 238.87*** 243.49***

The sample consists of foreign banks only. Estimates for the cost frontier are in Appendix 18. *, **, *** significant at the 0.1, 0.05, and 0.01 level. Sigma-squared, Gamma, and the LR-test give some information about the fit of the model. An explanation is given in Appendix 19.

��

��

Chapter 7Summary and conclusion

7.1 IntroductionThe aim of this thesis is to deepen our understanding of foreign bank behaviour and performance. The analysis consists of six chapters, organized in two parts. In Part I of this book we examine the entry and the nature of foreign bank activities in Central and Eastern Europe. In Part II we empirically study the impact of foreign ownership on bank performance.

The first research question is i) What is the role of foreign banks in the financial sector in Central and Eastern Europe? Chapter 1 consists of an analysis based on macro data from a national bank survey. Chapter 2 performs a comparative analysis based on micro data. The second research question is: What are the reasons for foreign banking in Central and Eastern Europe and how do these reasons relate to the mode of entry and the organizational form of the foreign owned bank? The available evidence is scarce and based on a few country studies. In chapter 3 conclusions are based on a multiple case study. Finally, this book focuses studies the question: Does foreign ownership affect bank performance? Chapters 4 and 5 present results on the relationship between foreign ownership and profitability based on OLS and GMM estimations. Chapter 6 presents the evidence on the relationship between foreign ownership and bank efficiency using the stochastic frontier technique. The remainder of this chapter rounds up the main findings of the chapters and provides for some suggestions for further research.

The aim of the rest of this chapter is to summarize the main findings of the analyses and propose some directions for future research.

7.2 Summary of the main findingsIn chapter 1 we start our study on foreign banking by focusing on the role of foreign banks in the financial development of Central and Eastern Europe and specifically on their role as providers of credit to the private sector. The chapter starts out amending the global advantage theory of Berger (2001) that assumes that foreign banks have a comparative advantage over domestic banks due to their ability to export at low cost best-practice banking. Focussing on Central and Eastern Europe, this theory may be not hold due to a lack of track record of company performance, the problems good firms in this region have to signal that they are a ‘good firm’ and the quality of the judiciary. Next, the chapter measures foreign bank presence in Central and Eastern Europe in two ways. Data show that in 1995 one in four banks was foreign owned and by 2004 more than 60 per cent of the banks were foreign owned. However, measuring the importance of foreign banks by their relative number underestimates

��

their relevance in the financial system: by 2004, foreign bank’ assets had accounted for more than 80% of total banking assets in Central and Eastern Europe, while in 1995 foreign banks’ assets accounted for less than 5% of total banking assets. It appears that foreign banks in European transition economies are attracted mainly from Austria, Belgium, Germany and Italy countries. To a lesser extent, foreign banks from the US, France and the Netherlands play a role. Intraregional financial investment is virtually non-existent. Data do not reveal a one-on-one relationship between growth in foreign bank presence and growth in credit to the private sector. In Central and Eastern Europe, credit to the private sector remained unchanged between 1995 and 2001 on a 23 percent of GDP level and only started to take off from 2002 to 34 percent of GDP in 2004. Foreign banks have changed their lending behaviour over time. From 1993-1998 foreign banks extended more credit to the public sector than to the private sector. Only from 1998, credit to the private sector by foreign banks exceeded their credit to the public sector. However, both findings are not unique for foreign banks as they also apply to credit supply by domestic banks. Only from 2000, domestic banks and foreign banks the objectives of foreign and domestic banks differ, as the latter primarily focus on extending credit to the private sector. The chapter finds no evidence for the global advantage theory as up to 1998 foreign banks underserviced the private sector in Central and Eastern Europe. Even in 2000, foreign banks underserviced the private sector in Croatia, the Czech Republic, Lithuania and Poland. However, foreign banks seem to be more profitable in terms of ROE and ROA although the data do not suggest that foreign banks incurred less overhead costs or earned higher net interest revenues.

Chapter 2 highlights the growth in foreign bank size in Central and Eastern Europe if compared to foreign bank size in the former EU15 countries and in less developed European transition economies. Aggregated loan data from bank balance sheets as available in the BankScope database confirm the findings in chapter 1 that intermediation in CEE remained relatively stable during the late 1990s and takes off only from 2002. Further examination of the balance sheets data finds evidence that foreign banks are more risk avers than domestic banks as problem loans and loan loss reserves of foreign banks in all three regions exceed those of domestic banks. The data also suggest that foreign banks in transition economies have more non-earning assets than foreign banks in developed economies. Next, we find evidence that foreign banks in both the EU15 and Central and Eastern Europe have more loan and lease contracts with the corporate sector than domestic banks.

In chapter 3 the results of an interview project show that foreign banks primarily entered Central and Eastern Europe both as a result of high expected profit and because of reasons related to reputation and status. Regarding the latter, the majority of banks seem to have been motivated more by following the competition than by the need to follow their existing customers. Chapter 3 observes some dynamics in customer focus. Banks that entered Central and Eastern Europe in an early stage, usually the global banks, did follow non-financial investment. Over the years, customer focus of foreign banks diversified because of i) decreasing margins in the corporate sector

��

Summary and conclusion

due to increased competition and ii) privatisations of former state owned banks. The shift in customer focus was also accompanied by a shift in the mode of entry. While the early enterers preferred to establish greenfield investments, foreign banks that entered at the end of the nineties preferred to take over existing banks as the increased competition made the greenfield alternative less attractive.

In Part II, the book shifts the attention to empirical analyses on the effect of foreign ownership on bank performance. Following the method of Demirgüç-Kunt and Huizinga (1998), we test whether foreign ownership has an effect on bank profitability. We consider two variables of foreign ownership: i) a continuous variable and ii) a dummy variable. The results in chapter 4 show that in European transition economies foreign ownership is negatively related to banks’ net interest revenues, overhead costs and profitability. Several other specifications confirm the robustness of the negative relationship between foreign ownership and bank profitability. First of all, estimation results show that economic development does not influence the negative relationship between foreign ownership and bank profitability. Neither the competitive conditions in the banking sector, nor the presence of foreign banks have any impact on this relationship.

In chapter 5, we extend the analysis of chapter 4 with a BankScope 1998-2001 panel of banks worldwide. The data show that worldwide one third of the banks are foreign owned. Differences in foreign bank presence arise when we group countries according to economic development. We find evidence that banks prefer to enter new markets in upper middle income countries, for example the CEE region, and that their market share, measured by relative assets, is highest. This would suggest that the competition in mature banking markets in high income countries and the risk of doing business in low income countries both deter foreign banks from being able to similar market shares as they have in the upper middle income countries. We use OLS and GMM estimations to examine the effect of foreign ownership on banks’ net interest revenues and profitability. The results confirm the findings of chapter 4 regarding an inverse relationship between foreign ownership and bank performance.

Chapter 6 focuses on the relationship between foreign ownership and bank efficiency. Up to now, the scarce foreign bank efficiency literature has found that foreign banks in developing countries are more efficient than domestic banks, while foreign banks in developed countries are less efficient than domestic banks. Besides the evidence on the relevance of host country characteristics for foreign bank efficiency, Berger et al. (2000) and Havrylchyk (2006) argue that home country characteristics matter for foreign bank efficiency. Based on a 1998-2001 panel of banks in 93 countries worldwide we find using the stochastic frontier approach, that foreign ownership is negatively related to cost efficiency. With additional estimation results we show that the improvement in the institutional context of the host country positively affects foreign banks’ cost efficiency. Also we show that an increase in the quality of the institutional context in the home country raises foreign bank matters. Finally, we provide evidence that similar quality of the institutional environment between the host and the home country positively affects foreign bank efficiency.

��0

Chapter 7

��0 ��

7.3 Policy implications, suggestions for further research and conclusionSince the 1990s, Central and Eastern European banks have increasingly become foreign owned. The evidence on foreign bank behaviour in Central and Eastern Europe shows that several years past before their relative credit supply to the private sector met their relative presence. Several factors seem to underlie this reserve. First, the entry strategy of the first foreign banks entering Central and Eastern Europe in the early 1990s entailed a focus on existing clients from Western Europe and on the bigger companies in the host country. Though this strategy might have confirmed some in their criticism that foreign banks underservice the lower segments in the credit market, this strategy was also determined by some specific features of banking in transition economies and developing countries. Assuming that banks prefer safe banking to risky banking, banks i) prefer to be able to distinguish a ‘good borrower’ from a ‘bad borrower’, i.e. the adverse selection problem and ii) need to be able to seize the collateral in cause of default or bankruptcy. Foreign banks in Central and Eastern Europe seem to have been suffering from the inability to overcome the adverse selection in all credit segments, to price collateral and to seize it in case of default. We notice that in Central and Eastern Europe credit to the private sector in terms of BNP remained stunted until the early 2000s and argue that the diversification of banks’ loan portfolio towards the SME and retail sector in the late 1990s seems to have been induced primarily by the competitive conditions in corporate banking rather than by a voluntary strategy to resolve credit constraints in the lower segments of the market.

We conclude that if governments and financial authorities in transition economies and developing economies decide to allow foreign banks to enter their national financial sector with a physical presence, policy should be directed at mitigating the obstacles associated with (foreign) banking in such economies. First we argue that, as the adverse selection problem in Central and Eastern Europe was largely due to the lack of historical data of borrowers, governments should consider initiating an institution that pools financial data of companies nationwide and provides access for banks to these data. In addition, an official rating office would enhance the ability of a ‘good borrower’ to signal that she is a ‘good borrower’ thereby increasing her chances of obtaining a bank loan.

A second challenge of foreign banking in transition economies, besides overcoming the adverse selection, has been to deal with the time-consuming and unpredictable process of going to court to seize collateral in case of loan default or bankruptcy. In this respect, policy should be directed at providing for laws that protect both debtors and creditors according to the rules of reasonableness and fairness. In many countries in Central and Eastern Europe the incentive to such policy was the prospect of EU accession. At the same time policy should be initiated to i) enhance the enforcement of the law, e.g. by monitoring the relationship between bankruptcy administers and judges and retrain judges with respect to new law, and to ii) eliminate corruption within the judiciary in general and of judges in particular, e.g. by raising salaries.

��0��0 ��

Summary and conclusion

The result from our central bank survey that in Central and Eastern Europe foreign banks have been more profitable than domestic banks in the period 1995-2000 confirms results in Bonin, Hasan and Wachtel (2005). Majnoni, Shakar and Várhegi (2003), Claessens et al. (2001), Kraft and Tirtiroglu (1998) and Sabi (1996) An interesting venue for further research would be to empirically assess the correlates of this superior foreign bank performance, as we show that, in this timeframe, net-interest revenues of domestic banks seem to have been higher than that of foreign banks and that overhead costs of foreign and domestic banks have been of comparable size. One could consider including the quality of loan portfolios in line with Berger and DeYoung (1997), who focus on the relationship between problem loans and cost efficiency in commercial banks.

Though foreign banks might have been more profitable in the early transition period in Central and Eastern Europe, our empirical analysis of foreign bank performance shows that, in 2001, the superior profitability of foreign banks has disappeared. More important is that, in an additional study with data on foreign banks worldwide, we find that in general foreign ownership is negatively related to bank profitability, irrespective of the economic development of the host country in which the foreign bank is active. This result contrasts with the findings of a study into the determinants of commercial bank profit of Demirgüç-Kunt and Huizinga (2000) with data from 1998-1995, that foreign banks are more profitable than domestic banks in developing countries but less profitable and that the former are less profitable in developed countries. Additional research could explain whether the sample period influences the way in which the economic development of the host country affects the relationship between ownership and profitability. To complement the literature on the nexus between ownership and bank performance, additional research could be directed at the correlation between foreign ownership and concentrated ownership and the correlation between domestic ownership and dispersed ownership.

Complementary evidence on the relevance of the institutional environment is shown by our research into the determinants of banks’ cost efficiency. We find that an increase in the quality of institutions enhances foreign bank efficiency. We argue that governmental policy aimed at improving the institutional context would benefit bank efficiency especially in those countries where the banking sector is dominated by foreign banks, e.g. in Central and Eastern Europe, Sub-Saharan Africa, Latin America and the Caribbean. Examples of such policy would consist of policy directed at eliminating people’s receptiveness toward corruption, and raising the effectiveness and predictability of the judiciary. Other examples would be to aim for increasing the quality of banking supervision, eliminating bureaucracy at the governmental level and provide for media independence.

Regarding the relationship between foreign ownership and bank efficiency we find with 1998-2001 data of banks worldwide that are negatively related in both developed economies and developing countries. This result contrasts with the findings of studies as Grigorian and Manole (2002), Jemric and Vujcic (2002), Nikiel and

��

Chapter 7

Opiela (2002), Yildirim and Philippatos (2003), Hasan and Marton (2003), Matousek and Taci (2004), Weill (2003), Bonin, Hasan and Wachtel (2005), Fries and Taci (2005) and Havrylchyk (2006) that find that foreign bank efficiency exceeds domestic bank efficiency in Central and Eastern Europe. We conclude that next to foreign banks’ relative superior profitability vis-à-vis domestic banks, their efficiency advantage has evaporated as well by the end of the 1990s. Future research in this field could be directed at focusing on the relationship between foreign bank efficiency and the mode of entry. Does the efficiency of a bank being acquired by a foreign bank differ from the efficiency of a foreign bank established as a greenfield? Studies focusing on the profitability of foreign banks could also include the mode of entry as an additional explaining variable. Results could further inform policy makers on the externalities of foreign bank entry regulation.

In sum, we conclude that in the 1990s, non-financial FDI as well as financial FDI and optimistic profit expectations have moved foreign banks, predominantly from Western European countries, to invest in one or more entities in Central and Eastern Europe. Second, we conclude that while the foreign banks that entered in the early transition period preferred to set up greenfields in the shape of a representative office, a branch or a joint-venture with other banks and that the ‘late comers’ would take the opportunity to bid for state owned (saving)banks that governments appointed for privatisation. Third, we conclude that the growing presence of foreign banks, based rather on their relative assets than on their relative numbers, increased the competition in the banking sector. Fourth, we argue that the increased competition decreased the margins in corporate banking and forced banks to diversify their loan portfolio. Fifth, we conclude that in an early transition period foreign banks are more profitable and efficient than domestic banks, but that eventually domestic banks are able to show better performance. Finally, we suggest that the inverse relationship between foreign ownership and bank performance gives cause for a new strand of literature that, instead of focusing on the effects of cross-border mergers, investigates profitability and efficiency effects of foreign banks after having been acquired by national domestic owned competitors. Ex ante we would expect positive effects.

��

Bibliography

Aigner, D., Lovell, C. A. K. and P. Schmidt (1977). Formulation and estimation of stochastic frontier production function models, Journal of Econometrics, 6, 21-37.

Allen, F. and D. Gale (2000). Comparing financial systems. Cambridge, Mass. and London: MIT Press.

Allen, F., and A.M. Santomero (2001). What do financial intermediaries do?, Journal of Banking & Finance 25, 271-294.

Arellano, M. and S. Bond (1998) Dynamical panel data estimation using DPD98 for Gauss: A guide for users. http://www.cemfi.es/~arellano/

BACA (2005). BA-CA Comparison of banks: International banks in the new Europe 2004/2005, www.ba-ca.com.

Barisitz, S. (2001). The development of the Romanian and Bulgarian banking sectors since 1990, Focus on Transition, 1, OeNB, Vienna.

Battese, G. E., and T.J. Coell (1995). A model for technical inefficiency effects in a stochastic frontier production function for panel data, Empirical Economics 20, 325-332.

Beck, T. and R Levine (2003). Legal institutions and financial development. NBER Working Paper 10126, Cambridge MA.

Beck, T., Demirgüç-Kunt, A. and R. Levine (2000). A new database on the structure and development of the financial sector, World Bank Economic Review, 14, 597-605.

Berger, A.N, DeYoung, R., Genay, H. and G. Udell (2000). Globalisation of financial institutions: evidence from cross-border banking performance, Brookings-Wharton Papers on Financial Services, 3, 23-158.

Berger, A.N. and R. DeYoung (1997). Problem loans and cost efficiency in commercial banks, Journal of Banking & Finance, 32, 849-870.

Berger, A.N. and D.B. Humphrey (1997). Efficiency of financial institutions: International survey and directions for future research, European Journal of Operational Research, 98, 175-212.

Berger, A.N. and G. Udell (2002), Small business credit availability and relationship lending: The importance of bank organisational structure, The Economic Journal, 112, 32-53.

Berger, A.N. and L.J. Mester (1997). Inside the black box: what explains differences in the efficiencies of financial institutions?, Journal of Banking & Finance, 21, 895-947.

Berger, A.N., Buch, C.M., DeLong, G. and R. DeYoung (2004). Exporting financial institutions management via foreign direct investment mergers and acquisitions, Journal of International Money and Finance, 23, 333-366.

Berger, A.N., Dai, Q., Ongena, S. and D.C. Smith (2003). To what extent will the banking industry be globalised? A study of bank nationality and reach in 20 European nations, Journal of Banking & Finance, 27, 383-415.

��

Berger, A.N., DeYoung, R. and G.F. Udell (2001). Efficiency barriers to the consolidation of the European financial services industry, European Financial Management, 7, 117-130.

Berger, A.N., DeYoung, R., Genay, H. and G.F. Udell (2000), Globalization of financial institutions: Evidence from cross-border banking performance, Brookings-Wharton Papers on Financial Services, 3, 23-158.

Berger, A.N., Hancock, D. and D.B. Humphrey (1993). Bank efficiency derived from the profit function, Journal of Banking & Finance, 17, 317-347.

Berger, A.N., Klapper, L.F. and G.F. Udell (2001). The ability of banks to lend to informationally opaque small businesses, Journal of Banking & Finance, 25, 2127-2167.

Berger, A.N., Miller, N.H., Mitchell, A.P., Rajan, R.G. and S.C. Stein (2005). Does function follow organizational form? Evidence from the lending practices of large and small banks, Journal of Financial Economics, 76, 237-569.

Berglöf, E. and P. Bolton (2002). The great divide and beyond. Financial architecture in transition, Journal of Economic Perspectives, 16, 77-100.

Bhattacharya, A., Lovell, C.A.K. and P. Sahay (1997). The impact of liberalisation on the productive efficiency of Indian commercial banks, European Journal of Operational Research, 98, 332-345.

Bhattacharya, S. and A.V. Thakor (1993). Contemporary banking theory, Journal of Financial Intermediation, 3, 2-50.

Bikker, J. and S. Wesseling (2003). Intermediation, integration and internationalisation: A survey on banking in Europe, De Nederlandsche Bank Occasional Studies No. 3, Amsterdam.

Blundell and Bond (1998), Initial conditions and moment restrictions in dynamic panel data models, Journal of Econometrics, 87, 115-143

Bondt, G. de (2000). Financial structure and monetary transmission in Europe. Cheltenham: Edward Elgar.

Bonin, J.P. and I. Abel (2001). Retail banking in Hungary: A foreign affair?, Background paper for World Development Report 2002: Institutions for Markets (New York, Oxford University Press). Middletown, Connecticut: Wesleyan University.

Bonin, J.P., Hasan, I. and P. Wachtel (2005). Bank performance, efficiency and ownership in transition countries, Journal of Banking & Finance, 29, 31-53.

Bonin, J.P., Mizsei, K., Székely, L.P. and P. Wachtel (1998). Banking in transition economies, developing market oriented banking sectors in Eastern Europe, Cheltenham: Edward Elgar,

Boot, A.W.A. and A.V. Thakor (1997). Financial system architecture. Review of Financial Studies 10, 693-733.

Bratkowski, A., Grosfeld, I. and J. Rostowski (2000). Investment and finance in de novo private firms: empirical results from the Czech Republic, Hungary and Poland, Economics of Transition, 8(1), 101-116.

Brealey, R.A. and E.C. Kaplanis (1996). The determination of foreign banking location, Journal of International Money and Finance, 15, 577-597.

Brown, K. and M. Skully (2003). International studies in comparative banking: A

��

Bibliography

survey of Recent Developments, SSRN Working Paper 365920.Buch, C.M. (1997). Opening up for foreign banks: How Central and Eastern Europe

can benefit. Economics of Transition, 5, 339-336.Buch, C.M. (2000). Why do banks go abroad? - Evidence from German data, Financial

Markets Institutions and Instruments, 9, 33-67.Buch, C.M. (2003). Information or regulation: what drives the international activities

of commercial banks?, Journal of Money, Credit and Banking, 35, 851-869.Buch, C.M. and G. DeLong (2004). Cross-border bank mergers: What lures the rare

animal?, Journal of Banking & Finance, 28, 2077-2102.Buch, C.M., Kleinert, J. and F. Toubal (2004). The distance puzzle: on the interpretation

of the distance coefficient in gravity equations, Economics Letters, 83, 293-298. Buch, C.M., Kleinert, J. and P.Zajc (2003). Foreign bank ownership: A bonus or threat

for financial stability? In: Securing Financial Stability: Problems and Prospects for New EU Members, SUERF Study 2003/4. http://www.suerf.org/download/studies/study20034.pdf

Caprio, G. and D. Klingebiel (2003). Episodes of systemic and borderline financial crises. Manuscript. World Bank.

Cerutti, E., Dell’Ariccia, G. and M.S. Martínez Pería (2005). How do banks go abroad, branches or subsidiaries?, World Bank Policy Research Working Paper No. 3753 World Bank, Washington D.C.

Chang, C.E., Hasan, I. and W.C. Hunter (1998). Efficiency of multinational banks: An empirical investigation, Applied Financial Economics, 8, 689-696.

Christensen, L. R., Jorgenson, D. W. and L.J. Lau (1973). Transcendental logarithmic production frontiers, Review of Economics and Statistics, 55, 28-45.

Claessens, S. (2006). Competitive implications of cross-border banking. World Bank Policy Research Working Paper No. 3854, World Bank, Washington D.C.

Claessens, S., Demirgüç-Kunt, A. and H. Huizinga (2000). The role of foreign banks in domestic banking systems, in Claessens, in S., Jansen, M. (Eds.), The internationalisation of financial services: Issues and lessons for developing countries. Boston: Kluwer Academic.

Claessens, S., Demirgüç-Kunt, A. and H. Huizinga (2001). How does foreign entry affect domestic banking markets?, Journal of Banking & Finance, 25, 891-911.

Claessens, Stijn and van Horen, Neeltje, “Location decisions of foreign banks and competitive advantage” (June 2006). Available at SSRN: http://ssrn.com/abstract=904332.

