Banking Reform in Central Europe and the Former Soviet Union 9789633865576

Citation preview

Banking Reforrn in Central Europe and the Former Soviet Union Edited by

Jacek Rostowski

►

CE U

CENTRAL EUROPEAN UNIVERSITY PRESS

First published in 1995 by Central European University Press 1051 Budapest Nador utca 9 Distributed by Oxford University Press, Walton Street, Oxford OX2 6DP Oxford New York Athens Auckland Bangkok Bombay Toronto Calcutta Cape Town Dar es Salaam Delhi Florence Hong Kong Istanbul Karachi Kuala Lumpur Madras Madrid Melbourne Mexico City Nairobi Paris Singapore Taipei Tokyo Toronto and associated companies in Berlin Ibadan Distributed in the United States by Oxford University Press Inc., New York © Jacek Rostowski 1995 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, without the prior permission of the copyright holder. Please direct all enquiries to the publishers. British Library Cataloguing in Publication Data A CIP catalogue record for this book is available from the British Library ISBN l-85866-038-6 Hardback ISBN 1-85866-039-4 Paperback Library of Congress Cataloging in Publication Data A CIP catalog record for this book is available from the Library of Congress Printed and bound in Great Britain by Biddies of Guildford Ltd. UK

CONTENTS

List of Contributors

V

Ack.nowledgements

vi

I

2 3

4

Introduction

Jacek Rostowski

The Banking System, Credit and the Real Sector in Transition Economies Jacek Rostowski

16

Financial Sector Design, Regulation and Deposit Insurance in Eastern Europe Arnaud W. A. Boot and Sweder van Wijnbergen

42

Lessons from Bad Loan Management in the East Central European Economic Transition for the Second Wave Reform Countries Kalman Mizsei Comment on Mizsei's 'Lessons' Rumen Dolninsky

5 6

7

I

Banking Privatization Policy in Poland and Czechoslovakia Kate Mortimer Payment System Reform in Poland

Andrzej Topinski and Artur Struzynksi

Reforming the Banking System in Estonia

Arda Hansson

58 84 89 113

142

Contents

iv

8

The First Stage of Banking Reform in Russia is Completed: What Lies Ahead? El,ena Zhuravskaya

166

Banking Reform in Ukraine A.R Latter

183

10

Banking Reform in the Republic of Georgia Lado Gurgenidze

197

11

Banking in Romania Daniel Daianu

212

9

Index

227

LIST OF CONTRIBUTORS

Arnoud W.A. Boot University of Amsterdam Daniel Daianu Chief Economist, National Bank of Romania Lado Gurgenidze Institute of EastWest Studies, New York Ardo Hansson Former Economics Adviser to the Prime Minister of Estonia; Board Member, Bank of Estonia A.R. Latter Head of International Division (Developing World), Bank of England; Former IMF Resident Adviser to the National Bank of Ukraine Uune 1992 to March 1993) Kalman Mizsei Deputy Director, Institute for East-West Studies; Member, Supervisory Board of Budapest Bank Kate Mortimer Finance Sector Adviser, UK Government Know How Fund Jacek Rostowski Board Member, Polish Development Bank; Former Economic Adviser, Ministry of Finance, Warsaw; School of Slavonic and East European Studies, University of London Artur Struzynski Head and Founder, Automated Inter-Bank Settlement System (KIR), Poland Andrzej Topinski Chairman, PKO BP; Former Acting Governor of the National Bank of Poland Sweder van Wijnbergen Lead Economist, East European Department, World Bank (to February 1993); Lead Economist, Mexico Department, World Bank (to February 1992); Professor of Economics, University of Amsterdam; Centre for Economic Policy Research E/,ena Zhuravskaya Director, International Centre for Research into Economic Transformation, Moscow

ACKNOWLEDGEMENTS

The Centre for Economic Performance at the London School of Economics financed the initial workshop on 'Lessons for the Former Soviet Union of Banking Reform in Central Europe' which took place on 8 May 1993 at the LSE. CEP also funded the follow-up conference 'Banking Reform in the Former Soviet Union and Eastern Europe: Lessons from Central Europe', held in Budapest on 14-15 January 1994. The Editor is very grateful to Richard Layard, the Director of CEP (an ESRC funded research centre), for his support and encouragement for this project, and to Nigel Rogers of CEP for extremely efficient administrative assistance. The Editor is also grateful to Janice Bell for research and editorial assistance during the production stage of this book. Acknowledgements are also extended to Heather Bliss and the staff at CEU Press.

CHAPTER 1

INTRODU CTION Jacek &stowski

Across the great land mass from Brandenburg to the Bering Straits the business of banking is reviving, as market economies are established.1 A process that took hundreds of years in Western Europe and decades in Japan is being telescoped into a few years in Central and Eastern Europe and the Former Soviet Union (FSU). The policy issues are of exceptional importance, as decisions made now will shape the financial systems of the region for decades to come. The developments also give researchers a unique opportunity to engage in comparative analysis, as the various countries have approached the building of their banking systems in very different ways. At the beginning of the transformation process, the 'first wave' of Central European reformers looked to the West for a model of how to transform their state-controlled monobanking systems so as to make them suitable for a market economy. The reform of the Central European banking systems, though far from complete, is now sufficiently far advanced for lessons to be drawn from the - quite different - experience of the three countries concerned. These conclusions are important for the 'second wave' reformers in the FSU and the Balkans. For this reason, the book starts with five thematic chapters which analyse the Central European experience. These are followed by five chapters reviewing the experience of banking reform so far in some of the 'second wave' countries (Estonia, Georgia, Romania, Russia and Ukraine), and considering what the implications of the Central European experience are for each country. 1 Many other professions that atrophied under communism are also reviving: accountancy, commercial law, advertising, etc.

2

R.ostowski

I. Deficit balance sheets: causes and systemic solutions Bad loans plague banks in all the countries in the region that have undergone macroeconomic stabilization. As a result, given implicit deposit insurance, bad loans plague the governments and taxpayers of these countries. Rostowski (Chapter 2) traces this to the fact that under communism loans were not allocated on a commercial basis. Thus, it is hardly surprising that in the new commercial environment many of them should turn out to be unrecoverable. Furthermore, bank staff, having had no experience in allocating loans commercially, will continue to misallocate credit for a significant period after hard budget constraints have been imposed on state­ owned enterprises (SOEs). Such misallocation of credit also occurred in Poland in 1989 and in Russia in 1990-4, where some loans were notionally allocated on a market basis, but real interest rates were so massively negative, due to very high inflation, that only the most incompetent borrower could fail to repay his debt at maturity. Banking skills and bank supervisory skills of necessity take a long time to acquire, yet if the threat to the stability of the financial system from bad loans is to be avoided, then these skills need to be learned instantaneously upon the transformation of the old communist monobank into a two-tier banking system or upon stabilization.2 This is the paradox of the 'jump from bad credit to good credit', which the process of banking reform that has recently been followed in Central Europe seems to require. A number of 'systemic' solutions designed to prevent the bad debt problem from appearing in the early transition period have been proposed. They all involve major departures from the usual process. One is the complete wiping out of bank debt at the beginning of the transition (Begg and Portes 1991). However, this is a partial solution as it deals only with the stock of bad debt inherited from the previous system, and does not help to avoid new bad debt which builds up rapidly as banks start allocating credit on a commercial basis for the first time in a highly unstable macroeconomic and microeconomic environment.3 2 The formal creation of a two-tier system may precede stabilization by a number of years. In that case, it is the latter event that will bring the bad debt problem to the surface. 3 In the West at the inception of banking in the fifteenth century, skills were also low - but the quantity of bank money, notes and deposits, was infinitesimal compared with coin. Postcommunist countries have a low level of banking skills, but a high share of deposits in total money.

Introduction

3

Another systemic solution is that of narrow banking (Boot and Wijnbergen, Chapter 3). This system has the advantage that it runs with the grain of already existing institutions, especially in the FSU. Under this architecture of the banking system, deposit insurance is limited to household deposits at the state savings bank, all other deposits being uninsured. The argument is that the private firms and SOEs that have deposits in the commercial banks are sufficiently large for it to be worth their while monitoring the performance of the banks they deposit with (and sufficiently sophisticated to be able to do so as well as anyone in that economy could). As a result, the usual argument for deposit insurance - that depositors face high individual costs and a free-rider problem in monitoring their banks disappears. Banks monitored by sophisticated depositors will have an interest in guarding their reputations, which requires conservative banking practice. 4 The problem remains of what to do with the state savings bank, where the bulk of household deposits are kept. Boot and Wijnbergen suggest that in order to enforce competition for household deposits, the state savings bank should be split up. The resistance of the Savings Bank in Romania to an increase in the massively negative real interest rates on deposits, which have obtained since the beginning of the transition in that country, reinforce that view. This is described by Daianu ( Chapter 11). This particularly interesting aspect of the chapter by Boot and Wijnbergen highlights the crucial importance of the inter-bank market in the architecture of new banking systems (see Section 5). We shall return often during this overview to the key role of this market within the banking system. A more far-reaching proposal is put forward by Rostowski (Chapter 2), who suggests that in the initial years of the transition, banks covered by deposit insurance should have 100 per cent reserve requirements imposed upon them. These banks would then have to concentrate primarily on providing payments services. They could cover the costs of these services either by investing deposits in government paper, 5 or by charging for them. As skills increased, banks in the payments system could be allowed to gradually shift away 4 S. Vassilev, deputy Economy Minister of Russia and a leading reform economist, has put the essence of this idea very succinctly: the bad loan problem in Russia is naturally resolved by the fact that the commercial banks which have the bad loans also have 'bad liabilities' (i.e., enterprise - rather than household deposits). 5 An alternative would be for the central bank to pay a (low) interest rate on reserves held with it.

4

Rostowski

from 100 per cent reserve requirements. It is interesting that in Poland at present over half of bank assets are in government paper, 6 which, it could be argued, is about the desirable level some five years into the transition. However, in Poland this occurred after a very sharp expansion in real terms of credit to the non-government sector (CNG) in 1990, which was partly reversed as a result of the large budget deficits in 1991 and 1992. Arguably, Poland had quite the wrong pattern of CNG expansion, with an initial boom followed by a sharp decline, rather than a slow but steady expansion, which would have allowed banks to develop credit allocation skills. If households and only households choose to deposit at the banks participating in the payments system, the architecture proposed by Rostowski differs little from that of Boot and Wijnbergen. This is particularly so as Boot and Wijnbergen expect a large part of the savings banks' funds to be invested in government paper. 7 However, given that one of the key problems facing firms in the very early transition is the long period the clearance of non-cash payments takes, it seems likely that - at least initially - firms, even more than individuals, would wish to be able to take advantage of a payments system which was both rapid and safe. Furthermore, there is a problem with Boot and Wijnbergen's scheme in the extent to which savings banks invest in commercial bank - rather than government liabilities. Deposit insurance for savings banks means that their asset management has to be supervised. If they lend significantly to commercial banks, then the quality of their loans to the individual commercial banks has to be assessed, and this in turn requires supervision of these commercial banks and their loans (which is exactly what the scheme aims to avoid, so as to save on highly scarce supervisory skills) . 8 6

As government paper does not count towards obligatory reserves, the reserves/deposits ratio is about 60 per cent. 7 A possible criticism of Boot and Wijnbergen is that this creates a captive pool of funds for the government to borrow, encouraging it to run a budget deficit. Rostowski (1993a) suggests that 100 per cent reserve banking should be combined with a currency board system in which high powered money is 100 per cent backed with international reserves. As a result, deposits in such 'safe banks' would effectively be held abroad and would not be available to the government to srend. Boot and Wijnbergen's hope that the firms which are the commercial banks' other depositors will do the job seems misplaced. It is possible that over time only good commercial banks will survive, because only they will receive deposits from commercial firms. Once this happens, supervision of the savings banks lending to commercial banks will become unnecessary. However, until

Introdudion

5

2. Policies to deal with bad debts If bad loans of a systemic magnitude9 cannot be avoided during the transition, then they must be dealt with as the transition proceeds. Mizsei (Chapter 4) contrasts the approaches adopted in Hungary and Poland. In the former, the authorities adopted a 'sticking plaster' approach, giving priority to fiscal considerations, and providing the minimum amount of financial support to the banking system. The result has been that three successive bail-outs of the banks have been necessary so far, and the total cost of these approaches $3 billion, or 9 per cent of Hungary's annual GDP. Poland, on the other hand, attempted a once and for all solution to the problem, the so-called 'Enterprise and Bank Financial Restructuring Progamme' (EBFRP). The novel aspects of this scheme were: (1) ex ante capitalization of the banks involved to bring them up to 12 per cent capital adequacy (according to the Cook Committee criteria) as of end 1991, a date well before the beginning of the development of the EBRFP, so as to reduce moral hazard; (2) a requirement for the banks involved in the programme to create specialized 'bad debt' departments and for these to liquidate all bad debts identified as of mid-1991 (and thus subject to the programme), by April 1994; (3) a special simplified and accelerated 'conciliation' procedure, which allowed the banks effectively to perform the function of bankruptcy courts, as far as writing down debts in exchange for recovery programmes by debtors is concerned. The overall cost of this scheme (funded mainly by Western donors) has been about $600 million. The key question is whether the programme will indeed be a one-off event, or the first in a series, as in Hungary. In this context, the recapitalization of the Bank for Food Economy (BGZ) in 1994 to the tune of $850 million, and without the stringent criteria of the EBFRP being imposed, causes concern (this was part of the spoils taken by the victorious Peasant Party after the 1993 elections). then, as long as savings banks benefit from deposit insurance they will need supervision and, as we have seen, this will in tum require the supervision of the commercial banks as well. 9 I.e., of a level which threatens the solvency of the banking system in the absence of government intervention.

6

Rostowski

A third approach, described by Hansson (Chapter 7), and adopted in Estonia, is what one might call 'hard supervision'. Banks have been allowed to go bankrupt, and depositors - including households have been required to bear a significant loss. The Estonian view is that such an establishment of reputation by the authorities early on in the transformation has sharply reduced the pressure for bail-outs, or even the need to ensure effective supervision (as it has demonstrated that implicit deposit insurance in Estonia is very limited). In the Czech Republic, it seems that the profitability of commercial firms is being sustained by the highly restrictive wages policy (which applies to both the private and the state sectors). This in turn ensures that most bank loans to firms remain apparently good. However, it is impossible to sustain a tight wages policy indefinitely, so the problems of the Czech banking sector are likely to surface some time. Once this happens, the fiscal aspect of the problem may be far larger than it is in Poland or Hungary, because of the far larger ratio of credit to business or GDP. In the Czech Republic this figure is over 70 per cent, as compared with about 40 per cent in Hungary and 20 per cent in Poland. Io

3. Bank privatization In the long run, the accumulation of new bad loans is likely as long as banks remain unprivatized. In most Western economies, banks are subject to three kinds of control: by supervisory authorities, by depositors and by owners. Supervisory authorities exist to strengthen protection of depositors' interests in the face of the free-rider problem (see Chapter 3). In the absence of private ownership, bank equity remains 'weak' (see Chapter 2), and supervisors have to protect both depositors and owners (that is the state). Even highly skilled supervisors are unlikely to be able to do this successfully, as the recently revealed massive losses at Credit Lyonnais (almost $8 billion) indicate. What is more, by the use of suitable reserve requirements supervisors can ensure that bank capital relative to risk-bearing assets is high.II In the presence of high capital/risk asset 10 The order corresponds to the inflation levels recorded in the three countries before the collapse of communism. 11 See Fernandez and Guidotti (1994) for a description of how high reserve requirements ensure high capital/asset ratios. High reserve requirements have the advantage over minimum capital/asset ratios of not requiring a sophisticated auditing system capable of assessing loan quality.

Introdudion

7

ratios, the interests of real private owners and depositors become more congruent, and owners will have an interest in doing much of the work of avoiding bad loans otherwise left to supervisors. 12 Banking systems have evolved very differently in the FSU (together with Bulgaria) and in Central Europe. In Central Europe, what has been called elsewhere the 'main sequence' of banking reform (Rostowski 1993a), the usual process consisted initially in the splitting up of the monobank into a number of state-owned commercial banks (SOCBs) on a regional and/or functional basis. Therefore, the main problem in Central Europe is how to privatize the SOCBs, which, in spite of the existence of new private banks (both domestic and foreign) form the bulk of the system. Mortimer (Chapter 5) deals with the contrasting approaches adopted in the Czech Republic and Poland. In the former, all the main banks have been privatized through the mass voucher privatization process. 13 Although this is a great achievement, 35-40 per cent of former SOCB shares are still held by the National Property Fund (that is effectively by the state), with much of the remaining shares being held by investment privatization funds (IPFs) managed by the main banks themselves. 14 There is thus a severe lack of outsiders in the banking system. 15 Poland, on the other hand, has attempted classical British-style privatization of its banks. The intention was initially to privatize two banks per year (out of an original stock of fourteen SOCBs). Three have already been privatized, but the share price manipulation scandal surrounding the privatization of Bank Sh,l.ski has politicized the issue to such a degree that even the government's need for revenues has not been sufficient to prevent a slowing down of the process (rather than the required acceleration). Even worse, the resulting politicization of bank privatization is likely to make Western banks - which are a vital component in the Polish strategy, because of the capital and skills they are expected to contribute - far less willing 12 Because equity is the first to be wiped out by loan losses. 13 A major exception is the Sporitelny Banka, the savings bank, whose ownership was handed over to local authorities. 14 Of the seven top investment funds, all belong to banks except the independent Harvard Fund and one which belongs to the former state insurance company! The Harvard Fund's attempt to gain influence in the Czech Republic's largest bank, Komercny Banka, seems to have been related to the enforced exile of its chairman, Kozeny. 15 However, the banks are interested in increasing their base through public offerings of shares, which will, with time, dilute the NPF and bank-controlled IPFs' holdings.

B

Rostowski

to involve themselves. The jury is thus still out as regards the choice between the Czech and Polish approaches. Finally, bank privatization in Hungary is not examined in detail because that country has failed to privatize any of its SOCBs. 1 6 This failure is due to the parlous financial position of the Hungarian SOCBs after the bankruptcy law and stabilization of 1991, when their Cook ratios 1 7 became either negative or fell to between 1 and 3 per cent (Chapter 4). This raises an important point: bank privatization may be necessary to avoid the re-appearance of bad loans, but the resolution of the bad loans problem is necessary for privatization. This is why Poland sold its three healthy banks first, and then undertook its EBFRP so as to be able to privatize the next tranche. The Czech Republic seems to have privatized its banks before the magnitude of its bad loans problem has been made manifest, as we noted in the previous section. In the FSU, in contrast, the 'main sequence' of bank reform was not followed. 18 Instead, the commercial operations of the monobank fractured into the three spetsbanki, 1 9 and then into hundreds of independent banks. At the same time, hundreds of new banks were established by SOEs (the so-called 'sectoral banks')20 and by private firms (Zhuravskaya, Chapter 8). The branches of the spetsbanki became independent banks initially through simple decentralization, and were then partly privatized through the issue of new shares. 21 In 1993 and 1 994, the Russian mass privatization programme indirectly privatized the post-spetsbanki and the 'sectoral banks' by privatizing their state enterprise owners. As a result, Russia has a largely private banking system, and in terms of the transformation of ownership in the banking sector it, together with the Czech Republic, leads the region. 16

Al though it has only five larger SOCBs. First tier capital to risk weighted assets (government paper has a zero wei§ hting) . 1 Developments in Bulgaria were closer to those in the FSU, those in Romania to Central Europe. 1 9 Promstroi, Agroprom and Zhilsots. These could finance themselves by writing debits on the 'inter-branch turnover' ( mezhdufialny oborot or MFO) , i.e. , effectively b y creating money. Thus, they were i n practice central banks. This leaves out of account the central banks of the other Soviet republics - later post-Soviet states - which could also create roubles after 1989. 20 These were a form of 'indirect state ownership' , whereby SOEs were the shareholders. 21 As the spetsbanki fragmented, their larger offshoots initially remained on the MFO, increasing the number of central banks in Russia into dozens by the end of 199 1 . 17

Introduction

9

This does not mean, of course, that there are no problems relating to bank ownership in Russia. The most important is the large amount of connected lending which banks - often established for this very purpose - have undertaken. The key ownership issue is, therefore, how to ensure either the divestment of ownership by large connected borrowers, or the transfer of their borrowing to other lenders. Here Central European experience is not much use. In the remaining countries of the FSU, in which mass privatization programmes have not been undertaken, the authorities need to consider whether they should follow the Czech, Polish or Russian paths as regards bank privatization. Much depends on the number of banks involved in a particular country. The Czech and Polish paths would require that banks which are at present in 'indirect state ownership' be reclaimed by the state before being privatized. Furthermore, it seems unlikely that voucher holders would be able to assess a large number of very small banks, or that the state would be capable of arranging the sale of many such banks either through flotations on the stock market or through trade sales. Only the largest banks would thus qualify for either one of the two Central European approaches. The small banks would have to be dealt with in one of three ways: (1) enforce divestment by existing state-owned entities (the main instrument in such a policy would be legislation forbidding connected lending); (2) encourage amalgamation by a tightening of licensing require­ ments, so as to eliminate very small banks; (3) wait for the privatization of the current owners (this would have to be through a mass privatization scheme, as all other methods are far too slow) . It would in any case not solve the connected lending problem.

4. The architecture of the banking system The key issue with regard to the 'architecture' of the banking system is the scope of banks' activity. One of the most basic of these activities is the provision of a non-cash payments system for the economy. Rostowski (Chapter 2) suggests that in second wave countries, banks involved in providing payments services should be obliged to hold 100 per cent reserves against deposits used for payments purposes.22 22 It is too late for such an approach to be adopted in Central Europe, but this does not mean that it should not have been (Rostowski 1993a) .

IO

Rostowski

Banks engaged in other activities would be regulated only lightly, if at all. Boot and Wijnbergen (Chapter 3), although they support the 'narrow bank' view, see households' deposits at savings banks as the ones which require the protection of insurance and therefore regulation. The question comes down to whether one believes that implicit deposit insurance results from the political need to protect households, or from the technical and economic need to protect the payments system. Topinski and Struzynski (Chapter 6) describe how Poland established a new payments system suitable for a market economy and a two-tier banking system. The old monobank payments system was based on automatic debiting of payers' accounts before payees were credited. Two instruments were used for this purpose. The first was the so-called 'payment demand order', by which a supplier sent his invoice via his own bank (previously his own monobank branch) to the payer's bank (branch). These payments were automatically effected in the order in which they arrived, and if no money was available on the account the demand order was filed in a 'second file' ( kartoteka dva) to await payment in chronological order when funds were paid in. The system made sense under the monobank system when the bank, the payer and the payee all belonged to the state and every firm had one bank account at one branch of the monobank. 23

In Central Europe, this 'payment demand order' system was abolished either very early on in the transition (Czechoslovakia) or even before the transition began (Poland) . 24 In the FSU, the system survived until mid-1992. This led to confusion between arrears in the payment of 'payment demand orders', which resulted from the voluntary provision of trade credit by suppliers in the context of an incredible stabilization programme (they could have demanded cash in advance) , and the breakdown of the payments system, for which the authorities could have been justifiably considered responsible (Rostowski 1993b). The result in the summer of 1992, was a surrender by the authorities to political pressure to resolve the 'non­ payment crisis': massive amounts of Central Bank of Russia credit were pumped into the economy and a multilateral clearing of the 'non-payments' was undertaken. 23 This account was often divided into a number of earmarked sub-accounts (for example, for wages). 24 However, payment demand orders survive in Hungary without any apparently noxious effects.

Introduction

11

The second instrument used for debiting payers in post-communist countries is the 'payment order', which is similar to the West European giro. 25 Struzynski and Topinski describe how after the introduction of two-tier banking, this system created a liability of the central bank corresponding to the amount of money in transit. In many countries, a rapid growth of transactions, which occurred at the beginning of the transition as a result of a massive increase in the number of businesses, caused severe payments delays, and a sharp increase in the amount of money in transit. This reduced the money multiplier, making monetary policy control more difficult. By contrast, the introduction of cheques resulted in the central bank providing a float to the banking system for payments purposes (this was an asset of the central bank, as payees were credited before payers were debited). The result was a massive kiting scam in Poland (the Art-B affair), resulting in losses to the National Bank of Poland to the tune of some $400 million. 26 Not only did this further complicate the execution of monetary policy, as base money declined while broad money increased during the scam (increasing the money multiplier), but it also brought to the surface the question of settlement risk, as a private commercial bank involved in the fraud proved incapable of meeting its obligations to other banks when the bubble burst. 27 Under the monobank system, of course, such a situation could not have arisen. According to Boot and Wijnbergen, all banks except savings banks should be largely free to undertake what business they wish. According to Rostowski, the same is true of those banks which do not provide payments services. However, the question remains whether banks, both in Central Europe and in the 'second wave' countries, should be encouraged to become universal banks. Boot and Wijnbergen, approaching this question from the point of view of banking regulation, point out that the decision has effectively already been taken, in the sense that banks throughout the region are free to become universal, as laws allowing them to do so have been passed in all countries. However, the question remains whether the post­ communist countries should encourage, rather than just permit, their 25 It is sent by the payer via his bank to the payee's bank. 2 6 One of the reasons why the scam was profitable was that there were very high interest rates in Poland together with a fixed exchange rate, in the aftermath of the stabilization programme. See Chapter 6, Section 3.4 for the Art-B affair. 2 7 This resulted in the wiping out of an equal amount of central bank and commercial bank liabilities, this time reducing the money multiplier.

12

Rostowski

banks to become universal banks of the German kind (which have powerful insider control over the bulk of major companies) in order to base the allocation of capital in their economies upon them. Rostowski argues that this would be neither desirable nor feasible. In the first place, state-owned universal banks would be a grave threat to the marketization of the economy. They could become the centres of large state-owned holding companies, which would both inhibit competition and could extract subsidies from the state (and might well hinder subsequent privatization). Secondly, the value of such banks' equity assets would be very opaque, as they would consist mainly of non-quoted stocks. 28 This would make the role of supervisors in protecting the interests of depositors and the state (as owner and deposit insurer) far harder. It would also increase the likelihood of soft budget constraints creeping into the financial system through the subsidization of firms by banks buying the unquoted equity of the former at inflated prices. Banks must, therefore, be privatized before they can be allowed to become German-type universal banks. As we have seen, in most countries this is likely to be a lengthy process, making this kind of banking unsuitable for the first phase (at least the first five years) of the transition. In my view, even the banking system of the Czech Republic is insufficiently privatized for that country to be able to introduce German-type universal banks safely (see Chapter 4). Fortunately the Klaus government shows no desire to do so, although the fact that the top seven Investment Privatization Funds (mainly controlled by the major banks) have extensive control over much of Czech industry, means that this may happen in spite of the authorities' preferences. Russia may be the country whose banking system comes closest to the required degree of privatization, but here we come up against another problem: German-style universal banking involves a high proportion of long-term lending in total bank credit. First, long-term lending is by definition riskier than short-term lending, and harder to do efficiently. Post-communist banks with low skills will therefore have even more bad loans than they already do if they become German­ type universal banks prematurely. Secondly, providing long-term 28 An abundance of quoted stocks for universal banks to invest in would imply a high level of development of the stock market. Yet one of the main arguments put forward for German-style universal banks is that stock exchanges will not develop sufficiently rapidly to provide commercial firms with the capital they need.