Clarke, G., Cull R., and M.S. Martinez-Peria (2001). Does foreign bank penetration reduce access to credit in developing countries? Evidence from asking borrowers. World Bank Policy Working Paper No. 2716, World Bank, Washington D.C.

Clarke, G., Cull, R., Martinez Peria, M.S. and S.M. Sánchez (2003). Foreign bank entry: experience for developing economies, and agenda for further research, The World Bank Research Observer, 18, 25-59.Clarke, G., Cull, R., Martinez-Peria, M.S. and S.M. Sánchez (2005). Bank lending

to small businesses in Latin America: Does bank origin matter?, Journal of Money,

��

Credit, and Banking, 37, 83-118.Coelli, T. (1996). A Guide to FRONTIER Version 4.1: A Computer program for

stochastic frontier production and cost function estimation, CEPA Working Paper 96/07, Centre for Efficiency and Productivity Analysis.

Coelli, T., Rao, D.S.P. and G.E. Battese (1998). An introduction to efficiency and productivity analysis. Kluwer Academic Publishers.

Crystal, J., Dages, B. and L. Goldberg (2002). Has foreign bank entry led to sounder banks in Latin America?, Current Issues in Economics and Finance, 8, 1-6, Federal Reserve Bank of New York.

Dages, B., Goldberg, L. and D. Kinney (2000). Foreign and domestic bank participation in emerging markets: Lessons from Mexico and Argentina, Federal Reserve Bank of New York Economic Policy Review, 17-36.

Das, U.S., Quintyn, M. and K. Chenard (2003). Does regulatory governance matters for financial stability? An empirical analysis, paper prepared for the Bank of Canada Conference The evolving financial system and public policy, 4-5 December.

De Haan, J. and I.J. Naaborg (2004). Financial intermediation in accession countries: the role of foreign banks in, D. Masciandaro (ed.), Financial Intermediation in the New Europe, Edward Elgar, Cheltenham, UK, 181-207.

De Haas, R.T.A. and I.J. Naaborg. Foreign banks in transition countries. To whom do they lend and how are they financed?, Financial Markets, Institutions and Instruments, forthcoming.

Dell’Ariccia, G. and R. Marquez (2004). Information and bank credit allocation, Journal of Financial Economics, 72, 185-214.

DeLong, G.L. (2001). Stockholder gains from focusing versus diversifying mergers, Journal of Financial Economics, 59, 221-252.

Demirgüç-Kunt, A. and H. Huizinga (2000). Determinants of commercial bank interest margins and profitability: Some international evidence, World Bank Economic Review, 13, 379-408.

Demirgüc-Kunt, A., Laeven, L. and R. Levine (2003). The impact of bank regulations, concentration, and institutions on bank margins, World Bank Policy Research Paper 3030. World Bank, Washington.

Detragiache, E., Tressel, Th. and P. Gupta (2006). Foreign banks in poor countries: Theory and Evidence. IMF Working Paper, IMF, Washington D.C.

DeYoung, R., and D.E. Nolle (1996). Foreign-owned banks in the US: Earning market share or buying it?, Journal of Money, Credit, and Banking, 28, 622-636.

Dikova, D. (2005). Studies on foreign direct investments in Central and Eastern Europe, Ridderkerk, Ridderprint.

Dobosiewicz, Z. (1995). Foreign banks in the Polish banking system, in E. Miklaszewska (ed.), Competitive Banking in Central and Eastern Europe, 97-107, Jagiellonian University, Krakow.

Du, J., (2003). Why do multinational enterprises borrow from local banks?, Economics Letters, 78, 287-291.

��

Bibliography

Dubauskas, G. (2002). Motivation of foreign banks entry into the emerging economies: the case of Lithuania, Working papers TTUWPE No 02/69, Tallinn.

EBRD (1998). Transition Report 1998, London.EBRD (2001). Transition Report 2001, London.EBRD (2003). Transition Report Update- May 2003, London.EBRD (2005). Transition Report 2005, London.ECB (2001). Financial Sector Development and Convergence in Accession Countries:

An Overview. Frankfurt am Main.ECB (2002). Financial Sectors in EU Accession Countries. Frankfurt am Main.ECB (2004). Adherence of the central banks and banking supervisors of the new

Member States to the Memoranda of Understanding on high-level principles of co-operation in crisis management situations and between payment systems overseers and banking supervisors, http://www.ecb.int/press/pr/date/2004/html/pr040721.en.html.

ECB (2005). Banking Structures in the New EU Member States, Frankfurt am Main.

EFC (2001). Report on financial crisis management, number 156, http://ec.europa.eu/economy_finance/ publications/economic_papers/2001/ecp156en.pdf.

EIB (2002a). The Financial Integration of an Enlarged EU: Banking and Capital markets. Cahiers Papers, 7, 1.

EIB (2002b). The Financial Integration of an Enlarged EU: Macroeconomic Interlinkages. Cahiers Papers, 7, 2.

Farabullini, F. and G. Ferri (2004). Eastern passages for Italy’s banks and industrial districts: Are the two related or independent? in Masciandaro, D. (Ed.), Financial Intermediation in the New Europe: Banks, Markets and Regulation in EU Accession Countries. Edward Elgar, Cheltenham.

Florescu, G. (2002). Foreign banks entry into Romania, Working Paper TTUWPE No 02/71, Tallinn.

Focarelli, D. and A.F. Pozzolo (2001). The patterns of cross-border bank mergers and shareholdings in OECD countries, Journal of Banking & Finance, 25, 2305-2337.

Fries, S. and A. Taci (2002). Banking reform and development in transition economies, EBRD Working Paper, no. 71, EBRD, London.

Fries, S. and A. Taci (2005). Cost efficiency of banks in transition: Evidence from 289 banks in 15 post-communist countries. Journal of Banking & Finance, 29, 55-81.

Fries, S., Neven, D. and P. Seabright (2002). Bank performance in transition countries, EBRD Working Paper, no.76, EBRD, London.

Galac, T. and E. Kraft (2000). What has been the impact of foreign banks in Croatia?, Croatian National Bank Survey S-4, Zagreb.

Goldberg, L.G. (1992). The competitive impact of foreign commercial banks in the United States., in Gilbert, R.A. (Ed.), The Changing Market in Financial Services, Proceedings of the Fifteenth Annual Economic Policy Conference of the Federal Reserve Bank of St. Louis. Kluwer Academic, Norwell.

��

Goldberg, L.G. and D. Johnson (1990). The determinants of US banking activity abroad, Journal of International Money and Finance, 9, 123-137.

Goldberg, L.S. and A. Saunders (1981). The determinants of foreign banking activity in the United States, Journal of Banking & Finance, 5, 17-32.

Golodniuk, I. (2005). Financial systems and financial reforms in CIS countries. CASE-Center for Social and Economic Research Studies and Analyses 306, Warsaw.

Gorton, G. and A. Winton (2002) Bank liquidity provision and capital regulation in transition economies, In Meyendorff, A. and A.V. Thakor (Ed). Designing Financial Systems in Transition Economies: Strategies for Reform in Central and Eastern Europe, MIT Press, Cambridge, Massachusetts.

Green, J.C., Murinde, V. and I. Nikolov (2004). Are foreign banks in Central and Eastern European countries more efficient than domestic banks?, Journal of Emerging Market Finance, 3, 175-205.

Grigorian, D.A. and V. Manole (2002). Determinants of commercial bank performance in transition: An application of data envelopment analysis, IMF Working Paper No. 02/146. IMF, Washington D.C.

Gros D. and A. Steinherr (2005). Economic Transition in Central and Eastern Europe: planting the seeds, Cambridge University Press, Cambridge.

Grosse, R. and L.G. Goldberg (1991). Foreign bank activity in the United States: An analysis by country of origin, Journal of Banking & Finance, 15, 1093-1112.

Grubel, H. (1977). A theory of multinational banking, Banca Nazionale del Lavozo Quarterly Review, 123, 349-363.

Gulamhussen, M.A. (2004). Market entry of foreign banks under uncertainty: a real option theory approach, EFMA Basel Meetings.

Hainz, C. (2002). Are transition countries overbanked? An answer from industrial organization, forthcoming, University of Munich.

Hainz, C. (2003). Bank competition and credit markets in transition economies, Journal of Comparative Economcis, 31, 223-245.

Harzing, A.W.K. (2002). Acquisitions vs. greenfield investments: International Strategy and management of entry modes, Strategic Management Journal, 23, 211-227.

Hasan, I. and K. Marton (2003). Development and efficiency of the banking sector in a transitional economy: Hungarian experience. Journal of Banking & Finance, 27, 2249-2271.

Hasan, I. and W.C. Hunter (1996). Efficiency of Japanese multinational banks in the United States, in Andrew H. Chen (ed.), Research in Finance, 14, Greenwich, CT and London: JAI Press, 157-173.

Havrylchyk, O. (2006). Efficiency of the Polish banking industry: Foreign versus domestic banks, Journal of Banking & Finance, forthcoming.

Heinkel, R.L. and M.D. Levi (1992). The structure of international banking, Journal of International Money and Finance, 11, 521-272.

Hermes, N. and R. Lensink (2004). Foreign bank presence, domestic bank performance and financial development, Journal of Emerging Market Finance, 3, 207-229.

Hultman, C.W. and L.R. McGee (1989). Factors affecting the foreign banking

��

Bibliography

presence in the US, Journal of Banking & Finance, 13, 383-396. Hymer, S., (1960/1976). The international operations of national firms: A Study of direct

foreign investment. MIT Press: Cambridge, MA.Iakova, D.M. and N.L. Wagner (2001). Financial sector evolution in the Central

European Economies: Challenges in supporting macroeconomic stability and sustainable growth, IMF Working Paper, 141. IMF, Washington D.C.

IMF, International Financial Statistics.International Monetary Fund (2000). International Capital Markets: Developments,

Prospects and Key Policy Issues, Washington DC, IMF.Jaffee, D. and M. Levonian (2001). Structure of banking systems in developed and

transition economies, European Financial Management, 7, 161-181.Jemric, I. and B. Vujcic (2002). Efficiency of banks in Croatia: A DEA approach.

Comparative Economic Studies, 44, 69-193.Jeon, Y., Miller, S.M. and P.A. Natke (2006). Do foreign bank operations provide a

stabilizing influence in Korea?, The Quarterly Review of Economics and Finance, 46, 83-109.

Kaufmann, D., Kraay, A. and M. Mastruzzi (2005). Governance Matters IV: Governance Indicators for 1996-2004, World Bank Policy Research Working Paper Series No. 3630, World Bank, Washington D.C.

Khan, M.S. and A.S. Senhadji. (2000). Financial development and economic growth: An overview, IMF Working Paper, 209. IMF, Washington D.C.

Kim, Suk-Joong and Wu, Eliza, “Sovereign credit ratings, capital flows and financial sector development in emerging markets”. Available at SSRN: http://ssrn.com/abstract=921263

Kindleberger, C.P., (1983). International banks as leaders of followers of international business, Journal of Banking & Finance, 7, 583-595.

King, R.G. and R. Levine (1993). Finance, entrepreneurship and growth: theory and evidence, Journal of Monetary Economics, 32, 513-42.

King, R.G., and R. Levine (1993). Finance and growth: Schumpeter might be right. Quarterly Journal of Economics, 108, 717-737.

Kodde, D.A. and F.C. Palm (1986). Wald criteria for jointly testing equality and inequality restrictions, Econometrica, 54, 1243-1248.

Koivu, T. (2002). Do efficient banking sectors accelerate economic growth in transition countries?, Bank of Finland, BOFIT Discussion Papers, no. 14.

Konopielko, L. (1999). Foreign banks entry into Central and East European markets: Motives and activities, Post-Communist Economies, 11, 463-485.

Kowalski, T., Uiboupin, J. and V. Vensel (2002). Motives of foreign banks entry into the transitional economies: the comparative empirical evidence from Estonia and Poland, Working paper TTUWPE No. 02/78.

Kraft, E. (2004). Foreign banks in Croatia: reasons for entry, performance and impacts, Journal of Emerging Market Finance, 3, 153-174.

Kraft, E. and D. Tirtiroglu (1998). Bank efficiency in Croatia: A stochastic-frontier analysis, Journal of Comparative Economics, 26, 282-300.

��0 ��

Kuenzle, M. (2005). Cost efficiency in network industries: Application of stochastic frontier analysis, Universität Zürich.

Kumbhakar S.C. and C.A.K. Lovell (2000). Stochastic frontier analysis, Cambridge University Press.

Lang, L.H.P. and R.W. So (2002). Bank ownership structure and economic performance, preliminary draft, Hong Kong.

Lensink, B.W. and J. de Haan (2002). Do reforms in transition economies affect foreign bank entry?, International Review of Finance, 3, 213-232.

Lensink, B.W. and N. Hermes (2004). The short-term effects of foreign bank entry on domestic bank behaviour: Does economic development matter?, Journal of Banking & Finance, 28, 553-568.

Lensink, B.W., and I.J. Naaborg (2006). Does foreign ownership foster bank performance? Applied Financial Economics, forthcoming.

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

Levine, R. (1999). Law, finance, and economic growth. Journal of Financial Intermediation 8, 36-67.

Levine, R. (2000). Bank-based or market-based financial systems: Which is better?, Paper presented at the World Bank conference on Financial Structure, Washington D.C.,

Levine, R. (2004). Finance and growth: Theory and evidence. NBER Working Paper 10766.

Levine, R., Loayza, N. and T. Beck (2000). Financial intermediation and growth: Causality and causes, Journal of Monetary Economics, 46, 31-77.

Lízal, L. and J. Svejnar (2002. Financial conditions and investment during the transition: Evidence from Czech firms, in Meyendorff, A. and A.V. Thakor (Eds), Designing Financial Systems in Transition Economies: strategies for reform in Central and Eastern Europe, MIT tress, Cambridge, Massachusetts.

Magri, S., Mori, A. and P. Rossi (2005). The entry and the activity level of foreign banks in Italy: An analysis of the determinants, Journal of Banking & Finance, 29, 295-1310.

Majnoni, G., Shankar, R., and E. Várhegyi (2003). The dynamics of foreign bank ownership: Evidence from Hungary, World Bank Policy Research Working Paper no. 3114, World Bank, Washington D.C.

Mamatzakis, E., Staikouras, C. and N. Koutsomanoli-Fillipaki (2005). Competition and concentration in the banking sector of the South Eastern European Region, Emerging Markets Review, 6, 192-209.

Martinez Peria, M.S., Powell, A. and I. Vladkova Hollar (2002). Banking on foreigners: The behavior of international bank lending to Latin America, 1985-2000, World Bank Policy Research Working Papers 2893. World Bank, Washington D.C.

Martinez Pería, S. and A. Mody, (2004). How foreign participation and market concentration impact bank spreads: Evidence form Latin America, Journal of Money, Credit and Banking, 36, 533-539.

��0 ��

Bibliography

Mathieson, D.J. and J. Roldos (2001). The role of foreign banks in emerging markets. Paper prepared for the IMF-World Bank Brookings Institution conference on Financial Markets and Development, 19-21 April, New York.

Matousek, R. and A. Taci (2004). Banking efficiency in transition economies: Empirical evidence from the Czech Republic. Economics of Planning, 37, 225-244.

Meeusen, W. and J. van den Broeck (1977). Efficiency estimation from Cobb-Douglas production functions with composed error, International Economic Review, 18, 435-444.

Mero, K. and M.E. Valentinyi (2003). The role of foreign banks in five Central and Eastern European countries, Magyar Nemzeti Bank Working Paper, No. 2003/10.

Merton, R.C. (1995). A functional perspective of financial intermediation, Financial Management, 24, 23-41.

Mervart, J. (1995). The Czech banking sector and its efficiency in comparison with chosen EU banking, Institute of Economy Working Paper No. 47, Czech national bank, Prague.

Mian, A. (2003). Foreign, private, domestic and government banks: New Evidence from emerging market, Working paper, University of Chicago Business School.

Mian, A. (2006). Distance constraints: The limits of foreign lending in poor economies. The Journal of Finance, 61, 1465-1505

Micco, A., Panizza, U. and M. Yañez. Bank ownership and performance. Does politics matter? Journal of Banking & Finance, forthcoming.

Miller, S. R. and A. Parkhe (1998). Patterns in the expansion of US banks foreign operations, Journal of International Business Studies, 29, 359-390.

Miller, S.R. and A. Parkhe (2002). Is there a liability of foreignness in global banking? An empirical test of banks X-efficiency. Strategic Management Journal, 23, 55-75.

Morosini, O., Shane, S., and H. Singh (1998). National cultural distance and cross-border acquisition performance, Journal of International Business Studies, 29, 137-157.

Moshirian, F. (2001) International investment in financial services, Journal of Banking & Finance, 25, 317-337.

Naaborg, I.J. and B.W. Lensink. Banking in Transition Economies: Does foreign ownership enhance profitability? The European Journal of Finance, forthcoming.

Naaborg, I.J. and L.J.R. Scholtens (2006). A comparative study of banks balance sheets in the European transition countries, 1995-2003, International Finance Review, 6, 157-178.

Naaborg, I.J., Bol, H., Haan, J. de, Scholtens, B. and R.T.A. Haas, de (2004). How important are foreign banks in the financial development of European transition countries?, Journal of Emerging Market Finance, 3, 99-123.

Nikiel, E.M. and T.P. Opiela (2002). Customer type and bank efficiency in Poland. Implications for emerging market banking. Contemporary Economic Policy, 20, 255-271.

Noyer, C., (2001). Banking in the Euro Area, Speech by the Vice-President of the

��

European Central Bank, at the 30th Anniversary of the Association of Foreign Banks Representatives, Frankfurt, 8 November.

Papi, L. and D. Revoltella (2000). Foreign direct investment in the banking sector: A transitional economy perspective, in Claessens, S. and M. Jansen (Eds.), The Internationalization of Financial Services –Issues and Lessons for Developing Countries, Kluwer Law International, London/The Hague/Boston.

Peek, J. and E.S. Rosengren (2000a). Collateral damage: effects of the Japanese bank crisis on real activity in the United States, American Economic Review, 90, 30-45.

Peek, J. and E.S. Rosengren (2000b). Implications of the globalization of the banking sector: The Latin American experience, New England Economic Review, September/October, 45-63.

Pomerleano, M. and G.J. Vojta (2001). What do foreign banks do in Emerging Markets? An institutional study, in Litan, E.R., Masson, P., Pomerleano, M. (Eds), Open Doors: Foreign Participation in Financial Systems in Developing Countries. Brookings Institution Press, Washington DC.

Raiffeisen (2005). CEE – Banking environment, Press Conference, 13 July 2005.Rugman, A.M. and A. Verbeke (1992). A note on the transnational solution and

the transaction cost theory of multinational strategic management, Journal of International Business Studies, 23, 761-771.

Sabi, M. (1995). What’s behind the growing presence of foreign banks in the U.S.? A test of the conventional hypothesis, Journal of International Markets, Institutions and Money, 5, 51-59.

Sabi, M. (1996). Comparative analysis of foreign and domestic bank operations in Hungary, Journal of Comparative Economics, 22, 179-88.

Schardax, F. and T. Reiniger (2001). The financial sector in five Central and Eastern European countries: An overview, Focus on Transition, 1, OeNB, Vienna.

Scholtens, B. (2000). Financial regulation and financial system architecture in Central Europe, Journal of Banking & Finance, 24, 525-553.

Sealey, C.W. and J.T. Lindley (1977). Inputs, outputs, and a theory of production and cost at depository financial institutions. Journal of Finance, 32, 1251-1265.

Seth, R., Nolle, D.E. and S.K. Mohanty (1998). Do banks follow their customers abroad? Financial Markets, Institutions, and Instruments, 7, 1-25.

Stanley, T.O., Roger, C. and B.L. McManis (1993a). A measure of the strategic loan portfolio management of full service U.S. banks with foreign owners, International Review of Economics and Finance 2, 59-80.

Stanley, T.O., Roger, C. and B.L McManis (1993b). The effects of foreign ownership of US banks on the availability of loanable funds to small business, Journal of Small Business Management, 31 (2), 51-66.

Steinherr, A. (1997). Banking reforms in Eastern European countries, Oxford Review of Economic Policy, 13, 106-125.

Sturm, J.E. and B. Williams (2004). Foreign bank entry, deregulation and bank efficiency: Lessons from the Australian experience, Journal of Banking & Finance, 28, 1775-1799.

��

Bibliography

Ter Wengel, J. (1995). International trade in banking services, Journal of International Money and Finance, 14, 47-64.

The Economist (2002). After the chaos: A survey of finance in Central Europe, September 14th, 2002.

The Economist (2005). Open wider: A survey of international banking. May 21st, 2005.

Tirole, J. (2006). The Theory of Corporate Finance. Princeton University Press, Princeton, New Jersey.

Tschoegl, A.E. (2002). Entry and survival: the case of foreign banks in Norway, Scandinavian Journal of Management, 18, 131-153.

Tschoegl, A.E. (2004). Who owns the major US subsidiaries of foreign banks?, A Note. International Financial Markets, Institutions & Money, 14, 255-266.

VanderVennet (1996). The effect of mergers and acquisitions on the efficiency and profitability of EC credit institutions, Journal of Banking and Finance 20, 1531-1558.

VanderVennet, R. and G. Lanine (2005). Microeconomic determinants of acquisitions of Eastern European Banks by Western European Banks, University of Leuven.

VanTassel, E. and S. Vishwasrao (2005). Asymmetric information and the mode of entry in foreign credit markets, Florida Atlantic University.

Wang, H-J. and P. Schmidt (2002). One-step and two-step estimation of the effects of exogenous variables on technical efficiency, Journal of Productivity Analysis, 18, 129-144.

Weill, L. (2003). Banking efficiency in transition economies: the role of foreign ownership, The Economics of Transition, 11, 569-592.

Wezel, T. (2004). Foreign bank entry into emerging economies: an empirical assessment of the determinants and risks predicted on German FDI data, Deutsche Bundesbank Discussion Paper No 01/2004.

Williams, B. (1997). Banking: Eclectic theory versus internalisation theory, Journal of Economic Surveys, 11, 71-100.

Williams, B. (2002). The defensive expansion approach to multinational banking: Evidence to date, Financial Markets Institutions & Instruments, 11, 127-203.

Williams, J. and N. Nguyen (2005). Financial liberalisation, crisis and restructuring: A comparative study of bank performance and bank governance in South East Asia, Journal of Banking & Finance, 29, 2119-2154.