Introduction

13

credit is impossible at the high rates of inflation experienced throughout the region (in excess of 10 per cent per annum), and it is quite unthinkable in countries which have not stabilized such as Russia (inflation rates in excess of 5 per cent per month). Apart from anything else, stabilization results in a sharp change in the relative profitability of various firms. Long-term commitment (supposedly the virtue of the German system) cannot be undertaken by banks under these circumstances without threatening their solvency even more than happens upon stabilization at present, in a situation where the banks are only providing short-term working capital. The third main issue affecting the architecture of the banking system is the structure of the inter-bank market. In many of the countries still undergoing very rapid inflation - including Russia household deposits have practically disappeared as a result of the refusal of the monopolistic savings bank to raise nominal interest rates sufficiently. In this situation, commercial banks obtain their funds from two sources: from the central bank and directly from commercial firms in the form of (relatively high interest bearing) deposits. In these countries, there is therefore no particular problem resulting from the structure of the inter-bank market, except for the continuous supply of central bank credit, and any micro-distortions this may result in. 29 However, once stabilization occurs the amount of household deposits increases very sharply (Rostowski 1994). The bulk of these inevitably goes to the savings bank, which then obtains a very powerful position as a provider of funds on the inter-bank market. Chapter 3 describes the high degree of concentration of household deposits in the case of Poland, which has one of the least monopolized structures in the region in this respect. In Russia, the savings bank has 30,000 branches compared to 4,660 for the whole of the rest of the banking system (Chapter 8). Such an unbalanced inter-bank market has a number of serious consequences for banking system reform. With one institution being the dominant supplier of funds, one of three unwelcome results can occur: (1) the savings bank can develop an unacceptable sway over domestic interest rates, which then depend on its - presumably commercial - goals, rather than on the requirements of monetary policy; 29 This is the reason why the IMF has persuaded the Central Bank of Russia to auction part of its credit. Whether this is sensible given the quality of many Russian banks is another matter.

14

Rostowski

(2) the central bank takes over the asset/liability management policy of the savings bank and tailors this to its monetary policy requirements; (3) the central bank becomes the major player in gross terms on the market. Thus, whatever happened, the typical Western case in which the inter-bank market is largely balanced in gross terms, and the central bank influences interest rates by operating at the margin, would be unattainable. The third solution may not be a problem if there is already a large amount of central government or central bank paper on the market. However, if there is not, the authorities would have to increase their indebtedness, possibly by a very large amount, just to avoid one of the other two quite unacceptable outcomes. Boot and Wijnbergen propose the division of the savings bank, and this would seem to be a minimum requirement for the inter-bank market to operate adequately. In Poland, as a legacy of the reformist communist period of the 1980s, there are two savings banks, one holding foreign currency deposits and the other zloty deposits. This does not seem to be a prudent way to proceed, as it can cause a serious currency matching problem (of liabilities and assets) in each of the savings banks. In Russia, it might be best to proceed as Poland did with the SOCBs it created out of the state monobank: that is to create independent, regionally based savings banks, which would initially remain state-owned until a sufficient number of large and respectable commercial banks had come into existence for the savings banks to be put up for sale. In the meantime, the savings banks might well begin to encroach upon each other's territory, as the Polish SOCBs have done. Even if they did not, they would be competitive suppliers of funds on the inter-bank market in Moscow. A similar solution should be possible in Ukraine. For very small countries such as the Baltic states, which have a very large proportion of foreign trade in their GDP and which are, therefore, largely price takers in the world market, an Estonian-type currency board system with free movement of capital may be the solution to the problem of the monopolistic structure of the inter­ bank market. It is less clear how the problem should be resolved in the medium-sized countries of the FSU and the Balkans (and indeed of Central Europe), which might be too large to adopt a currency board and too small to split up their savings banks on a regional basis.30 3

°

Countries such as Bielarus, Uzbekistan, Romania or Poland.

Introduction

15

5. Conclusion Banking sector reform still has far to go, both in Central Europe and in the 'second wave' countries of the FSU and the Balkans. How successful such reform is will be decisive for the economic development of a fifth of the world's land mass, because of the position of the banking system at the nerve centre of a market economy. In the short term, banking is vital for the effective functioning of a modern economy's non-cash payments system. Also, in the early transition, resistance to stabilization becomes manifest in the banking system, either in the form of so-called 'payments arrears', or as bad debts and connected lending. The experience of the post-communist countries has shown that, in this field, following Western models is insufficient. Solutions lie in combining bank privatization with an architecture of the banking system which severely restricts deposit insurance and combines this with tough but not excessively sophisticated - supervision mechanisms (such as very high reserve ratios), and ensures competition in the inter-bank market. German-style universal banking is beyond the reach of post-communist countries at present, because of the sophisticated bank supervision it requires and because of these countries' high inflation rates. Those countries which already have severe bad debt problems are better advised to follow a Polish-style strategy of the kind embodied in that country's EBFRP, rather than the ad hoc approach adopted in Hungary, although it is unclear whether in Poland too the EBFRP will not turn out to have merely been the first in a series of bank bail-outs. References Begg, D. and Portes, R., 1991, 'Enterprise Debt and Economic Transformation: Financial Restructuring in Central and Eastern Europe', mimeo. Fernandez, R. and Guidotti, C., 1994, 'Regulating the Banking Industry in Transitional Economies' , paper presented to Pew Conference on Economic Reform: Lessons from Latin America for the Transition Economies, Washington, DC, May. Rostowski, ]., 1993a, 'Problems of Creating Stable Monetary Systems in Post-Communist Economies', Europe Asia Studies, Vol. 45, No. 3. Rostowski, ]., 1993b, 'The Inter-enterprise Debt Explosion in the Former Soviet Union', Communist Economies and Economic Transformation, Vol. 5, No. 2. Rostowski, ]., 1994, 'Dilemmas of Monetary and Fiscal Policy in Post-Stabilization Russia', in Economic Transformation in Russia, ed. A. Aslund, London, Pinter.

CHAPTER 2

THE BANKING SYSTEM, CREDIT AND THE REAL SECTOR IN TRANSITION ECONOMIES Jacek Rostowski

1. Introduction: good and bad credit Before the beginning of the transition to capitalism, all credit in Central Europe and the Soviet Union was 'systemically bad' in the sense that it was not allocated on the basis of commercial criteria, and therefore there was no reason to suppose that once ordinary market conditions were established, any particular loan could be serviced and ultimately repaid by the borrower. Bad credit of this kind is not really an asset belonging to the lender, but rather a transfer to the borrower, occurring at the expense of either the lender or of those financing the lender (the lender's lenders). 1 To the extent that part of any credit could be serviced and repaid, this was accidental and, to the extent that servicing is enforced, could be thought of as a random tax on particular borrowers. The reasons for the bad state of credit before the transition were different in those economies which were centrally administered right up to the loss of power by the communists, to the reasons in those economies where some degree of market relations were already in place. Albania, Bulgaria, Czechoslovakia, the German Democratic 1 A significant number of loans may be recoverable given normal market conditions even when credit is ' systemically bad' and, vice versa, many loans may be unrecoverable even when credit is systemically good. The difference depends on whether the mechanisms requiring and enabling banks to select loans which will be good under normal market conditions exist or not. Credit was 'systemically bad' in communist economies in two senses: in the first sense, there was nothing in the procedure of credit allocation which gave reason to suppose that any particular loan could be repaid; in the second, therefore, a large, but unknown, proportion of loans could not be repaid. It is this latter proportion which constitutes a transfer to the borrower.

The Banking System, Credit and the Real Economy

17

Republic and Romania belonged to the first group; Hungary, Poland, the Soviet Union and Yugoslavia to the second. In the first group, credit was allocated on the basis of planners' preferences, which had no correspondence to commercial criteria. Sometimes credit was supplied to compensate state-owned enterprises (SOEs) for the fact that profits had been siphoned off, a transfer which was intended to prevent the SOEs from accumulating inventories (Schmieding 1991).2 In the second group of countries, rapid inflation and sharply negative real interest rates had a similar effect. When loans are granted under such circumstances, it is almost impossible to know which of them could be serviced and recovered under normal market conditions. Furthermore, in semi-market socialist systems of this kind, the central bank effectively acted as guarantor of all bank and inter­ enterprise loans (Calvo 1991). Such implicit full credit insurance was sufficient in itself to ensure that credit would never be repaid. Thus a key component of the transition to market capitalism is the transformation of systemically bad credit into systemically good credit. In Section 2, we look at the microeconomic reasons for and difficulties with the switch from bad to good credit; in Section 3 we discuss the macroeconomic aspects of the problem; in Section 4 we discuss the role of the banking system in fostering growth during the transition to capitalism; in Section 5 we examine the debate regarding the suitability of universal banks for post-communist economies (PCEs) and the role of long-term credit, and in Section 6 we examine - with the example of Poland - the extent to which credit is important for economic transformation. Section 7 tries to draw together the conclusions which emerge from our analysis of events in the Central European 'early reformers', and to suggest what might be an 'appropriate banking sector technology' for second-wave PCEs in the former Soviet Union and the Balkans. 2. Microeconomic aspects of switching from bad to good credit The reasons for wishing to eliminate 'bad credit' early on in the transition are self-evident: bad credit is a transfer to the borrower; moreover, it usually takes place at the expense of the state or those who are holding money (i.e. savers). It is thus both the reason for the persistence of soft-budget constraints for SOEs and a potential source 2 If planners believed an enterprise to be accumulating excess inventories they could reduce its credit, something which could not be done as easily with its own funds.

18

Rostowski

of inflation. Since it occurs outside public and/ or parliamentary control, bad credit is allocated in a less accountable way than are budgetary subsidies and is, therefore, even less desirable than these. What then stands in the way of switching from bad to good credit? We first examine the case of the centrally administered economies of the first group, as this allows us to avoid the additional complicating question of high inflation at the moment of the switch over. Banks which have never made such choices before suddenly have to decide which enterprises to continue financing, and which to cut off from credit. They are likely to make many mistakes. 3 The costs of the switch are, therefore, likely to be high, with viable enterprises having to curtail their activity and even being liquidated, while continued lending to unviable enterprises undermines the capital base of the commercial banks, leading either to the failure of certain banks (and possibly of the banking system as a whole), or to their recapitalization - of which more below. It is not easy to imagine a 'gradualist' solution to the problem of the switch from good to bad credit - after all, banks either are or they are not responsible for the loans they make. Yet it is in this area that the costs of 'shock therapy' are likely to be particularly high, as almost completely inexperienced institutions are given the duty to decide about the fate of a large part of the productive sector of the economy. There seem to be two ways out of this acute dilemma: a complete write-off of debt and inflation. 4 The first of these solutions has been proposed by Begg and Portes (1991) and Schmieding (1991). It stresses the economic irrationality of the initial allocation of credit under central planning, points to the risk of liquidation of economically viable enterprises and of the bankruptcy of a significant number of banks, and therefore proposes the writing-off of all bank debt inherited from the previous regime. Furthermore, it is claimed that bad inherited credit will contaminate the new, potentially good, credit. This is because banks, unable to bankrupt existing unviable borrowers due to the effects of such a 3 They may be as likely to withdraw credit from viable as from unviable enterprises, and may be as likely to continue providing credit to unviable as to viable SOEs. 4 There may also exist what I call the 'Czech solution', which is to maintain sufficiently tight control of wages to ensure that after price liberalization a significant number of SOEs retain profitability, while at the same time discouraging excessive reallocation of credit from old to new customers by the commercial banks. Instead of banks subsidizing bad enterprises, workers (in all enterprises) do.

The Banking System, Credit and the Real Economy

19

move on their capital base, may allow such borrowers to capitalize interest and roll over principal payments due (so-called 'evergreening'), thus depriving new, potentially far better borrowers from access to credit. Once the bulk of bank assets consists of such bad credits, the true financial position of the banks will come to light, and the bad credits will have to be written off in any case, it is argued. The Begg-Portes-Schmieding solution faces two key problems. The first is that it posits the feasibility of a credible, once and for all, forgiveness of debt which leads to no moral hazard. Such a deus ex machina may be somewhat more believable in an economy which up to the beginning of the transformation was highly centrally administered, such as Czechoslovakia or the GDR, and where there was no discernible market rationale for the distribution of credit. It is very doubtful if the 'one off' nature of the debt write-off would be credible in an economy such as Hungary or Russia, which has been trying to shift from some kind of market socialism plagued by soft budget constraints to a true market economy, and in which enterprises are equally responsible for their debts, since they had actively lobbied for credit. A more serious weakness of the proposal is that it treats the bad debt problem as if it were only a stock problem, whereas it is in fact both a stock and a flow problem. The flow aspect arises from the lack of credit allocation skills in the new commercial banks at the beginning of the transformation, from the property rights structure of both banks and their main customers the SOEs, and from certain cultural factors (see Section 4). The existence of a flow problem is confirmed by the Polish experience, which shows that there are about as many qualified loans among those made to new private sector borrowers as there are among those made to the old SOE customers (Kawalec, Sikora and Rymaszewski 1993, and Gomulka 1993, Table 4). There are two possible solutions to the flow problem after the stock problem has been resolved by a complete debt write-off. The first, suggested among others by Schmieding (1991), is to admit foreign banks to the domestic market at the very beginning of the transition, as they have the requisite skills, traditions and property rights structure. Since foreign banks are not burdened with any bad debts, their entry has been opposed in a number of PCEs by domestic banks, on the grounds that they would 'cherry pick' the best customers. However, after a complete debt write-off this would not be a problem. The former GDR is the sole example of massive entry of Western banks on the domestic market of a PCE, and a study of the quality of new lending there should be most instructive. Polish

20

Rostowski

experience is said to show that Western banks have not performed very much better than domestic banks (we discuss some of the possible reasons for this in Section 4). In any event, in some countries the entry of Western banks on a large scale is politically infeasible (this has been very much the case in Russia since the beginning of the transition). The second solution to the flow problem is to abolish fractional reserve banking (McKinnon 1992, Rostowski 1993a). Having to hold 100 per cent of their liabilities as reserves at the central bank, commercial banks would be unable to lend, and so would be unable to accumulate bad loans. This would allow banks to concentrate on their primary function as suppliers of payments services to the economy. Subsequently, a gradual reduction in required reserve ratios would allow banks to learn by doing, as the economic volatility succeeding the collapse of communist power died down. 3.

Macroeconomic aspects of switching from bad to good credit

As we have noted, inflation helps to solve the problem of the stock of bad credit inherited from the centrally administered system. But it does so at the cost of maintaining the flow of bad credit, and by ensuring that when inflation is finally brought under control the problem of the stock of bad credit will re-emerge, without there having been much progress in solving the flow problem. Nevertheless, it seems likely that the inflation which has accompanied the economic transition from central planning everywhere, has been due in part to the desire of the authorities to evade the bad credit problem (both in its stock and flow aspects). 3. 1 The 'credit crunch ' hypothesis The potential role of bad credit in generating inflation is clear. Let us imagine a PCE in which the banking system finances the purchase of inventories. 5 Loans are granted only for the purchase of real inventories and have to be repaid once these have been utilized in the production process. However, once this happens new loans are obtained to purchase new inventories. Unless there is some anchor, such as a limit on total nominal credit or a positive real interest rate, then credit, money and prices can increase indefinitely. This is so even though no firm receives 'excessive credit', because each firm individually can fully repay its loans by liquidating its stock of 5 Calvo ( 1 991 ) develops a similar model which requires, however, that real inventories continuously accumulate for inflation to be generated.

The Banking System, Credit and the Real Economy

21

inventories. What we have is a post-communist version of a banking system applying the 'real bills' doctrine. As is well known, under such circumstances the price level is indeterminate. Credit thus becomes a money (or inflation) machine. Furthermore, as we have seen, once inflation is high and real interest rates are sharply negative, all credit is 'systemically bad': under such circumstances most loans, however badly allocated, are likely to be repayable. What stands in the way of switching off this machine? The machine can be deactivated either by introducing positive real rates of interest (together with the requirement that interest on bank loans not be capitalized), or by having a nominal domestic credit limit (Calvo 1991). However, in post-communist economies the additional consequence of deactivating the machine is that banks have to select - for the first time - which customers to continue lending to, and which to cut off from credit. Since the banks concerned have never previously had to make such decisions, they are likely to make many mistakes, with the result that a large part of bank credit will be revealed as bad in the accounting sense (that is, it will be revealed that borrowers cannot service it - see Figure 2.1). Stabilization thus transforms systemically bad credit into openly bad credit. The danger is that the possible ensuing insolvency of the banking system will lead to a recapitalization of the banks which, unless the way in which banks operate is changed, will lead to new bad debts, new recapitalization, and so on, with the result that bank recapitalization becomes the new money machine. Thus the two events, switching to commercially based credit allocation and macroeconomic stabilization, are aspects of the same process: without stabilization banks will not start to allocate credit commercially (they need to believe that previous conditions under which real interest rates were highly negative have come to an end) , but unless banks begin to allocate credit commercially stabilization may fail because of the need to recapitalize the banks. Calvo and Coricelli (1993) have pointed to the decline in real credit which accompanies price liberalization as a cause of the sharp drops in industrial output in post-communist economies. The solution they have proposed is that stabilization be accompanied by a once and for all increase in nominal central bank credit - so that total real credit will not decline (and possibly may even increase). 6 However, this solution presents a number of problems. 6 If the one-off nature of the increase in nominal credit is credible, then prices will increase no more than (and possibly less than) nominal credit.

22

Rostowski A: Before stabilization/liberalization ASSETS

LIABILITIES

RESERVES

CENTRAL BANK CREDIT 300

300

CREDIT (all systemically bad) 300

DEPOSITS

270 CAPITAL

30

B: After stabilization/ liberalization and a 50% increase in prices ASSETS

LIABILITIES

RESERVES

300

CENTRAL BANK CREDIT 300

CREDIT

DEPOSITS

300

270

----------------------------------Bad (in an accounting sense)

90

CAPITAL

30

Fig. 2. 1 Consolidated balance sheet of the commercial banks Note: Units express nominal values, height expresses real values.

First, banks would misallocate much of a once and for all nominal credit increase. As a result, they would soon be obliged to increase their lending margins and shorten their balance sheets, so as to improve their capital/asset ratios, bringing about the very 'credit

The Banking System, Credit and the Real Economy

23

After stabilization/liberalization, a 50% in prices and a 50% increase in the nominal value of central bank credit* ASSETS

LIABILITIES

RESERVES 450

CENTRAL BANK CREDIT 450

CREDIT

DEPOSITS

450

420

----------------------------------Bad (in an accounting sense)

90

CAPITAL

30

* It is assumed that in the balance sheet of the central bank cash = international reseives. Fig. 2. 2 Consolidated balance sheet of the commercial banks

Note: Units express nominal values, height expresses real values.

crunch' the policy was designed to avoid (only somewhat later).7 Thus, for any given level of the banks' capital base, the greater the one-off increase in bank credit, the greater the level of bad debt and the greater the need for credit retrenchment subsequently (see Figure 2.2). Therefore, in order for a credit crunch induced fall in output to be avoided, the proposed increase in bank credit would need to be accompanied by an increase in bank capital. The question then is whether a stabilization programme which began with an increase in central bank credit and bank capital could be credible, or whether it would be more likely to be perceived as providing finance for the bad old habit of crediting loss makers. The orthodox approach of reducing central bank credit creation at least minimizes the ratio of bad credit to bank capital and also the amount of bad credit. 7 If the misallocation was sufficient to render the banks insolvent, recapitalization of the banks would be necessary, increasing the likelihood of the re-activation of the 'money machine'.

24

R.ostowski

Secondly, the proposal is vulnerable to a lack of credibility vis-a-vis both borrowers (assumed to be mostly enterprises) and deposit holders (assumed to be mostly households). If the former believe that the credit expansion is not truly once and for all, they will act on the assumption that more nominal credit will be available in the future. In doing so, they will maintain at a high level the cost to society (in terms of output and employment) of disappointing these expec­ tations. However, deposit holders also have the power to frustrate policy makers' altruistic intentions. If they do not believe in the once and for all nature of the increase in the nominal money supply which accompanies the once and for all increase in nominal credit, then they will refuse to hold the additional real money balances, so that the intended once and for all increase in nominal money and credit will be offset by an equivalent increase in the price level. The authorities will then be faced with the choice of either sticking to their announced policy, and refusing to increase nominal credit further in the face of continuing inflation, even though this results in the very reduction of real credit which they wished to avoid or, alternatively, they can increase nominal credit, confirming the expectations of those who doubted the once and for all nature of the earlier nominal credit increase. Another way of looking at the same problem is that negative real interest rates, and the creation of a flow of new nominal credit by the central bank which is directed to the SOEs, mean that during high inflation the SOEs are the main beneficiaries of the inflation tax, which is paid by holders of money. Thus, as long as the credit machine operates, SOEs are beneficiaries of a net flow of real funds from the rest of the economy (mainly households). Stabilization, by introducing positive real interest rates on loans, stops or even reverses this flow of real funds, with the banking system (and thus in PCEs indirectly the budget) being the main beneficiary. Households also benefit from the reduction in the inflation tax, though probably to a lesser extent. Unless other taxes can be increased sufficiently to maintain the subsidies to the SOE sector, which seems improbable, stabilization is bound to entail a deterioration in the real financial situation of SOEs as a whole. 8 8 Calvo and Coricelli (1993) seem to be tending towards such an interpretation when they say that capitalized interest should not be included in any calculation of real credit. However, it is unclear why any interest on loans, be it capitalized or not, should be included in real credit. If a firm actually pays interest due, it has that much less liquidity available, and this should be

The Banking System, Credit and the R.eal Economy

25

This in turn is likely to lead to a decline in their output.9 One way of mitigating these phenomena would be to combine stabilization with the complete debt write-off and the introduction of a very high (possibly even 100 per cent) reserve requirement in the banking sector, as discussed in the previous section. The SOE sector will then lose its net inflow of funds, but will not suffer a real net outflow as it has in some stabilizing PCEs. It should be noted that such a net outflow of real funds from the enterprise sector could take place after stabilization even if Calvo and Coricelli's 'once and for all' increase in nominal credit were successfully applied, in such a way that real credit did not decline. Such a net outflow would occur as long as real credit increased less than the capitalization of interest rates would require. 3. 2 Bad bank debt in post-stabilization PCEs: an overstated probl.em ? A number of countries which experienced very high inflation followed by stabilization in the 1970s, 1980s and 1990s have exhibited a money multiplier which behaves in a characteristic pattern: as inflation accelerates, the ratio of reserve money (the monetary base) to the chosen money aggregate increases - that is, the money multiplier falls and the proportion of the money supply which comprises central bank liabilities becomes larger. After stabilization, this process is reversed, with the ratio of central bank liabilities to commercial bank liabilities in the composition of the money supply falling, as the increase in real interest rates increases households' and firms' willingness to hold bank deposits, so that the money multiplier increases. In PCEs, this is a particularly acute problem because commercial banks have so little experience in allocating credit on a commercial basis at the time of stabilization. Furthermore, real loan interest rates are often high in absolute terms in the immediate post-stabilization period, resulting in borrowers who might have been creditworthy under stable non-inflationary conditions being unable to service their subtracted from the additional liquidity obtained by another firm, which is now the recipient of a new loan partly made possible thanks to the interest payment made by the first firm. Only a net flow of funds from other sectors of the economy, or from the external sector, should count as a 'liquidity increasing increment in real credit', and therefore relevant in the Calvo and Coricelli framework. 9 Even non-commercially allocated credit is likely to be related in some way to the amount of output (and even short-term value added) which an SOE produces.

26

Rostowski

debts. This happens just at the time when increasing deposits provide the banks with finance, so that the temptation to 'evergreen' loans to dubious borrowers by rolling over interest payments due is strong. In this way, a significant stock of bad debt can be built up by the banking sector quite rapidly. In Poland, total credit to non­ government borrowers was the equivalent of about 8 per cent of GDP in January 1990. By December of the same year it was about 17 per cent, a share it has only moderately exceeded since that time. International audits in mid-1991 showed that about one-third of this credit was qualified. Firm level data analysed by Gomulka (1993) suggests that the proportion of problem loans may be closer to 45 per cent. 10 In Hungary, qualified loans are said to have been 27 per cent of credit to non-government borrowers (Nyers and Lutz 1993 and National Bank of Hungary 1993). If the true value of the bad loans in Poland is half their nominal value, then the additional debt which would accrue to government if it recapitalized the banks would amount to 3 to 4 per cent of GDP. 1 1 While this is not a very large sum, it could if recapitalization were mis-handled lead to severe moral hazard problems in the banking sector, with the government printing money for repeated bank recapitalizations. In Russia, the bank credit to GDP ratio is about 14 per cent, quite similar to the Polish level before stabilization, 12 so that one could expect any bad debt problem to be of a similar, limited, magnitude. In Hungary, the same calculation suggests that recapitalization would also require almost 5 per cent of GDP. 1 3 10 This is the share of debt owed by the 'very bad firms' in Gomulka's sample after co-operatives in this group (which are mainly housing co-operatives) are excluded. The reason for excluding the housing co-operatives is that they have very low sales (which makes them show up as bad on Gomulka's quarterly sales adjusted quick ratio), while they service their debts punctually thanks to government subsidies. I am grateful to Mark Schaffer for pointing this out to me. Gomulka also reports that 34-48 per cent of interest payments were capitalized in 1992, giving another indication of the level of bad loans in Poland. 11 The actual amount by which the Polish government intends to recapitalize the banks is about $1 billion, or 1.2 per cent of GDP. But this may prove inadequate. 1 2 I am grateful to Brigitte Granvelle for the Russian data. In March 1993, commercial bank credit to non-government borrowers amounted to 8.7 trillion roubles while monthly GDP was 5.2 trillion roubles. In Poland in December 1989, the ratio of credit to non-government borrowers to twelve times monthly GDP was about 12 per cent. 13 Although a smaller proportion of credit to non-government borrowers seems to be bad than in Poland, such credit constitutes a larger share of GDP.

The Banking System, Credit and the Real Economy

27

However, a country such as the Czech Republic, where bank credit is about 75 per cent of GDP, is potentially in a far worse situation. The best way of minimizing the bad bank debt problem is to oblige banks to maintain high obligatory reserves and capital adequacy ratios (Rostowski 1993a). The former limits the degree to which deposits are backed by bank loans in favour of a greater share being backed by central bank liabilities. The latter ensures that there is a cushion to absorb loan losses before either depositors or the taxpayer have to suffer.14 Since capital adequacy ratios require that the quality of the port­ folio be regularly examined so that necessary provisions be made, and since such skills are largely unavailable in Russia and other 'second wave' reformers at present, it seems that high reserve requirements can more easily play a useful role under their conditions.

4. Bank credit and economic growth in PC& In the Calvo and Coricelli approach, banks are seen as a source of liquidity in the context of sustaining SOE output in the short term. In the context of facilitating growth, it is their role in the allocation of capital (as selectors of good projects and rejectors of bad ones) which comes to the fore. A number of points needs to be borne in mind: (1) Since banks have not had to fulfil this function under central planning or market socialism, initially they will not perform it well. (2) Cultural factors, and in particular corruption, may seriously inhibit the fulfilment of this function.15 14 Of course, if bank capital belongs to the state or to SOEs, the owners may not be able to exercise effective supervision over managers to encourage the avoidance of loss (see next section). This is why bank privatization is so im Bortant. 5 An example is the situation in Poland during 1990 and 1991, when interest rate ceilings on loans resulted in excess demand for loans. It is said that at that time the customary side payment to loan officers was 1 0 per cent of the credit allocated. Once such customs appear they may not disappear after the introduction of market clearing interest rates (loan officers could take bribes to allocate credit to what they know to be unviable projects). Such behaviour would be more likely to continue if the major banks retained their state-owned character (point (4)). However, privatization need not immediately lead to the elimination of such practices, particularly if the banks pursue a credit rationing policy, which relies on an intentional maintenance of loan rates below market clearing levels (Stiglitz and Weiss 1981). Corruption in the allocation of loans could then continue. Interestingly, such corruption would at the same time deprive credit rationing of its effectiveness.

28

&stowski

(3) The high level of risk in the initial years of transition means that the degree of moral hazard resulting from deposit insurance (usually government provided and implicit) is high. 16 (4) This is augmented by the 'weak' character of bank equity in PCEs, resulting from state ownership. (5) The low level of and/or 'weakness' of equity in non-financial firms in PCEs means that the conflict of interests between lender and borrower is particularly acute. (6) The allocation of short-term credit is easier than the allocation of long-term credit.

5. Banks and corporate governance A number of authors with otherwise very different views on the transition in PCEs have postulated an important role for banks in the corporate governance of non-financial firms (e.g. Lipton and Sachs (1991) and Corbett and Mayer (1992)). The idea is that banks in PCEs should follow the model of universal banks in Germany: owning, and controlling as proxies, large equity stakes in non­ financial firms and having significant representation on their supervisory boards (most PCEs have a two-board structure of the German kind for their larger joint-stock companies). The case for the universal bank intimately involved with its non-financial customers has been made most eloquently by Gerschenkron himself (1968, p. 137), describing the situation in Germany in the second half of the nineteenth century: The inadequacy in the number of available entrepreneurs could be remedied or substituted for by increasing the size of plant and enterprise above what otherwise would have been an optimum size. In Germany, the various incompetencies of the individual entrepreneurs were offset by the device of splitting the entrepreneurial function: the German investment banks - a powerful invention, comparable in economic effect to that of the steam engine - were in their capital-supplying functions a substitute for the

16 McKinnon (1986) argues that a high degree of macroeconomic instability

causes a strong positive covariance in the default rates of bank borrowers, which generates expectations that the government will bail out defaulters when outcomes are unfavourable, thus creating moral hazard in banking. If bail-outs are expected to affect only depositors, the usual results by which deposit insurance leads banks to undertake riskier lending holds (Fry 1988) . If, in addition, the bulk of bank capital is state-owned as in Hungary and Poland point (4) - then equity is also effectively insured. The tendency to choose higher interest yielding, riskier projects will then be strengthened even more.

The Banking System, Credit and the Real Economy

29

insufficiency of the previously created wealth willingly placed at the disposal of the entrepreneurs. But they were also a substitute for entrepreneurial deficiencies. From their central vantage points of control, the banks participated actively in shaping the major - and sometimes even not so major - decisions of individual enterprises. It was they who very often mapped out a firm's path of growth, conceived far-sighted plans, decided on major technological and locational innovations, and arranged for mergers and capital increases. Thus, the universal bank is seen as an instrument thanks to which the backward country can catch up with, and maybe overtake, the most advanced - Anglo-Saxon - market economies. The universal bank achieves this aim through 'far-sighted plans' which it can elaborate successfully thanks to its 'central vantage point of control' and because its 'capital supplying function' compensates for the 'insufficiency of wealth placed at the disposal of the entrepreneur'. In other words, as was believed to be the case with central planning, the universal bank is supposed to make possible a greater mobilization of savings and their successful application to long-term projects. The aim and the method remain the same, only the institutional instrument is different. The wonder is that Eastern Europe is not awash with state-supported (and owned) banks, owning large and concentrated shareholdings in non-financial businesses, and claiming to be budding Deutsche Banks! How relevant, in fact, is the German-type universal bank to the conditions and needs of PCEs? We need to distinguish between 'German-type' universal banks which are intimately involved in the ownership and control of non-financial firms, and 'non-German' universal banks, which merely provide a wide range of financial services to their clients. Whether banks in PCEs should develop along the lines of 'non-German' universal banks is not one we shall address. The answer depends on whether there are greater economies of scale or scope in the 'financial supermarket' or 'specialist bank' approaches. It has implications for the non-financial sector of the economy, mainly to the extent that greater efficiency implies lower costs of financial services, and as such it is not a problem of transition but of final destination of the banking system in PCEs, and so one which can be addressed at a far later stage in the reform process. 5. 1 The suitability of German-type universal banks in the early stages of transition

Two key facts have to be remembered about PCE banks in the early stages of transition: very low banking skills and state

30

Rnstowski

ownership. 1 7 Furthermore, medium and long-term lending is harder than short-term lending because unpredictability (risk) is greater as the term of loans increases. Low banking skills have already led to significant loan losses in spite of the fact that, outside Hungary and the former Czechoslovakia, almost all lending is short term. If PCE banks attempted to engage in a large amount of long-term lending the scale of these losses would be even greater. Yet it is the provision of long-term credit which is supposed to justify the close bank-firm relationship of the German model. PCE banks therefore need to learn to allocate credit effectively by first learning to allocate short-term, self-liquidating credit for working capital. It is only once they have mastered this that they can prudently go on to medium and long-term financing, 18 and it is only once this stage of expertise has been reached that German-style close involvement with customers can be considered. A high gearing ratio causes a conflict of interests between lender and borrower, increases agency costs and reduces investment relative to the optimum level. Bernanke and Gertler (1987) show that the higher the leverage ratio in a project, the more the interests of the borrower and lender diverge. This is because the return on equity is higher on a riskier project the higher the leverage, while the expected yield to the lender is smaller. PCEs suffer from low levels of equity capital among non-financial firms for two reasons. SOEs may have large amounts of equity, but this equity is 'weak' because stakeholders' (the state's, managers' and workers') rights to it are poorly defined. 19 Such equity can act as collateral for loans, but it does not act adequately to motivate SOE borrowers to avoid excessively risky loans, as managers of such firms will not themselves lose net worth if a project which they have borrowed to finance proves unviable. New private sector firms, in contrast, simply have little equity compared to the amount of capital they could profitably invest in projects because transition opens up a vast array of profitable new projects for the private sector (Rostowski 1993b). 1 7 In Russia and Bulgaria where many of the large banks are effectively in private hands, they do not seem to be in the hands of 'fit persons'. 18 The attempt to introduce German-type universal banking in PCEs in the early stages of transition reminds one of other attempts in the region to 'speed up history' . Like those, it would be a high risk strategy. 19 To be undertaken, a project must not only be profitable, but there must also exist a way for the stakeholders to benefit from its profitability (something which is hampered by tax-based incomes policies, for example) .

The Banking System, Credit and the Real Economy

31

Bernanke and Gertler argue further that the conflict of interests between lenders and borrowers when leverage is high increases the agency costs of the investment process, and hence reduces the level of investment below its optimum level. As we have seen, in PCEs this is likely to be the case both in the new private sector and in SOEs. This may help to account for the widespread populist pressure for 'cheap money' which one observes in the PCEs. At the heart of the German model lies the idea that close control by the bank and its access to insider information regarding its customers makes up for the lack of equity capital in the borrowing business and reduces the agency costs of monitoring loans. However, the shortage of equity in SOEs in PCEs is not absolute, but is of a special sort which we have called 'equity weakness' . SOE manage­ ments are not to be expected to be as concerned about the net worth of their firms as are managers of private firms. Although German-type access to insider information by lenders might improve the running of SOEs if the lending banks were private, the actual situation in the early years of transition is that large commercial banks are almost all state-owned, so that the equity within these state-owned commercial banks (SOCBs) is just as 'weak' as the equity in the SOEs they lend to, and bank managements cannot be trusted to maximize the net worth of their institutions. In this situation, 'fuzzy' instruments (such as long-term non-quoted debt or non-quoted equity), which make it harder to assess the value of a firm's liabilities to a bank, but which are so important in the intimate bank-firm relationship of the German model, need to be avoided until well after SOCBs have been privatized. It is clearly harder to judge whether a loan is good if it is long term than if it is short term, making long-term credit 'fuzzier' . Even more dangerous would be the ability of SOCBs to take equity positions in their customers, a key element of the German model. Non-quoted equity is the ultimate 'fuzzy' form of financing, whose commercial nature or otherwise is highly opaque and whose true value is only revealed when the shareholder exits by selling his stake. An SOCB wishing to hide the fact that it has made a bad loan would find it very convenient if it could purchase part of the borrower's equity thus enabling the servicing, or even full repayment, of the loan. By contrast, quoted equity - which is not at all 'fuzzy' as regards its value - is a highly volatile asset, putting the bank's capital base at risk. Furthermore, the concentration of both the debt and equity liabilities of a non-financial firm in the hands of a given bank increases the riskiness of the latter's exposure.20

32

Rostowski

Altogether, the suggestion that PCEs adopt a German-style relationship between banks and enterprises fails to take into account the fact that, as was stressed in Section 1, the main challenge facing SOCBs in PCEs is to shift from providing 'systemically bad' (that is, non-commercial) credit to providing systemically good credit. Everything which reduces the clarity with which a judgement can be made as to whether loans are good or not hampers this vital transformation. If PCEs were to follow the German model, a real danger is that SOCBs, which have only recently been created out of departments of the state monobank and are only finding their feet as true commercial institutions, could become the focal points of large financial-industrial holdings - which, under pressure from govern­ ments and from non-financial enterprises, would altogether fail to switch their lending policies to a commercial basis. Furthermore, it has turned out to be significantly easier to privatize non-financial SOEs than to privatize the banks in most transition economies. With major banks remaining state-owned for quite a few years after the start of the transition in many countries, bank control of the equity of private sector and privatized firms might effectively mean a degree of re-nationalization of the economy, which is unlikely to be either economically desirable or politically acceptable. Finally, German-type universal banks cannot operate in an environment of high inflation (see the next section), and as late as 1993 no PCE had achieved inflation of significantly less than 20 per cent per annum. The unsuitability of German-type universal banks does not mean that the banking system in PCEs cannot evolve in the direction of efficiency and commercialism. SOEs often have a considerable amount of net worth which can be used to colateralize loans, so that arm's length, asset-based lending - particularly in the short term - is perfectly feasible. This is the kind of lending which can form the basis for banks in PCEs to develop the skills which they need. It is worth remembering that this is the kind of lending which first appeared at the beginning of capitalism.

5.2 The German model in the intermediate phase of economic transition Let us now turn to the intermediate phase of economic transition, when it is assumed that banks already have some skill in making 20 If the equity holdings are kept in a separate subsidi , for which the ary parent bank has limited liability, financial risk is reduced. This is not, however, the German model of the universal bank but rather the French one.

The Banking System, Credit and the Real Economy

33

medium and long-term loans.21 How strong are the arguments in favour of the German model in this context? It seems likely that this phase of the transition in the banking systems of PCEs will more or less correspond to the period of resumed growth of output by non­ financial businesses. This would be the time when enterprise adjustment through the cutting back of unwanted output and the cutting out of unnecessary capacity would be giving way, in terms of its overall effects, to accelerating growth in the output of new products and in the creation of new capacity. It would thus be a period of strong and growing demand for capital. On the one hand, an advantage of the German model in this context is that the average cost of capital would be reduced (debt being cheaper than equity). On the other hand, Aoki (1984) develops a model in which, when banks are shareholders and have control of the capital structure of firms, they opt for a higher level of debt than that which maximizes the value of the shares. In this model, banks have no advantage from holding all the shares of a firm. They choose the smallest stake which gives them control over the firm's borrowing decisions. Stakes of 5-20 per cent, such as are often observed in Germany and Japan, seem to optimize banks' returns. If control can be exercised through the use of proxy votes, as in Germany, then the benefit of control can be had without the cost of having to own shares. Moreover, Steinherr and Huveneers (1992) argue that in the German model the leverage ratio of non-financial businesses is raised, not just as a result of the reduction in the cost of debt but also due to the increase in the cost of external equity. Thus the German model provides cheaper long-term debt to firms,22 at the cost of more expensive external equity. This is because managers of firms controlled by banks have considerable leeway in allocating free cash flow (what are called 'hidden reserves' in Germany). Since there are few hostile takeovers (that is, the market for corporate control is weak), there is a major principal-agent problem, so that agency costs for equity holders are high, making equity relatively unattractive to those shareholders who do not participate in sharing out the free cash flow. One of the key problems of PCEs is SOE privatization. The German banking model would thus make privatization slower, though possibly 21 This might be five to ten years into the transition. 22 Al though such bank debt is more expensive than securitized debt in Germany (Steinherr and Huveneers 1992) .

34

Rostowski

'better', as flotations would become harder, leaving only 'trade sales' to other large (presumably mainly foreign) firms as the major vehicle for the state to sell SOEs. However, the political obstacles standing in the way of privatization sales to foreigners are well known. The same problem of the high cost of equity would face firms which had already been privatized by some 'non-equivalent' method, such as voucher sale or leasing, but which needed a capital injection. The feasibility of the German model also depends on the inflation record of the PCE concerned. Henderson (1993) points to the UK's weaker record on inflation as a cause of the greater importance of equity markets in financing firms in that country. Higher inflation means higher volatility of nominal interest rates, which means higher risk to both lender and borrower, independently of whether contracts mostly involve fixed or variable rates. Given higher inflation, variable nominal interest rates may (but need not) mean more stable real rates. Even if they do, this happens at the cost of greater real amortization of a loan when the nominal rate is high, which can put considerable strain on a borrower's cash flow. At the same time, fixed nominal interest rates in an environment of high (and therefore variable) inflation, mean highly variable real interest rates. Both suppliers and demanders of capital can protect themselves from these kinds of volatility by using equity rather than debt. Thus, the close bank-firm relationship in both Germany and Japan has depended not only on a larger share of long-term credit, but also on a larger proportion of fixed interest loans than in many other countries, and this has in turn been made possible by these countries' superior inflation record. Indeed, one can argue that the determination of the authorities in both countries to maintain very low inflation has been due to their realization that their bank-based financial systems are more vulnerable than others to the disruption even moderate inflation would cause. In PCEs, one can see the likely effects of a country's inflationary history on its ability to adopt the German model by looking at the ratio of bank credit to the non-government sector as a share of GDP. In Czechoslovakia, this share was about 75 per cent, 23 in Hungary it is about 40 per cent, in Poland it is about 20 per cent, while in most remaining PCEs it is below 20 per cent. Those countries with a history of very high inflation find that the population minimizes its holdings of domestic money, which reduces the amount of real credit in the 23 The same ratio holds for the two successor states.

The Banking System, Credit and the Real Economy

35

economy.24 High budget deficits and domestic government debt also contribute to the problem, crowding out the amount of credit available to the non-government sector. Moreover, experience shows that post-stabilization remonetization of the economy although significant, remains limited (for example, in Poland the domestic money supply rose from 12 per cent to 20 per cent of GDP). Countries with low credit/GDP ratios cannot implement the German model, simply because the credit available to their banking systems will be inadequate for the external financing of the capital needs of their firms. Thus, only the Czech Republic together with - possibly Hungary and Slovakia, could be potential candidates for German-style universal banking, even in the second phase of transformation, when inflation in these countries ought to be down to single figures per annum.

6. The role of bank credit in economic transformation of PCEs: the case of Poland In Poland, bank credit has gone to those to whom it should not have gone (mainly the state sector losers) and has not gone in sufficient quantity to the winners, in particular not to the rapidly growing private sector, for whom its lack has - interestingly - not proved an unsurmountable obstacle. Gomulka (1993) has shown that, in Poland, 43 per cent of credit ended up with the 10 per cent of firms which have the worst 'sales related quick ratios'.25 Moreover, the extremely dynamic private sector has had very little access to credit. Some 29 per cent of bank credit to the non-government sector went to private businesses (excluding co-operatives) in Poland at the end of 1992.26 At that 24 Unless a parallel banking system based on foreign currency is permitted. 25 SRQR = (Cash + Receivables - Bank loans - Payments due) /three months sales. The actual share of credit going to firms in this 'very bad' group is higher, but in the text I exclude housing co-operatives, which were able to service their debts promptly thanks to budget subsidies, but which had very low sales. I am grateful to Mark Schaffer for pointing out this phenomenon. 26 I arrive at this figure by taking the official figures for credit for the private sector as a whole at end-1992 (that is, 1 22.9 trillion zlotys) and subtracting the figures given by Gomulka ( 1 993) for credit to large co-operatives at the same date ( 40.9 trillion) as well as credit to households ( 1 2 trillion) . This leaves a total of 70 trillion, or 28. 7 per cent of credit to the non-government sector, which is the upper bound of credit to the non-co-operative private sector. Whether co-operatives generally ought to be classified as part of the private sector is a moot point. In the present context, given that a very large part of co-operative indebtedness is due to housing co-operatives (see fn. 25) they ought definitely to be excluded.

36

Rostowski

Tab/,e 2. 1 . The non-co-operative private sector's share in each industry in 1 989 and in 1 993

Industryt Construction Transport

1 989

1 993*

7.4 30.0 5.9

32.3 80.0 39.2

* First three quarters of 1993. Guesstimates for the non-co-operative private sector have been obtained by assuming that the co-operative sector's share of output remained constant after the last date for which we have information concerning it, that is Ql-Q3 of 1992. The co-operative sector's share was declining up to that time in industry, increasing in construction and roughly constant in transport (Rostowski 1993b). t Manufacturing plus extractive industries. S()U,rce. Rostowski 1993b and GUS, Informacja (1993) . time, the private sector (excluding co-operatives) accounted for about 45 per cent of GDP and for 60 per cent of employment outside the budget sector. 27 What is more, a very large proportion of credit received by the private sector was misallocated. It was lent to firms which Gomulka places in his 'very bad' category, that is to say with 'sales related quick ratios' of less than -1.5. Out of 22.5 trillion zlotys lent to large private firms at the end of 1992, 16.9 trillion zlotys - or 75 per cent! - was to firms in this 'very bad' category. Among medium-sized firms, results were not quite so bad; only 50 per cent of loans went to such firms. If the share of loans to 'very bad' businesses was about the same among loans made to small private firms and sole traders, the total amount of lending which went to private sector firms not in the 'very bad' category probably amounted to some 23 trillion zlotys, or under 10 per cent of credit to the non-government sector. Given that the ratio of credit to the non-government sector to GNP was about 20 per cent at this time, the ratio to GDP of credit going to private businesses which did not fall into Gomulka's 'very bad' category was about 2 per cent. 28 It seems very unlikely that this tiny amount of 'good credit' for the private sector could have played an important role in the exceptionally rapid growth of that sector in 1990-93 (Rostowski 1993b). During this period, the non-co-operative private sector's share in various branches of the economy increased as shown in Table 2.1. 27 An d over 50 per cent of all employment. The ' budget sector' is that part of the economy financed through the government's budget. 28 If we included 'good credit' to co-operatives in this total we might double the figure to 4 per cent of GDP.

The Banking System, Credit and the R.eal Economy

37

Moreover, this very rapid growth of the private sector was only marginally due to privatization of state enterprises. In mid-1993, privatized enterprises accounted for 4.4 per cent of the revenue of all current and previously state-owned enterprises (GUS, Informacja 1993). If revenue and output shares were roughly equal, and if in mid-1993 about 45 per cent of GDP was generated in the private sector, 29 then about 2.3 per centage points of this was accounted for by privatized firms. To the extent that the state sector has been privatized, it is through the privatization of its physical assets, and not through the privatization of organized state-owned businesses. In this process which has been the most important element in the transformation of the Polish economy - bank credit has played a very small role. Indeed, on the basis of the Polish case one might be able to formulate a hypothesis about the microeconomic unimportance of bank credit in the early stages of transition in a PCE. What, then, are the actual sources of capital accumulation and economic development in this early phase of economic transition, and what mode of financing is appropriate to them? It seems that in Poland, since 1989, private sector expansion has resulted from the · ploughing back of retained earnings in existing firms and the establishment of new private sector firms. The market as a selection process has operated through the goods market rather than through the capital market. Successful firms have made large profits and thus been able to expand; unsuccessful firms have failed and gone bankrupt. In this way, not only successful projects, but also successful entrepreneurs and successful business organizations have been selected, and the information capital of the whole economy under the radically new circumstances of the transition has been increased (Atkeson and Kehoe 1993). It is difficult to imagine that these selection processes, which are vital if the new capitalist economy is to function successfully in Central and Eastern Europe, could have been undertaken by capital markets, given the very low quality of publicly available information regarding projects and the almost complete absence of information regarding entrepreneurs and organizations at the beginning of the transition.

29

Allowing 5 per cent of GDP to be generated in the co-operative sector.

38

Rostowski

7. Conclusion: Building appropriate f'mancial systems in post-communist economies What are the lessons from five years of economic transition in PCEs as regards the banking sector? The first is surely that, given how little banking systems have contributed to transformation and economic restructuring in Central Europe, the main motto should be 'safety first'. But safety against what? The first threat is that of bad debts in the banking system. For old bad debts, the solution is either a complete debt write-off, combined with the introduction of 100 per cent reserve banking for banks which benefit from the write-off, or a Polish-style 'Enterprise and Bank Financial Restructuring Project' (see Chapter 4). For new bad debts, the answer is very high capital and/or reserve ratios in those parts of the banking system which benefit from implicit deposit insurance. Although such requirements are often thought of as a tax on banking services, they can also be considered as a premium paid for implicit deposit insurance, or as a requirement on the behaviour of the insured to reduce the insurer's (government's) exposure. 30 Of the two approaches, high capital ratios make less sense as long as bank capital is state-owned (unless, as in Poland, capital adequacy is a criterion for SOCB privatization, and bank managements are very interested in privatization), and as long as bank auditing and supervisory skills are rare (valuation of bank equity requires realistic loan classification). High reserve ratios are easier to apply, and as one approaches 100 per cent reserves on deposits the need for high levels of capital to provide a cushion against bad loans disappears. SOCBs should initially concentrate on developing their payments services, which it can be argued, are the most basic and important services they supply in a newly marketizing economy. The first liberalization would then be to allow them to extend the safest kind to credit available (namely, that for self-liquidating working capital). As skills develop, the kind of loan business permitted would become more varied and longer term, obligatory reserve ratios would be reduced while, initially, capital adequacy ratios would need to be increased (to provide a cushion for greater loan risk, and to prepare the SOCB for privatization). 31 30 Analogous to the requirement for fire doors in buildings covered by fire insurance. 31 Subsequently, after privatization, capital ratios could decrease from what might have been temporarily very high levels indeed (say 25 per cent).

The Banking System, Credit and the Real Economy

39

Secondly, there is the question of severe moral hazard developing as a result of successive recapitalizations of the banking system by the state. Much will depend on the success of the Polish EBFRP (Kawalec, Sikora and Rymaszewski 1993). If it fails to avoid the moral hazard problem, this will strengthen the arguments for very narrow banking in the first period of the transition. If it succeeds, this will strengthen the view that bad loan losses resulting from excessively 'wide' banking during early transition, though unfortunate, are a mistake which can be successfully corrected at a later stage. Of course, such a gradualist/ conservative approach to banking reform suggests that there is very little scope for the development of German-style universal banks in PCEs for many years to come. This is because the danger of loose financial control of borrowers seems to be augmented when banks become universal while they are still dominated by state-owned equity and have developed few skills even for short-term credit allocation, let alone for the long-term credit allocation and equity finance of clients. Moreover, German-type financial systems require very low rates of inflation indeed, something which is unlikely to be achieved in PCEs for as much as a decade after the beginning of the transition. The third danger is that of a high level of politicization of credit allocation (this danger persists as long as most banks are state-owned, which remains the case in Central Europe outside the Czech Republic).32 Again, high reserve ratios in the SOCBs reduce the scope for such distortions. Furthermore, if SOCBs were largely limited to providing payments services and some limited credit for working capital, they might be easier to privatize.33 Finally, if the SOCBs functions are limited in this way, their resistance to the entry of foreign banks (which would be kept out of the payments business) might be reduced.

32 Even here the National Property Fund owns up to 40 per cent of the shares of the leading Czech banks. 33 See Chapter 3 for a description of how such 'narrow banks' (or 'Money Warehousing and Payments Companies' as they are called in that chapter) might generate income.

40

Rostowski

References Aoki, M., 1984, The Ecanomic Analysis of theJapanese Firm, New York, North- Holland. Atkeson, A. and Kehoe, P., 1993, 'Industry Evolution and Transition: The Role of Information Capital', mimeo. Begg, D. and Portes, R., 1991, 'Enterprise Debt and Economic Transformation: Financial Restructuring in Central and Eastern Europe', mimeo. Bernanke, B. and Gertler, M., 1987, 'Financial Fragility and Economic Performance', Princeton University, Woodrow Wilson School of Economics, mimeo, March. Calvo, G., 1991, 'Financial Aspects of Socialist Economies: From Inflation to Reform', in Reforming Central and Eastern European Economies: Initial Results and Challenges, eds. V. Corbo, F. Coricelli and ]. Bossak, World Bank. Calvo, G. and Coricelli, F., 1993, 'Output Collapse in Eastern Europe', IMF Staff Papers, Vol. 40, No. 1, March. Corbett, J. and Mayer, C., 1992, 'Financial Reform in Eastern Europe: Progress with the Wrong Model', Oxford Review ofEconomic Policy. Fry, M., 1988, Money, Interest, and Banking in Economic Development, Johns Hopkins University Press. Gerschenkron, A., 1968, Continuity in History and Other Essays, Harvard University Press. Gomulka, S., 1993, 'The Financial Situation of Polish Enterprises 1992-3 and its Impact on Monetary and Fiscal Policies', paper presented to the IIASA Conference on Output Decline in Eastern Europe, Laxenburg, Austria. GUS, Informacja o Sytuacji Spokczno-Gospodarczej Kraju, Ql-Q2 1993, Warsaw. Henderson, R., 1993, European Finance, McGraw-Hill. Kawalec, S., Sikora, S. and Rymaszewski, P., 1993, 'Dealing with Bad Debts: The Case of Poland', paper presented to the IMF and World Bank Conference on Building Sound Finance in Emerging Market Economies, Washington, DC. Lipton, D. and Sachs, J., 1991, 'Privatization in Eastern Europe: The Case of Poland', in Reforming Central and Eastern European Economies: Initial Results and Chalknges, eds V. Corbo, F. Coricelli and ]. Bossak, World Bank. McKinnon, R., 1992, 'Taxation, Money and Credit in a Liberalizing Socialist Economy', in The Emergence of Market Economies in Eastern Europe, eds. C. Cleague and G. Rausser, Blackwell. McKinnon, R., 1986, 'Domestic Interest Rates and Foreign Capital Flows in a Liberalizing Economy', Stanford University Department of Economics, mimeo. National Bank of Hungary, 1993, Annual Report 1 993, Budapest. Nyers, R. and Lutz, G., 1993, 'Development of the Financial Sector in Hungary During the Transition Period', Budapest, mimeo. Rostowski, J., 1993a, 'Problems of Creating Stable Monetary Systems in Post­ Communist Economies', Europe-Asia Studies, May. Rostowski, J., 1993b, !The Implications of Rapid Private Sector Growth in Poland', Discussion Paper No. 159, Centre for Economic Performance, London School of Economics.