Wilson, B. (1980). The propensity of multinational enterprises to expand through acquisition, Journal of International Business Studies, 11, 59-65.

www.worldbank.org/data/www.worldbank.org/wbi/governance/govdata.Yamori, N. (1998). A note on the location choice of multinational banks: The case of

Japanese financial institutions, Journal of Banking & Finance, 22, 109-120.Yildirim, H.S. and G.C. Philippatos (2003). Efficiency of banks: Recent evidence from

the transition economies of Europe – 1993-2000, University of Saskatchewan, www.yorku.ca/yildirim/CEEefficiency.pdf

��

Zajc, P. (2006). A comparative study of bank efficiency in Central and Eastern Europe: The role of foreign ownership, International Finance Review, 6, 117-156.

Zoli, E., Tang, H. and I. Klytchnikova (2000). Banking crises in transition economies: Fiscal costs and related issues, World Bank Policy Research Working Paper, No. 2484.

��

AppendicesAppendix 1 Description of the central bank survey The survey includes 15 templates for the period 1990-2000:1 Deposit money banks assets as percentage of GDP 2 Private credit by deposit money banks as percentage of GDP 3 Deposit money bank assets to total financial assets4 Foreign banks’ assets as percentage of GDP5 Private credit by foreign banks as percentage of GDP6 Foreign banks assets as share of deposit money banks assets7 Number of foreign and domestic banks8 Number of foreign subsidiaries 9 Number of foreign branches 10 Number of banks with foreign participation 11 Return on assets (ROA) of foreign and domestic banks 12 Return on equity (ROE) of foreign and domestic banks 13 Net interest revenues of foreign and domestic banks 14 Overhead costs as share of total assets of foreign and domestic banks 15 After tax income (percent of total income) of foreign and domestic banks

Definitions:Deposit money banks: all financial institutions that have liabilities in the form of deposits transferable by check or otherwise usable in making payments. Assets refers to total domestic financial intermediation that the respective intermediary performs; see Beck et al. (1999), p.4 (IMF definition).Assets: see Beck et al. (1999), Appendix, p. 4 (line 22 in IFS).Foreign bank: bank with at least 50 percent of the shares in foreign hands.Banks with foreign participation: banks with foreign ownership between 5 and 50 percent.Net interest revenues: see Beck et al. (1999).Overhead costs: see Beck et al. (1999).

Source: National central banks from Bulgaria, Czech Republic, Estonia, Latvia, Lithuania, Poland, Romania, Slovak Republic, and Slovenia.

��

Appendix 2 Total number of banks in CEE countries, 1994-2004

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Bulgaria 41 41 42 28 34 34 37 35 34 35 35

Croatia 50 54 58 61 60 53 43 43 46 42 37

Czech Rep. 54 55 53 50 45 42 40 38 37 35 35

Estonia 22 19 15 13 6 7 7 7 7 7 9

Hungary 41 43 42 45 44 43 42 41 38 38 38

Latvia 0 42 35 32 27 23 21 23 23 23 23

Lithuania 22 15 12 12 12 11 13 13 14 13 12

Poland 82 81 81 83 83 77 73 69 59 58 57

Romania 27 24 31 33 36 34 33 33 31 30 32

Slovak Rep. 28 39 33 29 29 27 25 23 21 20 21

Slovenia 33 39 36 34 30 31 28 24 22 22 22

Total 401 446 434 420 404 380 360 347 331 324 321

Source: EBRD and national central banks

Appendix 3 Number of foreign banks in CEE countries, 1994-2004

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Bulgaria 2 3 3 7 17 22 25 26 26 25 24

Croatia 1 1 5 7 10 13 21 24 23 19 15

Czech Rep. 21 23 23 24 25 27 26 26 26 26 26

Estonia 5 4 4 3 3 4 4 4 4 6

Hungary 20 21 24 30 28 29 33 31 27 29 27

Latvia 1 14 15 15 12 12 10 9 10 9

Lithuania 0 0 3 4 5 4 6 6 7 7 6

Poland 11 18 25 29 31 39 46 46 45 43 44

Romania 12 8 10 13 16 19 21 24 24 21 23

Slovak Rep. 14 18 14 13 11 10 13 12 15 16 16

Slovenia 6 6 4 4 3 5 6 5 6 6 7

Total 87 114 129 150 164 183 213 214 212 209 203

Foreign–owned banks are defined as those with foreign ownership exceeding 50 per cent, end-of-year. Figures include commercial and savings banks and exclude cooperative banks. Source: EBRD and national central banks.

��

Appendices

Appendix 4 Asset share of state-owned CEE banks, 1995-2004

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Bulgaria n.a. 82.2 66.0 56.4 50.5 19.8 19.9 14.1 2.5 2.3

Croatia 51.9 36.2 32.6 37.5 39.8 5.7 5.0 4.0 3.4 3.3

Czech Republic

17.6 16.6 17.5 18.6 23.1 28.2 3.8 4.6 3.0 2.9

Estonia 9.7 6.6 0.0 7.8 7.9 0.0 0.0 0.0 0.0 0.0

Hungary 49.0 15.3 3.5 9.8 7.8 7.7 9.1 10.7 7.4 6.6

Latvia 9.9 6.9 6.8 8.5 2.6 2.9 3.2 4.0 4.1 4.0

Lithuania 61.8 54.0 48.8 44.4 41.9 38.9 12.2 0.0 0.0 0.0

Poland 71.7 69.8 51.6 48.0 24.9 23.9 24.4 26.6 25.8 19.4

Romania 84.3 80.9 80.0 75.3 50.3 50 45.4 43.6 40.6 7.5

Slovak Republic

61.2 54.2 48.7 50.0 50.7 49.1 4.9 1.9 1.5 1.3

Slovenia 41.7 40.7 40.1 41.3 42.2 42.5 48.9 13.3 12.8 12.6

Median 50.5 40.7 40.1 41.3 39.8 23.9 9.1 4.6 3.4 3.3

The state includes the federal, regional and municipal levels, as well as the state property fund and the state pension fund. State-owned banks are defined as banks with state ownership ex-ceeding 50 per cent, end-of year. In percent of total bank assets. Source: EBRD.

��

Appendix 5 Number and origin of banks in sample chapters 1 and 4

COM SAV COO INV MLT R&M Total %

Albania 2 1 3 1.3

Armenia 1 1 0.4

Azerbaijan 2 2 0.9

Belarus 1 1 0.4

Bosnia-H.na 5 1 6 2.7

Bulgaria 14 2 1 1 18 8.0

Croatia 21 1 22 9.8

Czech Rep. 17 2 1 1 21 9.4

Estonia 3 3 1.3

Hungary 18 1 19 8.5

Kazakhstan 3 1 4 1.8

Latvia 9 9 4.0

Lithuania 5 5 2.2

Macedonia 4 4 1.8

Poland 23 1 1 1 26 11.6

Romania 17 17 7.6

Russia 22 1 1 1 25 11.2

Slovak Rep. 12 1 1 1 15 6.7

Slovenia 11 2 13 5.8

Ukraine 4 1 5 2.2

Uzbekistan 1 1 0.4

Yugoslavia 3 1 4 1.8

Total 198 9 10 3 2 2 224 100.0

Bank categories include commercial banks (COM), savings banks (SAV), cooperative banks (COO), investment banks (INV), medium and long term credit banks (MLT), real estate and mortgage banks (R&M). Definitions are in line with the BankScope database. In the regression analyses we present the estimation results for the whole sample as well as for the subset of commercial banks only.

Appendix 6 Number and origin of banks chapter 2

1995 1996 1997 1998 1999 2000 2001 2002

CEE countries

Bulgaria 12 16 18 21 20 24 24 23

Croatia 24 33 40 35 34 37 36 32

Czech Rep. 22 25 24 21 22 24 24 22

Estonia 5 9 9 4 4 5 5 5

Hungary 16 22 24 23 27 29 25 23

Latvia 9 16 21 18 18 19 21 21

��

Appendices

1995 1996 1997 1998 1999 2000 2001 2002

Lithuania 3 6 8 8 8 9 9 9

Poland 28 37 40 38 40 42 41 35

Romania 4 5 8 21 24 27 25 24

Slovak Rep. 13 18 19 18 14 16 16 16

Slovenia 12 25 24 19 19 19 17 15

Benchmark set 1

Austria 54 70 119 133 141 163 177 166

Belgium 67 74 69 60 57 54 53 41

Denmark 35 90 90 93 93 100 93 88

Finland 4 8 9 9 9 10 9 8

France 320 333 319 309 332 333 331 275

Germany 1,651 1,822 1,804 1,988 1,942 1,825 1,700 1,457

Greece 4 11 17 15 15 15 15 16

Ireland 15 25 27 29 31 33 34 35

Italy 275 336 598 585 651 651 682 636

Luxembourg 107 118 120 114 120 110 98 89

Portugal 29 32 33 33 28 27 22 19

Spain 67 126 141 134 122 133 144 140

Sweden 6 11 8 10 12 18 99 97

The Netherlands 42 52 47 44 43 39 39 37

UK 104 145 142 146 139 134 136 129

Benchmark set 2

Albania 2 2 7 6 5 5

Armenia 3 6 6 7 6 5 4

Azerbaijan 1 5 7 8 8 11 11 10

Belarus 3 6 6 7 11 11 14 13

Bosnia-Herzegovina 3 5 9 14 17 17 17 13

Georgia 1 4 5 6 10 9 5 4

Kazakhstan 1 4 10 18 16 17 19 16

Kyrgystan 1 1 2 6 6 3

FYR Macedonia 6 7 7 7 11 11 12 10

Moldova 3 6 7 6 10 10 10 10

Russia 41 45 46 38 76 114 131 104

Serbia and Montenegro

10 14 14 10 11 14 23 18

Tajikistan 1 1

Turkmenistan 1 1 1 1

Ukraine 10 11 24 23 30 32 32 24

Uzbekistan 1 2 3 4 8 7 10 8

��0 ��

Appendix 7 Banking assets per country (% assets per region), 1995-2002

1995 1996 1997 1998 1999 2000 2001 2002

Poland 24.3 29.6 34.2 33.3 36.9 39.4 38.2 32.6

Czech Republic 35.7 33.4 27.1 27.0 20.4 21.3 21.7 23.0

Hungary 11.0 10.7 11.5 11.9 13.7 12.4 11.1 12.8

Croatia 3.8 6.9 7.4 6.2 5.9 5.9 6.7 7.5

Slovenia 1.5 5.7 5.4 6.5 6.7 5.6 5.8 6.3

Slovak Republic 13.2 9.2 8.3 6.9 6.5 5.3 5.6 6.0

Romania 4.6 2.3 0.6 2.8 4.0 4.0 4.1 3.8

Estonia 0.4 1.0 1.8 1.6 1.7 1.6 1.9 2.2

Latvia 0.2 0.5 1.0 0.7 1.0 1.5 1.8 2.1

Bulgaria 4.7 0.2 1.9 1.9 1.9 1.7 1.8 2.1

Lithuania 0.5 0.6 0.9 1.1 1.2 1.2 1.3 1.4

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

France 23.2 26.8 23.2 22.7 27.4 25.1 25.1 25.6

Germany 24.2 15.7 21.1 18.9 19.5 21.5 21.5 18.9

UK 10.8 12.8 11.3 12.6 12.5 13.7 13.7 15.0

Italy 14.2 13.0 13.7 12.8 10.8 9.5 9.5 9.7

The Netherlands 2.0 6.2 6.0 6.9 7.0 7.1 7.1 7.2

Spain 4.9 8.2 7.5 7.1 5.6 5.9 5.9 6.0

Belgium 8.8 6.0 6.1 6.7 4.7 4.3 4.3 4.0

Austria 2.8 1.3 1.0 2.0 2.0 2.4 2.3 2.4

Luxembourg 4.3 2.9 2.7 2.5 2.7 2.5 2.5 2.3

Ireland 0.7 0.8 1.0 1.2 1.4 1.6 1.6 2.0

Sweden 0.4 2.2 1.9 1.8 1.9 1.8 1.8 1.9

Denmark 0.5 1.1 1.2 1.3 1.3 1.6 1.6 1.8

Finland 1.1 0.9 1.4 1.2 1.2 1.4 1.4 1.3

Portugal 2.1 1.4 1.4 1.4 1.3 1.1 1.1 1.1

Greece 0.1 0.6 0.6 0.7 0.7 0.7 0.7 0.7

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Russia 50.2 43.2 38.2 19.9 57.1 70.6 71.9 74.0

Kazakhstan 0.1 1.9 2.2 5.0 2.7 3.7 5.0 5.9

Ukraine 7.5 5.4 6.9 6.8 3.4 4.0 4.8 4.5

Uzbekistan 4.9 9.3 5.8 9.9 6.6 5.1 4.7 3.2

Serbia and Montenegro

32.7 33.9 21.5 33.1 15.7 5.4 3.3 2.9

��0 ��

Appendices

1995 1996 1997 1998 1999 2000 2001 2002

Belarus 2.1 2.5 22.1 18.9 6.2 2.6 2.6 2.5

Albania 0.1 0.1 2.1 2.0 1.9 1.7

Bosnia-Herzegovina 0.5 0.7 0.9 2.7 1.1 1.0 1.5 1.6

Turkmenistan 2.4 2.6 1.8 1.3

Macedonia 1.6 1.5 0.8 1.6 1.3 1.2 1.2 1.1

Azerbaijan 0.0 0.8 0.6 0.8 0.5 0.8 0.5 0.5

Moldova 0.3 0.5 0.5 0.5 0.3 0.4 0.4 0.4

Armenia 0.2 0.2 0.4 0.3 0.2 0.2 0.2

Georgia 0.1 0.2 0.2 0.3 0.3 0.3 0.1 0.1

Kyrgystan 0.0 0.0 0.0 0.0 0.0 0.1

Tajikistan 0.0 0.0

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Appendix 8 Number of domestic and foreign banks in sample chapter 2

1995 1996 1997 1998 1999 2000 2001 2002

CEE countries

All banks 148 212 235 226 230 251 243 225

Domestic banks 9 15 16 17 14 6 47 44

Foreign banks 8 14 18 19 23 36 95 89

EU15

All banks 2,780 3,253 3,543 3,702 3,735 3,645 3,632 3,233

Domestic banks 144 200 206 205 203 212 346 303

Foreign banks 107 130 132 132 145 135 240 217

SEE/ CIS countries

All banks 82 114 148 149 223 272 301 243

Domestic banks 0 1 1 1 7 8 39 34

Foreign banks 2 2 3 3 3 5 26 26

Domestic and foreign banks include the majority owned domestic and foreign banks identified with ownership data provided by the BankScope database.

��

Appendix 9 Average bank size (in €mln), 1995-2002

1995 1996 1997 1998 1999 2000 2001 2002

Czech Rep. 1,591 2,029 1,811 2,458 1,860 2,356 2,908 3,383

Poland 849 1,214 1,372 1,675 1,849 2,484 2,991 3,013

Hungary 676 736 767 985 1,015 1,133 1,425 1,801

Estonia 71 165 316 765 852 858 1,243 1,419

Slovenia 125 348 361 655 712 780 1,102 1,360

Slovak Republic 996 774 705 729 924 879 1,125 1,221

Croatia 155 320 295 340 350 423 594 761

Romania 1,115 686 118 258 338 391 526 515

Lithuania 174 151 183 260 312 364 455 514

Latvia 24 47 75 76 113 208 275 324

Bulgaria 387 18 173 172 195 193 242 292

The Netherlands 3,103 14,640 20,085 26,392 31,283 40,360 43,753 46,540

Finland 18,127 13,597 23,639 23,161 25,981 29,840 36,566 39,842

UK 6,642 10,886 12,484 14,634 17,215 21,595 24,182 27,838

Belgium 8,413 9,970 13,771 18,988 15,667 17,645 19,569 23,293

France 4,649 9,916 11,405 12,407 15,780 16,410 18,275 22,277

Ireland 2,919 3,884 5,859 7,265 8,564 9,892 11,075 13,886

Portugal 4,670 5,501 6,870 6,987 9,096 9,390 12,160 13,470

Greece 1,170 6,663 5,618 7,432 9,558 10,477 11,343 10,533

Spain 4,648 8,039 8,306 8,974 8,724 9,785 9,852 10,181

Luxembourg 2,555 3,020 3,482 3,695 4,232 4,843 6,027 6,281

Denmark 866 1,473 2,053 2,407 2,668 3,586 4,034 4,761

Sweden 4,105 25,138 37,117 31,128 30,517 23,494 4,497 4,726

Italy 3,303 4,769 3,589 3,700 3,180 3,544 3,349 3,652

Austria 3,273 2,260 1,280 2,595 2,659 3,296 3,102 3,489

Germany 941 1,063 1,832 1,606 1,920 2,936 3,050 3,098

Turkmenistan 1,819 2,492 2,051 1,483

Russia 400 346 466 172 568 588 631 805

Uzbekistan 1,590 1,678 1,089 805 627 693 542 453

Kazakhstan 25 168 121 90 128 207 300 420

Albania 23 22 226 309 443 394

Belarus 225 148 2,0612 882 423 224 216 220

Ukraine 245 179 160 96 84 119 172 210

Serbia-M.Negro 1,071 874 859 1,080 1,081 369 163 184

Bosnia-Herz. 59 48 59 62 50 56 101 140

��

Appendices

1995 1996 1997 1998 1999 2000 2001 2002

Macedonia 87 75 62 76 90 103 119 119

Azerbaijan 452 55 48 34 47 71 53 53

Moldova 37 32 44 30 23 34 45 50

Armenia 22 19 21 28 39 44 43

Georgia 20 15 20 18 19 29 29 36

Kyrgystan 12 9 14 6 8 23

Tajikistan 12 17

��

Appendix 10 Banks, supervisory institutions, the representatives and their position

Country Representative(s) Position of representative

Parent bank

ABN AMRO Bank Netherlands O.P Lindeman Senior Relationship Banker CEE region

BA-CA Austria G. Smoley Investor Relations

ERSTE Bank Austria M. Wohlmuth B.Spalt

Head of Group Strategy General Manager Group Risk Management

ING Bank Netherlands D.C. Klaasse P.J. van Baar

Regional Head of Financial Markets CEE Director Counterparty Risk Management

KBC Belgium H. Agneessens CEO/ Chief Risk Officer

Raiffeisen Austria R. Kossmann Regional SME Risk Manager

Sampo Bank Finland J. Ohls, R. Tornivaara

Head of Baltic Banking Board Member

SEB Sweden M. Kjaer Deputy Head Baltic & Poland Division

Swedbank Sweden L. Lundberg Senior Vice President

Foreign owned local bank

ABN AMRO Bank Czech Rep. H. Halter Country Administrative Officer

ABN AMRO Poland

Poland F.J. Umbgrove President of the Management Board

Bank BPH Poland N. Lundorff Member of the Management Board

Česká Spořitelna Czech Rep. D. Baran B. Lintner

Vice Chairman Board/CFO Deputy Dep. Head Credit Controlling and Policy

ČSOB Czech Rep. P. Daems P. Knapp.

Member Board of Directors/ Senior Ex. Officer Member Board of Directors/ Senior Executive Officer

Hansabank Estonia I. Neivelt A. Sahlén

Chairman Board/ Group CEO, Chairman Council Hansabank/ Senior Advisor Swedbank

HVB Bank Czech Rep.

Czech Rep. U.Szekulics Board Member

HVB Bank Hungary

Hungary M. Kunsch CEO

ING Bank Śląski Poland K.Tuijnman Executive Vice President

ING Bank Rt. Hungary P. de Haes CEO

Kredytbank S.A. Poland G.Libot Deputy President/ Deputy CEO

K&H Bank Hungary P. Szabo Director of Risk Management

Raiffeisen Bank Rt. Hungary I. Vass F. Szabó

Sen. Manager Treasury/ Head Correspondent Banking Deputy Managing Director

SEB Eesti Ühispank Estonia V. Svensson CFO

��

Appendices

Financial Supervision

Eesti Pank Estonia J. Tõrs Head of Financial Sector Policy Division

Ceská Národní Banka Czech Rep.

K. Gabrhel Ms. Musilova

Director Off-Site Banking Supervision Division I Director Banking Analyses Division

Rahoitustarkastus FinlandJ.Savela M.Hansson

Head of Credit Institutions Department Banking Supervisor

Finansinspektionen SwedenG. Ahlberg M. Stenhammar

Senior Financial InspectorSenior Advisor.

Narodowy Bank Polski Poland

P. Samecki T. Parys P. Bednarski M. Golajewska

Director International Department Deputy Director Bank Licensing Division Banking Supervisor Head of Financial System Stability Unit.

Magyar Nemzeti Bank Hungary G. Tóth Analyst Banking Department.

Oesterreichische Nationalbank Austria

M. Hammer T. Reininger

Foreign Research DivisionForeign Research Division

Interviews with representatives at the parent banks took place between June 2003 and Febru-ary 2005, ING Bank was consulted twice. Interviews with representatives of foreign owned local banks in Central and Eastern Europe took place between June 2003 and February 2005. CSOB, Hansabank and Raiffeisen Bank Rt. were consulted twice. Interviews with representa-tives of financial supervisory institutions took place in 2003.

��

Ap

pen

dix

11

Init

ial e

ntr

y st

rate

gies

an

d t

he

seq

uen

ce o

f ev

ents

Year

Ban

k H

ost

Ent

ry

Det

erm

inan

t(s)

Ent

ry M

ode

Org

anis

atio

nal

Form

Sequ

ence

of

Eve

nts

1975

Cre

dita

nsta

ltH

UIn

form

atio

n co

sts;

rela

tive

profi

tabi

lity

Follo

win

g cu

stom

ers

Gre

enfie

ldR

epre

sent

ativ

e of

fice

1998

: Joi

nt v

entu

re (C

IB)

1990

: Est

ablis

hmen

t of

a 10

0 pe

r ce

nt s

ubsi

diar

y: C

redi

tans

talt

Rt.

The

Bud

apes

t B

ank

take

s a

20 p

er c

ent s

take

and

late

r se

lls.

1998

: Nam

e ch

ange

into

Ban

k A

ustr

ia C

redi

tans

talt.

1999

: Cre

dita

nsta

lt se

lls it

s sh

ares

in C

IB (C

IB is

cur

rent

ly w

holly

ow

ned

by

Ban

ca I

ntes

a).

2001

: Hyp

over

eins

bank

Hun

gari

a R

t abs

orbs

Ban

k A

ustr

ia C

redi

tans

talt

(Hun

gary

). N

ame

chan

ge in

to H

VB

Ban

k H

unga

ry R

t.