The Banking System, Credit and the Real Economy

41

Schmieding, H., 1991, 'Transforming the Financial System in Eastern Europe's Emerging Market Economies', Working Paper No. 497, Kiel Institute of World Economics. Steinherr, A. and Huveneers, C., 1992, 'Institutional Competition and Innovation: Universal Banking in the Single European Market', in European Banking, ed. A. Mullineux, Blackwell. Stiglitz, J. and Weiss, A. , 1981, ' Credit Rationing in Markets with Imperfect Information', American Economic R.eview, Vol. 71, No. 3, June.

CHAPI'ER 3

FINANCIAL SECTOR DESIGN, REGULATION AND DEPOSIT INSURANCE IN EASTERN EUROPE Amoud W.A. Boot and Sweder van Wijnbergen

With the exception of the former republic of Yugoslavia, regular commercial banks did not exist anywhere in Eastern Europe before as recently as 1987. Since then, banks have been created at breakneck speed; Poland has over ninety banks, Russia around 2,000, and so on. Every country in Eastern Europe now has a fully fledged two-tier banking system in place. In most cases, the larger banks are spin-offs from regional offices of the central bank, such as in Poland and the Czech Republic. This core of relatively large commercial banks is in many countries supplemented by a fast growing group of smaller, privately owned banks. But with emerging banks have come all the problems that plague banking systems in other parts of the world: unserviced loans, fraud, financial scams, panics, taxpayer-financed bail-outs, etc. While the problems may be familiar, the circumstances within which they have to be tackled are not. The regulatory framework and supervision mechanisms are almost everywhere in their infancy; trained personnel are lacking, both in the banks and in the regulatory agencies, and the legal framework within which contracts need to be enforced is often unclear and unfinished. Moreover, the uncertain environment, lack of a civil service tradition and severe decline in income which characterize the region has triggered serious problems of corruption and fraud, problems to which the financial sector by the nature of its business is singularly vulnerable. All this is serious; if the rest of Europe is anything to go by, banks much more than stock markets - will eventually have to be the main suppliers of funds to enterprises new and old. Banks will have to play a major role in enterprise governance; and banks will remain the main outlet for private savings for years to come. Thus, continuing

Financial Sector Design

43

fragility and crisis in the banking system could well develop into a major obstacle to the successful industrial development of Eastern Europe. Hence the importance of the issue of financial sector design, the topic of this chapter. The objective is clear in principle: a regulatory and legal framework that will allow banks to channel savings to enterprises efficiently and play their role in corporate governance, while minimizing the vulnerability of the system to fraud, corruption and financial crisis. The tendency has been simply to mimic institutions and rules from the West, without much thought about whether that is in fact appropriate. Yet it is far from clear that simply copying Western Europe's financial structure is the right way to go. To begin with, the whole system itself is increasingly being questioned in the West, as a spate of major banking crises has taken place or is threatening to take place in countries as diverse as the United States, Norway, Israel and Britain. Equally important is the fact that the circumstances in which the financial sector operates in Eastern Europe differ from the West in ways that have an immediate bearing on issues of financial sector design. Information problems loom much larger, with more dramatic change taking place on the borrowers' side while at the same time information systems are much less developed. The prevalence of state enterprises create incentive problems of its own; moreover, the shortage of skilled and experienced bank supervisors and auditors is extreme. It is against this background that we wish to raise the issue of financial sector design in Eastern Europe, and in particular the Former Soviet Union (FSU). While many of the problems in Central Eastern Europe (CEE) are similar, developments have gone faster there than in the FSU and many choices still open in the FSU have already been foreclosed by decisions made in the CEE region. Deposit insurance is of course a key question. With external capital markets inaccessible for many, and insufficiently accessible for all, domestic savings may well be the key to future growth. Providing a safe outlet for them is accordingly essential, but the issues go beyond the question of deposit insurance. Given the difficulty of implementing proper regulation in East European circumstances, is it possible to design a financial structure that greatly reduces the need for regulation? Those are the questions we will at least begin to discuss in this chapter. In the first part of this chapter, we start with a brief survey of the role of the financial system (Section 1.1). We then survey some of the

44

Boot and van Wijnbergen

recent problems in Western banking (Section 1.2 ) . This part of the chapter ends with a discussion of issues at stake in financial system architecture, and the implications for the design of regulatory mechanisms (Section 1.3 ) , against the background of the problems in Western banking laid out in the preceding section. The second part draws the implications for Eastern Europe. We first provide a brief summary of financial sector developments in Eastern Europe (Section 2.1) . The next section sets out some principles for financial system design in Eastern Europe and in particular the FSU and elaborates on a recommended concept, narrow banking. Section 2.3 concludes.

1.

Issues in the design of financial systems

1 . 1 The rok of the financial system: stability and competitiveness as joint objectives The primary function of the financial system is to facilitate the transfer of resources from savers to those who need funds. The objective is to have an efficient allocation and deployment of resources. Efficiency in this context is interpreted broadly and presumes both stability and competitiveness of the financial system. Stability is needed to guarantee the orderly flow, allocation and deployment of resources. It is generally recognized that fragility of the financial system comes at great cost, since disruptions have potentially severe consequences for the economy at large. An efficient financial system should also minimize transaction costs, interpreted broadly as resources that dissipate or evaporate in the process of allocating resources. This generally necessitates a certain degree of competitiveness. Yet stability and competitiveness are very likely to be conflicting rather than complementary objectives, thus presenting regulators with a difficult trade-off. In the popular view, restrictions on competition would improve banks' profitability, reduce failure rates and hence safeguard stability. The special status of banks has been called into question after a decade of record-breaking inflation and interest rates, and developments in information technology have undermined banks' protected franchises. These developments made existing regulatory constraints increasingly costly for banks, and simultaneously spurred the growth of non-banking financial institutions, which could largely circumvent the regulatory constraints. This poses an important challenge: how to design a sustainable regulatory environment in

Financial Sector Design

45

banking? We first focus on some of the lessons from the evolution of Western banking. Subsequently, we seek to draw some implications for the regulatory structure. 1 . 2 Evaluation of the competitive environment of Western banking

There are striking variations in the configurations of financial systems between the different countries in the West. In some countries, such as the United States and Britain, financial markets are very important for the allocation of resources, while in others, such as most continental European countries, banks have played a more prominent role and financial markets are poorly developed. Likewise, in many countries banks do not hold major equity stakes in industrial companies, while in some countries, notably Germany, banks are among the largest shareholders. These differences have a long history and could be purely coincidental, but, more likely, depend on the evolution of the industrial structure in each country. Government involvement could explain some of these differences. This is particularly true in the United States, where rigid regulatory structures have fragmented its banking system. The US regulatory structure is characterized by a government­ sponsored deposit insurance system, a separation of investment banking and commercial banking, and pervasive entry barriers including limitations on the establishment of branches both within each state and outside the home state. This structure dates bank to the 1930s and is contained in the Banking Act of 1933, also known as the Glass-Steagall Act. Complementary legislation sought to reduce competition even further. In particular, regulatory caps, known as Regulation Q, were for a long time imposed on deposit rates. The three pillars of this Act - federal deposit insurance, restrictions on bank business and entry barriers - guaranteed stability for over forty years. Recent environmental and competitive changes have, however, disturbed the balance provided by the Glass-Steagall Act. The volatile environment made regulatory caps on deposit interest rates too costly for bank depositors, prompting the diversion of savings to the largely unregulated money-market mutual funds that offered competitive interest rates. This forced banks to borrow at more costly market interest rates and was a real threat to the banks' protected franchises. Their traditionally best customers increasingly sought access to equity and bond markets, elevating the risk of the banks' remaining clientele. Higher and more volatile funding costs also coaxed the banks into the business of writing off-balance sheet guarantees and trading all manner of financial derivatives.

46

Boot and van Wijnbergen

Collectively, these changes elevated the banks' risks in virtually all aspects of their business. Advances in information technology facilitated the circumvention of regulation and tilted the competitive advantage away from the 'opaque' financial institutions, such as deposit takers and insurance companies, towards more 'transparent' intermediaries such as mutual funds, and also to direct financing in the capital markets. As a consequence, there has been a proliferation of specialized non-bank financial institutions. While oligopolistic practices - see our comments next on Europe may temporarily hide the competitive deterioration of traditional banking institutions, they soon need to face the new realities. The same is true for the regulatory framework. Under the earlier bank­ government nexus, public regulation inhibited both the establishment of new banks and the demise of failing institutions. The latter is still the case as governmental deposit insurance continues to deter bank failures under the banner of protecting depositors. This can no longer be sustained. With the rapidly decreasing costs of computing and communicating, all manner of non-bank financial institutions are successfully encroaching on the banks' traditional markets. Artificial life-support measures and the preservation of inefficient operations are becoming increasingly costly. With some notable exceptions, such as the Scandinavian countries, other West European countries were spared the banking turmoil. European banks are better diversified, both geographically and functionally, than their US counterparts. They typically operate nationwide, often have substantial cross-border operations, and engage in both commercial and investment banking activities. In addition, the greater concentration among European banks probably facilitates collusive pricing and other oligopolistic practices. Thus, Europe has not yet faced the unbridled competitive pressures that characterize US banking. Moreover, the most recent consolidation and widening of activities among European banks - especially in Spain, Denmark and the Netherlands - can be seen as a pre-emptive response to the threat of increased foreign competition. As a result, the market share of many European countries has reached unprecedented levels with the larger institutions absorbing smaller and often more specialized ones. This means that West European banks have not yet faced the entire effects of a more competitive environment and the imminent dissolution of monopoly rents. 'Europe '92' is likely to imply that they will, however, with all the attendant problems sketched so far.

Financial Sector Design

47

1 . 3 Deposit insurance: rationale and implications The regulatory interference that characterizes banking suggests that banks are considered 'special' or different from other firms. Obviously, regulation has made them special. But what is different about their operations that justifies this 'special' regulatory treatment? This question needs to be addressed before we can derive the structure of the optimal regulatory response, if any. A starting point is the observation that banks typically have a very fragmented deposit base; bank debt ('deposits') is typically held by many different actors, none of whom holds a very large fraction of the total debt of the bank. This creates a gap in corporate governance; while equity holders have sufficient incentive to monitor the managers when times are good, they do not have so when times are bad, since the benefits of monitoring and imposing corporate governance would in those cases mostly accrue to debt holders. With a normal debt structure, the latter fact will be enough of an incentive for debt holders to start monitoring management. However, with a very fragmented deposit base, obvious free-rider problems would prevent an active role of debt holders in monitoring to emerge. Thus, one should expect bank managers to engage in excessively risky behaviour when times are bad, as the fragmented nature of the deposit base destroys corporate governance mechanisms in those situations (Dewatripont and Tirole 1992). The special - fragmented - nature of bank debt highlights the lack of corporate governance. It is widely believed that the potential fragility of banks stems from another feature of bank debt, that is, their vulnerability to runs rooted in the withdrawal-upon-demand and sequential service constraint features of the deposit contract. The fear is that excessive withdrawals would force a bank to liquidate assets and incur substantial liquidation costs, which would undermine the bank's ability to honour its remaining deposits. The excessive withdrawals could be triggered by concern about the bank's well-being. However, the bank's demise could then become a self-fulfilling prophecy: once a depositor thinks that others will withdraw, he will withdraw too. This is due to the sequential service constraint feature of deposits: it pays to be first in line. Some have argued that these runs may trigger a system-wide collapse or a panic if runs are contagious. The potential vulnerability to runs of deposit-funded banks, and the banking system's vulnerability to panics, is often used as a reason for

48

Boot and van Wijnbergen

regulation, in particular deposit insurance (Diamond and Dybvig 1983). It is generally thought that private arrangements are beset with free-rider problems and, therefore, could not cope with these problems. So most countries have enacted lender of last resort and deposit insurance (DI) arrangements which guarantee that banks and certain other credit institutions can meet their commitments to depositors. As long as the insurance system is credible and fully guarantees each depositor's funds, bank runs will not materialize. But deposit insurance, while safeguarding depositors, widens the gap in corporate governance; depositors do no longer have any incentive to monitor the bank. It therefore exacerbates the problem of excessive risk-taking by bank managers since only the taxpayer the ultimate financier of the DI system - bears the consequences of any increase in down-side risk. Thus the existence of DI necessitates further regulation, in particular on the lending side to contain the risk-taking incentives. These arguments quite convincingly explain why extensive deposit guarantees - as observed throughout the world - have induced governments to severely regulate the banks' operations. Banks face restrictions on their activities, minimum capital and liquidity requirements and are subject to extensive examinations and supervision. The moral hazards created by a fixed-rate, risk-insensitive deposit insurance system are widely acknowledged. There also seems considerable support for the notion that these incentives have contributed to the financial crises as experienced in Western banking. However, this consensus seems at odds with the apparent stability of DI arrangements for most of the 1935-80 period. Various authors, see for example Keeley (1990), argue that the tendency towards risk was restrained for almost half a century by economic rents earned in banking. In recent decades, however, rents were eroded - witness the decade of inflation and volatile interest rates, as well as the technological advances alluded to earlier. This exposed the latent design flaws in deposit insurance. On a more fundamental level, we may conclude that a system of deposit insurance distorts the relation between a bank and its providers of funds. In particular, it reduces or undermines market disdpline. Depositors, knowing that their funds are insured, will feel little inclination to monitor their investment by evaluating the banks' activities. While, as we have emphasized, depositors are generally small and may not have a sufficient economic incentive to monitor even in the absence of deposit insurance, it is likely that in a world without deposit insurance market-rooted solutions would develop to

Financial Sector Design

49

facilitate monitoring. There would also be a real urgency because without these solutions funding might not be forthcoming. One potential solution is the concentration of ownership of debt. This could make (costly) monitoring economically viable. Another solution is rooted in the banks' incentives to develop a reputation; that is, credibility in the funding market. A sufficient reputation could convince the market that a bank would not exploit problems of non-transparency and moral hazard. The bank would then benefit and obtain funds of a lower rate. Once a reputation is established, a bank has a powerful incentive to behave prudently and thus preserve its reputation. While we have linked this mechanism to the funding of banks, it is generally applicable to all firms that seek access to the financial markets. For banks, however, the mechanism has been suppressed! The important observation is that the banks' reliance on deposit insurance fixes their costs of funds at the risk-free rate, and also guarantees the availability of funds. Reputation does not lessen the banks' costs or improve the availability of funds, and the banks' incentives to develop good reputations would accordingly be diminished. Their prudent operation would then depend primarily on the regulatory design. The conclusion is that monopolistic benefits provided banks with compelling incentives to follow low-risk strategies, despite the presence of deposit insurance. Market discipline was not necessary, and regulation and supervision were only of secondary importance; rents were the primary defence against moral hazard! With the disappearance of rents, rigid regulatory structures like the Glass-Steagall Act in the United States were subject to unique challenges. The viability of the financial system became fully dependent on regulation and supervision. As has become clear, the regulatory agencies could not cope with this task. The questionable sustainability of Western banking structures, held in check only by regulations, suggests that Eastern Europe should not necessarily seek to implement or imitate the Western banking structures as they exist today. Recent experience in Western banking suggests that the tradition of guaranteeing stability by reducing competitiveness is no longer viable. It is thus important to (re)examine the issues of competitiveness and stability. The objective is to design a banking structure and regulatory framework whose sustainability is minimally dependent on regulation and supervision; an objective that is even more urgent in Eastern Europe with its as yet shaky legal and regulatory structure.

50

Boot and van Wijnbergen

There are other reasons as well why Eastern Europe should not solely focus on Western banking models. In particular, it is important to realize that financial systems may well be history-dependent, namely, depend on the development of each country's industrial structure. This may also explain the striking variations in Western financial structures. Not surprisingly, it is therefore impossible to rank different financial systems in terms of efficiency or performance; more than one financial system might be optimal at the same point in time, and optimal configurations of financial systems may change over time.

2. Financial sector reform in the FSU 2. 1 Banking in Eastern Europe In communist Eastern Europe, the government received anci dispensed all moneys, and allocative efficiency was not a matter of concern. Thus there was no need for Western-style commercial banks, and no such banks existed (except in Yugoslavia, a maverick communist country in those days). A monolithic central bank, usually assisted by a savings bank to collect deposits, undertook both central bank functions (which were minimal in the absence of free prices and open international exchanges) and extended credit to industry. Almost immediately after the collapse of communism, and in some cases even preceding it (as in Poland), two-tier banking was established, typically by splitting off the regional branch offices and turning them into commercial banks. In much of Central Eastern Europe, central banks take their newly confined role seriously and refrain from direct lending to the enterprise sector. However, in most of the former Soviet Union and South Eastern Europe, central banks continue to hand out loans to favoured industrial or agricultural enterprises, often at rates far below inflation. This happens sometimes directly, as in the Ukraine, and sometimes indirectly through blanket rediscounting of commercial loans or through unconditional recapitalization of commercial banks endangered by poor loan servicing of state enterprises. The newly created commercial banks typically inherited a loan portfolio full of loans to troubled state enterprises. Although in some countries a bout of hyperinflation following price liberalization wiped the slate clean, existing customer relationships and government pressure conspired to recreate the same problem in a matter of years. Polish banks, where exactly such a sequence was played out, had

Financial Sector Design

51

about 30 per cent of their loans classified as bad or doubtful by international auditors as late as June 1992, although triple digit inflation in 1989 had pretty much provided them with a fresh start. More recently tighter owners' control over managers seems to have reduced the problems somewhat, but other countries are reported to be in worse shape (Dittus 1993). Widespread bankruptcies have not occurred (in many countries, of course, there is still no functioning bankruptcy mechanism). Central banks have as yet not imposed the loan classification schemes and capital adequacy requirements that would bring insolvency to light. Insolvency of private and/or public banks would mainly put corporate depositors at risk as most countries' household deposits still largely flow through a dedicated savings bank. Typically this savings bank is the only bank with a national network of retail outlets as it was the designated recipient of such deposits under the communist regime. This structure is increasingly diluted as commercial banks expand; but for example in the Ukraine, on the liability side of commercial banks, household deposits make up less than 10 per cent of corporate deposits. Considering the presence of (limited) deposit insurance for household deposits, solvency problems could thus have more serious public finance consequences in Central Europe than in the FSU. 2. 2 A regulatory structure for banking in Eastern Europe

The question that has occupied much of the Western literature on banking in Eastern Europe, whether Eastern Europe should go to universal banking or not, has become largely irrelevant as countries across the board chose that model. This poses obvious problems for the regulators, to which we will come back; but given that capital markets are likely to play a limited role for quite some time to come, if not indefinitely, allowing the banks a freer role in establishing corporate governance and restructuring has much to say for it and this becomes a great deal easier under the banking model chosen (cf. van Wijnbergen (1993) for a strong argument in this direction). But with that choice behind us, a question immediately arises, the answer to which will take the remainder of this chapter. Given the clear regulatory problems created by universal banking, and banks thus exposed to more risks and commensurately more difficult to regulate, should deposit insurance be adopted, as it universally has been in the West? Deposit insurance exists de facto for the larger commercial banks in most of the countries concerned as they are still owned either by the state or, as in the Ukraine, by state enterprises.

52

Boot and van Wijnbergen

As private banks expand and state banks are privatized, like in the Czech Republic and, haltingly, in Poland, should the state protect depositors in the private banks from the consequences of bank failure? Whichever way the answer to that question goes, a strong case can be made for a substantially larger direct element in bank regulation than can currently be found in Western Europe. 1 Direct regulation could stipulate that insurance activities and corporate restructurings be placed in separate subsidiaries, which will then fall under specialized regulatory agencies where necessary. Similarly, one could require that high-risk activities not be undertaken with insured deposits. This is in fact already the de facto situation in the FSU (and, until very recently, also Central Eastern Europe); as we already saw, most household deposits flow through a savings bank which typically only invests in government or central bank paper, or in other commercial banks. Most commercial banks rely largely on enterprise deposits and the central bank for their funding. This situation would considerably reduce the need for deposit insurance, and also reduce the contingent liability (the exposure) of the deposit insurer to a bare minimum. We identify at least two reasons for this. The first is that banks which do engage in risky activities have a much more concentrated deposit base than would be the case with household deposits. For one thing, the savings banks themselves - we envisage several of these in the future 2 - will become major depositors. Such a higher concentration is likely to greatly reduce the problems that make deposit insurance desirable to begin with. After all, the reason why a bank manager might be taking undue risks is that with a fragmented deposit base no individual depositor has enough of an incentive to monitor the management; and even if he had, free-rider problems would likely prevent him from doing so. 3 But a large depositor may be able to appropriate enough of the benefit of monitoring management to make the effort worthwhile; so a concentrated depositor base would greatly reduce the need for 1 See Boot and van Wijnbergen ( 1 994) for an elaboration. These could be spinoffs from the current savings bank, or newly created ones. 3 If any given depositor monitors management, all other depositors can 'free ride' on that effort; as a consequence no individual depositor may in the end invest in monitoring activity. This is known as the 'free-rider problem' in economic theory. 2

Financial Sector Design

53

deposit insurance. Similarly (and this is the second reason), the monitoring and discipline induced by a more stable and concentrated deposit base would mitigate the problem of bank runs. The reason is simply that with a concentrated deposit base co-ordination among depositors is easier and depositors are better informed about each other. Bank runs can be excluded altogether by relinquishing the special features of the deposit contract that give rise to them. Thus having savings banks as major players in the inter-bank market might well be a substitute for deposit insurance. Another consideration is that deposit insurance would then be granted to institutions - the savings banks - that are truly diversified; that is, deposit insurance rests with institutions that could diversify their funds over various, sometimes only regionally active, commercial banks. This would reduce the contingent liability of the deposit insurer and make it insensitive to regional economic downturns. As to implementation, the critical role of the savings banks presumes that the savings banks themselves should not be allowed to engage in risky activities; otherwise their management is subject to the very incentive problems that this whole arrangement is set up to solve. Thus cementing the current situation in much of the FSU in direct regulation would make the imposition of deposit insurance for commercial banks unnecessary and would thereby greatly reduce the regulatory burden on the government. Deposit insurance could be granted to the savings banks but their restricted activities would con­ tain supervisory checks as well as the exposure of the deposit insurer. Such a situation would create two categories of financial institutions, the commercial banks and the savings banks. The commercial banks would operate with little regulatory interference, but would not qualify for deposit insurance, while savings banks could obtain deposit insurance but would face substantial regulatory restrictions. We can summarize the minimum requirements of the structure that we envisage as follows: (1) restricting household deposits to the savings banks, or, at least, restricting deposit insurance to these banks; (2) requiring the savings banks to invest mostly in high grade assets, which for the time being means government paper and lending to commercial banks with a substantial capital base; (3) not allowing commercial banks to take household deposits, or, at least, publicly refusing deposit insurance to such deposits outside the savings bank.

54

Boot and van Wijnbergen

This set of proposals looks very much like a variant of the 'narrow banking' proposals currently debated in the Western finance and banking literature. Such a structure has a number of clear advantages. First, by putting distance between household deposits and risk activities, it reduces the incentive problems that may lead to excessive risk-taking by bank managers. It therefore reduces the need for deposit insurance and the associated need for lender regulation. A narrow banking structure along the lines proposed here will greatly reduce the regulatory burden on the governments in Eastern Europe. Moreover, it will keep incentives intact for banks to build reputations for reliability. Such incentives are destroyed by deposit insurance; after all, full deposit insurance makes all deposits equally attractive for depositors irrespective of the bank taking them. So no cost advantages can be obtained by building up a reputation for sound banking and any efforts in that direction will thus be greatly reduced if DI is introduced. This would be an unfortunate development since in many East European countries efforts to build up a reputation are clearly under way in a significant subgroup of banks. Most of the nine larger commercial banks in Poland have made strenuous attempts to upgrade the quality of their portfolios, as have the privatized banks in the Czech Republic. Similarly, Lamdany (1993) reports the emer­ gence of a group of about thirty medium-sized banks in Russia ( out of a total of about 2,000) that clearly are trying to signal their reliability through restricting their risk activities. Deposit insurance would cut such developments short. There are two potential disadvantages, both of which can ( and shoula) be accommodated. The first problem stems from the restrictions banks will face in their funding possibilities. Once large­ scale privatization takes place and more regular manufacturing activities pick up, expansion based on the private sector will need bank finance. Household bank deposits will most likely remain the predominant source of private savings for a long time to come. In this way, arrangements that would exclusively channel household deposits to the savings banks could become a serious hindrance for private sector industrial development. Developing an inter-bank market to which the savings banks could lend is thus an essential complement to the proposals made here for narrow banking. Such a mechanism would still keep savings banks' managers at arm's length from direct risk activities, which after all would reside with the commercial banks. It would add a major

Financial Sector Design

55

depositor to the commercial banks' deposit base, thus lessening the free-rider problems associated with a fragmented depositor base. Limiting access to such a market to properly capitalized banks will add further incentives for prudent management for bank managers. A second problem will be clear to anybody who has had to deal with institutions like the Ukrainian or Russian savings bank. Since they had an entirely captive clientele, customers were at their mercy. Competition on the deposit side clearly needs to be introduced to remedy this situation. We envision the creation of several savings banks. Commercial banks could also participate by setting up American-style money market funds. These funds channel deposits into government paper and other high investment grade assets, in a legal structure that protects those assets from problems in the rest of the bank (this also means that banks should not be allowed to use those assets as collateral, for example). This would make them into safe deposits without the need for deposit insurance. Banks would earn management fees and so have an incentive to compete with the savings banks for deposits. Such institutions would also help to create markets for Treasury paper, a market of immense macroeconomic importance. Moreover, this structure would bring the financial system in the FSU closer to the Western concept of a narrow bank, which envisages a symbiosis of safe and risky elements in one bank. In the future such restrictions on household deposits could be lifted, at which time the countries involved could evolve towards a traditional Western banking system; a necessary precondition for such a change is that the regulatory and supervisory institutions are fully staffed and capable of their task. Alternatively, the banking system could evolve towards the narrow banking concept Western-style. This would also involve lifting the restrictions on household deposits, but would require banks to separate their risky and safe activities and only use funds from household deposits to fund safe activities unless sufficient credit enhancement has taken place. However, the required degree of financial sophistication for the latter operation does not yet seem to exist currently in Eastern Europe.