1986

Rai

ffeis

en

HU

Inst

itutio

nal

cont

ext

Hun

gary

; try

-ou

t.

Gre

enfie

ld10

per

cen

t in

join

t ve

ntur

e U

nicB

ank

1998

: Nam

e ch

ange

Uni

cban

k in

to R

aiffe

isen

Uni

cBan

k.19

99: N

ame

chan

ge R

aiffe

isen

Uni

cBan

k in

to R

aiffe

isen

Ban

k H

unga

ry R

t.

1990

AB

N A

MR

OH

UFo

llow

ing

cust

omer

s;

inst

itutio

nal

cont

ext

Gre

enfie

ldR

epre

sent

ativ

e of

fice

1993

: AB

N A

MR

O B

ank

(Mag

yaro

rsza

g) R

t. es

tabl

ishe

d.19

97: A

BN

AM

RO

acq

uire

s 89

per

cen

t of

Mag

yar

Hite

l Ban

k in

pri

vatis

atio

n an

d ch

ange

s th

e na

me

of th

e ba

nk in

to A

BN

AM

RO

Mag

yar

Ban

k. M

agya

r H

itel B

ank

abso

rbs

AB

N A

MR

O b

ank

(Mag

yaro

rsza

g) R

t. 20

01: K

eres

kede

lmi e

s H

itelb

ank

RT

abs

orbs

AB

N A

MR

O (M

agya

r) b

ank

Rt.

AB

N A

MR

O b

ecom

es m

inor

ity o

wne

r of

Ker

eske

delm

i es

Hite

lban

k.20

05: A

BN

AM

RO

sel

ls its

min

ority

sta

ke to

maj

ority

sha

reho

lder

KB

C.

?B

ank

Aus

tria

HU

Rel

. pr

ofita

bilit

y;in

form

atio

n co

sts;

follo

win

g cu

stom

ers

Gre

enfie

ldJo

int v

entu

re

Eur

opea

n C

redi

tban

k

1998

: Ban

k A

ustr

ia s

ells

its p

art o

f E

urop

ean

Cre

ditb

ank

to C

itiba

nk.

1991

Ban

k A

ustr

ia

CZ

Rel

ativ

e pr

ofita

bilit

yG

reen

field

Subs

idia

ry B

ank

Aus

tria

CR

a.s.

1998

: Ban

k A

ustr

ia (C

R) a

.s. is

abo

rbed

by

Cre

dita

nsta

lt, c

reat

ing

Ban

k A

ustr

ia

Cre

dita

nsta

lt C

zech

Rep

ublic

.20

01: H

ypoV

erei

nsba

nk C

Z, a

.s. a

bsor

bs B

ank

Aus

tria

Cre

dita

nsta

lt C

zech

R

epub

lic, c

reat

ing

HV

B B

ank

Cze

ch R

epub

lic A

S.

��

1991

Cre

dita

nsta

ltC

ZR

elat

ive

profi

tabi

lity

Gre

enfie

ldSu

bsid

iary

(C

redi

tans

talt

A.S

., Pr

ague

)

1998

: abs

orbs

Ban

k A

ustr

ia (C

R) a

.s., c

reat

ing

Ban

k A

ustr

ia C

redi

tans

talt

Cze

ch

Rep

ublic

. 20

01: H

ypoV

erei

nsba

nk C

Z, a

.s. a

bsor

bs B

ank

Aus

tria

Cre

dita

nsta

lt C

zech

R

epub

lic, c

reat

ing

HV

B B

ank

Cze

ch R

epub

lic A

S.

1991

ING

H

UFo

llow

ing

cust

omer

sG

reen

field

Bra

nch

ING

Ban

k R

t 20

00: C

itiba

nk b

uys

ING

Ban

ks c

onsu

mer

bus

ines

s.

1991

AB

N A

MR

O

CZ

Tes

t the

mar

ket;

Follo

win

g cu

stom

ers

Gre

enfie

ldR

epre

sent

ativ

e of

fice

1993

: rep

rese

ntat

ive

offic

e ch

ange

d in

to b

ranc

h.19

95: o

peni

ng o

f se

cond

bra

nch.

Lat

er: s

ubsi

diar

y A

sset

Man

agem

ent.

1991

AB

N A

MR

O

PLFo

llow

ing

cust

omer

s;

shar

ing

risk

; te

stin

g th

e w

ater

Gre

enfie

ld15

per

cen

t in

join

t ven

ture

In

tern

atio

nal B

ank

in P

olan

d

1994

: Acq

uisi

tion

of a

sm

all b

ank

to o

btai

n lic

ense

1991

Cre

dita

nsta

ltPL

Follo

win

g cu

stom

ers;

re

lativ

e pr

ofita

bilit

y

Gre

enfie

ldSu

bsid

iary

Ban

k C

redi

tans

talt

SA19

97: a

cqui

sitio

n of

PB

K19

98: m

erge

r B

ank

Aus

tria

and

Cre

dita

nsta

lt: B

ank

Aus

tria

Cre

dita

nsta

lt Po

land

SA

.20

00: a

bsor

bed

by P

BK

SA

.20

00: fi

nal s

take

of

56 p

er c

ent i

n PB

K20

00: M

erge

r of

BA

CA

and

HV

B: B

PH P

BK

1992

Bay

eris

che

Ver

eins

bank

, A

G M

nich

ov.

CZ

Follo

win

g cu

stom

ers

Gre

enfie

ldSu

bsid

iary

B

ayer

isch

e V

erei

nsba

nk A

G

Mní

chov

until

31

Dec

embe

r 19

95 n

amed

Bay

eris

che

Ver

eins

bank

,AG

Mni

chov

, pob

očka

Pr

aha;

19

96: c

hang

ed in

to a

sub

sidi

ary

nam

ed V

erei

nsba

nk (C

Z) a

.s.

1999

: Ver

eins

bank

(CZ

) abs

orbs

HY

PO-B

AN

K C

Z a

.s.: H

ypoV

erei

nsba

nk (C

Z)

AS.

20

01: H

ypoV

erei

nsba

nk C

Z, a

.s. a

bsor

bs B

ank

Aus

tria

Cre

dita

nsta

lt C

zech

R

epub

lic, c

reas

ting

HV

B B

ank

Cze

ch R

epub

lic A

S.

1992

HV

BH

UFo

llow

ing

cust

omer

sG

reen

field

Subs

idia

ry

Hyp

oVer

eins

bank

H

ungá

ria

Rt

2001

Nam

e: H

VB

Ban

k H

unga

ry R

t.

��

Ap

pen

dix

11

(con

tinu

ed)

Year

Pare

nt B

ank

Hos

t E

ntry

Det

erm

inan

t(s)

Ent

ry M

ode

Org

anis

atio

nal

Form

Sequ

ence

of

even

ts

1992

ING

PL

Cir

cum

stan

tial

fact

ors;

follo

win

g cu

stom

ers

Gre

enfie

ldJo

int v

entu

re

Han

dlow

y H

elle

r S.

A

Han

dlow

y H

elle

r S.

A is

a jo

int v

entu

re o

f N

MB

-Hel

ler

Gro

up w

ith B

ank

Han

dlow

y C

onsu

ltant

s.19

93: I

NG

War

saw

offi

ce19

94: I

NG

acq

uire

s 25

.9 p

er c

ent i

n B

ank Śląs

ki.

1996

: IN

G b

ecom

es o

wne

r of

54.

08 p

er c

ent o

f th

e sh

ares

. 199

8: I

NG

ow

ns 5

4.98

per

cen

t of

the

shar

es.

2001

: IN

G g

radu

ally

rai

ses

stak

e to

87.

77 p

er c

ent.

2001

: nam

e ch

ange

: IN

G B

ank Śląs

ki S

.A a

nd m

erge

r w

ith I

NG

Ban

k,

War

saw

bra

nch.

1993

ER

STE

Ban

k C

ZPr

ofit f

rom

gro

wth

; ge

t mor

e m

arke

t sh

are,

lim

itatio

n on

gr

owth

hom

e m

arke

t, E

U A

cces

sion

.

Gre

enfie

ldSu

bsid

iary

E

rste

Ban

k Sp

arka

sse

CR

2000

: 52

per

cent

acq

uisi

tion

of Č

eská

Spo

řitel

na. Č

eská

Spo

řitel

na

abso

rbs

Ers

te B

ank

Spar

kass

e (C

R).

2001

: 71

per

cent

acq

uisi

tion

of p

refe

rent

ial s

hare

s20

02: Č

eská

Spo

řitel

na d

elis

ted

2003

: sta

ke r

aise

d fr

om 9

5 pe

r ce

nt to

98

per

cent

as Č

eská

Poj

isto

vna

purc

hase

s 3

per

cent

1993

KB

CH

UN

eed

to g

row,

cr

eate

a s

econ

d ho

me

mar

ket;

EU

A

cces

sion

Gre

enfie

ldSu

bsid

iary

A

rgos

z.

Gre

enfie

ld

Insu

ranc

e

1997

: 56

per

cent

acq

uisi

tion

toge

ther

with

Iri

sh L

ife o

f K

&H

, in

priv

atis

atio

n19

99: S

take

in K

&H

rai

sed

to m

ajor

ity20

01: m

erge

r of

K&

H B

ank

and

AB

N A

MR

O M

agya

r B

ank:

KB

C

maj

ority

ow

ner

1995

Hyp

o B

ank

PLSe

rvic

e ex

istin

g cu

stom

ers

Gre

enfie

ldSu

bsid

iary

H

ypo-

Ban

k Po

lska

SA

1999

: Ver

eins

bank

abs

orbs

Hyp

o-B

ank

Polsk

a SA

. Nam

e:

Hyp

oVer

eins

bank

Pol

ska

S.A

.19

99: H

ypoV

erei

nsba

nk P

olsk

a SA

is a

bsor

bed

by B

ank

BPH

.20

01: B

ank

BPH

abs

orbs

PB

K (o

wne

d by

Ban

k A

ustr

ia C

redi

tans

talt)

. N

ame:

BPH

-PB

K.

2004

: Nam

e ch

ange

Ban

k B

PH-P

BK

into

Ban

k B

PH S

A.

��

Appendices19

96Sw

edba

nk

EE

To

form

an

allia

nce;

tr

ansf

er c

ompe

tenc

e,

shar

e co

sts,

build

up

a ne

twor

k ar

ound

B

altic

Sea

; exp

ecte

d gr

owth

.

Acq

uisi

tion

33 p

erce

nt in

th

e pr

ivat

isat

ion

of H

oiup

ank

1998

: Mer

ger

Hoi

upan

k an

d ol

d H

ansa

bank

, Sw

edba

nk o

wns

4 p

er c

ent

in th

e ne

w H

ansa

pank

. Afte

r bi

ddin

g w

ar w

ith S

EB

, sha

reho

lder

of

old

Han

saba

nk, F

SB b

ecom

es s

trat

egic

ow

ner.

1999

: maj

ority

ow

ner.

2000

: Sw

edba

nk r

aise

s ow

ners

hip

to 5

8 pe

rcen

t.20

03: S

wed

bank

rai

ses

owne

rshi

p to

60

perc

ent.

2005

: Sw

edba

nk a

cqui

rers

100

per

cen

t and

Han

saba

nk is

del

iste

d.

1996

KB

CPL

Nee

d to

gro

w; t

o cr

eate

a se

cond

hom

e m

arke

t; in

stitu

tiona

l co

ntex

t

Acq

uisi

tion

10 p

erce

nt in

th

e pr

ivat

isat

ion

of K

redy

t Ban

k

1997

: Kre

dytb

ank

abso

rbs

Polis

h In

vest

men

t Ban

k.20

00: K

BC

acq

uire

s Pa

tria

.20

01: K

BC

is m

ajor

ity o

wne

r of

Kre

dytb

ank.

2002

: KB

C r

aise

s ow

ners

hip

in K

redy

tban

k to

57

per

cent

.20

03: K

BC

rai

ses

stak

e to

76

per

cent

.

1996

Ver

eins

bank

PLSe

rvic

e ex

istin

g cu

stom

ers

Gre

enfie

ldSu

bsid

iary

V

erei

nsba

nk

Polsk

a S.

A

1999

: Ver

eins

bank

abs

orbs

Hyp

o-B

ank

Polsk

a SA

. Nam

e:

Hyp

oVer

eins

bank

Pol

ska

S.A

.19

99: H

ypoV

erei

nsba

nk P

olsk

a SA

is a

bsor

bed

by B

ank

BPH

.20

01: B

ank

BPH

abs

orbs

PB

K (o

wne

d by

Ban

k A

ustr

ia C

redi

tans

talt)

. N

ame:

BPH

-PB

K.

2004

: Nam

e ch

ange

Ban

k B

PH-P

BK

into

Ban

k B

PH S

A.

1998

SEB

EE

Geo

grap

hica

l and

hi

stor

ical

ties

of

Swed

en w

ith E

ston

ia

Acq

uisi

tion

30 p

erce

nt in

pr

ivat

isat

ion

of

Ees

ti Ü

hisp

ank

2000

: SE

B r

aise

s ow

ners

hip

to 1

00 p

er c

ent.

2001

: Ees

ti Ü

hisp

ank

is d

elis

ted.

1998

Sam

poE

EE

xpec

ted

grow

th,

join

ing

in w

ith

Nor

dea.

Gre

enfie

ldR

epre

sent

ativ

e of

fice

1998

: Mer

ger

of E

STIB

and

Ees

ti Fo

reks

pank

.20

00: M

erge

r of

Opt

iva

Ban

k w

ith S

ampo

insu

ranc

e, n

ew n

ame

is A

S Sa

mpo

Pan

k.20

02: S

ampo

is d

elis

ted.

1999

KB

CC

ZN

eed

to g

row,

cre

ate

a 2n

d ho

me

mar

ket;

inst

itutio

nal c

onte

xt,

Follo

win

g cu

stom

ers

Acq

uisi

tion

66 p

erce

nt in

pr

ivat

isat

ion

of

ČSO

B

2000

: IPB

acq

uire

s al

l ass

ets

and

liabi

litie

s of

Inv

estic

ni a

Pos

tovn

i Ban

ka

AS-

IPB

.20

03: K

BC

ow

ns 8

2 pe

r ce

nt in

ČSO

B.

Sour

ce: b

ank

inte

rnet

site

s, in

terv

iew

s.

��0 ��

Appendix 12 Number and origin of banks in sample chapter 5

HIC No. Banks

UMC No. Banks

LMC No. Banks

LIC No. Banks

Australia 1 Argentina 18 Bolivia 3 Bangladesh 1

Austria 4 Botswana 1 Bulgaria 1 Indonesia 1

Bahrain 5 Brazil 20 China 1 Kenya 4

Belgium 5 Chile 7 Colombia 7 Nicaragua 1

Canada 10 Costa Rica 5 Dom Rep. 2 Nigeria 3

Denmark 4 Croatia 1 Egypt 6 Pakistan 1

Finland 1 Czech Rep. 1 El Salvador 3 Uganda 1

France 27 Hungary 2 Iran 1

Germany 37 Latvia 1 Namibia 1

Hong Kong 2 Lebanon 1 Paraguay 5

Iceland 1 Lithuania 2 Peru 5

Israel 2 Malaysia 2 Philippines 4

Italy 4 Mauritius 1 S-Africa 3

Korean Rep. 1 Mexico 10 Thailand 2

Kuwait 2 Oman 1 Turkey 1

Luxembourg 24 Panama 3

Macau, China 3 Poland 1

Malta 1 Puerto Rico 2

Netherlands 5 Saudi Arabia 3

New Zealand 4 Trinidad 1

Portugal 3 Venezuela 2

Singapore 3

Slovenia 1

Spain 28

Sweden 2

Switzerland 49

Taiwan 1

UAE 8

UK 31

USA 102

Total 371 85 45 12

Economies are categorized in income groups according to 2001 gross national income per ca-pita, in line with the World Bank World Development Report 2003. Groups are defined as: High Income Countries (> $9,205), Upper Middle income Countries ($2,976 - $9,205), Lower Mid-dle income Countries ($746 - $2,975) and Low Income Countries (< $745).

��0 ��

Appendices

Appendix 13 Macro-economic descriptives of chapter 5 bank sample

Country G GDPPC INF INT

Argentina -1.19 7,980 -0.56 16.18

Australia 3.84 23,441 2.79 5.75

Austria 2.58 32,343 1.62 5.37

Bahrain 3.53 10,753 -0.79 9.53

Bangladesh 5.33 366 5.22 11.19

Belgium 2.57 30,300 1.77 5.88

Bolivia 2.61 960 4.81 30.28

Botswana 6.94 3,717 7.67 8.32

Brazil 1.70 4,563 5.49 60.54

Bulgaria 3.90 1,540 6.75 5.88

Canada 3.89 22,359 2.00 4.92

Chile 2.50 5,278 3.96 9.60

China 7.63 773 -0.67 7.39

Colombia 0.12 2,309 12.44 13.59

Costa Rica 4.94 3,864 10.98 12.86

Croatia 2.47 5,156 4.50 7.34

Czech Rep. 1.49 5,337 5.34 3.82

Denmark 2.19 37,868 2.40 5.18

Domin. Rep. 6.25 2,692 4.05 13.36

Egypt 4.71 1,189 3.05 8.77

El Salvador 2.37 2,438 2.13 28.46

Finland 4.99 30,511 1.98 3.18

France 3.06 29,513 1.13 5.68

Germany 1.84 32,119 1.48 8.70

Hong Kong 2.08 23,461 -1.62 11.50

Hungary 4.50 5,239 10.78 4.66

Iceland 4.95 30,207 3.37 10.21

Indonesia 3.00 1,010 11.90 7.91

Iran 4.40 1,653 15.27 0.00

Israel 2.50 16,637 3.22 9.21

Italy 2.08 20,771 2.33 3.88

Kenya 0.96 333 7.05 13.89

Korean Rep. 4.14 12,483 3.67 9.36

Kuwait 1.10 13,871 1.65 8.28

Latvia 5.16 2,612 2.50 7.70

Lebanon 0.77 2,903 0.00 18.30

Lithuania 2.72 2,171 2.02 8.48

��

Country G GDPPC INF INT

Luxembourg 5.07 54,066 1.94 2.56

Macau -0.21 15,222 0.00 12.47

Malaysia 1.85 4,606 2.74 5.07

Malta 3.07 9,930 2.45 4.50

Mauritius 4.63 4,011 5.97 14.39

Mexico 5.10 3,657 14.00 8.78

Namibia 3.14 2,363 9.06 6.57

Netherlands 3.16 30,610 2.81 1.92

New Zealand 2.47 24,396 0.97 7.38

Nicaragua 4.07 437 12.12 7.60

Nigeria 2.25 253 9.87 9.66

Oman 2.52 5,857 -0.65 5.81

Pakistan 3.48 508 4.91 0.00

Panama 1.97 3,260 1.02 9.54

Paraguay 0.62 1,735 8.63 19.07

Peru 0.94 2,320 4.11 22.76

Philippines 2.56 1,143 6.72 4.63

Poland 3.03 3,643 7.65 11.71

Portugal 3.30 12,658 3.08 2.63

Puerto Rico 4.44 13,311 0.00 0.00

Saudi Arabia 2.59 5,331 0.36 2.32

Singapore 5.70 26,478 0.37 7.60

Slovenia 4.15 11,355 8.87 6.83

South Africa 2.12 3,995 5.46 9.42

Spain 3.83 17,025 2.79 1.62

Sweden 3.23 30,562 0.94 4.87

Switzerland 2.07 46,165 0.84 3.26

Taiwan 2.25 811 -0.69 4.19

Thailand -0.48 2,723 3.31 8.32

Trinidad 5.18 3,976 3.42 8.17

Turkey 1.91 3,097 68.14 0.00

Uganda 5.54 341 3.05 16.54

UK 2.59 21,938 2.43 3.38

Un. Arab Em. -1.94 835 2.64 37.87

USA 3.23 31,126 2.49 6.21

Venezuela -0.89 3,367 25.18 7.75

Data refer to average GDP growth (G), gross domestic product per capita (GDPPC), inflation (INF) and interest (INT) in the period 1998-2001. G, INF and INT are in percentages.

��

Ap

pen

dix

14

Des

crip

tive

sta

tist

ics

of c

hap

ter

6 b

ank

sam

ple

Fore

ign

bank

ow

ners

hip

(FO

R) i

s the

199

8-20

03 m

ean

and

scal

ed fr

om 0

-1. O

ther

var

iabl

es a

re a

vera

ges b

ased

on

avai

labl

e ob

serv

atio

ns. P

erso

n-ne

l exp

ense

s (P

E),

othe

r op

erat

ing

inco

me

(OO

I) a

nd e

quity

(EQ

) are

sca

led

to to

tal a

sset

s. In

tere

st e

xpen

ses

(IC

) are

the

sum

of

depo

sits

& s

hort

te

rm fu

ndin

g pl

us o

ther

fund

ing,

sca

led

to to

tal a

sset

s. T

otal

cos

ts (T

C) a

re th

e su

m o

f pe

rson

nel e

xpen

ses,

inte

rest

exp

ense

s an

d ot

her

oper

atin

g ex

pens

es, s

cale

d to

tota

l ass

ets.

Loa

ns (T

L),

secu

ritie

s (T

S) a

nd lo

an lo

ss r

eser

ves (

LL

R) a

re sc

aled

to to

tal a

sset

s. G

ross

dom

estic

pro

duct

per

cap

ita

(GD

PPC

) is

in U

SD. N

omin

al in

tere

st (I

NT

) is

scal

ed a

s hu

ndre

dths

of

perc

enta

ges.

Dom

estic

cre

dit

to t

he p

riva

te s

ecto

r (D

CPS

) is

scal

ed t

o gr

oss d

omes

tic p

rodu

ct. B

ank

conc

entr

atio

n (C

ON

) is m

easu

red

as th

e as

sets

of

the

thre

e la

rges

t ban

ks o

ver

tota

l ban

ks a

sset

s and

scal

ed fr

om 0

-1.