2.3 Conclusions At the very moment when Eastern Europe seems set to follow the West in its financial market regulatory design, a series of crises in the United States, Scandinavia, Spain and elsewhere has led many to question its adequacy. In this chapter, we have presented the

56

Boot and van Wijnbergen

analytical issues in financial systems architecture so as to shed light on the problems that have been noted recently in the West and draw conclusions for regulatory design in the East. An analysis of recent developments in the OECD suggests disconcerting conclusions for the trade-off between stability and competition in financial markets. As financial innovation and regulatory reform have swept most countries outside Western Europe, crises have followed; many observers believe that Western Europe has been spared these developments only because competition policy has yet to reach the banks. A comparison of Western Europe with the rest of the OECD suggests that monopoly rents rather than effective regulation have been the main force behind financial stability: protecting future rents is apparently a powerful force for conservatism. The apparent difficulty in regulating a truly competitive banking system leads us to commend a variant of the narrow banking concept. Such banking structures attempt to put some distance between household deposits and the funding of risky activities. The banking structure that in particular still exists in much of the FSU in fact makes that very easy to implement: with most of the deposits still channelled through a specialized savings bank, narrow banking takes place de facto if not de jure. Where household deposits are used to fund risk activities only indirectly, we show how the burden on regulatory authorities may be greatly reduced by essentially eliminating the incentive problems that gave rise to the need for regulation to begin with. Whether in the long run the variation on the narrow banking concept as proposed in this chapter will evolve towards a traditional Western banking system will depend on developments in the East and in the West; but since the concept is de facto implemented at present in most of the FSU, we strongly recommend that that structure be maintained until more competent supervision than currently exists has been put in place and shown to work.

References Boot, AW.A. and Greenbaum, S.I., 1993, 'Bank Regulation, Reputation and Rents: Theory and Policy Implications', in Capital Markets and Finandal Intermediation, eds. C. Mayer and X. Vives, Cambridge University Press. Boot, AW.A. and Thakor, AV., 1993, 'Self-Interested Bank Regulation', American Economic Review 83, 206-12.

Financial Sector Design

57

Boot, A.WA. and van Wijnbergen, SJ.G., 1994, 'Banking Reform and Regulatory Design Issues in Eastern Europe', mimeo, University of Amsterdam. Dewatripont, M. and Tirole, J., 1992, 'A Theory of Debt and Equity: Diversity of Securities and Manager-Shareholder Congruence', mimeo, Universite Libre de Bruxelles. Diamond, D. and Dybvig, P., 1983, 'Bank Runs, Deposit Insurance and Liquidity', Journal of Political Economy 91, pp. 401-19. Dittus, P., 1993, 'Finance and Corporate Governance in Eastern Europe', mimeo, Bank for International Settlements. Keeley, M.C., 1990, 'Deposit Insurance, Risk and Market Power in Banking', American Economic Review 80, 1183-200. Lamdany, R., 1993, 'Gradualism in Banking Reform', paper presented to the Sapir conference, Tel Aviv University. Wijnbergen, S. van, 1993, 'The Role of Banks in Corporate Restructuring: The Polish Example', Working Paper No. 898, Centre for Economic Policy Research, London.

CHAPTER 4

LESSONS FROM BAD LOAN MANAGEMENT IN THE EAST CENTRAL EUROPEAN ECONOMIC TRANSITION FOR THE SECOND WAVE REFORM COUNTRIES Kalman Mizsei 1. Introduction This chapter explores the Hungarian and Polish governments' treatment of the 'bad loan portfolio' of their major commercial banks so as to draw up policy lessons for the latecomer countries to the economic transition, i.e. those countries of Eastern Europe which for one reason or another have not yet taken up prudent macroeconomic policies and have not yet addressed the problem of insolvent enterprises. 1 Of course, the economic and political events of the last couple of years mean that there are not many potential recipients of these lessons. However, at least the three Baltic countries and some of the Balkan states can (with optimism) be perceived as such, though Russia, Ukraine and the other countries with similar policy characteristics will only arrive at that point when their macroeconomic policies resemble those of more normal economies. References to 'the East Central European experience' in this chapter explore the Hungarian and Polish governments' approaches to the bad loan problem. The other 'first wave' reformers did not develop any essentially different tools or ideas in approaching the problem. Therefore, it is legitimate to think of the two countries' strategies as representing 'the East Central European approaches'.

1 The author would like to note that this chapter overlaps significantly with another paper written for J.P. Bonin and I.P. Szekely, eds ( 1 994) , Development and Reform of the Financi,al System in Central and Eastern Europe (Edward Elgar) on enterprise bankruptcy and its relevance for the banking system.

Lessons for Bad Loan Management

59

2. Why did the bad loan problem emerge? The bad loan portfolio of the banks was a phenomenon that became an issue of policy concern in the countries of rapid economic transition almost immediately after their serious reforms started. The dynamics of the situation were relatively simple: these countries reformed their banking system by hiving off some commercial banks from the former central bank which used to supply the economy with credit; the economic reform established market prices for most goods and services and made the companies responsible for their own solvency. Simultaneously, government budgetary subsidies as well as directed credits with subsidized interest rates were drastically reduced. The combination of these changes ensured that the bad loan problem would appear because, on the one hand, banking is such a complicated operation that it takes a long time for a large organization to learn how to evaluate credit risk and, on the other hand, the dramatic changes in the economic environment fundamentally altered the relative profitability of different businesses. Furthermore, 'inherited bad loans' (that is loans which had been allocated by government decision and proved to be economically wrong) continued to burden the banks; it is very' difficult to draw the line after which it should be entirely the banks' responsibility how they lend. Yet establishing a more or less clear line is crucial from the point of view of the credibility of the government's policies towards its banks. To summarize, one of the preconditions for the emergence of the bad loan problem as a major policy concern is that the country has established, in the context of economic transition, a relatively solid macroeconomic regime. The western tier of the region (namely, Poland, the Czech Republic, Slovakia, Hungary and Slovenia) reached this state in the early 1990s. At the other extreme in Eastern Europe are countries of the CIS such as Ukraine (where lack of macropolicies has reached the point of being an international security concern) or internationally less visible new states such as Georgia. There is an important point, which Rostowski (1994) makes, concerning the relationship between bad debt and the macro­ economic situation. Countries where credit, because of previous bad policies, is a relatively small proportion of GDP, after stabilization (even if the banks have to be recapitalized) have less trouble than countries with larger credit/GDP ratios (such as the Czech Republic, Slovakia or Hungary), which have to bail out their banks. In this respect there is an important difference between Poland and Hungary, the two countries studied in this chapter: the level of credit

60

Mizsei

to the non-government sector in Poland is about 20 per cent of GDP; in Hungary this share is about 45 per cent. The Czech Republic needs to be even more careful since there this ratio is around 80 per cent. This message is optimistic: there are some areas where less advanced countries would be able to make a more efficient transition, if the situation were properly perceived. Another important variable is the systemic treatment of insolvent en terprises. Here one sees different approaches in the three countries which I have studied at various times. The Czechoslovak (later Czech) approach has been to postpone enterprise bankruptcies primarily so as to protect their ambitious voucher privatization project. This was achieved by two postponements of the enforcement of the bankruptcy law adopted in early 1992 and, after the enforcement of it in April 1993, application of two further 'grace periods'. The bankruptcy law did not start to threaten financially vulnerable enterprises until October 1993, and by May 1995 no big enterprise had yet been declared bankrupt by its creditors. The Czech government's concern for stability of the company structure has also been visible in its efforts to keep wages under strict control. In 1991 wages were virtually frozen, in 1992 they were strictly regulated and only in 1993 did liberalization of wages occur, but even then the government made a great effort to discourage companies from granting high wage increases. What we are dealing with here is a very deliberate sequencing of the economic reform: low inflation and privatization have been the focal points of this policy while other considerations have been consciously subordinated and in fact postponed. The two main goals of the government have been achieved, but the puzzle of the future will be what price the country will have to pay for delayed financial restructuring at the enterprise level. An underlying assumption of the policy is that by privatizing the economy quickly and after economic growth resumes, the liquidity of the enterprises will improve. Another assumption is that by privatizing the treatment itself can be decentralized. However, this strategy may be rather a trap: in order to secure the success of privatization, the government has become overprotective of the enterprises. The question is how long this overprotective attitude will last. Poland, in this respect, is a different case. In that country, the regulation of financial restructuring and of bankruptcy by the courts as it had been in 1934 was reinstated in 1989 after four decades of suspension. The two laws are elaborate legal acts2 and yet their use

2 For a detailed analysis of the policy aspects and application of the bankruptcy law, see Mizsei (1993) .

Lessons for Bad Loan Management

61

was very limited in the initial period of economic transition, i.e. in 1990-2. The primary reasons for this are that, on the one hand, the creditors did not expect efficient bankruptcy procedures so they tried in many other ways to recover at least part of what they were owed (some of these ways, such as administration, are also very circuitous just because of the poor state of the courts), and, on the other hand, that (at least initially) most of the creditors were themselves state-owned units where, in spite of spectacular improve- ments, profit considerations were not as sharp as is the case in a normal market economy. The creditors might reasonably suppose that the bankruptcy machinery would not work well. On the basis of this logic Fan and Schaffer (1993) justifiably question the relevance of the idea of creditor passivity; 3 at the very least one can say that creditor passivity is not the only, even probably not the overwhelming cause of the small number of bankruptcies. In my field research in Poland, I also found that the commercial courts system did not yet work in a way that would have encouraged creditors to take their claims to court. Courts were understaffed and there were no experienced bankruptcy judges; initially there were no trustees available and firms were not very familiar with their legal possibilities. In other words, it takes time for economic agents to learn to use the bankruptcy regulations. Furthermore, the incentives were not there to reward bankruptcy trustees (let alone court judges) and to attract qualified people into this highly difficult business. As a result, in the first years of the transition in Poland there were practically no bankruptcies. In 1990-2, many policy debates centred on how to handle the problem of mass illiquidity of enterprises. However, it took three years for the government to elaborate its programme of 'banking conciliation' as a reaction to the mounting bad debt problem. The way the Hungarian government approached enterprise bankruptcy is different from both the Polish and Czechoslovak policies. Hungary in a sense has taken the most radical approach so far to the problem. It adopted very radical bankruptcy legislation at the end of 1991. It is important to emphasize, however, that even in the 1980s quite a number of liquidations of enterprises and financial reorganizations (under a special procedure for state-owned enterprises) were orchestrated by the system on the basis of ad hoc regulations. In a way, Hungary had a small advantage by 1990 over 3

On creditor passivity, see Mitchell (1992) .

62

Mizsei

the other East European countries that in economic thinking bankruptcy was no longer an alien notion; moreover, it had gained by then some broader public acceptance as well (Mora 1 992). In 1990 and 1991, Hungarian economic policy had two simultaneous concerns. On the one hand, price liberalization and the elimination of price and production subsidies put a strong upward pressure on consumer price inflation. On the other, monetary austerity and the disintegration of COMECON-trade, while somewhat moderating inflationary pressures, had a strong contractionary effect on industries exporting to the Soviet Union and other communist countries. An important fact of the Hungarian political economy in that period is that these two considerations, very much as during the communist era, appeared as a battle between different sections of the government. It should have been different since the political situation was uniquely comfortable from the point of view of establishing a government without needing too many compromises in order to keep the coalition together. Besides agriculture (where the Smallholders' Party had strong particular interests), the leading governing party did not need to face any real coalition constraint in shaping its economic policies. Yet lack of experience and to some extent also of competence, as well as excessive emphasis on political and ideological issues rather than on problems of economic transition, meant that homogeneity of economic policy was not established. One part of the government followed the conservative IMF policies in good faith (i.e. in the belief that it was plainly the only policy option), while another one opposed it with rather obsolete pro-growth arguments. There was no intellectual centre in the government during this period that could have elaborated feasible alternatives to doing nothing with the financially troubled enterprises suffering from the rapid change of environment, including the collapse of the Eastern markets and sudden increases in energy and other basic input prices. In 1 990, enterprises were quasi-subsidized by receiving the price of their Soviet exports from the Hungarian authorities in spite of the fact that the Soviet Union even at that time regularly failed to meet its financial obligations. This solution in 1 990 again was more a result of a lack of coordination than of any conscious policy to maintain enterprise liquidity. In 1 991, automatic payment for East European exports by the Hungarian state was terminated; enterprises were pushed one step closer to the harsh realities of the market. In this year (and even in 1 990), enterprises were able to reorientate their trade relations towards the developed countries' markets. However, it is increasingly

Lessons for Bad Loan Management

63

clear that much of the reorientation was attributable to strictly temporary 'mobilization of reserves', i.e. unprofitable, low-priced exports of whatever was available to sell, and much less to the presence of competitive goods, better marketing, etc. Shortening of the enterprise managers' time horizon caused by rapid changes in the environment, expectations of privatization and so on, has contributed to a great extent to decisions to 'export at any price in order to survive for a little longer'. Enterprise profitability (already in a bad shape as a consequence of a few years austerity) quite obviously plummeted in this period. Inter-enterprise payment arrears were probably at their peak. Still, there was no institutional mechanism or any policy programme to break the chain of liquidity gaps in the period of export reorientation and price/trade liberalizations, i.e. in 1990-2. The Ministry of Finance fiercely opposed any enterprise bail-outs in 1991, arguing that the government would 'soften the budget constraint' by offering bridging finance for enterprises. It obviously did not help that consumer goods inflation was at its peak and inflationary fears were, probably justifiably, strong in that period. It is also a defendable argument that the knowledge and financial know­ how were simply not there to choose the right subjects and right tools to assist in restructuring the enterprises; all the troubled large Soviet export industries obviously could not have been chosen and nobody had any idea which would be the future 'winners' that should be supported. Retrospectively, it might be useful to draw two lessons from the inaction of the Hungarian government. (Let it again be emphasized that no one, including the author, had clear and right answers to the problem at the time; events developed too quickly and knowledge was limited.) One lesson is that lack of any enterprise bail-out and virtual inaction in the truly exceptional year of 1991 (because of the sudden loss of COMECON-markets) probably cost more in terms of later loan consolidation, etc., than a carefully designed support programme at the time would have cost. The situation was worsened by the government's inaction. The other lesson this story tells is that the situation would have been significantly better had some large industrial enterprises been put on sale in 1989 or 1990 when the interest in 'buying Hungarian' was quite enormous. Obviously, there was a limit to what the government apparatus could do in a given time (and voucher sale and other quasi-privatization methods obviously could not have solved the short-term problems either), but in those two years there would have been a real chance to sell big fishes, such as lkarus and Raba,

64

Mizsei

along with some major commercial banks. 4 However, there was a fear of 'selling the family silver' (though in most cases it was not silver but quite well worn aluminium) cheap to foreigners. Retrospectively, one can say that the Hungarian economy would have been significantly better off and the enterprises' financial crisis would have been much less severe had some of these large organizations been privatized early on. One has to remember, in order to keep things in perspective, that none of the post-communist reform countries reacted to the collapse of the COMECON any better than Hungary in 1991; arguably the contrary was the case in Czechoslovakia, where the illusion about clearing mechanisms in their Soviet trade prevailed much longer than in Hungary. Nor did the Polish economic policy react in time to the challenge of market collapse. The accumulating inter-enterprise indebtedness pressed the Hungarian government in 1991 to implement quite radical bankruptcy legislation in order to achieve a turnabout in the payment discipline of the firms. Lack of proper policy co-ordination in the government, mentioned earlier in this section, led to several unexpected phenomena in the economy. In 1992, the combined effect of three new regulations created a real shock in the Hungarian economy: the bankruptcy law, the consequences of the new accounting regulations and the banking law forced the economic agents to reveal the true state of their financial assets and to face the consequences. Many argue that it was a real 'shock therapy' to the extent that it exceeded the microeconomic effects of the Polish stabilization programme in 1990. While this statement probably exaggerates, it is true that the Hungarian conservative government showed a radicalism unprecedented in Eastern Europe in so far as they forced companies and banks to reveal the real state of their financial assets and to live up to the consequences. With respect to the financial sector, the Hungarian government and the public lived under an illusion until this operation and even for some time after. The big state banks showed large windfall profits 4 A 1993 smvey in the Financial Times also mentions two examples: that of the chemical concern TVK, which was approached by Eastman Kodak, as well as the pharmaceutical company Richter Gedeon, where Takeda of Japan would have been a buyer, both in 1990. The latter case, especially, is also good to illustrate how the 1991 final collapse of COMECON caused a sharp depreciation in the value of Hungarian (and generally East European) companies. In the case of pharmaceuticals, Russia was the largest market; its collapse made the companies in that sector much less attractive. See Denton (1993) .

Lessons for Bad Loan Management

65

in the early years of transition precisely because they were not confronted with the true quality of their asset portfolio. The Ministry of Finance was satisfied because this was an important source of revenue for the embattled state budget, and in the press there were frequent attacks on the overspending of the 'rich' banks while the nation was suffering. The overspending was true in the sense that if the banks had behaved according to their real financial situation they could not have afforded to expand their branch network so fast, nor pay their employees the salaries and other benefits that they handed out to management. In 1992, it came as a shock to the uninformed public (to which unfortunately the bulk of the government also belonged) that the banking sector's profits had dried up altogether; how to bail them out became a major policy issue if the consequences of one of the large state banks going bankrupt was to be avoided. How little the government sensed these problems is illustrated by the episode in early 1992 when the Minister of Finance wanted to force the banks to repay the money they had received back from the tax authority due to overpayment of the profit tax in the course of 1991. At that time, the Minister of Finance still believed that the banks were in good shape and were only hiding their wealth from the fiscal authorities. In fact, at that point, being faced with the new rules on provisions against doubtful assets, the government started its series of frequent programmes aimed at the banks' long-term reform. In many instances, this has also involved large-scale and costly state assistance.5 The scale of subsequent bail-outs is regarded by Oblath and Valentiny (1993) as more expensive operations (in terms of adding more to the public debt) in the 1 990s than the very large budget deficits. Here we embark upon a real reform sequencing dilemma to which we do not have ready answers at this point in time, because we are neither theoretically prepared nor have enough empirical evidence to solve the problem. The new macroeconomic regime has revealed (and maybe partly caused) the true financial state of the enterprise 5 Again the author would like to emphasize that this chapter does not enter into discussion about the regulation of banking failure. Especially avoided is discussion of the possibility of bankruptcy of the large state-owned banks in the environment of highly concentrated financial markets. The omission of this issue (and also of the discussion of the regulation of the importance of a deposit guarantee) does not mean, however, that it should not be addressed altogether in the context of this chapter's topic. It is especially important in countries where the financial market concentration is less strong as, for instance, is the case in Poland, Russia, or, until very recently, in Bulgaria.

66

Mizsei

sector in each of the fast-track East European countries, i.e. Poland, Hungary, the Czech Republic and Slovakia. In order to establish credible long-term financial discipline in the economy, one has to formulate rules to handle this problem. Simultaneously, other microeconomic reforms have been proceeding such as the privatization of state enterprises; other reforms could have been accomplished, at least theoretically, in the meantime. This refers especially to the reform - and mainly privatization - of the financial sector. I would refer at this point primarily to the Hungarian experience. As stated, Hungary adopted the toughest bankruptcy rules, together with accompanying banking and accounting legislation, in the whole region, at the beginning of 1992. This goes against the popular perception of Hungary's transition being 'soft', or 'too gradualist' as compared with the 'shock therapy' of Poland and the 'boldness' of the Czech approach. The Hungarian bankruptcy legislation did two things which made it radical. On the one hand, it instituted the 'automatic trigger', i.e. enterprises which could not meet their payments to any one of the creditors were obliged to file for financial restructuring or for liquidation. This is a crucial fact because the Polish example tells us that in the absence of such measures the introduction of bankruptcy as a real option becomes very slow. The Hungarian legislation did not even establish a certain minimum ceiling for the automatic trigger; the any delay triggered the requirement to file for bankruptcy at the court. The other aspect was that the Hungarian law made it very difficult to reach an agreement in court between the debtor and the creditors by requiring the unanimous consent of the creditors to the financial restructuring of the company's debt. 6 With such tough bankruptcy legislation, the policy makers expected that financial discipline would radically improve in the economy. They probably achieved that at least in the short run (see Varhegyi (1993)). Payment arrears have decreased in real terms in this period. Of course, one has to ask whether this was the only way to achieve that result, or if this was a legitimate policy concern at all. Some radical insolvency rules were certainly badly needed to gain credibility for economic policy; in Poland the initial phase of the Balcerowicz programme gained credibility and resulted in shrinking payment arrears in real terms. But even there, this initial credibility bonus 6 This chapter cannot give a full description of the philosophy of the Hungarian bankruptcy law. For a detailed analysis, see Mizsei ( 1 993) .

Lessons for Bad Loan Management

67

needed a reinforcement; ultimately, regulation of financial default and enforcement of the rules will carry the weight of disciplining the companies. In Hungary, there was no macroeconomic shock so the microeconomic reforms had to carry the weight of demonstrating credibility from the very outset. Also, the Hungarian business infrastructure was more ready for such a shock than the Polish economy was in 1992. Yet the harshness of the automatic trigger was clearly exaggerated and caused so many enterprise failures that the system could not credibly deal with it. The 'domino effect' of the 'brutality' of the regulation was also an issue of clear concern. One of the effects of the bankruptcy legislation was a huge wave of bankruptcy filings from April 1992, the starting date of the application of the law after a three months grace period. The system - courts, judges, trustees, banks, etc. - was unable to meet the time requirements of the law; even by 1994 only about one-third of the announced liquidation cases (and one-tenth of the filed cases) have been finalized and even a smaller share of filings has led to real liquidation. So, even in spite of the increased credibility of the system in the short run, it is quite questionable whether this credibility bonus will not be eroded over time. Also, other dysfunctions of the legal system related to bankruptcy, such as the slow and difficult enforcement of claims, have raised questions about the overall credibility of law enforcement in the economy. Lack of policy co-ordination was also very visible in this issue; in spite of the courts' (most notably the largest Budapest commercial court's) repeated appeal for investment in the technology and staffing of the courts, the Ministry of Finance refused to do so. The size of the requested investment was incomparably lower than the losses caused by delays and dysfunctionalities in the courts' processing of bankruptcy cases. This would have been a public investment with an unusually high rate of return. Furthermore, the weakness of the infrastructure strengthens Fan and Schaffer's above-mentioned argument that the low level of bankruptcy filings in post-communist economies is not necessarily caused by 'creditor passivity', but in many cases partly by the creditors' rational calculation that the liquidation value of firms will be lower than that which can be recovered without bankruptcy (Fan and Schaffer 1993). The broad picture is that, although regulation of reorganization was far too severe, at least half of the finalized cases actually ended successfully with agreement (see Table 4.1). Part of the explanation for this is, though, that the creditors were in a forced situation exactly because the expected liquidation value of firms was very low due to

68

Mizsei

the weak legal infrastructure and capital market sophistication. However, the other broad experience is that many of the agreements were too optimistic, and consequently after a few months those firms ended up in liquidation. The good news in the bad news is that these applications were obviously spread out over time and therefore were not an extreme burden on the courts' capacities and on the banks' own human resources, current accounts, etc. However, this slowness highlights the problem of lengthy liquidation procedures. The commonsense argument is that liquidations take a long time, even in countries with well-established economic and legal infrastructure. This is true; however, this seems to be yet another field where the East European countries cannot afford the slowness and inefficiencies of well-established market economies. There are ways to accelerate the process; establishing aggressive success fees and connecting them not only to the sales price but also to the speed of the liquidation is one such possibility. Tabl,e 4. 1. Financial reorganizations and liquidations in Hungary since the introduction of new bankruptcy regulations Financial restructuring

Filed Administrative end Announced Finalized Reorganization agreed Liquidation initiated

Liquidation

1 992

1 99 3

Total

1 992

1 99 3

Total

%

4,231 1,254 2,500 1,099 682 417

372 1 62 295 349 1 40 1 65

4,603 1,416 2,795 1,404 882 582

10,062 4,963 2,227 562*

2, 1 80 1,026 591 431

1 2,242 5,989 2,818 993

100 49 23 8

*Four counties' data are missing, therefore incomplete. Source : KOPINT ( 1993) , p. 160.

From the very beginning, the new bankruptcy rules were criticized on the grounds that they were too harsh (automatic trigger), too strict on the reorganization side and created too many victims given the absorptive capacity of the legal infrastructure. It was also said that the system followed the German bankruptcy philosophy too closely, while, at least at this stage of the Hungarian market economy, the US

Lessons for Bad Loan Management

69

concept might have been more applicable even if it was actually very questionable in the American setting. The requirement of unanimous consent in the 1 992 law was certainly counterproductive; speedy processing of liquidations required even more aggressive incentives than were sanctioned. 7 However, I would not recommend completely eliminating the automatic trigger; it would be enough to establish minimum debt levels (in absolute terms and/or in relation to the debtor's capital base) in excess of which filing for bankruptcy (or financial restructuring) would be mandatory. Changes came very late in the summer of 1993 (yet another illustration of the lack of co-ordination and prioritization in Hungarian economic policy) and only brought relief when unanimous consent was eliminated. However, filing for reorganization became unnecessarily complicated8 and no additional incentives have been introduced in order to accelerate liquidations which lag heavily behind, causing many delays and, in fact, abuses too. Moreover, the complete elimination (rather than mitigation) of the automatic trigger probably gave the wrong signals to businesses about the commitment of the government to maintaining the financial discipline in the economy that it had achieved. The overall lesson from the history of the recent Hungarian bankruptcy regulations is that the regulators should primarily protect the creditors already in the transitional period (otherwise financial discipline cannot be established) . However, chances of financial restructuring should exist since the transition causes an unusually large amount of enterprise failures which should be mitigated in a limited period of time. Thirdly, the regulators should refrain from extensive interference in forcing too many bankruptcies upon the system. The legal infrastructure cannot handle that situation and the recovery of assets will be very inefficient. The trick is to make the regulation creditor-friendly and yet to avoid too much interference.

7 Even though generally the success fee is built into the system, it is worth noting that in environments with strong law enforcement and good civil service traditions, the costs of building in success fees can easily exceed the benefits. Such an example is probably Britain. In Hungary and, as a matter of fact, elsewhere in Eastern Europe, the benefits - at least for the time being - strongly exceed the possible costs. 8 Even in accepting the application, the consent of two-thirds of the creditors is required.