Sour

ces:

Ban

kSco

pe, W

orld

Ban

k W

DI a

nd IM

F In

tern

atio

nal F

inan

cial

Sta

tistic

s. In

stitu

tiona

l ind

icat

ors i

nclu

de V

oice

and

acc

ount

abili

ty (V

AC

), Po

litic

al in

stab

ility

and

vio

lenc

e (P

IV),

Gov

ernm

ent e

ffect

iven

ess (

GE

F), R

egul

ator

y qu

ality

(RE

G),

Rul

e of

Law

(LA

W),

and

Con

trol

of

corr

uptio

n (C

OR

). So

urce

: ww

w.w

orld

bank

.org

/wbi

/gov

erna

nce/

govd

ata

Cou

ntry

Obs

Ban

ksFO

RPE

ICT

CO

OI

EQ

TL

TS

LL

RG

DPP

CIN

TD

CPS

CO

NV

AC

PIV

GE

FR

EG

LA

WC

OR

AL

BA

NIA

22

1.00

0.01

0.03

0.05

0.02

0.06

0.24

0.40

1.93

1,39

20.

140.

080.

80-0

.04

-0.7

3-0

.41

-0.2

2-0

.87

-0.7

8

AN

GO

LA

32

0.00

0.03

0.03

0.07

0.06

0.11

0.22

0.00

4.98

778

0.96

0.05

0.73

-1.3

3-1

.57

-1.2

7-1

.45

-1.4

9-1

.20

AR

GE

NT

INA

9739

0.26

0.04

0.06

0.13

0.04

0.16

0.45

0.23

10.1

47,

621

0.22

0.20

0.37

0.34

0.04

0.02

0.03

-0.2

2-0

.47

AR

ME

NIA

64

0.82

0.02

0.04

0.08

0.03

0.10

0.20

0.10

1.59

705

0.28

0.08

0.56

-0.3

9-0

.56

-0.5

9-0

.16

-0.4

9-0

.69

AU

STR

AL

IA21

90.

570.

010.

050.

070.

020.

060.

570.

260.

7820

,387

0.08

0.86

0.54

1.54

1.27

1.85

1.48

1.93

2.05

AZ

ER

BA

IJA

N7

30.

180.

030.

040.

110.

090.

290.

370.

1831

.39

755

0.18

0.06

0.83

-0.8

6-0

.99

-0.9

1-0

.70

-0.9

0-1

.04

BA

HR

AIN

246

0.56

0.01

0.04

0.05

0.01

0.10

0.49

0.19

10.0

611

,676

0.11

0.61

0.77

-1.0

60.

140.

661.

010.

890.

56

BA

NG

LA

DE

SH19

70.

070.

010.

060.

070.

020.

060.

590.

094.

7037

10.

150.

270.

46-0

.45

-0.6

5-0

.52

-0.6

1-0

.72

-0.7

7

BE

LA

RU

S12

40.

340.

030.

080.

150.

070.

160.

500.

066.

761,

311

0.41

0.11

0.84

-1.2

3-0

.04

-0.9

3-2

.09

-1.0

9-0

.51

BE

LG

IUM

94

0.22

0.01

0.04

0.04

0.01

0.04

0.45

0.32

1.03

22,3

840.

080.

770.

501.

331.

021.

741.

181.

521.

52

BE

LIZ

E2

10.

000.

010.

040.

060.

020.

150.

740.

031.

623,

532

0.15

0.61

1.00

0.85

0.59

0.01

0.18

0.07

-0.1

9

BO

TSW

AN

A17

40.

440.

020.

070.

100.

030.

090.

530.

203.

163,

188

0.15

0.16

0.81

0.76

0.85

0.83

0.78

0.67

0.81

BR

AZ

IL20

976

0.43

0.03

0.25

0.19

0.01

0.16

0.33

0.39

7.98

3,50

30.

690.

350.

330.

470.

00-0

.15

0.29

-0.2

00.

00

BU

LG

AR

IA16

110.

680.

020.

020.

040.

030.

150.

510.

175.

241,

748

0.10

0.21

0.41

0.56

0.44

-0.0

50.

55-0

.01

-0.1

3

CA

NA

DA

8731

0.78

0.01

0.05

0.07

0.01

0.08

0.63

0.18

2.09

22,8

000.

060.

800.

481.

351.

222.

031.

401.

922.

30

CA

PE V

ER

DE

21

0.87

0.01

0.04

0.05

0.02

0.04

0.33

0.17

11.1

51,

275

0.13

0.36

0.98

0.51

0.78

-0.1

2-0

.12

0.19

0.42

CH

ILE

5817

0.40

0.02

0.06

0.09

0.02

0.11

0.60

0.05

2.67

5.00

70.

120.

640.

480.

800.

851.

321.

391.

261.

45

��

Ap

pen

dix

14

con

tinu

edC

ount

ryO

bsB

anks

FOR

PEIC

TC

OO

IE

QT

LT

SL

LR

GD

PPC

INT

DC

PSC

ON

VA

CPI

VG

EF

RE

GL

AW

CO

R

CH

INA

-PE

OPL

E’S

RE

P.4

30.

030.

000.

020.

030.

000.

040.

510.

003.

2195

50.

061.

330.

58-1

.40

0.07

0.20

-0.3

5-0

.31

-0.3

7C

OL

OM

BIA

3614

0.39

0.04

0.14

0.23

0.06

0.13

0.57

0.19

7.66

1,99

70.

230.

290.

33-0

.47

-1.7

6-0

.23

0.14

-0.7

1-0

.47

CO

STA

RIC

A23

70.

220.

020.

100.

130.

020.

170.

730.

122.

354,

187

0.25

0.25

0.37

1.23

1.13

0.57

0.87

0.75

0.89

CR

OA

TIA

6724

0.50

0.02

0.03

0.07

0.02

0.14

0.52

0.08

9.76

4,30

10.

130.

440.

540.

250.

470.

230.

270.

070.

02

CY

PRU

S15

31.

000.

010.

060.

070.

010.

060.

670.

052.

5511

,951

0.08

1.50

0.70

1.07

0.49

1.07

1.16

0.89

1.02

CZ

EC

H R

EP

7222

0.58

0.01

0.05

0.06

0.02

0.08

0.37

0.09

6.59

5,48

90.

080.

460.

611.

000.

950.

710.

870.

640.

37

DO

MIN

ICA

N R

EP.

24

80.

310.

030.

080.

130.

030.

110.

580.

063.

572,

327

0.27

0.36

0.46

0.22

0.13

-0.4

00.

17-0

.26

-0.4

1

EC

UA

DO

R25

110.

270.

030.

100.

140.

030.

090.

510.

1315

.41

1,32

90.

210.

260.

67-0

.02

-0.7

8-0

.92

-0.2

6-0

.67

-0.9

0

EL

SA

LVA

DO

R4

20.

000.

020.

070.

100.

010.

090.

660.

053.

902,

095

0.15

0.44

0.63

0.14

0.33

-0.0

71.

29-0

.30

-0.2

2

EST

ON

IA17

40.

860.

020.

030.

070.

020.

100.

610.

103.

354,

169

0.09

0.26

0.95

0.94

0.93

0.81

1.31

0.71

0.69

FIJI

21

0.51

0.03

0.01

0.07

0.03

0.08

0.60

0.13

3.64

2,29

00.

080.

301.

00-0

.01

0.20

-0.0

5-0

.16

-0.3

30.

08

GA

MB

IA1

11.

000.

020.

030.

090.

070.

090.

370.

005.

4631

00.

240.

171.

00-1

.03

0.56

-0.8

2-0

.54

-0.5

3-0

.74

GE

OR

GIA

RE

P. O

F1

10.

000.

010.

020.

080.

050.

400.

530.

026.

7372

90.

320.

090.

68-0

.32

-1.4

9-0

.78

-0.7

2-1

.03

-0.9

7

GE

RM

AN

Y14

80.

180.

010.

030.

040.

020.

060.

290.

162.

6622

,486

0.09

1.19

0.18

1.40

1.25

1.84

1.38

1.84

1.90

GR

EE

CE

52

0.60

0.02

0.06

0.10

0.02

0.13

0.46

0.24

6.46

11,0

310.

080.

680.

611.

020.

740.

791.

040.

760.

62

HO

ND

UR

AS

32

0.67

0.03

0.07

0.12

0.04

0.21

0.39

0.08

3.31

920

0.30

0.40

0.46

0.09

-0.0

7-0

.33

0.52

-0.6

5-0

.72

HO

NG

KO

NG

9125

0.60

0.01

0.08

0.06

0.01

0.15

0.48

0.14

3.91

24,1

620.

071.

570.

58-0

.12

1.08

1.43

1.65

1.50

1.55

HU

NG

AR

Y38

130.

930.

010.

060.

090.

020.

100.

540.

091.

544,

678

0.13

0.33

0.48

1.15

1.00

0.77

1.16

0.83

0.68

ICE

LA

ND

72

0.00

0.01

0.05

0.07

0.01

0.05

0.34

0.01

4.34

29,3

630.

150.

890.

671.

441.

582.

051.

271.

982.

35

IND

IA54

260.

080.

010.

060.

080.

020.

050.

420.

395.

6147

60.

120.

310.

310.

39-0

.72

-0.0

9-0

.31

0.09

-0.3

1

IND

ON

ESI

A48

0.40

0.01

0.11

0.12

0.04

0.04

0.44

0.25

17.9

773

70.

220.

270.

53-0

.71

-1.5

8-0

.48

-0.3

5-0

.93

-1.0

3

IRE

LA

ND

2810

0.93

0.00

0.05

0.05

0.00

0.10

0.42

0.27

1.08

24,3

940.

051.

040.

501.

391.

411.

831.

631.

781.

77

ISR

AE

L31

100.

170.

010.

070.

080.

010.

070.

720.

083.

1817

,661

0.13

0.87

0.70

0.81

-0.8

61.

010.

890.

991.

18

ITA

LY10

30.

090.

020.

040.

050.

010.

090.

560.

143.

8018

,603

0.06

0.76

0.27

1.11

0.88

0.90

0.92

0.90

0.87

JAM

AIC

A13

60.

150.

030.

090.

150.

070.

150.

210.

487.

783,

103

0.23

0.22

0.80

0.61

-0.0

3-0

.20

0.41

-0.2

9-0

.32

JAPA

N5

30.

100.

010.

010.

030.

020.

080.

550.

212.

8837

,665

0.02

1.55

0.20

0.99

1.19

1.12

0.97

1.44

1.23

JOR

DA

N5

11.

000.

010.

040.

050.

020.

090.

410.

036.

991,

755

0.11

0.74

0.83

-0.3

3-0

.09

0.43

0.37

0.46

0.14

��

Ap

pen

dix

14

con

tinu

edC

ount

ryO

bs.

Ban

ksFO

RPE

ICT

CO

OI

EQ

TL

TS

LL

RG

DPP

CIN

TD

CPS

CO

NV

AC

PIV

GE

FR

EG

LA

WC

OR

KO

RE

A R

EP.

OF

2711

0.13

0.01

0.05

0.06

0.01

0.04

0.58

0.24

4.36

10,9

530.

090.

930.

360.

700.

430.

720.

570.

750.

29

KU

WA

IT7

20.

000.

010.

050.

050.

010.

110.

410.

2015

.03

16,8

330.

080.

680.

54-0

.28

0.61

0.11

0.02

1.07

0.99

LA

TV

IA25

130.

780.

020.

030.

070.

020.

100.

480.

123.

273,

442

0.11

0.21

0.59

0.82

0.76

0.45

0.75

0.28

0.04

LE

BA

NO

N78

240.

320.

010.

070.

080.

010.

100.

310.

0311

.64

3,86

90.

140.

860.

35-0

.50

-0.5

4-0

.19

0.02

-0.1

1-0

.41

LIT

HU

AN

IA23

70.

750.

030.

030.

070.

030.

140.

550.

133.

333,

461

0.10

0.14

0.86

0.90

0.73

0.44

0.67

0.33

0.23

MA

CA

U5

41.

000.

010.

040.

050.

010.

120.

320.

274.

8115

,244

0.07

0.65

0.63

0.42

0.56

0.86

0.71

0.73

-0.0

7

MA

CE

DO

NIA

(FY

RO

M)

21

0.22

0.03

0.06

0.12

0.09

0.19

0.77

0.01

36.1

61,

717

0.17

0.19

0.74

-0.2

3-0

.96

-0.3

2-0

.12

-0.4

5-0

.67

MA

LAY

SIA

9131

0.48

0.01

0.05

0.05

0.01

0.09

0.56

0.18

8.39

3,79

30.

081.

470.

26-0

.27

0.38

0.83

0.48

0.60

0.42

MA

LTA

84

0.25

0.01

0.06

0.07

0.02

0.08

0.36

0.37

3.08

9,52

10.

071.

120.

871.

351.

300.

970.

680.

820.

52

MA

UR

ITIU

S11

30.

470.

010.

070.

070.

010.

130.

660.

162.

493,

833

0.21

0.61

0.81

1.00

1.15

0.57

0.56

0.88

0.44

ME

XIC

O22

110.

590.

040.

490.

310.

040.

230.

500.

164.

635,

726

0.19

0.20

0.37

0.06

-0.1

60.

270.

66-0

.36

-0.3

4

MO

LD

OV

A R

EP.

OF

54

0.54

0.03

0.04

0.10

0.08

0.24

0.46

0.07

6.92

345

0.24

0.18

0.57

-0.2

9-0

.16

-0.6

9-0

.37

-0.5

5-0

.88

MO

NG

OL

IA7

20.

290.

020.

060.

100.

030.

070.

430.

1310

.05

403

0.31

0.16

0.76

0.57

0.88

-0.0

60.

150.

24-0

.12

MO

RO

CC

O6

30.

580.

010.

030.

050.

020.

090.

610.

065.

411,

222

0.13

0.56

0.43

-0.3

9-0

.08

0.03

0.08

0.14

0.11

MO

ZA

MB

IQU

E15

50.

760.

040.

050.

140.

020.

080.

320.

1222

.93

224

0.23

0.09

0.84

-0.2

1-0

.07

-0.3

1-0

.29

-0.7

2-0

.64

NA

MIB

IA12

30.

510.

020.

090.

120.

030.

080.

740.

033.

141,

807

0.16

0.47

0.79

0.35

0.12

0.26

0.34

0.80

0.51

NE

PAL

73

0.46

0.01

0.05

0.06

0.02

0.09

0.47

0.09

5.24

231

0.12

0.29

0.67

-0.0

6-0

.88

-0.8

2-0

.36

-0.3

0-0

.58

NE

TH

ER

LA

ND

S13

60.

230.

010.

050.

050.

010.

090.

640.

141.

3223

,158

0.04

1.39

0.72

1.56

1.48

2.18

1.86

1.90

2.30

NE

W Z

EA

LA

ND

205

0.33

0.01

0.05

0.06

0.01

0.05

0.78

0.07

0.53

17,9

090.

101.

120.

611.

491.

391.

771.

572.

002.

40

NIG

ER

IA15

70.

240.

020.

080.

130.

060.

130.

380.

0511

.92

335

0.21

0.15

0.38

-0.8

8-1

.42

-1.1

3-0

.74

-1.2

4-1

.14

NO

RW

AY7

30.

000.

010.

060.

080.

010.

070.

830.

032.

9440

,132

0.08

0.83

0.54

1.57

1.49

1.78

1.28

1.95

2.06

OM

AN

93

0.27

0.01

0.05

0.06

0.01

0.18

0.75

0.05

3.63

8,36

10.

100.

420.

66-0

.66

1.00

0.96

0.64

1.12

0.87

PAN

AM

A17

70.

710.

010.

050.

060.

010.

100.

630.

052.

484,

049

0.10

0.97

0.28

0.57

0.44

-0.0

40.

81-0

.03

-0.2

6

PAPU

A N

EW

GU

INE

A8

40.

760.

030.

080.

120.

050.

090.

430.

305.

1567

80.

160.

160.

890.

02-0

.57

-0.7

2-0

.53

-0.5

9-0

.78

PHIL

IPPI

NE

S48

200.

260.

010.

060.

080.

020.

160.

520.

076.

7099

30.

120.

420.

380.

31-0

.41

0.04

0.31

-0.4

2-0

.43

POL

AN

D45

190.

820.

020.

070.

100.

030.

120.

500.

175.

344,

322

0.17

0.27

0.43

1.09

0.75

0.60

0.68

0.59

0.43

��

Ap

pen

dix

14

con

tinu

edC

ount

ryO

bsB

anks

FOR

PEIC

TC

OO

IE

QT

LT

SL

LR

GD

PPC

INT

DC

PSC

ON

VA

CPI

VG

EF

RE

GL

AW

CO

R

POR

TU

GA

L12

70.

220.

010.

050.

050.

020.

090.

450.

063.

569,

958

0.06

1.09

0.37

1.37

1.46

1.38

1.16

1.26

1.53

RU

SSIA

N F

ED

ER

AT

ION

103

430.

330.

030.

070.

110.

050.

200.

450.

117.

081,

819

0.25

0.16

0.55

-0.4

2-0

.60

-0.5

1-0

.78

-0.8

2-0

.88

SEY

CH

EL

LE

S2

10.

220.

000.

020.

030.

010.

070.

150.

450.

007,

118

0.11

0.29

1.00

0.13

1.01

-0.2

4-0

.67

0.32

0.27

SIE

RR

A L

EO

NE

103

0.27

0.04

0.02

0.10

0.09

0.16

0.14

0.40

33.2

513

30.

240.

030.

91-1

.11

-1.5

2-1

.19

-1.2

1-0

.98

-0.7

8

SLO

VA

K R

EP.

3713

0.66

0.01

0.06

0.09

0.01

0.09

0.41

0.20

9.83

3,87

60.

150.

450.

680.

810.

860.

310.

540.

290.

20

SLO

VE

NIA

2310

0.67

0.02

0.05

0.08

0.02

0.10

0.53

0.22

6.35

9,68

10.

140.

370.

571.

021.

150.

830.

760.

950.

96

SOU

TH

AFR

ICA

6016

0.12

0.02

0.09

0.13

0.02

0.13

0.73

0.08

5.67

2,94

10.

161.

350.

330.

88-0

.34

0.44

0.34

0.21

0.45

SPA

IN41

110.

210.

010.

030.

040.

010.

090.

580.

151.

8313

,770

0.05

0.99

0.48

1.19

0.84

1.80

1.32

1.28

1.58

SRI

LA

NK

A7

20.

050.

020.

060.

080.

020.

060.

510.

168.

2486

90.

150.

290.

67-0

.23

-1.4

5-0

.23

0.36

-0.0

1-0

.15

SWA

ZIL

AN

D5

20.

790.

040.

060.

120.

040.

080.

600.

062.

321,

343

0.15

0.14

0.84

-1.2

00.

32-0

.49

-0.2

2-0

.40

-0.2

9

SWE

DE

N21

70.

240.

010.

030.

040.

020.

050.

650.

150.

6826

,765

0.06

0.92

0.55

1.57

1.46

1.86

1.44

1.92

2.40

SWIT

ZE

RL

AN

D34

910

40.

550.

030.

030.

050.

050.

190.

460.

135.

2033

,825

0.04

1.61

0.65

1.61

1.69

2.32

1.46

2.18

2.32

TA

NZ

AN

IA1

11.

000.

010.

010.

030.

050.

090.

390.

211.

3530

90.

140.

080.

58-0

.38

-0.3

1-0

.43

-0.5

2-0

.48

-0.7

7

TH

AIL

AN

D39

110.

300.

010.

040.

060.

010.

060.

550.

2212

.54

2,05

20.

091.

160.

450.

200.

290.

220.

400.

31-0

.28

TR

INID

AD

AN

D T

OB

AG

O10

30.

500.

010.

050.

080.

030.

130.

370.

112.

036,

432

0.15

0.43

0.67

0.66

0.32

0.55

0.75

0.37

0.15

UK

RA

INE

2612

0.38

0.03

0.06

0.11

0.04

0.13

0.57

0.04

7.65

694

0.34

0.15

0.44

-0.4

9-0

.19

-0.7

8-0

.87

-0.7

8-0

.95

UN

ITE

D A

RA

B E

MIR

AT

ES

125

0.14

0.01

0.04

0.04

0.01

0.17

0.48

0.07

10.6

321

,036

0.09

0.51

0.44

-0.6

11.

070.

660.

591.

310.

78

UN

ITE

D K

ING

DO

M22

465

0.40

0.01

0.04

0.05

0.02

0.13

0.42

0.14

3.51

24,4

920.

051.

340.

241.

420.

922.

131.

691.

882.

13

UR

UG

UAY

32

1.00

0.02

0.20

0.20

0.02

0.09

0.80

0.11

1.10

6,27

40.

560.

480.

670.

820.

710.

681.

030.

570.

50

USA

669

240

0.11

0.01

0.03

0.06

0.02

0.09

0.61

0.28

1.92

34,0

420.

072.

290.

101.

290.

941.

761.

511.

781.

83

VE

NE

ZU

EL

A21

70.

400.

040.

070.

100.

030.

170.

450.

235.

344,

799

0.31

0.11

0.49

-0.2

4-0

.72

-0.9

6-0

.43

-0.8

8-0

.79

VIE

TN

AM

104

0.20

0.01

0.05

0.05

0.01

0.10

0.66

0.03

0.34

412

0.11

0.34

0.72

-1.5

00.

45-0

.27

-0.6

4-0

.64

-0.6

7

ZA

MB

IA25

80.

670.

060.

070.

160.

070.

120.

270.

076.

6433

40.

410.

070.

62-0

.27

-0.2

2-0

.66

-0.0

4-0

.45

-0.7

8

��

AppendicesA

pp

end

ix 1

5 C

orre

lati

on m

atri

x of

inef

fici

ency

cor

rela

tes

(for

eign

ban

ks

only

)E

QV

AC

host

PIV

host

GE

F host

RE

Gho

stL

AW

host

CO

Rho

stV

AC

hom

ePI

Vho

me

GE

F hom

eR

EG

hom

eL

AW

hom

eC

OR

hom

eSi

mila

rity

EQ

1

VA

Cho

st0.

088

1

PIV

host

0.10

60.

815

1

GE

F host

0.13

00.

804

0.83

31

RE

Gho

st0.

066

0.75

70.

791

0.88

51

LA

Who

st0.

106

0.80

80.

880

0.97

90.

889

1

CO

Rho

st0.

113

0.80

70.

846

0.97

60.

869

0.98

31

VA

Cho

me

-0.1

570.

137

0.07

1-0

.040

-0.0

24-0

.034

-0.0

331

PIV

hom

e-0

.183

0.07

30.

077

-0.0

090.

002

-0.0

03-0

.009

0.72

41

GE

F hom

e-0

.205

0.04

80.

039

-0.0

38-0

.034

-0.0

36-0

.039

0.76

30.

808

1

RE

Gho

me

-0.1

860.

034

0.01

4-0

.075

-0.0

37-0

.077

-0.0

820.