70

Mizsei

3. Government action to consolidate the financial position of banks in Hungary One of the consequences of the Hungarian bankruptcy regulation was that the asset portfolio of the banks appeared to be much worse than had been previously thought. The shock effect of the set of new rules, but especially of the bankruptcy law, in the beginning of 1992 revealed the quality of the asset portfolios of the Hungarian banks. One question is then whether a different sequence or less stringent measures could have made the books of the banks brighter in the long run. Furthermore, another question is whether the Hungarian government handled its large banks' difficulties correctly and whether there was any alternative method. First, one has to underline again the impact of macropolicies on the liquidity of enterprises and, consequently, on the banks' financial positions. A tough wage policy can ease the pain of transition and the exchange rate policy also matters. In both fields, the Hungarian economic policy has avoided big disasters but has not been very imaginative. Wages and, even more importantly, non-wage labour costs have increased rather rapidly, making the financial situation of enterprises more vulnerable. As far as exchange rate policy is concerned, real appreciation had its merit in 1990 and in 1991 but it has increasingly become a large burden on the enterprises' finances. In terms of the government's attitude towards the banking sector, again, the reaction time was very slow. Even in 1991, it was clear that the asset portfolios of the major commercial banks were in a bad and deteriorating shape and, moreover, that the worst was still to come. The Ministry of Finance still lived under the naive illusion that it was enough to squeeze the banks and they would continue delivering fat tax payments. The Ministry's attitude towards the enterprise sector was similar. While the conventional argument about 'industrial policy' does not have much relevance for East Central Europe, it is valid to question whether the government should not have done more to help the enterprises' trade reorientation with additional financing of credible programmes right after the COMECON crisis, even though it is clear that the government was by and large unable to check the feasibility of restructuring programmes on the company level. These (the exchange rate policy and bridging loans) are examples of inflationary finance where I believe the short-term and long-term inflationary impacts probably clashed. This leads to the issue of the unexplored opportunities - in fact duties - of banking supervision in a situation where the institutional

Lessons for Bad Loan Management

71

preconditions of a hard budget constraint are still very fragile, especially due to widespread state ownership of banks and companies. In this kind of situation, the supervisory authority has to force the banks to maintain prudent policies so that the stability of the financial sector is protected. In other words, in this period the role of banking supervision should be partly to substitute for the lack of management control by the banks' owners. A clear trade-off exists between the quality of banking supervision and the need for the government to intervene through costly operations in order to protect the financial sector's stability. However, as the banking skills were absent so were those to control the banks' activities in the government. Also, the management of the banks should be made personally responsible for irresponsible banking practices, i.e. for lack of maintaining financial stability of their institutions' operation. In reality, banking supervision was not up to its job and co-ordination (again) between the Ministry of Finance and banking supervision was poor in the critical period 1991-3 in Hungary. Was the government's management of the emergency situation, which followed the microeconomic shock in 1992, proper? In mid-1992 it had already become obvious to those dealing with economic policy that something had to be done with the banks' balance sheets. I would look back and emphasize again that the crisis of 1992 could have been much smaller had some of the large banks already been privatized in a credible way, i.e. a way which would minimize the need for a future state bail-out. There was a chance in 1989-90 to privatize some of the large banks (again, I am not being a purist, the government could have decided to privatize only some banks, even one of the four large banks would have been much more than none); later, there was a chance to privatize at least the Foreign Trade Bank, which, because of its earlier specialization, had a very clean portfolio. In fact, in 1991 and 1992, and even in 1993 the government was very slow and overcautious and did not get good advice on the philosophy of privatization of the banks. Despite the fact that there was no outright bank privatization (some privatization of the banking business occurred through the growing market share of smaller, partly foreign-owned, private banks; in this respect the government's policy was admirable), the loan consolidation action in 1992 still could have been much better. In fact, this was one of the worst economic policy actions in the period of the Antall government. The government could not create a meaningful strategy at all in 1992, and at the end of that year it let itself be blackmailed by the top managers of the large banks on

72

Mizsei

grounds of the urgency of cleaning the balance sheets if the government wanted to privatize the banks in the foreseeable future. The initial fallacy of this idea is that it is quite clear that the two largest banks could not be privatized without a more fundamental improvement in their balance sheets than this action allowed. The second problem was that the terms required for swapping debt for government bonds were so unfavourable to the banks. Although the immediate threat of inadequate capital resources was eliminated, the new assets would generate such a low rate of return that it might well depress the price of the banks in case of privatization. Thirdly, as a consequence of the action, recovery of the poor quality assets is more doubtful than before, since these were taken over by an organization (the Hungarian Investment and Development Corporation) which neither has any practice in such activity nor possesses a similar depth of information as the previous owners of the assets. Fourthly, it is impossible to match the asset portfolio with a similar liability duration structure which makes the banks vulnerable against future inflation. Fifthly, and most importantly, the action does not reflect in any way the efforts of the bank managers themselves; the policy makers did not do anything either to eliminate the potentially huge moral hazard problem involved in relieving the bank managers of the responsibilities for lending decisions or to bring privatization of the banks any closer (Abel and Bonin 1993). The loan consolidation programme involved assets worth 1 04 billion forints ($1.3 billion) which were replaced by 80 billion forints ($975 million) in so-called government consolidation bonds. However, the bonds' unfavourable interest rates decreased their real value substantially. 9 Anecdotal evidence says that the banks, knowing about the early 'debtor consolidation' plans, did not put some of their worst assets into the package, so increasing the likelihood of demands for further bail-out. The government could have supported the privatization of one of its large banks, the Budapest Bank, by a more comprehensive assistance programme because of the relatively better performance of that bank's management at that time as well as the quite good international reputation and interest in investing into that bank. Helping out the two other large banks did not bring their 9 For a description of the terms of the loan consolidation, see Varhegyi ( 1 993) , as well as Abel and Bonin ( 1 993) . The latter authors estimate that the net present value of the bonds is somewhere between one-third to two-thirds of their face value.

Lessons for Bad Loan Management

73

privatization any closer, but sent the wrong message concerning future relationships with the government, i.e. the operation increased the probability of future government generosity. It is especially true since bad performance did not bring about any consequences for the top management. The situation was very similar at the 1993 recapitalization of the banks: its hasty preparation was not justified by any urgency related to privatization. Without that perspective, recapitalization will again ease the pressure on the banks themselves. Selectivity, serving the privatization of those banks which can be privatized more immediately, would be badly needed; this need is especially acute in the light of a sharply increasing domestic debt. Instead, the government has prepared a recapitalization package which does not solve anything, again does not bring bank privatization closer and increases the moral hazard. By adopting the principle that each large state-owned bank will be recapitalized to the extent that their net risk-adjusted capital position will be zero, it sent the same bad signal to the banking community: performance does not matter, on the contrary, it worsens one's chance of getting a state subsidy. Another part of the 1993 recapitalization episode is that as part of this the Ministry of Finance is going to take over a 'golden share' of ownership from the State Holding Company and the recapitalization contract gives extensive rights to the Ministry to execute control over the banks. 10 However, the Ministry does not possess the skills to control the functioning of the banks and by duplicating other organizations' overseeing functions, clear responsibilities will be impossible to establish. As with the loan consolidation episode, there is no sign that the state has appreciated these consequences concerning the banks' management. Anyway, politicization of state control over the banks intensified tremendously on the eve of the 1994 elections. The bank recapitalization involves nine partly or entirely state-owned banks and its estimated cost is about 150 billion forints ($1.6 billion). The state bonds, issued to recapitalize the banks, have a twenty-year maturity with interest rates equal to the interest on 90day discount Treasury bills, adjusted twice a year. This means that the market value of the bonds will be much higher than that of the 10 According to my very latest information, the Budapest Bank has been able to get some concessions in terms of the conditions of participating in the programme which reflect the government's (still somewhat hesitant) will to privatize this bank in the foreseeable future.

74

Mizsei

bonds issued at the first loan consolidation. However, this obviously also means that this issue will also increase the state debt much more than the previous one. The other component of the programme, making it even more expensive, is that the participating banks will have to notify their 'bad debtors' that they can sign up for a debt restructuring programme. It will be a centralized programme in the sense that a government commission will generate the negotiations. If the creditor bank does not accept the deal brokered by the other creditors and the state representative, the bank has to sell the asset on 'net value', i.e. it will lose the sum it put aside against the bad asset. The 'debtor consolidation' programme has been structured in such a way that, first, an 'accelerated reorganization' of fifty-five (mostly the largest) firms would happen; the deadline for agreements for these firms was 31 March 1994, i.e. the schedule for these very complicated cases was extremely tight. This would cause many poorly prepared and therefore very costly deals. Another tier of debtors to sign up for consolidation contains approximately 200 companies, while the avalanche of state subsidies might not stop even at that point; there is talk about a third tier of several thousand enterprises. The only hope is that the government will run out of time and after the May 1994 elections the new government will stop this very dangerous escalation of state subsidies to enterprises. The debtor consolidation programme might be very expensive for the state in two ways. Its estimated direct cost is somewhere between 60 and 100 billion forints ($500 and $980 million). The other, indirect, cost is that the action relaxes financial discipline in the enterprise sector once again. Anecdotal evidence reveals that after the debtor consolidation plan was leaked, payments discipline deteriorated. Even if the government takes measures in order to avoid these kind of abuses (by announcing that arrears before the leakage can be involved in the programme) the sequence presented above of 'soft budget constraints' has sent a hugely demoralizing message to the enterprise sector. The debtor consolidation programme is very broad and generous and clearly intended to serve the election purposes of the governing coalition. One of the most immediate needs of the new government after the election will be to reshape the country's banking policies and to establish order and payment discipline in the recently confused situation. Regaining credibility for the government in this respect should be one of the primary goals of its operation. To sum up, one's intuitive reading of the Hungarian lesson is this:

Lessons for Bad Loan Management

75

the policies towards the enterprise sector as well as the bankruptcy regulation could have been somewhat softer (perhaps by helping out the enterprises selectively in 1991 and by adopting bankruptcy regulations without or with a much less harsh automatic trigger in 1992) and macropolicies could have been explored better during the years of transition. Furthermore, the government has lost several good opportunities to privatize parts of the financial sector. In this respect, a case-by-case approach to bank privatization still seems to be the most feasible one, of course now in somewhat worse conditions than if the process had started earlier. The government's handling of the bad loan problem was characterized by the lack of a clear strategy, i.e. a well-designed effort to avoid large-scale rent-seeking behaviour of the banks and of large enterprises. In fact, this is one of the fields of Hungarian economic policy after democratization where government performance was the least satisfactory. Therefore, the new government badly needed to have a 'new beginning' after the 1994 elections. That is, it needed to have built up credibility that it would only see that banks maintained the rules of the game and the prudential regulations, so that the stability of the banking sector could be maintained without repeated government financial support and, as an intimately linked issue, without intense politicization of the banking industry. The unpopular conservative government was unseated in the May elections, and replaced by a Socialist ( ex-communist) and Free Democrat coalition. The government's ambitious reform programme gained credibility under the guidance of liberal Finance Minister Bekesi. However, actual progress was slow, with key pieces of legislation delayed. The resignation of Bekesi in late January 1995, along with delays in appointing a new privatization minister and central bank chief, threw his programme into doubt, raising fears that Hungary's government would divide over economic policy, and perhaps succumb to more populist pressures. However, in the meantime the macroeconomic situation markedly worsened in many respects: the current account turned into a deep deficit in 1993 and consequently the national debt began to grow again significantly in 1993 and 1994. Inflation bottomed out, lessening only slowly, from 22.5 per cent in 1993 to 21.1 per cent in 1994. Economic decline stopped, however, and growth in 1994 was estimated to be in the area of 4 per cent, comparable to the Czech Republic and Poland. Had the economic policy, including policies towards enterprises in difficulties and towards the banking sector, been better focused in the last three years of the HDF government,

76

Mizsei

and had liberals been more influential in the unsteady coalition, the prospects for a solid macroeconomic base and sustained growth would, no doubt, have been somewhat better.

4. The Polish banking conciliation project The Polish banking conciliation project is interesting because it offers an alternative approach to the interconnecting problems of enterprise liquidity crisis and excessive bad loans in a transition economy. East European countries mainly followed the pattern shown in the previous section; this certainly applies to Czechoslovakia (see Brom and Orenstein ( 1993)) and Slovenia, and later Bulgaria seemed to follow suit. 1 1 Poland took quite a long time to come up with a comprehensive government approach to the bad loan problem, but the Polish project seems to be superior in many ways to that in which other countries tackled the issue. First of all, one needs to ask whether government inaction in tackling the bad loan issue was necessarily a bad thing in the first three years of transition. Without claiming to have any ultimate answer, one can say that many 'activist' approaches have cost much in terms of public debt. In Hungary, the current value of the three above-mentioned loan and debtor consolidation programmes is probably about 250 billion forints ($245 million), i.e. about 8 per cent of Hungarian GDP; the expected cost of the Bulgarian operation is about $2 billion, i.e. about 25 per cent of GDP, without any guarantee that the economic mechanisms leading to a replication of the problem will cease to exist. There is no hard empirical evidence to answer the above question, unlike in the case of inter-enterprise arrears, where available evidence strongly supports the non-activist approach. In Poland, the government started to work intensely on the related issues of inter-enterprise indebtedness and bad banking assets right after the success of stabilization partially cleared the government's agenda. The idea of the programme of banking conciliation was born in the Ministry of Finance and fine-tuned in cooperation with the World Bank because the government wanted to cover some of its costs by means of international assistance. The draft legislation was practically ready in early 1992 but because of the political difficulties 1 1 I am relying on Rumen Dobrinsky's description (in the pages immediately following this chapter) of the current Bulgarian debates, for which I am grateful to him.

Lessons for Bad Loan Management

77

and complexities of legislation in Poland in that period, it only became the 'Enterprise and Bank Financial Restructuring Law' in early 1993. It aimed to integrate the solution of the two problems (bad debts and enterprise indebtedness) into one coherent procedure. One of the instruments, sanctioned by the law, is the 'banking conciliation procedure'. It empowers the principal banks of heavily indebted firms to conduct financial restructuring of the firm (for the period of the validity of the law replacing the courts in conducting such exercises), should the indebted company initiate it. In the first step, the Ministry of Finance has recapitalized the state banks so that their risk-weighted capital adequacy reaches 12 per cent, i.e. theoretically they became significantly overcapitalized. After that, for a limited period of time (until the end of March 1994), the banks could accomplish the restructuring of those state enterprises that had applied. For an agreement to be valid, the enterprises had to obtain the consent of the owners of more than 50 per cent of the credit involved. Contrary to the general regulation on enterprise financial restructuring, in this temporary one there is no provision concerning the agreement of a minimum share of creditors. So in this respect the law is more permissive or debtor- friendly than any such regulation in 'normal' situations. Only firms above a minimum level of indebtedness could apply. The participating firms (which had to be 100 per cent state-owned or have a majority state ownership) had to apply for commercialization (transformation into corporate legal status). The Ministry of Finance (MoF) supported the agreements by contributing a similar proportion of its overdue receivables to other creditors. The banking conciliation law specifically permitted debt-for-equity schemes as part of the agreements. Also, after expiration of the March 1994 deadline, the principal bank had to sell its overdue claims on the market (the affected banks had already started to create the market infrastructure for those transactions) or to announce bankruptcy of the debtor according to either of the two laws dating from the 1930s (Groszek 1993). The promising thing about the Polish approach is that it embraces the problem in its whole complexity. It would have been even better to have launched the programme one or two years earlier, but at that time the Ministry of Finance (MoF) was not prepared for it. The political process (parliamentary acceptance of the banking conciliation law) also took a long time because of the confused political situation in Poland at that time.

78

Mizsei

At the end of August 1 993, 214 initiated cases were registered at the ten largest banks. 12 However, by February 1 994, only a very few sporadic cases had actually been finalized. From the intermittent information available about the process, I can only point out some possible pitfalls or difficulties of the programme. One of the issues that emerged was whether the March 1 994 deadline had not been too tight for the banks to finalize all the cases. Some calculations put the number of applications on the average at more than 200 in each bank involved. 1 3 Clearly, there was a shortage of expertise in the financial restructuring of enterprises not only in the banks but in Poland at large. This becomes an acute problem when such a concentration of restructuring needs emerges. 1 4 The law permitted the minority creditors to attack the agreements in court; this, on one hand, was necessary to protect them from the threat of dishonest agreements on the part of the majority creditors, but, on the other hand, made the procedure more vulnerable to delays. Since one could expect only a relatively small proportion of cases to have been finalized before the end of March 1 994, the other possible pitfall was that after this deadline the principal bank would have had to proceed against clients which had not succeeded in getting approval for their financial restructuring. The bank would either initiate bankruptcy or sell the bad assets on the market. In this case, however, the question is whether there would be buyers on such a massive scale as the procedure might generate, especially where the 'commodities' for sale were not standardized and information would be expected to be minimal about them. 1 5 One possible solution is that the government would initiate the extension of the deadline and it might, as many feared, 1 6 exert political pressure on the creditors to be lenient towards the restructuring plans of debtor enterprises. Either of these could, as in the Hungarian case, have caused a dangerous precedent and might have resulted in an avalanche of restructuring applications difficult to resist. Another problem with the secondary market was that the sale of the enterprise debt meant the bank would have to pay 40 per cent tax on the provisions, based on 12 See 'Oddluzanie dla najlepszych', Nowa Europa, 23 November 1993. 1 3 Eleven banks were recapitalized in the spring of 1993 within the framework of the programme. 14 See Henryka Bochniarz' opinion on lack of expertise in 'Dlugi pod spe� alnym nadzorem' , Nowa Europa, 28-30 January 1994. 1 Lado Gurgenidze and Istvan Szekely drew my attention to this particular problem with selling debt the Polish way. 1 6 See, for instance, Balicka ( 1 993) .

Lessons for Bad Loan Management

79

the qualified asset. This might significantly reduce the supply of debt on the secondary market, and consequently might have a negative impact on this path to privatization. 17 One can legitimately ask whether the Polish programme brings us any closer to bank privatization. It certainly intends to contribute to enterprise privatization by forcing the firms to transform into corporate structures and by pushing the banks first towards creating a secondary market of enterprise debt and then towards debt-equity swaps. The programme might also help bank privatization in the sense of making the banks more attractive because of a stronger capital base behind the asset portfolio; however, World Bank assistance itself does not require subsequent bank privatization. Moreover, political changes in 1994 might make it likely that bank privatization would be further delayed. This would be a mistake because with recapitalization and the subsequent debt write-offs and with the acquisition of some vital work-out skills, some of the Polish banks at least should be attractive to investors. The next logical long-term question was whether the banks would need frequent and regular financial assistance from the government. There were contradictory reports about the evolution of the bad debt problem. Gomulka (1993) estimated that by the end of 1992, about 50 per cent of bank assets were doubtful. This is a very high share and the Hungarian experience tells us that aggressive pressure on the economy to force bankruptcies of insolvent enterprises would make the bank balance sheets look even worse. Yet, the MoF, based on an independent audit, seemed to be confident that economic growth in 1 993-4 eased the bad debt problem in the banks and believed that the banks were in reality overcapitalized. The estimated cost of the programme was about $2.5 billion (Groszek 1993). This sum roughly corresponded to the theoretical costs of bad loan write-offs, based on Gomulka's assumptions as to the extent of the problem loans and Rostowski's very rough assumptions concerning the possible state involvement (3 to 4 per cent of GDP in Poland) in recovering bad assets (see Rostowski (1994)). In sum, Polish banking conciliation was a uniquely intelligent project in the recent history of East European government loan consolidation programmes. It tried to address the incentive issue (assets remain in the banks' books) and the moral hazard problem (up-front recapitalization) better than any other government programme in the region had done. It also invited banks to take 17 See 'Dlugi pod specjalnym nadzorem', Nowa Europa, 28-30 January 1994.

80

Mizsei

more responsibility for the management of companies which were seriously indebted to them. However, some danger still existed that this would not be the last programme as there were still open questions and one could not be sure that the government would pursue the same economic philosophy with the same degree of discipline as its predecessor. 18

5. Alternatives to loan consolidation One can say with reasonable safety that the Polish banking conciliation action is superior to any other loan consolidation arrangements because it decentralizes the selection process and the restructuring procedure to the level where information is most likely to exist about the efficient recovery of the debt. The incentive structure also seems to be superior to other schemes' structures. The only broad question is whether the complexity of the programme does not have too great a danger of derailing itself by changing a political environment that in Eastern Europe would be seen as the natural situation for at least some years to come. 19 However, even if one assumes (as seems at least a reasonable assumption in the context of Poland) that the political process will not change the plan fundamentally, the 'activist' nature of the programme undoubtedly bears some extra costs: the banks are pushed to work out their problem loans in a very strictly limited time period and then are pushed to adopt certain, mainly much less optimal, alternatives. Ideas have emerged lately for mechanisms which would use the market even more strongly than the Polish scheme. This section does not discuss these alternatives in detail but attention is drawn to the fact that one train of thought suggests a minimum state engagement in the recovery of bad debt by generating a market for it and by creating an infrastructure of risk distribution. One of these ideas has been suggested by Szekely (1993) for countries where some minimum 'real' banking experience is already in place. The programme would involve the Treasury of the country putting up the sum of money it intends to contribute to loan consolidation. Then the banks, on one 18 The government crisis of early 1994 over the policies also illustrates how the 'left turn' in Polish unpleasant risks for economic policy and might derail well. 19 Douglas Kruse has repeatedly drawn my attention aspect.

government's banking politics contains some banking conciliation as to the relevance of this

Lessons for Bad Loan Management

81

side, would be asked to price assets they want to offer for sale in the framework of the programme. On the other side, potential buyers would be asked to do the same pricing exercise with assets they intend to buy. The bid would be valid for those assets which had selling and buying offers close enough to be matched by the implicit rate established by the amount of money put up by the Treasury. The remaining money from the Treasury pool would be spent on assets in an increasing order of proportional losses. The government would act under the understanding that this was a once-and-for-all action. Szekely also suggests that participating firms on the buyers' side could spread the risk by issuing high risk, high yield securities. The potential weakness of this strategy is that because of underdeveloped capital markets and techniques and lack of proper information about the 'commodities' offered, the interest in buying could be extremely low, compromising the whole exercise. It is also difficult to imagine in those circumstances that the programme could cover a large enough share of the problem loans to make it credible, unless it had a large cash contribution from the state up-front. However, the analysis of the secondary market of problem loans and a kind of market-based criterion for state financial support make the scheme worthy of further scrutiny. Another idea also aiming to achieve market-based risk distribution relies on modern Western techniques of asset securitization (Gurgenidze 1993). The proposal would pool together inherited bad loans (i.e. ones for which the banks bear relatively little or no responsibility because the assets were acquired under the old economic regime) in a government agency and securitize them with government backing. The newly created asset-backed securities would then replace the pooled assets in the banks' balance sheets. Aggressive action to recover the assets would be secured by leaving a certain share (let us say, 25 per cent) of the bad assets in the banks and by contracting out the recovery of the pooled assets to their original owners. Securitization would already be arranged with a certain discount which would share the burden and decrease the costs of the programme for the state budget. The author sees the potential advantages of the programme in the above-mentioned incentives and in the very nature of spreading out the risk of problem loans, as well as in the indirect benefits of the banks' acquiring modern technical skills. An unsolved problem of the scheme is that it does not address the issue of the enterprise debt burden. Also, the scheme needs scrutinizing about its technical complexity if it were to be used in a country with an underdeveloped financial sector.

82

Mizsei

However, solutions that mean relatively less state involvement and, it is hoped, cause less additional public debt should continue to be sought as it becomes increasingly clear that the poor credit capabilities of financial sectors in the newly emerging market economies have a huge potential to damage the public finances of these countries with a lasting adverse impact on long-term growth prospects.

6. Conclusions The purpose of this volume is to draw lessons from the experiences of financial sector reform in the fast-track East Central European countries for the benefit of the less advanced reformers. Concerning the issue of bad debts, the Hungarian and Polish examples show us that the governments of the region need a strategy to address the problem. The strategy should be relatively simple, partly because of the changing political environment; it must be relatively cheap, engaging the actors of the economy with proper incentives to work out the problem loans. It should appropriately address the problem of moral hazard (especially acute in the highly concentrated financial sector), i.e. it should carry a credible message that the economic policy is going to be tough (hard budget constraint), that there is no future state support to come, and that there will be general financial sector reform, especially bank privatization. According to these criteria, the Polish scheme is superior to the Hungarian one, and in fact is superior to any existing East European programmes during the transition to a market economy, though it also has some disadvantages. Therefore, the search for even cheaper and more effective solutions is legitimate. Finally, one should emphasize again that the question is relevant only when economic reform starts to address the basic macroeconomic issues, i.e. government support in the form of subsidies and directed credits ceases to exist or gets narrowed down sufficiently, in other words when enterprise liquidity starts to depend on market performance. References Abel, I. and Bonin, ]., 1993, 'Financial Sector Reform in the Economies in Transition: On the way to Privatizing Commercial Banks', paper presented to the International Conference on Development and Reform of the Financial System in Central and Eastern Europe, October, Vienna, Austria.

Lessons for Bad Loan Management

83

Balicka, M. , 1993, 'Amnestia dla molochow', Wprost, 3 December. Brom, K. and Orenstein, M. , 1993, 'Restructuring and Corporate Governance in the Czech Republic', draft working paper, Institute of EastWest Studies, Prague, December. Denton, N. , 'Utilities take up the torch', Financial Times, 17 November 1993. Fan, Q. and Schaffer, M.E. , 1993, ' Government Financial Transfers and Enterprise Adjustments in Russia, with Comparisons to Central and Eastern Europe', Discussion Paper No. 191, Centre for Economic Performance, London School of Economics. Gomulka, S. , 1993, 'The Financial Situation of Polish Enterprises 1992-93 and its Impact on Monetary and Fiscal Policies', mimeo, London School of Economics. Groszek, M. , 1993, 'Og6lna charakterystyka problemu "zlych" dlug6w', mimeo. Gurgenidze, L. , 1993, 'Securitization of non-performing loans in transitional economies', mimeo. Hlavacek, J. and Tuma, Z., 1993, 'Bankruptcy in a Privatizing Economy: The Case of the Czech Republic', Reform Round Table Working Paper No. 8, CERGE, Prague, April. KOPINT, 1993, Konjunkturajeumtes, No. 2, June. Mizsei, K. , 1993, Bankruptcy and the Post-Communist Economies of East Central Eurape. Institute for EastWest Studies, New York, Prague and Budapest. Mitchell, J., 1992, 'Creditor Passivity and Bankruptcy: Implications for Economic Reform', mimeo. Mora, M. , 1992, 'A vallalati valsagok kezelese 1990-91-ben', mimeo. Oblath, G. and Valentiny, A. , 1993, Seignorage, injlacios ado es allamadossag: nehany makrogazdasagi osszefugges magyarorszagi alkalmazasa, KOPINT-DATORG. Rostowski, J. , 1994, 'Dilemmas of Monetary and Financial Policy in Post-Stabilization Russia', in Economic Transformation in Russia, ed. A. Aslund, Pinter, London. Szekely, I. , 1993, 'Economic Transformation and the Reform of the Financial System in Central and Eastern Europe', Discussion Paper No. 816, Centre for Economic Policy Research, London. Varhegyi, E., 1993, 'The Second Reform of the Hungarian Banking System', paper presented to the International Conference on Development and Reform of the Financial System in Central and Eastern Europe, October, Vienna.

COMMENT ON MIZSEl'S 'LESSONS' Rumen Dolmnsky The bad loans issue itself and the strategy for bad loans management cannot be analysed on their own, disregarding their interactions and interrelations with other aspects of economic transition and systemic transformation (see Mizsei Chapter 4). On the contrary, one should assess the success and failures of policy measures related to the bad loans issue only in the context of the whole process of fundamental change in a given country. Accordingly, the strategy for working out the bad loans in every country should be tailored to the specific macroeconomic, legislative and institutional environment in that country. This is also true with regard to the lessons which can be drawn from the experience of the leading reform countries by the second wave countries. Clearly, there are no ultimate and universal solutions to the problem. Another important point which is explicitly made in this chapter is that the magnitude of the problem is different in different countries, i.e. countries may have to face the problem from different starting points. In making this point, Mizsei refers to Rostowski's evaluation of the credit/GDP ratios for Hungary and Poland (45 per cent and 20 per cent, respectively). For comparison, this ratio for Bulgaria in 1991 was 70 per cent and in 1992, 68 per cent. 1 About half of all credits extended are assessed to be substandard or non-performing (Dobrinsky 1994). One obvious consequence of this finding is that even if each country follows an optimal strategy towards working out the bad loans problem, the price which will have to be paid will be inherently different across countries. However, one should also take into account the fact that problems of different magnitude might essentially suggest different approaches towards their working out (this point will be returned to later). Another point which one can implicitly derive from Mizsei's 1 Evaluation based on data from the News Bulletin of the Bulgarian National Bank.