730

0.75

50.

898

1

LA

Who

me

-0.2

100.

071

0.05

7-0

.003

-0.0

070.

003

-0.0

040.

755

0.86

50.

938

0.85

51

CO

Rho

me

-0.2

170.

064

0.05

7-0

.014

-0.0

14-0

.007

-0.0

120.

768

0.83

30.

965

0.87

30.

963

1

Sim

ilari

ty0.

046

0.21

60.

252

0.47

80.

348

0.46

20.

475

-0.1

11-0

.057

-0.0

37-0

.042

-0.0

55-0

.053

1

Ban

k eq

uity

(EQ

) is s

cale

d to

tota

l ass

ets.

The

inst

itutio

nal i

ndic

ator

s inc

lude

voi

ce a

nd a

ccou

ntab

ility

(VA

C),

polit

ical

inst

abili

ty a

nd v

iole

nce

(PIV

), go

vern

men

t effe

ctiv

enes

s (G

EF)

, rul

e of

law

(LA

W) a

nd c

ontr

ol o

f co

rrup

tion

(CO

R).

Defi

nitio

ns a

re in

Tab

le 1

. The

sub

scri

pts

host

an

d ho

me

deno

te w

heth

er th

e in

stitu

tiona

l var

iabl

e re

fers

to th

e ho

st o

r ho

me

coun

try

of th

e fo

reig

n ba

nk.

��

Appendix 16a Estimation of the cost frontier (model specification Table 6.4a)LN(TC/PL)

(1) (2) (3) (4) (5) (6) (7)

Outputs and input prices with trend

Intercept 4.11 4.19 4.09 4.18 4.20 4.13 4.11

(23.44)*** (23.72)*** (22.7)*** (23.41)*** (23.57)*** (23.11)*** (23.17)***

Ln(PF/PL) 0.18 0.18 0.18 0.18 0.19 0.18 0.18

(5.45)*** (5.48)*** (5.34)*** (5.41)*** (5.65)*** (5.53)*** (5.28)***

Ln(loans) 0.31 0.30 0.31 0.30 0.29 0.30 0.30

(11.55)*** (11.3)*** (11.51)*** (11.21)*** (11.09)*** (11.48)*** (11.53)***

Ln(Securities) 0.38 0.38 0.38 0.38 0.38 0.38 0.39

(21.92)*** (21.72)*** (21.78)*** (21.75)*** (21.75)*** (21.76)*** (21.89)***

1/2(ln(PF/PL))2 0.15 0.15 0.15 0.15 0.15 0.15 0.15

(26.27)*** (26.33)*** (25.77)*** (26.08)*** (26.57)*** (27.19)*** (26.14)***

1/2(ln(loans))2 0.10 0.10 0.10 0.10 0.10 0.10 0.10

(35.42)*** (35.03)*** (34.43)*** (35.57)*** (34.93)*** (34.53)*** (34.36)***

1/2(ln(securities))2 0.07 0.07 0.07 0.07 0.07 0.07 0.07

(36.28)*** (36.18)*** (35.61)*** (36.26)*** (36.3)*** (35.89)*** (36.09)***

ln(PF/PL) x ln(loans) 0.02 0.02 0.02 0.02 0.02 0.02 0.02

(5.76)*** (5.46)*** (5.7)*** (5.48)*** (5.38)*** (5.67)*** (5.72)***

ln(PF/PL) x ln(Securities) 0.001 0.001 0.001 0.002 0.002 0.001 0.001

(0.25) (0.49) (0.29) (0.51) (0.55) (0.45) (0.42)

ln((PF/PL) x Year -0.01 -0.01 -0.01 -0.01 -0.01 -0.01 -0.01

(-2.6)*** (-2.52)** (-2.59)*** (-2.57)** (-2.68)*** (-2.71)*** (-2.62)***

ln(loans) x ln(Securities) -0.08 -0.08 -0.08 -0.08 -0.08 -0.08 -0.08

(-38.32)*** (-38.07)*** (-37.68)*** (-38.5)*** (-38.13)*** (-37.95)*** (-37.92)***

ln(loans) x Year -0.004 -0.004 -0.004 -0.004 -0.004 -0.004 -0.004

(-1.66)* (-1.66)* (-1.62) (-1.67)* (-1.69)* (-1.66)* (-1.52)

ln(Securities) x Year 0.001 0.002 0.001 0.002 0.002 0.002 0.001

(0.71) (0.74) (0.66) (0.76) (0.92) (0.73) (0.65)

Year 0.04 0.04 0.04 0.04 0.04 0.04 0.04

(1.69)* (1.62) (1.67)* (1.58) (1.54) (1.65)* (1.54)

1/2Year2 0.0006 0.0008 0.0006 0.0009 0.0002 0.0006 0.0008

(0.13) (0.17) (0.13) (0.2) (0.05) (0.15) (0.17)

Country-level variablesGDPPC (x 1,000) -0.00001 -0.00001 -0.00001 -0.00001 -0.00001 -0.00001 -0.00001

(-10.91)*** (-10.71)*** (-10.58)*** (-10.46)*** (-10.48)*** (-10.42)*** (-10.44)***

INTEREST (nominal) -0.20 -0.20 -0.20 -0.21 -0.21 -0.20 -0.20

(-4.46)*** (-4.63)*** (-4.47)*** (-4.68)*** (-4.75)*** (-4.76)*** (-4.56)***

CONCENTRATION 0.16 0.16 0.17 0.16 0.16 0.16 0.17

(5)*** (4.84)*** (5.05)*** (4.91)*** (4.86)*** (5.06)*** (5.09)***

DCPS (% of GDP) 0.09 0.09 0.09 0.09 0.09 0.09 0.09

(6.03)*** (5.86)*** (5.97)*** (5.9)*** (5.83)*** (5.9)*** (5.89)***

Control for variation in riskLOAN LOSS RESERVES (% gross loans)

0.003 0.003 0.003 0.003 0.003 0.003 0.003

(3.32)*** (3.23)*** (3.23)*** (3.39)*** (3.22)*** (3.38)*** (3.3)***

OTHER INCOME (% total assets)

0.01 0.02 0.01 0.01 0.02 0.01 0.01

-0.12 (0.17) (0.1) (0.15) (0.23) (0.12) (0.12)

��

Appendix 16b Estimation of the cost frontier (model specification Table 6.4b)LN(TC/PL)

(1) (2) (3) (4) (5) (6)


Intercept 3.48 3.52 3.51 3.53 3.47 3.46

(11.12)*** (10.98)*** (11.32)*** (11.41)*** (11.11)*** (11.02)***

Ln(PF/PL) -0.01 0.00 0.01 0.02 0.01 0.02

(-0.11) (-0.07) (0.25) (0.31) (0.14) (0.44)

Ln(loans) 0.50 0.50 0.49 0.49 0.50 0.49(11.13)*** (10.95)*** (11.17)*** (10.95)*** (11.05)*** (10.97)***

Ln(Securities) 0.32 0.32 0.32 0.32 0.32 0.32

(11.3)*** (11.24)*** (11.34)*** (11.38)*** (11.38)*** (11.32)***

1/2(ln(PF/PL))2 0.10 0.10 0.10 0.10 0.10 0.10

(8.6)*** (8.45)*** (8.51)*** (8.35)*** (8.46)*** (8.23)***

1/2(ln(loans))2 0.08 0.08 0.08 0.08 0.08 0.08

(18.01)*** (17.58)*** (17.85)*** (17.59)*** (17.61)*** (17.54)***

1/2(ln(securities))2 0.07 0.07 0.07 0.07 0.07 0.07

(22.48)*** (22.14)*** (22.21)*** (22.32)*** (22.23)*** (22.32)***

ln(PF/PL) x ln(loans) 0.04 0.04 0.04 0.04 0.04 0.04

(7.06)*** (6.91)*** (6.7)*** (6.65)*** (6.76)*** (6.55)***ln(PF/PL) x ln(Securities) -0.004 -0.004 -0.004 -0.004 -0.004 -0.004

(-0.73) (-0.78) (-0.78) (-0.77) (-0.79) (-0.75)

ln((PF/PL) x Year -0.004 -0.004 -0.005 -0.005 -0.004 -0.005(-0.67) (-0.73) (-0.78) (-0.94) (-0.66) (-0.94)

ln(loans) x ln(Securities) -0.07 -0.07 -0.07 -0.07 -0.07 -0.07(-21.91)*** (-21.37)*** (-21.39)*** (-21.48)*** (-21.37)*** (-21.38)***

ln(loans) x Year -0.01 -0.01 -0.01 -0.01 -0.01 -0.01

(-2.28)** (-2.31)** (-2.3)** (-2.21)** (-2.31)** (-2.28)**

ln(Securities) x Year 0.01 0.01 0.01 0.01 0.01 0.01(2.95)*** (2.93)*** (2.93)*** (2.99)*** (2.87)*** (2.94)***

Year -0.003 0.001 0.003 -0.004 0.006 0.006

(-0.09) (0.02) (0.07) (-0.09) (0.14) (0.15)

1/2Year2 0.01 0.01 0.01 0.01 0.01 0.01(0.94) (0.88) (0.82) (0.82) (0.78) (0.73)

Country-level variables

GDPPC (x 1,000) -0.000004 -0.000005 -0.000003 -0.000001 -0.000004 -0.000002

(-0.74) (-1.05) (-0.67) (-0.32) (-0.82) (-0.41)

NOMINAL INTEREST RATE -0.24 -0.23 -0.24 -0.24 -0.23 -0.24

(-4.41)*** (-4.21)*** (-4.28)*** (-4.42)*** (-4.21)*** (-4.29)***

CONCENTRATION RATE 0.26 0.25 0.25 0.25 0.25 0.26

(4.52)*** (4.4)*** (4.4)*** (4.44)*** (4.45)*** (4.65)***

DCPS (% GDP) -0.03 -0.03 -0.03 -0.03 -0.03 -0.02

(-1.1) (-1.04) (-0.91) (-0.98) (-0.86) (-0.84)Control for variation in riskLOAN LOSS RESERVES(% GROSS LOANS)

0.002 0.002 0.002 0.002 0.002 0.002

(2.19)** (2.26)** (2.13)** (1.86)* (2.14)** (1.9)*

OTHER INCOME(% TOTAL ASSETS)

0.45 0.44 0.44 0.44 0.44 0.43

(3.76)*** (3.69)*** (3.68)*** (3.66)*** (3.64)*** (3.68)***

��0 ��

Appendix 17 Estimation of the cost frontier (model specification Table 6.5)LN(TC/PL)

(1) (2) (3) (4) (5) (6)


Intercept 4.24 4.06 4.08 4.28 4.10 4.15(11.22)*** (12.03)*** (10.51)*** (11.33)*** (10.49)*** (10.53)***

Ln(PF/PL) -0.08 -0.09 -0.09 -0.09 -0.08 -0.09

(-1.25) (-1.36) (-1.44) (-1.46) (-1.34) (-1.42)

Ln(loans) 0.26 0.28 0.28 0.26 0.28 0.27(4.89)*** (5.42)*** (5.2)*** (4.75)*** (5.13)*** (4.96)***

Ln(Securities) 0.50 0.51 0.51 0.51 0.51 0.51

(13.77)*** (14.16)*** (13.96)*** (14.02)*** (13.82)*** (13.96)***

1/2(ln(PF/PL))2 0.11 0.11 0.11 0.11 0.11 0.11

(11.37)*** (11.35)*** (11.2)*** (11.36)*** (11.36)*** (11.31)***

1/2(ln(loans))2 0.11 0.11 0.11 0.11 0.11 0.11

(18)*** (17.89)*** (17.91)*** (18.03)*** (17.9)*** (17.96)***

1/2(ln(securities))2 0.07 0.07 0.07 0.07 0.07 0.07

(20.82)*** (20.82)*** (20.66)*** (20.76)*** (20.68)*** (20.81)***

ln(PF/PL) x ln(loans) 0.03 0.03 0.04 0.03 0.03 0.03(5.12)*** (5.38)*** (6.09)*** (5.33)*** (5.5)*** (5.36)***

ln(PF/PL) x ln(Securities) 0.01 0.01 0.01 0.01 0.01 0.01

(1.82)* (1.78)* (1.62) (1.78)* (1.6) (1.73)*

ln((PF/PL) x Year -0.01 -0.01 -0.01 -0.01 -0.01 -0.01(-1.06) (-1.57) (-1.24) (-1.03) (-1.27) (-1.2)

ln(loans) x ln(Securities) -0.09 -0.09 -0.09 -0.09 -0.09 -0.09

(-20.99)*** (-21.88)*** (-21.04)*** (-21.08)*** (-20.85)*** (-21.11)***

ln(loans) x Year -0.004 -0.004 -0.004 -0.003 -0.004 -0.004(-0.79) (-1.03) (-0.97) (-0.66) (-0.92) (-0.86)

ln(Securities) x Year 0.01 0.01 0.01 0.01 0.01 0.01

(1.65)* (1.93)* (1.94)* (1.69)* (1.87)* (1.8)*

Year -0.05 -0.04 -0.05 -0.05 -0.05 -0.05(-0.97) (-0.87) (-0.99) (-1.13) (-0.98) (-1.04)

1/2Year2 0.01 0.01 0.01 0.01 0.01 0.01

(0.93) (0.81) (1.02) (0.86) (0.98) (1.01)

Country-level variablesGDPPC (x 1,000) -0.00001 -0.00001 -0.00001 -0.00001 -0.00001 -0.00001

(-3.96)*** (-4.01)*** (-4.13)*** (-4.07)*** (-4.16)*** (-4.02)***

NOMINAL INTEREST RATE

-0.50 -0.48 -0.50 -0.49 -0.50 -0.49

(-6.51)*** (-6.85)*** (-6.46)*** (-6.52)*** (-6.38)*** (-6.34)***

CONCENTRATION RATE 0.02 0.02 0.03 0.02 0.02 0.02(0.26) (0.41) (0.41) (0.29) (0.34) (0.26)

DCPS (% GDP) -0.03 -0.03 -0.03 -0.03 -0.03 -0.03

(-1.08) (-1.05) (-1) (-0.94) (-1.04) (-1.02)Control for variation in risk

LOAN LOSS RESERVES (% GROSS LOANS)

-0.001 -0.001 -0.001 -0.001 -0.001 -0.001

(-0.74) (-0.81) (-0.61) (-0.93) (-0.59) (-0.77)OTHER INCOME (% TOTAL ASSETS)

0.56 0.55 0.57 0.58 0.57 0.59

(3.05)*** (3.13)*** (3.23)*** (3.23)*** (3.21)*** (3.24)***

��0 ��

Appendices

Appendix 18 Estimation of the cost frontier (model specification Table 6.6)LN(TC/PL)

(1) (2) (3) (4) (5) (6)

Outputs and input prices with trendIntercept 4.18 4.18 4.20 4.06 4.23 4.12

(10.47)*** (10.92)*** (11.17)*** (11.13)*** (11.34)*** (10.27)***

Ln(PF/PL) -0.09 -0.09 -0.08 -0.09 -0.10 -0.09

(-1.39) (-1.43) (-1.24) (-1.39) (-1.54) (-1.47)

Ln(loans) 0.27 0.27 0.27 0.28 0.26 0.27

(4.82)*** (4.87)*** (5.01)*** (5.3)*** (4.93)*** (4.99)***

Ln(Securities) 0.51 0.51 0.51 0.52 0.51 0.52

(13.86)*** (14.05)*** (14.07)*** (14.3)*** (14.04)*** (14.09)***

1/2(ln(PF/PL))2 0.11 0.11 0.11 0.11 0.11 0.11

(11.79)*** (11.42)*** (11.57)*** (11.49)*** (11.23)*** (10.88)***

1/2(ln(loans))2 0.11 0.11 0.11 0.11 0.11 0.11

(18.12)*** (18.27)*** (18.31)*** (18.26)*** (18.38)*** (18.35)***

1/2(ln(securities))2 0.07 0.07 0.07 0.07 0.07 0.07

(20.6)*** (21.17)*** (20.83)*** (21.47)*** (20.55)*** (21.47)***

ln(PF/PL) x ln(loans) 0.03 0.03 0.03 0.03 0.03 0.03

(5.06)*** (5.41)*** (4.93)*** (5.36)*** (5.37)*** (5.31)***

ln(PF/PL) x ln(Securities) 0.01 0.01 0.01 0.01 0.01 0.01

(2.07)** (1.89)* (2.11)** (1.92)* (1.96)** (2.12)**

ln((PF/PL) x Year -0.01 -0.01 -0.01 -0.01 -0.01 -0.01

(-1.16) (-1.28) (-1.36) (-1.55) (-1.43) (-1.57)

ln(loans) x ln(Securities) -0.09 -0.09 -0.09 -0.09 -0.09 -0.09

(-20.86)*** (-21.03)*** (-21.17)*** (-21.62)*** (-20.95)*** (-21.38)***

ln(loans) x Year -0.004 -0.004 -0.003 -0.003 -0.003 -0.003

(-0.81) (-0.86) (-0.66) (-0.75) (-0.69) (-0.71)

ln(Securities) x Year 0.01 0.01 0.01 0.01 0.01 0.01

(1.79)* (1.85)* (1.73)* (2.01)** (1.89)* (1.91)*

Year -0.05 -0.05 -0.06 -0.06 -0.07 -0.07

(-1.11) (-1.04) (-1.28) (-1.23) (-1.37) (-1.35)

1/2Year2 0.01 0.01 0.01 0.01 0.01 0.01

(1) (0.86) (1.08) (0.85) (1.06) (1.09)

Country-level variablesGDPPC (x 1,000) -0.00001 -0.00001 -0.00001 -0.00001 -0.00001 -0.00001

(-3.6)*** (-3.88)*** (-3.7)*** (-3.77)*** (-3.52)*** (-3.38)***

NOMINAL INTEREST RATE -0.51 -0.51 -0.55 -0.54 -0.54 -0.53

(-7.02)*** (-6.61)*** (-6.86)*** (-6.91)*** (-6.76)*** (-6.94)***

CONCENTRATION RATE 0.02 0.02 0.01 0.02 0.01 0.01

(0.26) (0.4) (0.18) (0.26) (0.19) (0.22)

DCPS (% GDP) -0.04 -0.03 -0.04 -0.03 -0.03 -0.04

(-1.2) (-1.09) (-1.13) (-1.01) (-1.1) (-1.16)

Control for variation in risk

LOAN LOSS RESERVES (% GROSS LOANS)

-0.001 -0.001 -0.001 -0.002 -0.002 -0.002

(-1.02) (-1) (-1.04) (-1.19) (-1.15) (-1.17)

OTHER INCOME(% TOTAL ASSETS)

0.55 0.56 0.60 0.59 0.62 0.59

(3.03)*** (3.17)*** (3.35)*** (3.27)*** (3.44)*** (3.25)***

��

Appendix 19 Generalised likelihood ratio tests

To test the impact of efficiency on our model, we perform a generalized likelihood ratio (LR) test. Our model has two types of variances, the variance of the error term and the variance of the efficiency term. These are sv

2 and su2 respectively. The total

variance of the model is:

s2 = sv2 + su

2 (a)

The inclusion of the inefficiency term adds a certain amount of variance to the model. A measure of this amount, with respect to the total variance, is:

g = su

2

(b) s2

The other effects of the inclusion of efficiency in the model can be determined by the coefficients d0 up to di . The total effect of efficiency can be tested with the null hypothesis that there is no effect of efficiency on the model.

h0: g= d0 = d1=...= 0 (c)

This can be tested with the generalized LR test:

LR=-2{lnL[(h0)]-ln[L(ha)]} (d)

which can be calculated by obtaining the log likelihood of a standard ordinary least squares (OLS) model (h0) and the log likelihood of the Battese and Coelli (1995) model. This LR test follows a mixed chi-squared distribution with degrees of freedom equal to the number of restrictions (Coelli, Rao, and Battese, 1998). The significance level is determined with the use of table 1 in Kodde and Palm (1986).

��

Samenvatting en Conclusie (Summary in Dutch)

InleidingHet onderzoek van dit proefschrift richt zich op de aard en de resultaten van buitenlands bankieren door middel van een fysieke aanwezigheid in het buitenland. De internationale standaardisering van bankregulering tussen jurisdicties heeft ertoe geleid dat banken uit verschillende landen in principe kunnen concurreren op een level playing field. Waarneembaar is dat banken, daar waar zij voorheen hun internationale kredietverlening vooral verzorgden vanuit de historische thuismarkt, ook wel cross-border banking genoemd, in toenemende mate zijn overgegaan tot het opzetten van één of meerdere fysieke vestigingen in het buitenland. Een dergelijke buitenlandse vestiging kan de vorm aannemen van een nieuw kantoor, een zogenoemde greenfield. Ook kan een buitenlandse aanwezigheid worden gerealiseerd door de aankoop van een al bestaande bank. Buitenlandse banken zijn vooral actief in het Caraïbisch gebied, Centraal- en Oost-Europa, Centraal Azië en Sub-Sahara Afrika. Dit proefschrift wil meer duidelijkheid verschaffen over de mogelijke verschillen in gedrag en prestatie tussen banken waarvan de eigenaren zich in het binnenland bevinden en banken waarvan de eigenaren zich in het buitenland bevinden. De sterke aanwezigheid van buitenlandse banken in Centraal en Oost-Europa geeft aanleiding een belangrijk deel van de analyse uit te voeren aan de hand van gegevens in deze regio. De focus hierbij ligt op buitenlands bankieren in de landen die op 1 mei 2004 tot de Europese Unie zijn toegetreden, of dit naar verwachting op korte termijn zullen doen, te weten: Bulgarije, Estland, Hongarije, Letland, Litouwen, Kroatië, Polen, Roemenië, Slovenië, Slowakije en de Tsjechische Republiek. Aan dit proefschrift liggen drie onderzoeksvragen ten grondslag. Ten eerste, wat is de rol van buitenlandse banken in de bancaire sector van Centraal- en Oost-Europa? Ten tweede, waarom zijn buitenlandse banken activiteiten begonnen in deze regio en hoe houden deze argumenten verband met de manier waarop ze hun aanwezigheid vormgeven? Ten derde, wat is het effect van buitenlands eigendom op de winstgevendheid en de efficiency die zich hierdoor kenmerken? De vragen worden beantwoord aan de hand van beschrijvende statistiek, interviewresultaten en empirisch onderzoek.

SamenvattingIn Hoofdstuk 1 staat centraal de rol van buitenlandse banken in de bancaire sector van Centraal- en Oost-Europa. De nexus tussen de aanwezigheid van buitenlandse banken en krediet aan de private sector in transitielanden is tot nog toe in de literatuur onderbelicht gebleven. De global advantage theorie (Berger, 2001) veronderstelt dat, met name in opkomende economieën en transitielanden, buitenlandse banken, doorgaans afkomstig uit westerse economieën, gezien hun ervaring over effectievere intermediatietechnieken beschikken dan binnenlandse banken, met name op het

��

gebied van screenen, monitoren en het prijzen van risico. Er zijn echter factoren inherent in het bankieren in Europese transitielanden die deze relatieve voordelen van buitenlandse banken tegenwerken. Zo maakt het ontbreken van betrouwbare historische gegevens van locale bedrijven het vooral voor buitenlandse banken moeilijk om het probleem van averse selectie, het verlenen van krediet aan bedrijven met slechte vooruitzichten, op te lossen. Het is zelfs waarschijnlijk dat binnenlandse banken de zogenoemde soft information (meningen, visie, rapporten, aantekeningen) van locale bedrijven beter kunnen achterhalen en interpreteren waardoor de kwaliteit van hun kredietbeslissing die van buitenlandse banken overtreft. Het gevolg kan zijn dat buitenlandse banken zich onthouden van kredietverlening aan bedrijven die voor hen in eerste instantie minder transparant zijn.

Sectie één van het hoofdstuk laat de groei in de totale activa onder beheer van buitenlandse banken in Centraal- en Oost-Europa zien. Duidelijk is dat zich hierin binnen de regio grote verschillen hebben voorgedaan. De toename van het aantal buitenlandse banken in Centraal- en Oost-Europa heeft doorgezet tot 2000. Na 2000 stabiliseerde het totale aantal buitenlandse banken en vanaf 2002 is ook het relatieve aantal buitenlandse banken stabiel op 61 procent van het totaal van binnenlandse en buitenlandse banken. In 1995 bedroeg het relatieve aantal buitenlandse banken 25 procent.

De relatieve activa onder beheer van buitenlandse banken overtreft het relatieve aantal buitenlandse banken. Vanaf 2002 beheren de buitenlandse banken 84 procent van alle activa binnen de bancaire sector in Centraal- en Oost-Europa. In 1995 bedroeg dit percentage 4 procent. De regionale verschillen in de toetreding van buitenlandse banken doen zich mede voor in de tijd. In Hongarije en Letland beheerden buitenlandse banken in 1996 al een groot deel van de bancaire activa. In 1997 deed dit zich voor in Litouwen, in 1998 in Estland en in 1999 in Polen. De gemiddeld grootste toename in activa onder beheer van buitenlandse banken deed zich voor in 2000, vooral in Bulgarije, Kroatië, Tsjechië, Slowakije en Slovenië. Een uitzondering op deze groeiontwikkeling vormt Letland waar het aandeel van activa van buitenlandse banken vanaf 1998 is afgenomen van 81 procent tot 49 procent in 2004.

Het hoofdstuk vervolgt met onderzoek naar de herkomst van buitenlandse banken. In 2002 merkte het tijdschrift The Economist op dat grote Amerikaanse en Engelse banken ontbreken in Centraal- en Oost-Europa. Uit mijn analyse blijkt dat anno 2005 de Oostenrijkse, Italiaanse, Belgische en Duitse banken nog steeds de belangrijkste investeerders zijn in de toonaangevende banken in deze regio. Intraregionale investeerders in de bancaire sector komen uit Estland, Hongarije, Tsjechië en Polen maar hun aandeel staat in schril contrast met het aandeel van westerse banken: slechts 1.3% van de bank activa wordt beheerd door intraregionale buitenlandse investeerders terwijl 62% van de bancaire activa in handen is van westerse banken en instituties als de EBRD en het IFC.

Vervolgens vindt hoofdstuk 1 dat de omvang van de bancaire sector als deel van het bruto nationaal product in Centraal- en Oost-Europa, ondanks de licht stijgende

��


trend, nog steeds de helft bedraagt van wat gemiddeld gangbaar is in landen die de euro voeren. Ten aanzien van het niveau van kredietverlening aan de private sector wordt zelfs gemeten dat deze gedurende de periode tussen 1993 en 2000 onveranderd laag is gebleven. Gemeten als percentage van het bruto nationaal product ligt het niveau van krediet aan de private sector op een vierde van dat in het eurogebied. Ook gemeten als percentage van het totale kredietaanbod is het krediet aan de private sector in deze periode onveranderd gebleven in Centraal- en Oost-Europa. Pas sinds 2002 lijkt het krediet aan de private sector gemiddeld toe te nemen. Dit gemiddelde maskeert echter grote intraregionale verschillen. De kredietgroei doet zich vooral voor in de Baltische staten, Kroatië, Hongarije en Slovenië. In Tsjechië en Slowakije daalt het krediet aan de private sector al gestaag sinds 1997. Gebruik makend van een model dat meet in hoeverre de kredietgroei in Centraal- en Oost-Europese landen afwijkt van wat gangbaar is in economieën met vergelijkbaar BNP per capita, blijkt dat de landen waarin de kredietverlening sinds 2002 groeit, ook de landen zijn waarin de gemiddelde afstand tot de kredietbenchmark het kleinst was tussen 1995 en 2001. Uitzonderingen vormen Tsjechië en Slowakije waar de afstand tot de benchmark sinds 1999 onophoudelijk toeneemt.

De analyse legt bloot dat er geen duidelijke één op één positieve, dan wel negatieve relatie is tussen de toetreding van buitenlandse banken en groei van krediet aan de private sector. Op de vraag wat dan wel de rol is van buitenlandse banken als intermediairs in de bancaire sector vindt het hoofdstuk twee antwoorden. Ten eerste nemen buitenlandse banken vanaf 2000 de rol van kredietaanbieders over van binnenlandse banken. De reden hiervoor is eenvoudigweg dat het aantal buitenlandse banken vanaf dat moment het aantal binnenlandse banken overtreft, ook gemeten in relatieve activa onder beheer. Daarmee is nog niet de vraag beantwoord hoe buitenlandse banken zich gedragen als kredietverleners. De tweede waarneming van de analyse betreft het feit dat buitenlandse banken pas sinds 1997 krediet verlenen in een mate die overeenkomt met hun relatieve aanwezigheid. Voor 1997 verstrekten binnenlandse banken meer krediet aan de private sector dan men zou verwachten op basis van hun relatieve aandeel in de bancaire sector. Zelfs in 2000 verstrekten buitenlandse banken in Kroatië, Litouwen, Polen en Tsjechië minder krediet aan de private sector dan men zou verwachten op basis van de relatieve omvang van activa onder beheer.

Het hoofdstuk besluit met een prestatievergelijking tussen buitenlandse en binnenlandse banken. Het blijkt dat de winstgevendheid van buitenlandse banken, gemeten naar rendement op aandelen en rendement op totale activa die van binnenlandse banken overtreft heeft gedurende 1995-2000. Na deze periode is sprake van convergentie. De hogere winstgevendheid lijkt echter niet te worden veroorzaakt door een lager kostenniveau. Ook blijkt dat de hogere winstgevendheid van buitenlandse banken niet veroorzaakt wordt door hogere netto-rente inkomsten omdat deze gedurende deze hele periode lager zijn dan die van binnenlandse banken. Aangemerkt moet worden dat de analyse geen uitsluitsel geeft over de statistische significantie van de prestatieverschillen tussen binnenlandse en buitenlandse banken.

��

De schattingsresultaten in hoofdstukken 4-6 geven nader uitsluitsel over de relevantie van buitenlands eigendom met betrekking tot de prestatie van banken.

Het beeld dat in dit hoofdstuk naar voren komt is als volgt. Allereerst kan vastgesteld worden dat de buitenlandse banken in Centraal-Europa tot de landen in de regio niet in dezelfde mate en in dezelfde periode zijn toegetreden. Ten tweede komen de aanwezige buitenlandse banken voornamelijk uit de landen uit het eurogebied. Er lijkt daarmee een relatie tussen afstand tussen het gastland en het thuisland. Vooralsnog lijkt een eerste conclusie dat deze banken hun bestaande klanten zijn gevolgd naar Centraal- en Oost-Europa waarschijnlijk. Dit beeld wordt bevestigd door de initiële terughoudendheid van buitenlandse banken om krediet te verstrekken aan de private sector. Hoofdstuk 3 zal in het bijzonder ingaan op de hypothese dat buitenlandse banken in eerste instantie buitenlandse vestigingen openen, zij het kantoren, zij het dochterondernemingen, om hun bestaande klanten te kunnen bedienen.

Een andere oorzaak van de initiële terughoudendheid van buitenlandse banken kan gelegen zijn in de verschillen tussen de instituties in het gastland en die in het thuisland. Men denke dan aan langduriger juridische procedures in het gastland dan in het thuisland om beslag te leggen op onderpand van de wanbetaler. Deze juridische procedures kunnen bovendien worden bemoeilijkt door het bestaan van corruptie binnen het apparaat van beslagleggers en de rechterlijke macht. Het lijkt waarschijnlijk dat buitenlandse banken in Centraal- en Oost-Europa soms moeilijk beslag konden leggen op onderpand, als al aanwezig, en de markt voor dit onderpand niet altijd liquide was.

De conclusies op basis van dit hoofdstuk zijn in overeenstemming met die van Mian (2006) die stelt dat er belangrijke beperkingen zijn aan de kracht van kapitaalmobiliteit om financiële onderontwikkeling op te lossen. De analyse uit dit hoofdstuk werpt nadere vragen op. Ten eerste: in welk opzicht zijn buitenlandse banken in Centraal- en Oost-Europa anders dan buitenlandse banken in het eurogebied of buitenlandse banken in Europese transitie-economieën die geen zicht hebben op EU-toetreding? Wat is de reden dat buitenlandse banken in Centraal- en Oost-Europa zich pas sinds eind jaren negentig meer lijken te focussen op kredietverlening aan de private sector? Wat kan over het algemeen nu precies gezegd worden over de invloed van buitenlands eigendom van banken op de prestaties van deze banken? En ten slotte: in hoeverre heeft de institutionele omgeving invloed op de werkwijze van buitenlandse banken? Met behulp van een microanalyse zal hoofdstuk twee een antwoord proberen te vinden op de eerste vraag door de balansen van buitenlandse banken in de drie gebieden te onderzoeken. De tweede vraag zal in hoofdstuk drie beantwoord worden door buitenlandse banken zelf te vragen wat de strategie achter hun toetreding tot Centraal- en Oost-Europa is. De andere vragen komen empirisch aan bod in Deel II van dit boek.

In Hoofdstuk 2 breiden we de analyse van de rol van buitenlandse banken in Centraal- en Oost-Europa in verschillende opzichten uit. Het hoofdstuk zet ontwikkelingen van buitenlands bankieren in Centraal- en Oost-Europa af tegen enerzijds buitenlands

��


bankieren in ontwikkelde economieën en anderzijds buitenlands bankieren in landen met een lage economische ontwikkeling. In het bijzonder gaat dit hoofdstuk in op de volgende vragen: (i) hoe verhoudt zich de grootte van een buitenlandse bank tot de grootte van een bank waarvan de aandeelhouders zich bevinden in het binnenland? (ii) waaruit bestaan de activa van een buitenlandse bank? (iii) wat is de structuur van de kredietportefeuille, en (iv) hoe zijn buitenlandse banken gefinancierd? De eerste benchmark bestaat uit de vijftien EU lidstaten die tot mei 2004 de EU vormden; de tweede uit 16 transitielanden in Zuidoost Europa en het Gemenebest van Onafhankelijke Staten (GOS). De onderzoeksperiode start in 1995, toen de locale overheden in Centraal- en Oost-Europa nog het merendeel van de bancaire activa bezaten. Gedurende acht jaren worden bankbalansen in de drie regios gevolgd tot en met 2002, toen het krediet aan de private sector in de regio het begin van groei laat zien. De analyse is gebaseerd op 31.319 bank observaties van commerciële banken, spaarbanken en coöperatieve banken in de drie regio’s.

De bevindingen in dit hoofdstuk zijn als volgt. Allereerst is het verschil in omvang tussen een gemiddelde CEE bank en een bank in West-Europa tussen 1995 en 2002 toegenomen. In die periode groeide een Centraal- en Oost-Europese bank van gemiddeld €560mln tot ruim €1.3bn. In Tsjechië, Polen en Hongarije bevinden zich gemiddeld de grootste banken, in Bulgarije, Letland, Litouwen en Roemenië de kleinste. Het balanstotaal van een gemiddelde West-Europese bank bedroeg in 1995 ruim €4.6bn en groeide tot €15.6bn in 2002. Terwijl vanaf 1997 de omvang van Centraal- en Oost-Europese banken begon toe te nemen, groeiden banken in Zuidoost Europese en GOS landen tussen 1995 en 2002 nauwelijks. Uit de analyse komt naar voren dat in Centraal- en Oost-Europa buitenlandse banken in omvang zijn toegenomen en binnenlandse banken in omvang zijn afgenomen. Als verklaring hiervoor gelden de intraregionale fusies en overnames en de privatiseringen van voormalige staatsbanken en buitenlandse investeerders. De Centraal- en Oost-Europese banksector onderscheidt zich hiermee van die in West-Europa en de Zuidoost Europese en GOS landen, alwaar buitenlandse banken doorgaans kleiner zijn dan binnenlandse banken.

De bevindingen in het hoofdstuk bevestigen die van het eerste hoofdstuk ten aanzien van een ondermaatse intermediatie in Centraal- en Oost-Europa tussen 1995 en 2002. Het aandeel van leningen in de totale activa van een gemiddelde bank is in deze regio lager dan in West-Europa; pas in 2002 stijgt de leningen/activa ratio. In banken in de Zuidoost Europese en GOS landen doet zich dezelfde situatie voor. Daarnaast moet vermeld worden dat het aandeel niet-renderende activa afneemt in banken in de transitieregio’s evenals het aandeel van de vaste activa binnen de totale activa. Als verklaring geldt de verbeterde macro-economische vooruitzichten, een afname van informatieassymetrie ten aanzien van de kredietvrager en verbeterde werking van instituties. Ten aanzien van de het lagere aandeel van de vast activa kan worden aangemerkt dat bankkantoren worden gesloten, samengevoegd of geherstructureerd. Ook de looptijd van de leningen neemt toe in Centraal- en Oost-Europa evenals de diversificatie in looptijd van de gemiddelde kredietportefeuille.

��

De drie ontwikkelingen wijzen op een convergentie van de karakteristieken van de kredietportefeuilles West-Europese en CEE banken. In West-Europa heeft 40 procent van de leningen een looptijd van minder dan een jaar, in CEE landen is dit 50 procent en in Zuidoost Europese en GOS landen meer dan 60 procent.

Uit de balansgegevens van banken in de Baltische Staten blijkt ook dat het aandeel van leningen aan de publieke sector is afgenomen vanaf 1995. Ook deze bevinding staaft die uit het voorgaande hoofdstuk dat zich in de jaren negentig een crowding out effect heeft voorgedaan in de kredietmarkt in Centraal- en Oost-Europa, dat tot een halt komt vanaf 2002. Balansgegevens bevestigen daarnaast dat het gebruik van lease constructies door banken in Centraal- en Oost-Europa erg populair is. Afgezien van belastinggerelateerde motieven liggen redenen aangaande de (on)mogelijkheid tot opeisen van onderpand en de afwezigheid van historische gegevens van kredietvragers hieraan ten grondslag.

De passivazijde van de balans van banken in Centraal- en Oost-Europese landen, Zuidoost Europa en GOS landen vertonen meer overeenkomsten onderling dan met die van banken uit West-Europa. Zo bestaat in transitielanden 90% van de verplichtingen uit deposito’s en slechts enkele procenten uit geldmarkttitels, in West-Europa bedragen deze percentages 70 procent en bijna 9 procent.

Ten slotte onderzoeken we in dit hoofdstuk in hoeverre de balansen van buitenlandse en binnenlandse banken verschillen en of dit verschil karakteristiek is voor Centraal- en Oost-Europa. De eerste conclusie is dat tussen 1995 en 2002 er in Centraal- en Oost-Europa geen verschil was tussen de hoeveelheid krediet die buitenlandse en binnenlandse banken op de balans hadden staan. Dit bevestigt het beeld uit hoofdstuk 1 dat buitenlandse banken pas vanaf 1998 krediet verstrekten dat in overeenstemming was met hun aanwezigheid, terwijl zij voor dat jaar relatief weinig krediet verstrekten. Ook in West-Europa is er geen verschil tussen de hoeveelheid krediet verstrekt door buitenlandse en binnenlandse banken. In de groep van lager ontwikkelde transitielanden is dit verschil er wel: binnenlandse banken verstrekken meer krediet dan buitenlandse.

Ook komt uit de balansgegevens een verschillend risicoprofiel van buitenlandse en binnenlandse banken naar voren: in zowel West-Europa als Centraal- en Oost-Europa als Zuidoost Europa en de GOS landen hebben binnenlandse banken meer probleemleningen op de balans en ook meer verliesvoorzieningen. Opmerkelijk is dat buitenlandse banken in Centraal- en Oost-Europa meer niet-renderende activa op de balans noteren dan binnenlandse banken. Dit reflecteert de ondermaatse intermediatie in Centraal- en Oost-Europa gedurende deze periode.

Uit de balansgegevens komt verder naar voren dat buitenlandse banken zich bij voorkeur richten op het verstrekken van bedrijfskredieten en lease activiteiten. Binnenlandse banken in West-Europa richten zich voornamelijk op leningen aan de publieke sector en hypotheken. De kredietportefeuille van buitenlandse banken in transitielanden kenmerkt zich door een bredere waaier aan looptijden. Dit fenomeen kan ook worden opgevat als de introductie van nieuwe producten. Een ander resultaat van de analyse is dat buitenlandse banken, meer dan binnenlandse banken, leningen

��


met langere looptijden verstrekken. Hoofdstuk 3 gaat nader in op de redenen van buitenlandse banken om activiteiten te ondernemen in Centraal- en Oost-Europa. Wegens het niet voor handen zijn van data zijn, in een project met De Nederlandsche Bank, interviews afgenomen met talrijke bankiers en financiële toezichthouders. Tot de selectie van de bezochte hoofdkantoren behoren (i) de vier belangrijkste buitenlandse banken in deze regio, te weten de Oostenrijkse ERSTE Bank, Raiffeisen International, HVB-Group/BaCA en het Belgische KBC (ii) twee grote global banks, te weten de Nederlandse ABN AMRO en ING Bank, en (iii) de grootste Scandinavische banken in de Baltische Staten, te weten Swedbank en Skandinaviska Enskilda Banken. Daarnaast zijn bankiers van vijftien dochterbanken en medewerkers van zeven financiële toezichthouders in de regio bezocht.

Uit analyse van de data blijkt dat het leeuwendeel van de banken hoge winstverwachtingen had. Een deel van de banken geeft aan dat zij haar bestaande cliëntèle volgde naar Centraal- en Oost-Europa. Doorgaans waren dit de global banks, die als één van de eersten tot de regio toetraden. De grootste regionale buitenlandse banken geven aan dat de hoge concurrentie, en daarmee de grenzen aan groei, in het thuisland meespeelde in de overwegingen activiteiten te ontplooien in Centraal- en Oost-Europa. Voor de banken die relatief laat actief werden in Centraal- en Oost-Europese landen was de vooruitgang in het functioneren van instituties een belangrijk aspect. Ingegeven door het vooruitzicht van toetreding tot de EU, verbeterden de governance en moderniseerde de wet. Dat buitenlandse banken zich vooraleerst in Hongarije vestigden is deels te danken aan de liberale politieke situatie van het land. Sommige banken geven aan dat de verslechterende institutionele context voor hen reden is om een land weer te verlaten: voorbeelden hiervan vormen corruptie in Letland en de Oekraïne.

Uit de interviews komen andere, nog niet eerder opgetekende redenen naar voren. Het merendeel van de banken geeft aan dat reputatie en status redenen waren voor de banken om concurrentie te volgen naar Centraal- en Oost-Europa. Deze zogenoemde peer pressure vindt plaats zowel tussen de global banks onderling als de grote regionale spelers. Hieraan gerelateerd is de bevinding, dat het meegaan met de massa voor sommige banken belangrijk is; soms zelfs belangrijker dan de rationele afweging op basis van risico en rendement. Een derde overweging die uit de gesprekken naar voren kwam kan ik vooralsnog alleen als hypothese aantekenen en kenschetsen als de stepping stone strategie. Deze strategie geeft het toekomstig buitenlands bankieren in land B als reden voor het nu betreden van land A. De geografische en culturele afstand tussen Land A en land B is doorgaans klein; de institutionele afstand doorgaans groot. De opgedane kennis en ervaring in land A heeft tot doel een comparatief voordeel te ontwikkelen ten opzichte van andere buitenlandse banken, dat kan worden aangewend zodra de institutionele context van land B dat toelaat.

Ten aan zien van de keuze tussen het opzetten van een greenfield en een acquisitie vindt het hoofdstuk het volgende. Global banks zetten bij voorkeur nieuwe kantoren, zogenoemde greenfields op, vaak in de vorm van branches of joint-ventures in

��0 ��

samenwerking met andere buitenlandse banken. Banken die zich niet exclusief richten op zakelijk bankieren verwerven bij voorkeur bestaande banken. Uit de ervaringen van banken in Centraal- en Oost-Europa blijkt ook dat de keuze van banken voor het opzetten van greenfields onaantrekkelijker wordt naarmate de concurrentie toeneemt. Een derde variabele voor banken die een buitenlandse vestiging overwegen zijn de kosten van het alternatief, een acquisitie, en de daarmee gepaard gaande integratie van de gekochte bank in de moederbank. Bij deze kosten moet gedacht worden aan de standaardisatie van risicomanagement, rapportage en informatietechnologie (IT). Een andere kostenpost, die specifiek bij de privatisering van staatsbanken in Centraal- en Oost-Europa een rol speelde was de informatie asymmetrie ten aanzien van de omvang van de slechte leningen van voormalige staatsbanken

Concluderend vindt het hoofdstuk bewijs voor een nieuwe pecking order aangaande de redenen voor buitenlands bankieren. De conclusie is dat de avant-garde van buitenlandse banken die vanaf het begin jaren negentig vestigingen in Centraal- en Oost-Europa oprichtten, zich voornamelijk concentreerden op bestaande cliëntèle en locale grote bedrijven. Het midden en kleinbedrijf en de microbedrijven werden in veel mindere mate door banken bediend. Door de toename van het aanbod van zakelijk bankieren door buitenlandse banken, de daarmee de dalende marges binnen dit segment en de privatiseringen van voormalige staats(spaar)banken, vond eind jaren negentig diversificatie van customer focus plaats en werd de aandacht uitgebreid naar retailbankieren.