Comment on Mizsei, 's 'Lessons '

85

chapter is that the appearance of the bad loans problem is the result of the rational economic behaviour of economic agents under the unique circumstances in which they were placed in the initial phase of the transition process. At the same time, the strategy for the management of bad loans can also (at least partly) be regarded as a rational expectations game between policy makers and the involved economic agents (mainly enterprises and banks). The success (or failure) of such a strategy will to a very large extent depend on its credibility as well as on whether it transmits the right signals to the right agents about the expected and desired type of behaviour and especially about the required change in behaviour (including incentives and sanctions). If these are not in place, the economic agents (behaving quite rationally and maximizing their own utility) will simply follow their previous policies and the bad loans problem will reappear in the future. I would also like to suggest an alternative view on some issues raised by Mizsei. In analysing the genesis of the bad loans problem, Mizsei disregards one very important circumstance which (at least from a moral point of view) might justify a more 'activist' involvement of the state in working out the problem. This is the fact that in many cases the bad loans inherited from communist times were not generated by the economic agents which own them at present (the enterprises and the banks) but rather by the state itself when bank financing was enforced by the central planning authorities on certain projects instead of direct equity financing by the state (Szekely 1993). This might be considered as an argument (at least from the moral point of view) that the one who generated the problem cannot stand aside from its solution. Another point of possible argument is Mizsei's proposition that 'other "first wave" reformers did not develop any essentially different tools or ideas in approaching the problem. Therefore, it is legitimate to think of the two countries' strategies as "representing the East Central European approaches" '. I think that one cannot apprehend the full picture without paying some tribute to the Czechoslovak (now Czech) approach and experience with this problem. Although similar to the Hungarian one, it is by no means identical to it. Following this line, one can see that in comparing the Hungarian and Polish strategies for dealing with the bad loans problem Mizsei is critical of the Hungarian approach and obviously favours the Polish one. Once again, I would like to stress that such a policy package should be assessed mainly against the specific macroeconomic, institutional and legislative background of the country concerned. As

86

Dobrinsky

pointed out by Balcerowicz (1995), 'a particular policy should always be evaluated by comparing the phenomena which can be ascribed to it with the phenomena which would have arisen as an effect of a realistic alternative policy, conducted under similar initial conditions and in similar circumstances'. In this regard, the real question is whether the Polish approach could have been applied in Hungary and whether it would have generated better results than the current Hungarian strategy. If I was to be asked whether the Polish approach could have been applied in the case of Bulgaria I would probably say 'no', at least as concerns bad loans inherited from the previous regime. The main arguments stem from what I have said in the beginning: the magnitude of the problem as well as the quality of the bad loans in the two countries are substantially different; the economic environment is different (Poland has already bottomed out and reports positive growth for a second consecutive year whereas Bulgaria still has not reached the bottom); the macroeconomic situation is different (Bulgaria does not have the stabilization fund which Poland could use for these purposes), etc. Continuing in this direction, I would like to describe briefly the Bulgarian approach to dealing with the bad loans issue. As mentioned above, Bulgaria appears as the country in the region where the bad loans problem is most severe in relative terms. This is partly due to the fact that a large share of old investment credits (allocated in the 1980s) were denominated in convertible currencies and were not eroded by the high inflation in 1991-3 (Dobrinsky 1994). However, mismanagement and inconsistent policies in the transition period also contributed to the aggravation of the bad loans problem. Probably the most serious mistake was the intention (which was incorporated in a series of economic policy measures) to apply a gradualist, step-by-step approach to the cleansing of the banks' balance sheets. Thus in 1991, the government announced its decision to replace about 25 per cent of the non-performing bad loans allocated before 1990 (at that time they were estimated at about 20 billion leva - $1.13 billion or 15 per cent of GDP) with government securities, and that later on it would continue to deal with the remaining part in a similar manner (eventually in 1992-3, another equal share of bad loans was also replaced with government securities). However, this move gave the wrong signals and expectations to the enterprises of an 'all-forgiving' policy on behalf of the state, the outcome of which was that practically all state-owned enterprises stopped servicing all their old bank credits. By the middle

Comment on Mizsei 's 'Lessons '

87

of 1992, the total sum of substandard and non-performing bank credits allocated before 1990 was evaluated at about 70 billion leva ($3 billion) (Dobrinsky 1994). The result was a rapid deterioration of the balance sheets of the commercial banks to an extent which created a threat for the stability of the whole financial system. The principle outcome of the inconsistent government policy was that the expectations which it generated became a self-fulfilling prophecy. Faced with the dangers of a snowballing financial instability, in December 1993 the authorities adopted a large-scale bad loans management programme (its coverage is estimated to be over $2 billion). Its principal features are as follows. The state issues long-term government bonds to the amount of the outstanding bad loans allocated before the deadline date of 31 December 1990. Banks holding such loans replace them with these bonds. Two types of bonds will be issued: ones denominated in leva and ones denominated in US dollars to replace bad loans denominated in domestic or in convertible currencies, respectively. The interest paid on the leva-denominated bonds is variable, phasing into the basic interest rate of the Bulgarian National Bank within six years, starting from one-third of the basic rate and reaching the basic rate in the seventh year. The interest paid on the US dollar-denominated bonds is equal to the six month LIBOR but is paid in leva on the current leva value of the bond, at the current exchange rate. When enterprise bad loans are cleansed, only interest arrears are written off the enterprise balance sheet (they are transformed into reserves) , whereas the principal is transformed into collectable debt to the state. Enterprises whose debt has been cleansed have to present a business plan of their future activities (including expected solvency position). This will be a precondition for such enterprises to re-enter the commercial credit market. One distinctive and unique feature of the Bulgarian approach is that the bonds replacing the bad loans are designed as quasi-'convertible' in the sense that they can be used as a means of payment in the process of privatization of state-owned enterprises. This mechanism is not finalized yet but it will be based on the idea of the formation of a secondary market for these bonds where they would be traded at (supposedly) market prices. The rationale is as follows. Since the bonds carry interest below the market rate, they are still low quality assets for the banks. The banks will be interested to exchange them (eventually at some loss) for good quality assets. However, it is assumed that the bonds will be

88

Dolninsky

attractive to privatization investors because they would provide them with a discount on their investment. An active secondary market will have two positive effects. The fresh investors' capital will be directed to the banks and indirectly will recapitalize them (by increasing their risk adjusted capital adequacy ratio). At the same time, it will give an impetus for speeding-up the process of privatization. When an investor uses the bonds to acquire stakes in state-owned enterprises that are going to be privatized, the bonds are redeemed. Thus, as more and more bonds are to be channelled through this procedure, the financial burden on the state in terms of interest paid on the bonds will decrease proportionally. Eventually, within several years the whole bond issue could be redeemed in this way (especially if this were to be combined with a more intensive inflow of foreign capital into the country). Of course, it remains to be seen how this mechanism will be implemented in practice.

References Balcerowicz, L., 1995, 'Common Fallacies in the Debate on the Economic Transition in Central and Eastern Europe', in Capitalism, Socialism, Transf g g z z C C "' "' ,.._� "' �

0

I

I

"' "' !il "' "

pl

pl CD

C

I

I pl C:

cu

"' ,.._"' 0

Fig. 1 0. 2 The coupon/US dollar exchange rate, April 1993-January 1994

Banking Reform in Georgi,a

203

against both the rouble and the dollar since its inception. The cash spot dollar/coupon rate reached 100,000 by year-end 1993, soared to about 400,000 in March 1994, exceeded 1,000,000 by early May, and decreased to roughly 800,000 in June (see Figure 10.2). 3 As the coupon depreciated sharply in real terms, a ban on domestic transactions in other currencies has proved unenforceable, and the phenomenon of currency substitution has increasingly emerged under these circumstances. While hard currency remained the only 'inflation-proof financial asset available to the population, the rouble is used in most domestic retail transactions. The coupon almost ceased to circulate, and could hardly be used as a unit of exchange in retail trade, except for bread and a very limited list of other goods and services. This trend toward 'decouponization' of the Georgian economy threatens the stability of the financial sector.

3. Banking reform The Law on Banks and Banking Activity and the Law on Monetary and Credit Regulation were adopted by the Parliament in August 1991, thereby establishing, alongside the Law on the National Bank, a legal framework for the establishment of a two-tier banking system. The Law on Banks and Banking Activity provided a legal basis for the nationalization of local branches of the former Soviet specialized banks and their transformation into independent banking entities, and also allowed the establishment of private commercial banks. To date, these laws comprise the backbone of banking legislation in Georgia. The preparation of a new draft of the Law on Banks and Banking Activity is under way. The Georgian banking system consists of five state commercial banks - Savings Bank, Eximbank, Agroprombank, Industriyabank and Zhilsotsbank. These banks were formed out of the Georgian branches of five specialized banks of the former Soviet Union, whose assets and liabilities in Moscow were taken over by the NBG. The Savings Bank has the largest branch network (eighty-six branches and about 1,100 affiliated offices) and is poised to dominate retail banking now that it is allowed to set deposit and lending rates 3

The recent appreciation of the coupon is conceivably attributable to the market's expectation of a 1:1 conversion rate for the coupon if Georgia were to be reintegrated into the rouble zone, prompted by the Belorussian precedent and repeated official announcements by the authorities that such reintegration is being considered.

204

Gurgenidze

independently. Until December 1 993, the bank's interest rate policy and day-to-day operations were guided by the NBG. Until January 1 994, deposit withdrawals were severely limited. Sharply negative rates of return on deposits and virtual illiquidity of the bank essentially precluded an inflow of new deposits in the last two years. The Savings Bank, however, continues to have an appeal for risk-averse depositors, since its deposits are guaranteed by the state. The bank has evolved into a major player in the nascent inter-bank market, granting loans to private commercial banks. In the absence of a general deposit insurance scheme, other banks are forced to offer substantially higher rates on deposits to attract funding. Corporatization of the Savings Bank began in April 1 994, whereby depositors were offered a chance to convert their deposits into equity. The process is not completed at the time of writing, and it remains to be seen how successful it will be. Eximbank, which is a successor of the local branch of Vneshekonombank-USSR, is the only former specialized bank to have completed the corporatization process. The bank has experienced a gradual erosion of its franchise in foreign exchange operations and external trade financing, due to growing competition from private banks. Corporate banking is still dominated by the three specialized lending banks. Refinancing credits from the NBG comprise the lion 's share of their liabilities, with public sector enterprise deposits accounting for the rest. Despite strongly negative (coupon) lending rates (in real terms) , these banks remain profitable due to the availability of refinancing at even more negative real interest rates. Exposure of Agroprombank, Industriyabank and Zhilsotsbank to the private sector has been gradually increasing, and loans to private firms account for about 30 per cent of their loan portfolios. Lenient licensing procedures, low minimum capital requirements and high profitability of the banking sector in the early years of the transition caused a rapid proliferation of private and quasi-private commercial banks. Currently, over 200 commercial banks are believed to be operating in Georgia. The overwhelming majority of these banks are undercapitalized and fail to comply even with the extremely low minimum capital requirements established by the NBG (see below) . The banks we refer to as ' quasi-private ' are those founded by state enterprises, often primarily for the purpose of obtaining cheap financing. This results in suboptimal credit allocation, as such banks are pressured to lend to their shareholders at low rates which do not reflect the credit risk of the loans. The banks attempt to compensate for this opportunity cost by maintaining

Banking Reform in Georgia

205

excessively high spreads on their retail banking activities. Such inefficiencies notwithstanding, a tier of relatively large and well­ capitalized private banks with sound loan portfolios is beginning to emerge. Such banks usually hold general licences for foreign exchange operations, have rather extensive and growing correspondent account networks, and maintain a competitive edge by means of offering innovative products and services. Overall, banking remains one of the most profitable and dynamic fields of business activity in Georgia. Due to drastic depreciation of the coupon in 1993 and 1994 and the resultant 'decouponization' of the economy, commercial banks conduct an increasing share of their business in hard currency. Virtually all retail lending is in US dollars. Accordingly, dollar deposits are accepted by the majority of banks. Furthermore, a host of non-banking firms, which are engaged in taking hard currency deposits has emerged. These businesses, which usually dub themselves 'investment companies', fall predominantly in the grey area, for they do not possess proper licences to conduct banking activity. Most of these companies either make short-term retail loans, primarily to individuals engaged in 'shopping tourism', or finance their own consumer goods trading activities with the deposited funds, or are pure pyramid schemes. Interest rates on hard currency deposits, offered by such companies, are very high and currently range from 10 per cent to 25 per cent per month. In order to compete, commercial banks are forced to offer deposit rates in the range of 2 per cent to 15 per cent per month. Comfortable spreads of 3 to 10 percentage points are maintained. The NBG is responsible for banking supervision and regulation in Georgia. Since the establishment of the NBG in 1991, some progress has been made in streamlining and consolidating the banking supervision and regulation function within the Banking Supervision Department, created in 1993. None the less, the supervisory capacity of the NBG remains inhibited by lack of adequate staffing, unsatisfactory professional qualifications of poorly trained and inexperienced supervisors, and, most importantly, a deficient regulatory framework. Deficiencies of the regulatory framework are as follows. First, licensing requirements are too lenient. Basic provisions include a minimum capital requirement of 500 million coupons, raised in January 1994 .from 200 million (which was equivalent to roughly $2,000 at a cash spot rate as of end December 1993), a nominal processing fee, and a limit of 35 per cent on an equity stake of a

206

Gurgenidze

single shareholder. Secondly, the existent framework of prudential regulations is obsolete and, consequently, inadequate. All five main prudential regulations have existed prior to the break-up of the USSR, and have hardly been amended since. These are: the capital to liabilities ratio of 5 per cent, household deposits to capital ratio (50 per cent), single borrower lending limit of 50 per cent of the bank's capital (30 per cent for shareholders and bank management), a liquidity ratio of 30 per cent, and a long-term liquidity coefficient, meaningless in the current highly inflationary environment. Both single borrower and insider lending limits seem too high, especially the latter, given the widespread incestuous relationships between banks and state enterprises. The rationale for limiting the banks' deposit-taking ability through a household deposits to capital ratio is unclear. The liquidity ratio puts a heavy toll on banks, since neither cash nor the banks' correspondent accounts (i.e. liquid assets, included in the formula) bear interest. Banks would, therefore, be induced (or, indeed, forced) to maintain high interest rate spreads in order to compensate for an opportunity cost of complying with the liquidity ratio. As far as the banks' capital adequacy is concerned, to this date no provisions have been made for the adoption of risk-adjusted capital guidelines in the letter and spirit of the Basie Agreement on International Capital Standards, although some larger private banks with international exposure have reportedly declared their voluntary commitment to satisfy the Basie risk-adjusted capital requirements by a specified future date. A deficient accounting framework, largely unchanged since the restoration of independence, exacerbates the obsolescence of prudential ratios. For instance, lack of clear guidelines for recording and accounting treatment of paid-in capital in excess of par (value of common stock) and of treasury stock, and failure to recognize subordinated debt and other forms of supplemental capital as such, on the one hand, and exclusion of off-balance sheet liabilities from the formula, on the other hand, make the capital to liabilities ratio, as it is presently calculated, almost meaningless. (One might also wonder what is the rationale of formulating the ratio as capital to liabilities, rather than capital to assets). Most importantly, neither existing prudential regulations, nor even the reserve requirements (raised from 15 to 20 per cent in 1992) have been properly enforced by the NBG. Clearly, unless the NBG is willing to enforce compliance strictly with the ratios by means of charging severe non-compliance penalties (rather than by temporary

Banking Reform in Georgia

207

closures of the violator bank's correspondent account), the overall safety and soundness of the Georgian banking system will not be improved. As a first step, prudential regulations should be made applicable to state commercial banks, which account for over 80 per cent of all lending. So far, they have not even been required to report, let alone comply with, their prudential ratios. Most banks in Georgia have unsound loan portfolios. Total enterprise arrears to banks have increased from 16.7 billion roubles on 1 April 1993 to 27.9 billion roubles on 1 July 1993, growing much faster than inter-enterprise arrears. 4 The proportion of non­ performing loans in portfolios of commercial banks is currently believed to be within a 10 to 40 per cent range. The deficient accounting framework conceals the true magnitude of the problem. The problem of an increase in non-performing loans is common to all transitional economies, although magnitudes vary from country to country. Reasons for the emergence of the bad debt problem include a decline in output, deterioration of payments discipline, inheritance of sizeable amounts of non-performing loans from the communist past, slow development of credit risk assessment skills on the part of lending officers and the extension of credit based on non-commercial considerations (i.e. lending by state banks to unprofitable state enterprises, backed by central bank refinancing). The problem is further exacerbated by poor sectoral and geographic diversification of the banks' portfolios, especially in case of state commercial banks, which largely retain their specialized profile. The fact that the bad debt problem has already reached the aforementioned magnitude in Georgia is particularly alarming, given the macroeconomic instability in the country and the persistence of a soft budget constraint. Given the experience of Central and Eastern European countries, the problem is very likely to reach critical proportions once progress is made with regard to macroeconomic stabilization, and subsidies to the state enterprise sector are diminished considerably or eliminated altogether, resulting in a credit squeeze. None the less, we believe that provided an appropriate regulatory and accounting framework is in place, most banks are capable of accumulating adequate loan loss reserves. Current accounting and taxation principles, however, do not establish clear guidelines and 4 The Committee for Social and Economic Information of the Republic of Georgia, which published these figures, switched to reporting in Georgian coupons from August 1993, making direct comparison difficult.

208

Gurgenidz.e

strong incentives for loan loss provisioning on the part of the banks. The most apparent deficiencies in this regard are as follows. Loans that become overdue as to principal or interest are reclassified as non-performing, unless they are rescheduled on different terms and collateralized. Lack of transparency of this process induces fraud and corruption. As a result, bad loans are routinely rescheduled on terms which include under-the-table arrangements between borrowers and bank managers. Thus, the true condition of the banks' balance sheets is obscured, and bank managers reap illegal profits at the expense of shareholders. Currently used accounting rules do not provide for the accumulation of loan loss reserves. Rather, loan losses, alongside other losses, are charged off against the reserve fund (which is equity, rather than a loan-loss provision or contra-asset account) after a delinquency period of ninety days. Both annual contributions to the reserve fund and the overall size of the reserve fund are limited to a certain percentage of operating income and to a certain proportion of paid-in capital, respectively. These limits are stated in the banks' by-laws. Annual contributions to the reserve fund are deducted from net (rather than pre-tax) income, and thus there are no tax incentives for the banks to build up adequate loan loss reserves. Moreover, in highly inflationary economies such as that of Georgia, the rate of growth of common equity of banks normally lags far behind both the rate of inflation and, importantly, the rate of growth of assets (i.e. loans). Most banks are chronically undercapitalized. Deriving a maximum allowable size of the reserve fund from common equity, rather than from the amount of gross loans, which grow much faster, severely limits banks' ability to build adequate loan loss reserves and increase them alongside the growth of the loan portfolio. Specifically, present constraints essentially prevent banks from using a reserve method of loan loss provisioning. The reserve method seems to be more appropriate than the experience method in the context of rapidly changing conditions in a transitional economy, or a specific charge-off method, which is likely to result in under-provisioning as a result of corrupt loan rescheduling practices, widespread in Georgia. In 1991, two state commercial banks, Agroprombank and lndustriyabank, which had serious bad loan problems, were bailed out by the government at a cost of about 3 billion roubles, or $1.8 billion (the government assumed liability for these banks' non-performing loans). Such bail-outs, as well as various loan consolidation schemes, carried out in more advanced transitional

Banking Reform in Georgia

209

economies (e.g. debt-for-debt swaps and bank recapitalizations), cause severe moral hazard problems, exacerbate fiscal imbalances, and have potential inflationary consequences. Thus, it is desirable to avoid bail-outs of the banking system in the future. 5

4. Concluding remarks Considering the collapse of the real sector of the economy, the volatile macroeconomic environment, the slow pace of structural reforms and the political disarray, some progress with regard to banking reform has been achieved in the Republic of Georgia. A legal framework for the functioning of a two-tier banking system has been established, rudimentary foreign exchange and inter-bank markets have emerged, many new commercial banks have been created, and an array of products and services offered to consumers has expanded considerably. Yet, many critical problems remain unresolved and should be addressed in the near future. First, the National Bank of Georgia should refrain from determining allocation of credit through directed refinancing and should allocate available credit resources through the discount window and by means of auctions. Secondly, steps should be taken to eradicate the distinction between cash and non-cash funds. Thirdly, the regulatory framework should be brought in line with inter­ national standards, reserve requirements and prudential ratios should be strictly enforced, without giving any preference to state commercial banks, and the supervisory capacity of the NBG should be strengthened considerably. Fourthly, accounting and taxation changes, necessary to attract adequate provisioning for loan losses, should be made. Fifthly, temptation to bail out state commercial, let alone private, banks, overburdened with non-performing assets, should be resisted. Sixthly, strict deadlines for corporatization and subsequent privatization of state commercial banks should be set. Finally, once an appropriate regulation and supervisory system is in place and adequate disclosure is achieved, an explicit deposit insurance scheme should be established in place of the current implicit system. 5 However, if government bail-outs are unavoidable, the least distortional scheme must be identified and implemented. A discussion of the relative merits of conventional loan consolidation schemes and a description of an alternative scheme, based on the concept of asset securitization, can be found in Gurgenidze (1993), ' Securitization of Non-Performing Loans in Transitional Economies', publication pending.

210

Gurgenidze

This is by no means an exhaustive list of issues that need to be addressed within the context of banking reform in the Republic of Georgia. It should be emphasized that neither these nor any other steps toward improving the health of the banking sector will succeed unless macroeconomic reform is undertaken in a timely and decisive manner.

References Conway, P. and C. Pant, 1993, 'Georgia: Defining National Independence Within the Context of Economic Interdependence', mimeo. 'Decree of the Head of the Georgian State on Certain Extraordinary Measures of State Monetary and Credit Regulation in the Republic of Georgia', 1993, Svobodnaya Gruziya, 1 December (in Russian) . Government of the Republic of Georgia, 1993, Medium-Term Program of Macroeconomic Stabilization and Systemic Changes in the Republic of Georgia, Tbilisi. Gurgenidze, L., 1993, 'Securitization of Non-Performing Loans in Transitional Economies', manuscript. Helsinki Union of Georgia, 1990, 'Economic Recovery Program of Independent Georgia', manuscript (in English) . Ilia Chavchavadze Society, 1992, 'An Outline of the Economic Program of the Ilia Chavchavadze Society', unpublished campaign material, Atlanta, September. International Monetary Fund, 1993, Georgia: Economic Review, Washington, DC, June. International Monetary Fund, 1992, Georgia: Economic Review, Washington, DC, May. Onoprishvili, D. and Tserteli, M.G., 1993, 'Georgia', in Economic Consequences of Soviet Disintegration, ed. J. Williamson, Institute for International Economics. Orlowski, L., 1993, 'Indirect Transfers in Trade Among Former Soviet Union Republics: Sources, Patterns, and Policy Responses in the Post-Soviet Period', Europe-Asia Studies, Vol. 45, No. 6. Radio Free Europe/Radio Liberty, 1993, Daily Report No. 241, 17 December. 'Regional Aspects of the Soviet Economy', 1991, PlanEcon Report, 15 January.

Report on Social and Economic Condition in the Republic of Georgia in January-October 1 993, 1993, Committee for Social and Economic Information of the Republic of Georgia, Tbilisi, November (in Georgian) .

Report on Social and Economic Condition in the Republic of Georgia in January-July 1 993, 1993, Committee for Social and Economic Information of the Republic of Georgia, Tbilisi, August (in Georgian) .

Report on Social and Economic Condition in the Republic of Georgia in January-June 1 993, 1993, Committee for Social and Economic Information of the Republic of Georgia, Tbilisi, July (in Georgian) .

Banking Reform in Georgia

21 1

'Russian Economic Monitor', 1993, PlanEcon Report, 19 December. Universal Banking Corporation IberiaBank, 1993, Annual Report 1 992, Tbilisi. World Bank, 1993, 'From Crisis to Recovery: A Blueprint for Reforms', Georgi,a Country Economic Memorandum, World Bank, October. World Bank, 1993, 'Georgia: A Blueprint for Reforms', World Bank Country Study, Washington, DC, October.

CHAPTER 1 1

BANKING IN ROMANIA Daniel Daianu

1. Introduction One of the major paradoxes and challenges facing the economies undergoing post-communist transformation is the assumed status and functioning of banks as agents of change. Apart from building up institutions, change means basically the restructuring of economies characterized by the huge misallocation of resources - as a legacy of the command system. It is beyond contention that restructuring is a prerequisite for sustainable stabilization in these economies, bearing in mind the existing massive cross-subsidization and entrenched behavioural patterns and routine of enterprises. A crucial question, therefore, emerges: should banks be in the forefront of change, and are they capable of providing the thrust and the technical back-up for restructuring? The immediate answer would be affirmative since the financial system is presumed to be the main mechanism for resource allocation in a market environment. Besides, capital markets are very much in their infancy (or non-existent) and non-bank financial institutions are so crude (or, likewise, non-existent) that banks get a high profile in any case. 1 A sceptical answer would, however, pay attention to the inherent primitiveness and fragility of the banking sector and would question the realism of vesting such a role (allocative function) in it. 2 1 'The objection that banks do not have the skills to do that right is less compelling than at first sight. The point is not that banks are good at doing this, but that they are likely to be better than anybody else' (Van Wijnbergen 1992) . A similar view is held by J. Corbett and C. Mayer ( 1 991 ) . 2 As J. Rostowski ( 1 993) put it, 'The lack of banking skills in PCEs thus introduces a fundamental instability into their monetary systems. Until this problem is resolved, macroeconomic stability will remain hard to attain. ' Similarly sceptical is R. McKinnon ( 1 991 ) . See also D.H. Scott ( 1 992) .