Deel II van het proefschrift richt zich op de vraag: “Wat is de invloed van buitenlands aandeelhouderschap op de financiële prestatie van een bank?” In het bijzonder onderzoeken de drie hoofdstukken of de winst en de efficiëntie van een bank significant worden beïnvloed door binnenlands versus buitenlands eigendom. Winstgevendheid van een bank is relevant voor de continuïteit en de stabiliteit van het financiële systeem. Efficiëntie is relevant voor het eigendom van een bank. Een minder efficiënte bank vormt een overname-object. Het probleem wordt theoretisch onderlegt door de vraag of de buitenlandse eigenaar genoeg kennis heeft van de locale markt, haar cultuur en haar institutionele context om de locale dochter bank in dezelfde mate winstgevend en efficiënt te besturen als binnenlandse aandeelhouders/eigenaren dit zouden doen? De theorie neemt aan dat de overige managementkwaliteiten van buitenlandse en binnenlandse aandeelhouders dezelfde zijn en voorspelt dat binnenlandse banken beter presteren dan buitenlandse banken. In het geval van de prestaties van buitenlandse banken in transitielanden of ontwikkelingslanden wordt de theoretische voorspelling geproblematiseerd. Indien een buitenlandse bank zich vestigt in minder ontwikkelde landen is het waarschijnlijk dat zij naar hun dochterbanken kennis exporteren, die in omvang en kwaliteit niet in dezelfde mate aanwezig is in concurrerende locale banken. Het nadeel dat buitenlandse banken dus, volgens de theorie, doorgaans plegen te ondervinden zou heel goed kunnen worden (over)gecompenseerd door de best-practice praktijken die zij importeren in de nieuwe buitenlandse vestiging. Het maakt dan geen verschil of de buitenlandse bank een nieuwe bank, een greenfield of een de novo

��0 ��


bank opzet, of een locale bank overneemt. Vooraf is een voorspelling aangaande de relatieve winstgevendheid en kostenefficiëntie van een bank met buitenlands eigendom dan onmogelijk.

Hoofdstuk 4 onderzoekt of banken met buitenlands eigendom in Europese transitie-landen beter presteren dan binnenlandse banken. De analyse, gebaseerd op lineaire regressies, met behulp van de Kleinste Kwadraten Methode, richt zich op bankprestaties van 244 commerciële banken, spaarbanken, coöperatieve banken en investment banks in 20 Europese transitielanden in 2001. In concreto gaat de analyse na of het niveau van buitenlands eigendom van invloed is op (i) de netto rente inkomsten, (ii) de winst voor belastingen, en (iii) overhead kosten. Vervolgens testen we of de resultaten robuust zijn wanneer we de variabele van buitenlands eigendom een dummy karakter geven. Ten slotte testen we de relatie tussen buitenlands eigendom en bank prestaties op (a) de invloed van de economische ontwikkeling van het gastland, (b) de concentratie in de bancaire sector in het gastland, (c) de aanwezigheid van buitenlandse banken gemeten naar relatieve activa in het gastland, en (d) de aanwezigheid van buitenlandse banken gemeten naar aantal in het gastland.

De resultaten laten zien dat in transitielanden buitenlands eigendom een negatieve invloed heeft op de netto rente inkomsten van een bank. Dit kan wijzen op een agressieve strategie ter vergroting van het marktaandeel of op een structureel nadeel van de bank met buitenlandse eigendom ten opzichte van haar concurrenten die in binnenlandse handen zijn. Een tweede conclusie is dat in transitielanden de overhead kosten lager zijn naarmate een buitenlandse bank meer in buitenlandse handen is. Dit tweede resultaat vormt bewijs voor de hypothese dat, althans in transitielanden, buitenlandse banken een efficiëntere bedrijfsvoering kennen dan hun concurrenten die de best-practice kennis uit meer ontwikkelde banksectoren missen, of nog niet hebben kunnen toepassen. De tegengestelde effecten van buitenlandse eigendom op netto-rente inkomsten en overhead kosten kunnen niet verhinderen dat buitenlandse banken uiteindelijk minder winstgevend zijn. De resultaten zijn robuust: voor de deelverzameling van 198 commerciële banken worden dezelfde resultaten gevonden.

Daarnaast worden de volgende effecten gevonden van bovengenoemde factoren die de relatie tussen buitenlands eigendom en bankprestatie zouden kunnen beïnvloeden. Ten eerste is het niveau van economische ontwikkeling in het gastland niet van invloed op de gevonden relaties. Wel is het zo dat in rijkere transitielanden banken (i) minder netto-rente inkomsten hebben en (ii) minder overhead kosten maken. Als verklaring hiervoor zou de grotere concurrentie in rijkere landen kunnen worden opgevoerd. Ten tweede vinden we dat de mate van concentratie in de bancaire sector in transitielanden niet van invloed op de gevonden relaties. Het blijkt dat bank concentratie überhaupt geen sterke invloed heeft op de winstgevendheid van banken. Ten derde leidt een sterkere aanwezigheid van buitenlandse banken, gemeten naar activa, tot meer winst voor buitenlandse banken. In landen waar buitenlandse banken een kleinere rol vervullen in het bancaire systeem, behalen buitenlandse banken een

��

lagere winst. De oorzaak hiervan lijkt eerder te liggen in hogere netto rente inkomsten dan in lagere overhead kosten. Ook in landen met relatief veel buitenlandse banken maken buitenlandse banken meer winst dan in landen met weinig buitenlandse banken. De oorzaak hiervan is niet te herleiden op hogere rente-inkomsten of lagere overhead kosten.

De gevonden resultaten zijn het eerste bewijs voor een zogenoemd thuisvoordeel voor binnenlandse banken in Europese transitielanden. Om tegemoet te komen aan het mogelijke endogeniteitsprobleem kan de gehanteerde econometrische techniek verbeterd worden door toepassing van de gegeneraliseerde momentenmethode. Hierin voorziet hoofdstuk 5.

Hoofdstuk 5 onderzoekt het verband tussen buitenlands eigendom en de financiële prestaties van 511 banken wereldwijd. Het panel bestaat uit eigendom- en resultatengegevens voor de periode 1998-2001. Wanneer we de bancaire sector van de 73 landen in de sample bestuderen op de aanwezigheid van buitenlandse banken, dan zien we dat het gemiddelde aantal buitenlandse banken wereldwijd een derde bedraagt van het totaal aantal banken. Er zijn echter grote verschillen tussen landen, wanneer we deze landen groeperen naar economische ontwikkeling. Van meest ontwikkelde naar minst ontwikkelde categorie definieert de Wereldbank deze als landen met (i) hoog inkomen, (ii) hoger-midden inkomen, (iii) lager-midden inkomen en (iv) laag inkomen. Buitenlandse banken zijn relatief het meest talrijk in landen met een hoger-midden inkomen; een bruto nationaal inkomen van tussen de $3.000,- en $9.200 per hoofd van de bevolking. Tot deze landen behoren vooral de Centraal- en Oost-Europese landen en Zuid- en Midden-Amerikaanse landen als Argentinië, Brazilië, Chili, Mexico en Venezuela. In deze landen zijn gemiddeld twee van de vijf banken buitenlands en bezitten buitenlandse banken een derde van ’s lands activa (hoofdstuk 1 gaf aan dat deze ratio’s voor Centraal- en Oost-Europa zelfs nog hoger liggen). Vergeleken met de bancaire sector van landen met een hoger-midden inkomen, spelen buitenlandse banken een minder belangrijke rol in het banksysteem van landen met een hoog inkomen; die met een bruto nationaal inkomen per hoofd van de bevolking van boven de $ 9.200. Hoewel één op de drie banken buitenlands is, bezitten zij gemiddeld slechts rond de 10% van de bancaire activa. De rol van buitenlandse banken in landen met een lager-midden inkomen, zoals China, Colombia, Egypte en Iran, en landen met een laag inkomen zoals Indonesië, Kenia, Nigeria en Pakistan is verwaarloosbaar. In de eerste groep landen bezitten buitenlandse banken rond de 4% van de activa en in de laatste groep rond de 1%.

Ook in dit hoofdstuk laten de schattingsresultaten zien dat buitenlands aandeelhouderschap een negatieve invloed heeft op zowel het niveau van de netto rente inkomsten als op de winst voor belasting. Gezien de vergelijkbare resultaten van de Kleinste Kwadraten Methoden als de Gegeneraliseerde Methode van Momenten, kunnen de resultaten als robuust gekwalificeerd worden. Opnieuw ook is een conclusie dat het niveau van economische ontwikkeling van een land niet van invloed is op deze relatie. Kortom, ongeacht of een buitenlandse bank zich vestigt in een rijk land

��


of een arm land, de netto rente inkomsten en de winst zijn lager dan die van de concurrent in locale handen. Een andere conclusie is dat voor zowel buitenlandse als binnenlandse banken geldt dat naarmate een land rijker is, minder netto rente inkomsten worden gegenereerd. Ook deze resultaten komen overeen met die in hoofdstuk 4. Gezien het feit dat hoofdstuk 2 toont dat banken in rijke landen meer leningen op de balans hebben staan, kan de reden voor het negatieve verband tussen economische ontwikkeling en netto-rente inkomen niet zijn dat banken in rijkere landen zich minder met kredietverstrekking zouden bezighouden. Veeleer lijkt het erop dat in landen met een hoger BNP per hoofd van de bevolking de marge(s) tussen de uit- en inleenrente(s) veel kleiner zijn dan die in landen met een lager inkomen per hoofd.

Hoofdstuk 6 onderzoekt de invloed van buitenlands eigendom op de kostenefficiëntie van een bank. Kostenefficiëntie meet de technische én allocatieve efficiëntie van het gebruik van productiemiddelen. Onder een technisch efficiënt gebruik van de productiemiddelen wordt verstaan dat gegeven de omvang van de productie, zo weinig mogelijk kosten worden gemaakt aan de inzet van de productiemiddelen. Onder een allocatief efficiënt gebruik van productiemiddelen wordt verstaan, dat gegeven de output, geen betere mix van inputs kan worden samengesteld, die tot lagere kosten zouden kunnen leiden. De in dit hoofdstuk gebruikte methode meet in hoeverre de kosten van een bank afwijken van de kosten die een best-practice bank zou maken voor dezelfde productie onder dezelfde omstandigheden. Het is juist déze afstand, die wij hier onder het begrip kosteninefficiëntie verstaan. Ter bepaling van deze inefficiëntie veronderstelt het gebruikte model twee bankproducten, zogenoemde outputs: het verstrekken van kredieten en het kopen van schuldtitels. Deze schuldtitels bestaan uit schuldpapier van de publieke sector, tegoeden bij (centrale) banken, obligaties en aandelen, en handelspapier. Het model veronderstelt tevens twee productiemiddelen: de kosten van arbeid en de prijs van ‘funding’, die bestaat uit het totaal van rentekosten verschuldigd over de aangetrokken deposito’s en rentekosten verschuldigd over andere aangetrokken gelden.

De gebruikte data bestaat uit 3.850 observaties van 1.363 banken in 93 landen. Bij de bepaling van de productie van best-practice banken wordt rekening gehouden met landspecifieke kenmerken als het niveau van economische ontwikkeling, de concentratie binnen het bancaire systeem en de ontwikkeling van de financiële sector, en bankspecifieke kenmerken als de omvang van de reserveringen voor slechte leningen en de variatie in bancaire activiteiten. De gebruikte methode ter bepaling van de relevantie van buitenlands eigendom voor bank(in)efficiëntie is de Stochastische Frontier Analyse. Tijdens deze analyse wordt vastgesteld: (i) de omvang van de kosteninefficiëntie van de bank en (ii) de omvang en significantie van de mate van buitenlands eigendom ter verklaring van de onder (i) vastgestelde kosteninefficiëntie. De schattingsresultaten ten aanzien van de inefficiëntieverklarende variabelen geven aan dat buitenlandse banken minder kostenefficiënt zijn dan banken met binnenlands eigendom.

��

Als verklaring van de lagere efficiëntie van buitenlandse banken onderzoekt het hoofdstuk vervolgens het effect van het institutionele raamwerk. Het institutionele raamwerk wordt in het hoofdstuk door zes indicatoren benaderd. De door Kaufmann (2005) onderscheiden indicatoren liggen op het terrein van: (i) burgervrijheden en onafhankelijke media, (ii) politieke stabiliteit en geweld, (iii) effectiviteit van de overheid, (iv) kwaliteit van de regelgeving, (v) heersend recht, en (vi) inperking van corruptie. De theoretische grondslag voor het (apart) opnemen van deze variabelen is de veronderstelling dat buitenlandse banken in landen met een ‘slechter bestuur’ minder kostenefficiënt opereren dan hun concurrenten met binnenlands eigendom die meer bekend zijn met de institutionele context en de wijze hoe daar zo kostenefficiënt mogelijk mee om te gaan. De schattingsresultaten geven aan dat verbeteringen op alles zes de gebieden van goed bestuur de buitenlandse bankinefficiëntie verlaagt; vooral op gebieden van regelgeving en corruptie. Aanvullende schattingen leveren bewijs voor de hypothese dat ook het institutionele raamwerk in het thuisland van invloed is op de efficiëntie van de buitenlandse bank. Ook hier geldt dat door verbeteringen op de onderscheiden zes gebieden van goed bestuur de efficiëntie van de buitenlandse bank toeneemt. Ten slotte concludeert het hoofdstuk dat hoe meer de institutionele indicatoren van het thuisland en die van het gastland op elkaar lijken, hoe kostenefficiënter de buitenlandse bank kan opereren.

ConclusieSinds de jaren negentig zijn Centraal- en Oost-Europese banken in toenemende mate in handen gekomen van buitenlandse investeerders, doorgaans banken. Het bewijs over de kredietverlening door buitenlandse banken in Centraal- en Oost-Europa laat zien dat er verscheidene jaren overheen gingen alvorens de relatieve omvang van de kredietverlening aan de private sector door buitenlandse banken overeenkwam met hun relatieve aanwezigheid. Verscheidene factoren lijken aan deze terughoudendheid ten grondslag te liggen. Ten eerste toont het bewijs over de strategieën van buitenlandse banken die zich als eersten vestigden in Centraal- en Oost-Europa, dat zij zich primair richtten op bestaande klanten uit West-Europa en op de grotere ondernemingen in het gastland. Ondanks dat deze strategie de kritiek leek te bevestigen dat buitenlandse banken de lagere segmenten in de kredietmarkt niet of nauwelijks bedienen, werd deze strategie mede ingegeven door enkele specifieke factoren die kenmerkend zijn voor het bankieren in transitielanden en ontwikkelingslanden. Aannemende dat banken veilig bankieren verkiezen boven risicovol bankieren, betekent dit dat zij i) verkiezen om een goede kredietnemer te kunnen onderscheiden van een slechte kredietnemer, het averse selectieprobleem, en ii) het noodzakelijk achten beslag te kunnen leggen op onderpand in het geval van wanbetaling of faillissement van de kredietnemer. Buitenlandse banken in Centraal en Oost-Europa lijken onvoldoende bij machte geweest om het averse selectie probleem op te lossen, om onderpand te prijzen en het liquide te maken. Wij constateren dat in Centraal- en Oost-Europa het niveau van het krediet aan de private sector in termen van BNP onveranderd vlak is gebleven tot

��


het begin van de 21e eeuw, en dat de diversificatie binnen de kredietportefeuilles van banken met krediet richting MKB en de retail-sector in de late jaren negentig primair lijkt te zijn ingegeven door de toegenomen concurrentie in het topsegment van het zakelijkbankieren, eerder dan door een vrijwillige strategie de lagere segmenten van de kredietmarkt te gaan bedienen.

Wij concluderen dat wanneer overheden en financiële autoriteiten in transitielanden of ontwikkelingslanden besluiten buitenlandse banken toe te laten tot hun nationale bancaire markt met een fysieke vestiging, het beleid van eerstgenoemde gericht zou moeten zijn op het mitigeren van de obstakels die verband houden met het (buitenlands) bankieren in dergelijke landen. Ten eerste betogen wij dat, aangezien het adverse selectieprobleem in deze landen voornamelijk te wijten is aan het ontbreken van historische gegevens van kredietnemers, overheden in deze landen zouden kunnen overwegen het initiatief te nemen tot het opzetten van een organisatie die financiële gegevens verzamelt van bedrijven en hier desgevraagd toegang tot verleent aan kredietgevers. Daarnaast zou het oprichten van een rating bureau ertoe bij kunnen dragen dat een ‘goede kredietnemer’ aan een kredietgever het signaal kan afgeven een ‘goede kredietnemer’ zijn, waardoor haar kansen op een banklening vergroten.

Naast het averse selectieprobleem vormt een tweede uitdaging van het buitenlands bankieren in transitie-economieën de vaak tijdrovende en onvoorspelbare gang naar de rechtbank in geval van wanbetaling of faillissement. In dit opzicht zou overheidsbeleid erop gericht kunnen zijn tot het aannemen van wetgeving die zowel debiteuren en crediteuren beschermt volgens de principes van redelijkheid en billijkheid. In veel Centraal en Oost-Europese landen vormde de prikkel hiertoe het vooruitzicht hierdoor te kunnen toetreden tot de Europese Unie. Nochtans betekent dit niet dat geen aandacht zou moeten worden gericht op i) terugdringing van de ontvankelijkheid voor omkoping binnen het justitiële apparaat in het algemeen, en binnen de rechterlijke macht in het bijzonder, bijvoorbeeld door salarisverhoging en op ii) de handhaving van de wet, door bijvoorbeeld het controleren van de zuiverheid van de relaties tussen curatoren en rechters en bijscholing van laatstgenoemden naar aanleiding van nieuwe wetgeving.

Het resultaat uit onze centrale bank enquête dat in Centraal- en Oost-Europa buitenlandse banken ten opzicht van binnenlandse banken winstgevender zijn geweest in de periode tussen 1995 en 2000 bevestigt resultaten van Bonin, Hasan and Wachtel (2005). Majnoni, Shakar and Várhegi (2003), Claessens et al. (2001), Kraft and Tirtiroglu (1998) en Sabi (1996). Interessant nader onderzoek zou zich kunnen richten op het empirisch vaststellen welke de oorzaken zijn geweest van deze superieure buitenlandse bank winstgevendheid, aangezien onze gegevens suggereren dat de netto-rente inkomsten van binnenlandse banken die van buitenlandse banken overtroffen en dat daarnaast de kosten van overhead van buitenlandse en binnenlandse banken in deze periode weinig verschilden. Men zou kunnen overwegen hierbij te betrekken de kwaliteit van de kredietportefeuilles in navolging van een studie van Berger and DeYoung (1997), die de relatie onderzoekt tussen probleemleningen en

��

kostenefficiëntie binnen commerciële banken. Ondanks dat buitenlandse banken winstgevender lijken te zijn geweest in

de vroege transitieperiode, laat onze empirische analyse zien dat in 2001 deze superieure winstgevendheid van buitenlandse banken in Centraal- en Oost-Europa is verdwenen. Belangrijker nog is dat wij in een vervolgstudie met gegevens uit de periode 1998-2001 van banken wereldwijd vinden, dat buitenlands eigendom in het algemeen negatief gerelateerd is aan de winstgevendheid van een bank, ongeacht de economische ontwikkeling van het land waarin deze buitenlandse bank actief is. Dit resultaat staat in contrast met de bevindingen van een studie naar de determinanten van bankwinsten in de periode 1988-1995 van Demirgüç-Kunt and Huizinga (2000), die vindt dat in ontwikkelingslanden buitenlandse banken winstgevender zijn dan binnenlandse banken en dat in ontwikkelde landen binnenlandse banken winstgevender zijn. Aanvullend onderzoek zou uitsluitsel kunnen geven of, en in hoeverre, de onderzoeksperiode van invloed is op het effect van de economische ontwikkeling van een land op het verband tussen eigendom en winst van een bank. Om de literatuur aangaande de nexus tussen eigendom en winstgevendheid te completeren zou aanvullend onderzoek kunnen nagaan in hoeverre buitenlands eigendom en geconcentreerd eigendom, en binnenlands eigendom en verspreid eigendom gecorreleerd zijn.

Nader bewijs ten aanzien van de relevantie van de kwaliteit van de institutionele omgeving wordt geleverd door de analyse van de kostenefficiëntie van buitenlandse banken. We vinden dat een betere kwaliteit van de institutionele omgeving de efficiëntie van buitenlandse banken verhoogt. Het lijkt erop dat overheidsbeleid gericht op verbetering van de institutionele omgeving vooral zou kunnen bijdragen aan de gezondheid van de bancaire sector, wanneer deze sector wordt gedomineerd door buitenlandse banken, zoals het geval is in Centraal- en Oost-Europa maar ook in Sub-Sahara Afrika en Latijns-Amerika en de Caraïben. Men denke hierbij aan beleid, behalve gericht op het terugbrengen van de ontvankelijkheid voor omkoping, dat als doel heeft een effectieve en voorspelbare rechtspraak. Ook kan men denken aan beleid waarmee men bijdraagt aan een betere kwaliteit van financiële toezichthouders. Ten slotte lijkt het erop dat ook beleid gericht op het terugdringen van overheidsbureaucratie en het bevorderen van onafhankelijke media positieve effecten heeft op de kostenefficiëntie van buitenlandse banken.

Ten aanzien van de invloed van buitenlands eigendom op de efficiëntie van een bank vinden wij, voor een 1998-2001 sample, dat beide negatief gerelateerd zijn en dat zowel in ontwikkelde landen als in ontwikkelingslanden. Dit resultaat contrasteert met de bevindingen van studies die een superieure efficiëntie van buitenlandse banken ten opzichte van binnenlandse banken vinden in Centraal- en Oost-Europa zoals de studies van Grigorian and Manole (2002), Jemric and Vujcic (2002), Nikiel and Opiela (2002), Yildirim and Philippatos (2003), Hasan and Marton (2003), Matousek and Taci (2004), Weill (2003), Bonin, Hasan and Wachtel (2005), Fries and Taci (2005) en Havrylchyk (2006). Wij concluderen dat behalve de relatieve superieure

��

university of groningen foreign bank entry and …...remco van der molen, thanks for being a great...

Documents