Banking in Romania

213

The 'narrow' vs. 'universal' banks debate mirrors this concern pertinently. This theoretical debate is still in vogue and presents a certain operational relevance for Russia and other former Soviet republics. For Central and Eastern Europe, its relevance in terms of institutional design has been much diminished under the flow of real events. 'The model issue thus was in fact not a real issue from the very beginning,' asserts Szekely (1993). 3 None the less, in practice - the way banks actually operate and the functions they perform - the tension between the two major viewpoints is easily discernible all over the region: the real functioning of banks is shaped more by the dynamics of the environment than by the intentions of policy makers. For instance, when inflation is very high and real interest rates are negative, when the propensity to save is feeble and long-term investments are rare, banks hardly reallocate resources: ironically, they help to preserve a pattern of resource misallocation. But as such, they do not fit the image of genuine, well-performing universal banks. However, a qualification is needed in this respect: the less attuned to a market environment the economy is, the more ineffective the would-be 'universal' banks are and the more inadequate the process of resource allocation is. Romania's experience validates the last statement. The banking system is confronted not only with relative retardation as to human skills and institutional reform, with the lack of a 'banking culture', with the common problems of segmentation, concentration and an antiquated payments system, but also with the tremendous pressure exerted by the magnitude of resource misallocation (illustrated by the size of implicit subsidies provided via the monetary policy and the exchange rate system). Moreover, the failed attempts at stabilizing the economy added to uncertainty over property rights have not induced major foreign banks to set foot in Romania and help develop solid banking activity. The challenge for local banks and for the Romanian policy makers is daunting. Due to highly negative real interest rates (which help to subsidize the inefficient parts of the economy) and large spreads, banks became surprisingly 'solvent' over the last two years. But, this apparent solvency is very misleading and masks the actual situation. 3 Szekely continues: 'On the other hand, it is not an issue any more, as the so far introduced legal regulations in each country in the region opted basically 'for universal banking, following the EC regulations, and the intentions in the remaining countries point to the same direction' (Szekely (1993)).

21 4

Daianu

Were real interest rates to become positive, an explosion of enterprise bad debt would put most of the banks under terrible strain. One can conjecture that commercial banks have developed a powerful vested interest in preserving negative real interest rates and high spreads and, thereby, they impede restructuring (instead of enhancing it), harm the long-term prospects of the real economy and harm the private sector in particular. This situation has been caused by a 'captive syndrome' on the part of the commercial banks, their oligopoly power and the indecisiveness of the central bank. The real test for the banking system will come when real interest rates turn positive and restructuring will really start biting.

4 2. Highlights on banking reform As has been mentioned, the common blueprint - as part of conventional wisdom in the making - was applied to banking reform in Romania. The fundamental idea of setting up a two-tier system fell into shape during 1991. The initial step was taken in December 1990 when the National Bank spun off its commercial operations, which were taken over by a newly created bank - the Romanian Commercial Bank. Legally, the operation was concluded in March 1991: 5 the new legal framework bestowed 'universal' banking attributes on commercial banks6 whereas the National Bank (as the monetary authority) was granted a high degree of independence and made accountable to the legislative body of the country. The sectoral specialization of banks was done away with and all state-owned banks were turned into commercial companies. There are three factors that have shaped the evolution (the anatomy and physiology) of the Romanian banking system: institutional design; economic policy and the real economy. The 4 For a good overview of banking reform in Romania, see also E. Ghizari (1992) . 5 The new legal set-up consisted of the 'Law Regarding the Status of the National Bank of Romania' (as a central bank) and the 'Law on Banking Activity' securing entry into this sector. 6 Apart from the National Bank, three specialized 'commercial banks' were functioning in Romania at the beginning of 1990: the Romanian Bank for Foreign Trade, the Investment Bank (which turned later into the Romanian Development Bank) and the Bank for Agriculture and Food Industry (currently, the Agricultural Bank) . There was also the Savings Bank (CEC) , which collected household savings and made housing loans. One should mention also the local branches of two major foreign banks: Manufacturers Hanover Trust (currently Chemical Bank) and Societe Generale.

Banking in Romania

215

specific features of the banking system are discussed in these terms in the following sections. 2. 1 Institutional design A most conspicuous trait of the banking system - as it was drawn up is the independence of the National Bank (the central bank). The intention of the policy makers was to set up a monetary authority which, by virtue of its legal status, would be able to pursue the stabilization of the economy unflinchingly and fight inflation. What actually happened in Romania teaches us several lessons. First, monetary policy is easily emasculated if it is not embedded in a properly constructed stabilization policy - if it is not adequately supported by the other components of the policy mix. Thus, in spite of its own inconsistencies (maintenance of cheap credit for certain activities), monetary policy showed clear signs of tightening in 1993-4. None the less, the failure of the overall policy to impose a modicum of the required financial discipline and to initiate enterprise restructuring and the burgeoning of inter-enterprise credits (arrears), have undermined the effectiveness and credibility of monetary policy. Secondly, it is hard to be bold and make good use of the independence provided by a legal status in the absence of a supportive environment. When co-ordination and coherence are missing in the context of the overall policy mix and hurdles are proliferating, it may make sense for those in charge of monetary policy to lower their aspirations and 'accommodate' their measures. That this accommodation creates a policy path that is detrimental to credibility is undeniable, but the trade-off exists and, in most of the cases, partial 'accommodation' is what seems to be a preferred policy response on the part of the monetary authority. Thirdly, the central bank needs the authentic support of the commercial banks in pursuing the goal of stabilization. As the Romanian experience clearly indicates, commercial banks are tempted to play foul and use their oligopolistic domination of the local market, thereby running counter to the policy promoted by the central bank. One could argue that, as commercial entities, banks optimize, which does not necessarily imply moves congruent with the aims of the intended monetary policy.7 On the other hand, policy makers need to be 7 For instance, the National Bank started to auction money to be lent at base rates for refinancing commercial bank needs in the second part of 1993. The starting rate was set at 130 per cent as an indication of a tightening of monetary policy. For several months, although demand exceeded supply by a large margin, banks colluded in appropriating the whole supply of this money at the initial

216

Daianu

cognizant of this state of affairs and formulate decisions accordingly ('when markets do not work, [their expected results] should, maybe, be simulated'). Finally, I am ever more convinced that, especially during transition, however important the institutional framework is, the quality and the team work of the top decision makers is paramount in good policy making, and in reducing the number of major policy blunders, which can have a lasting negative impact. 8 'Underbanking' is another feature of the Romanian environment, even when judged according to regional standards, which is not related to the wait-and-see attitude of major foreign banks only. It can justifiably be argued that due caution has prompted a deliberate attempt to restrict the entry of new banks. 9 Moreover, competition has been reduced and the existing banks have had more leeway to capture the field. Furthermore, segmentation and concentration both in deposits and loan markets continue to plague the system heavily. 10 The case of the State Savings Bank (CEC) is notorious for its deleterious effects on the effectiveness of monetary policy. Aside from the poor quality of its services and due to its virtual monopoly of savings deposits - which made it the main supplier of resources of the inter-bank market and a very obsolete operational and administrative capacity, the CEC undercut the rediscount policy of

rate. Later on, the National Bank itself had to raise the starting rate by starting at a substantially higher level - around 220 per cent in December 1993. 8 The tension which has been developing between the National Bank and high profile members of the government as to the conduct of monetary policy is an open secret (the National Bank has been advocating a much firmer stance on pursuing stabilization). One can make a parallel with the dispute between the Governor of the Central Bank of Russia, V. Geraschenko, and B. Fyodorov, the Finance Minister, and emphasize the striking difference in roles. Therefore, what counts ultimately for good policy are competence and team work. 9 The Governor of the National Bank stated: 'We prefer to have strong banks. We are cautious so as to prevent the mushrooming of small banks and a chain of bankruptcies in the banking sector' (Romanian Bank for Foreign Trade, Country Report, July 1993, p. 11). IO In June 1993, the four biggest (state-owned) commercial banks - the Romanian Commercial Bank, the Romanian Bank for Foreign Trade, the Agricultural Bank (Agrobank) and the Romanian Development Bank accounted for more than 80 per cent of total credits - as against 94 per cent in December 1991. Concurrently, the State Savings Bank (CEC) accounted for over 95 per cent of sight deposits by individuals and 40 per cent of its customers' term deposits.

Banking in Romania

217

the National Bank over the last two years. 1 1 Only in the last quarter of 1993 was this issue addressed in the sense of making the government the main borrower of CEC funds and, thereby, forcing banks to turn to the National Bank for their refinancing needs. 'Underbanking' can be viewed in both a broad and a specialized sense. In a broad sense, it refers to the whole gamut of banking services. In a specialized sense, it means the undersupply of 'universal' banking services. The setting up via a grand design (privatization) of the five Private Property Funds is far away from compensating for the poor performance of the commercial banks as 'universal' banks. The very inadequate payments system is also an outcome of bad design, or 'malign neglect'. The very primitive interbank clearing system as well as the potential benefit commercial banks get by using the 'float money', cause excessive delays (above four weeks is the norm), much resentment among customers and complications for the monetary control exerted by the National Bank. 2. 2 The policy impact Two major policy-related effects are to be pointed out. One has already been mentioned and refers to the way foreign banks perceive the national economic environment. Chemical Bank (formerly the Manufacturers Hanover Trust) and Societe Generale had been operating in Romania since 1975 and 1980, respectively; yet no major foreign banking institution made a similar move after December 1989. This is quite unfortunate for the development of the local banking industry and, further, for the scope and pace of economic restructuring in general (keeping in mind the limited specialized skills of the local human capital). One should mention also that both Chemical Bank and Societe Generale focus on trade financing, paying less attention to project financing. Merchant banking, which is so badly needed in an economy in transformation, is still a very 'scarce commodity' in Romania. 12 1 1 In an effort to stabilize the exchange rate and achieve positive real interest rates, the National Bank raised its rediscount rate to 80 per cent in mid-1992. During that period the CEC was paying 25 per cent and 50 per cent for its sight and time deposits, respectively (a rate of 6 per cent was paid in 1 99 1 ) . 12 A breakthrough occurred i n 1993 when Capital SA, a joint venture investment bank, was formed. The partners are the US merchant bank Wasserstein Perella (27 per cent) ; the EBRO (27 per cent) ; the Romanian Bank for Foreign Trade and the Ion Tiriac Bank (together, 27 per cent) ; Asset Management, a US fund management company (9 per cent) ; and the IFC ( 1 0 per cent) . The venture operates a $15 million initial investment fund.

218

Daianu

The second and main effect is the outcome of overall reform policy. An ineffective stabilization demarche built up with basically no industrial restructuring policy - which would have enhanced the imposition of financial discipline and mitigated the costs of the required adjustment - has not induced banks to behave as active creditors; the structure of incentives has not been altered decisively to fit the goals of transformation. In spite of lending less and of becoming more prudent, the banking system has retraced the old circuit of resource allocation. In this way, the conditions for the reproduction of bad debt have been maintained. One needs to mention also the impact of the overall policy environment on how banks can perform credit analysis. The comparative disadvantage of Romanian banks in this respect is constantly accentuated by high inflation that distorts financial ratios, by negative real interest rates, by inter-enterprise arrears, by incomplete price liberalization, etc. Policy, via the functioning of the banking system, has paradoxically contributed to the fragmentation of the economy. This may sound strange should one think of the integrative functions of the market mechanism. The fact is, however, that the 'rationing' which still characterizes the functioning of the official foreign exchange market has brought about a strong hard currency constraint for many domestic producers - some of them, though positive value-added producers, are forced out of the game (temporarily, at least). It is true that the full retention rights (for hard currency) granted to exporters in May 1992 1 3 have enabled the tying of export and import transactions and, thereby, have created the so-called 'grey market', where the supply of foreign exchange equates its demand at an implicit quasi-equilibrium exchange rate. However, the grey market lacks transparency and the information and transaction costs for many agents can be so high that the latter are not able to sell at competitive prices, or are not able to sell at all. Besides, in a highly inflationary environment hard currency is seen as a very reliable financial asset and many producers (exporters) are much tempted to hoard it: this hoarding puts additional pressure on the foreign exchange market and complicates stabilization policy. 14 13 As a way to stem capital flight and restore some credibility to economic policy after the failed attempt to introduce internal convertibility (in November 1991) and after what was considered - by exporters - as confiscation of their foreign exchange holdings. 1 4 This hoarding should be judged in the context of an increasing dollarization of the Romanian economy - as a result, essentially, of highly negative real interest rates. Thus, in March 1993, 27 per cent of company and

Banking in &mania

219

Apart from the impact of a fall in real wages on the propensity to save, the functioning of the banking system - its inability to help to contain high inflation and, related to this, the discouraging negative real interest rates (see footnote to Figure 11.1) - bears also responsibility for the diminution of this propensity. 15 The relationship between the functioning of the banking system and investment can be assessed from two perspectives. Thus, very high inflation favours short-termism. Additionally, much uncertainty over property rights undermines investors' confidence. It is indisputable that, where restructuring and privatization are very slow, uncertainty is very high. Therefore, where the banking system fares relatively poorly on both accounts - which is the case of Romania investment will be very inadequate. 2.3 The real impact on the economy

The evolution of the banking system cannot escape 'the grip of the real economy'. Though a clever and determined reform of the banking system can bring it to the forefront of change, there is a 'power of structure' 16 of the entire system that keeps things 'in balance' in the end. This power moderates the development of the banking system and is the source of its 'captivity', of the 'creditor passivity' syndrome (Begg and Portes 1991). The bad debt problem is the most telling expression of this relationship which, unless the real sector witnesses deep restructuring and radical changes in behavioural patterns, tends to reproduce itself. This is why addressing the bad debt problem means attacking both its stock and flow aspects. Trying to find a solution to this problem is doubly overwhelming since its roots are to be found in the congenital inefficiency of the former command system, the institutional primitiveness of the emerging market environment, and in the strain under which the transforming economies operate now: this strain is connected with the magnitude of the required structural adjustment, granted the inherited misallocation of resources. Inter-enterprise arrears, also, are an embodiment of this strain; they can be viewed as a perverse defence reaction of a system whose adjustment capacity

household deposits were held in foreign exchange accounts (up from 5 per cent in December 1991) . 15 In relation to GDP, the savings deposits of the population dropped from 30 �er cent at the end of 1990 to 6 per cent at the end of 1992. 1 See also Daianu (1993) .

220

Daianu

may be overwhelmed by external and internal shocks. 1 7 Inter­ enterprise arrears should be of much concern when they cover persistent losses which, sooner or later, become a liability to the government. Consequently, a sensible policy - that aims to enhance the performance of the banking system too - has to take into account the existence of this unusual strain.

3. Bad debt in Romania Two features connected with the bad debt phenomenon stand out in the Romanian case: the relative neglect of the importance of dealing with the real economy, and the current deceptive financial prospects of the commercial banks. To say that enterprise restructuring is painful and politically very sensitive is almost a truism. The danger for policy appears when the magnitude and, relatedly, the difficulty of the task becomes an excuse for inaction. By the latter, is meant the immobilism of a government in devising a strategy for breaking the 'power of structure', the lack of an industrial policy aimed at triggering or enhancing restructuring. One could argue that what counts most is a restructuring viewed as a spontaneous process, undertaken at a grassroots level and not subject to political pressures. 1 8 It is also true that an effective and credible economic policy can turn restructuring into an essentially self-organizing movement when a relatively successful stabilization and a working new legal framework speed up real change, via the effects of capital inflows as well. But when the burden of the past is very heavy, when the policy track record is not favourable and capital inflows are not substantial, a 'hands-off' attitude on the part of authorities invites stagnation. For too long, Romanian policy makers have been dragging their feet in acknowledging operationally the organic link between enterprise reform and the performance of the banking system. Repeated attempts have been made to 'clean the slate' of banks 19 17 Inter-enterprise arrears can be seen as 'temporary quasi-inside money' which endogenizes, partially, the money supply and thereby undermines monetary policy. 1 8 This point can be made forcefully with reference to the Polish experience. A centralized scheme has its own pitfalls which need to be considered. 19 Over the last few years, there have been several rounds of bad debt cancellation dealing with unserviceable loans incurred both before and after December 1989 (see also E.V. Clifton and M.S. Khan (1993), p. 684). It should be stressed that so far no uniform legal rules concerning bad loans classification

Banking in Romania

22 1

encumbered with large non-performing loans, 20 but enterprise restructuring seems to remain an unattainable policy goal. Only lately, however, some real action has been initiated to set up the necessary organic link. The idea is to address the bad debt problem by isolating the main generators of bad debt (and inter-enterprise arrears) and by putting them under surveillance . In this way, it is hoped that banks would be able to behave more like active creditors, 21 and that the restructuring of big offenders will be made easier. To this end, a Restructuring Agency was set up in December 1 993. Prior to this (in june 1 993) , a number of big enterprises in dire financial distress had been put under the regime of financial surveillance. 22 Strikingly, this policy breakthrough (in terms of intentions) is taking place when the financial statements of commercial banks reveal a conflicting picture. Whereas the entrenched view was that banks are undercapitalized and with a large proportion of bad debt, 23 1 993 revealed their return to ' solvency' during the previous year. 24 and provisioning against them have been adopted. Although no official figures on bad debt are available, estimates made by World Bank experts put the total bank non-performing loans in September 1990 at about 7.5 per cent of GDP (over 580 billion lei). This estimate includes loans made before 1991 which the banks assessed as bad (according to Law 7 which stipulated that the government will take over 90 per cent of such loans), as well as loans classified by foreign auditors at end-1991 as non-serviceable, and estimated overdue global compensation loans in 1992. Other rough estimates put current bad debts at more than 11 per cent of credit to the non-government sector. 20 S ome aggregate estimates made for the four major state-owned commercial banks put the size of bad debt, in 1992, at about 30 per cent of their loans portfolio. 21 ln order to deal with the bad debt problem banks will set up special workout units. Within the framework of 'Enterprise Restructuring and Private Sector Development' (EPRD), teams of foreign experts with skills in corporate analysis, legal and accounting expertise, and debt for equity swaps will augment the resources of the banks. 22 By the end of 1993, thirty companies with large arrears to banks have been isolated from the banking system. The Restructuring Agency plans to identify another 250 companies and fifty regies autonomes, which are together responsible for half of the losses in the state sector, in order to restructure them. But the struggle will be extremely fierce; powerful industrial lobbies are already closing their lines of defence. 23 With the exception of the Romanian Bank for Foreign Trade, all state-owned banks were clearly insolvent at the end of 1991. 24 According to foreign auditors' reports, for two of the largest commercial banks, the ratio of equity (after auditors' provisions) to risk-weighted assets is 8 per cent or more, while for two other banks, that ratio turned from negative levels to close to 8 per cent.

222

Daianu

This is a deceptive image because the current financial state of the banks is unsustainable should real interest rates become positive and spreads turn to normal levels. As a McKinsey report indicates, 'The banks have returned to solvency because net interest income has soared, because real loans outstanding have fallen and because the assessed quality of the loan portfolio has improved dramatically. ' All this has occurred against the background of unhealthy trends in the economy. Thus, the growth of net interest income was entailed by very high spreads that burden the real economy. Concurrently, the fall in real loans to companies (by more than a third in 1 992 as compared to 1 99 1 ) 25 meant financial crisis due to falling supplies of bank funds (as depositors attempt to protect their assets when confronted with highly negative real interest rates) and to falling demand as firms prefer to use non-bank credit. As the same report emphasized, 'High inflation and negative real interest rates result in a major subsidy flow from depositors to borrowers, and act to recapitalize companies even if their underlying commercial activity is economically unsound. Perversely, indeed, poor companies' assessed creditworthiness improves much more if they are given additional credit. ' Consequently, the 'solvency' of the banks is unsustainable unless this large-scale type of subsidization is presumed to become a lasting behavioural pattern for the economy - an assumption that is hardly acceptable. 26 It is to be hoped that the measures taken by the government would enhance enterprise reform and would also help to contain inter-enterprise arrears in the struggle against high inflation.

4. The banking system: structure, privatization, deposit insurance and supervision The current structure of the banking system in Romania is shown by the flow chart and the list (see Figure 1 1 . 1 ) . Although the system is overwhelmingly dominated by state-owned banks, private banks have been increasing in number and are making their presence ever more 25

Thereby reducing the risk assets against which equity is required.

26 McKinsey recommended balance sheet restructuring for several banks, the

creation of a restructuring agency to take over bad loans to selected customers in three industrial sectors (metallurgy, machine-building, petrochemicals) and the issue of government bonds in substitution for banks' bad assets. The cost of recapitalization of the banks would have to be covered by the state budget by the issue of government securities of equal face value, so that the budget would initially bear the interest cost of the securities only. To provide true recapitalization, the securities should be marketable after a short period.

Banking in Romania

223

felt. One big advantage of private banks is that they do not carry the burden of large bad debts, since they have avoided lending to big, troubled enterprises. Confronted with limited loan-granting capacity, private banks now concentrate their efforts on expanding their domestic network of branches and attracting foreign partners. A few banks - both state-owned 27 and private 28 - have made equity investments and, in certain cases, act, or plan to act, as strategic investors.

COMMERCIAL BANKS

(25)

FORIEGN BANKS

(2)

AUTHORIZED AND OPERATING

(7)

MAJORITY STATE OWNED BANKS (6)

I

SAVINGS BANK

(1)

I

JOINT VENTURE BANKS

(7)

LOCAL BANKS

( 1 5)

AUTHORIZED AND OPERATING

AUTHORIZED AND PREPARING TO START UP

(1 1 )

(2)

JOINT CAPITAL (STATE AND PRIVATE) BANKS

I

MERCHANT BANK 'CAPITAL'

(1)

REPRESENTATIVES OFFICE

( 1 5) I

AUTHORIZED AND PREPARING TO START U P

(4)

PRIVATE BANKS

(1)

Regulatory and supeivisory authority:

(1)

I

NATIONAL BANK OF ROMANIA

I

Fig. 1 1 . 1 The structure of the Romanian banking system in 1994 Foreign banks: Chemical Bank (Manufacturers Hanover Trust), Sociere Generale. Joint-venture banks: (a)Authorized and operating: Frankfurt Bucharest Bank AG, Banque Franco-Roumaine SA, MISR Romanian Bank SAE, Interconfessional Bank, Bank for Industrial and Commercial Credit; (b) Authorised and preparing to start up: Turkish-Romanian Bank, Bucharest Bank. Local banks: (a) Authorized and operating majority state-owned banks: Romanian Bank for Foreign Trade, Romanian Commercial Bank, Romanian Agricultural Bank, Romanian Development Bank, Postbank, Eximbank; (ii) Authorized and operating joint-stock (state and private) banks: Dacia Felix Bank; (iii) Authorized and operating private banks: Bankcoop, Ion Tiriac Bank, Credit Bank, Mindbank; (b) Authorized and preparing to start up: Bancarom (Romanian Bank), Credcom (Romanian Commercial Credit), Patronbank, Romexterra; Representative office: ltalo-Romena. 2 7 The Romanian Bank for Foreign Trade seems to benefit from the excellence of its management team and a relatively highly qualified staff. 2 8 The Tiriac Bank, in which the EBRD of London has a stake.

224

Daianu

The Romanian authorities have planned to privatize two major banks by 1 995. To this end they are relying on foreign technical and financial assistance (EU PHARE, the World Bank and the EBRD ) . In the first quarter of 1 994, financial audits and portfolio assessments for the previous year are programmed for the state-owned banks, which will supply significant details in deciding the sequence of privatization. 29 A programme for the restructuring and modernization of the Savings Bank (CEC) had not been undertaken by May 1 995 despite intents to do so. Regarding the payment system, a memorandum, which was concluded between the National Bank and the commercial banks in December 1 993, provides for establishing a new settlement system ( that will reduce delays) and full computerization. The Law on Banking (No. 33/ 1 99 1 ) stipulates that commercial banks have to observe the prudential regulations issued by the National Bank regarding capital adequacy, banking supervision and lending exposure. The prudential rules issued so far refer to large exposures (loans to a single borrower may not exceed 20 per cent of capital and reserves of the lending bank) , limits on investments in non-banking companies (up to 20 per cent of capital and reserves of the respective company) , and liquidity ratio (reserve requirements of 1 0 per cent of non-government deposits in domestic currency) . Comprehensive regulations have already been issued covering minimum capital for setting up a bank, licensing new banks, foreign currency exposure, prohibited operations, ownership of banks by non-banks, and insider lending. Law No. 33/ 1991 asks banks to assign to their reserves 20 per cent of yearly gross profits, until the reserve fund equals the capital and then up to 1 0 per cent of profits until the reserve fund is twice the capital. After reaching that level, the reserves will be increased from net profits as decided by the board of directors of banks. The BIS capital adequacy standards are in the process of being adopted. State guarantees are provided for household deposits with the Savings Bank (CEC) . A deposit insurance scheme has not yet been devised. However, Law No. 33/ 1 99 1 mentions the possibility of setting up one or more funds for the insurance of household deposits held by banks.

29 Eximbank (the Export-Import Bank) will not be privatized.

Banking in Romania

225

5. Concluding remarks Those who argue against 'universal' banking for transforming economies, as a way of making the system less unstable, use a strong line of reasoning worthy of consideration. For the case of Central Europe, there is another train of thought which puts forward counter-arguments that cannot be easily dismissed. In my view, the jury is still out on the case of front-runner transforming economies. Very much will depend on the evolution of corporate governance in the banking sector. The case against universal banking becomes more persuasive for the countries which started reforms later, where the institutional ingredients of an advanced market environment are largely lacking. But even there, the flow of events may overtake institutional design 'from above'. None the less, a strategy of 'damage control' by strengthening prudential rules and supervision is more than urgent. Questions can be posed related to this debate and some of them will be mentioned. Thus, which is more important for 'modelling' banks: banking skills or the history of inflation? For there is the experience of the Czech Republic, with very low inflation (until recently and a high ratio of bank credit to non-government loans, 30 as against that of Hungary and Poland. It can hardly be argued that the average Czech banker is better trained than his Hungarian and Polish counterparts; the reverse is more likely to be true. Banks do make many mistakes, but at the same time they have shown much more prudence in lending and a great reluctance to engage in long-term project financing. It seems that what banks do now is, comparatively, less related to the skills of their staff (capabilities to appraise projects) than to what was called 'the power of structure' and government policy. Therefore, on the one hand, banks need to be 'liberated' from 'non-banking constraints'. On the other hand, they should be under the tight rein of prudential regulations and supervision. If one assumes that skills can improve quite significantly and that twinning with foreign banks (a strategy pursued intensely by the Romanian banks in 1993) can compensate - or complement - local resources, the uncertainty in the environment turns out, 30 According to Rostowski, 'Countries with a very low credit to GDP ratio cannot implement the German model' because they cannot rely on debt for the 'external financing of the capital needs of their firms' (see Chapter 2 of this volume) .

226

Daianu

comparatively, to be a difficulty that is harder to overcome. As the experience of foreign banks operating in the region shows, even they can perform surprisingly poorly. Besides, they do not possess inside information. If banks are to be restricted formally in allocating 'resources', solutions have to be sought for providing project finance, because some institutions will have to undertake this activity. Would any institution be able to perform better than 'liberated' banks? How likely is it that this 'liberation' would occur? How to 'liberate' banks effectively is, perhaps, the crucial question to raise.

References Begg, D. and Portes, R., 1991, 'Enterprise Debt and Economic Transformation: Financial Restructuring in Central and Eastern Europe', Discussion Paper 695, Centre for Economic Policy Research, London. Corbett, J. and Mayer, C.P., 1991, 'Financial Reform in Eastern Europe: Progress with the Wrong Model', Discussion Paper No. 603, Centre for Economic Policy Research, London. Clifton, E.V. and Khan, M.S., 1993, 'Interenterprise Arrears in Transforming Economies', IMF Staff Papers, Vol. 40, No. 3, September. Daianu, D., 1993, 'Inter-Enterprise Arrears in Post-Command Economies Thoughts from a Romanian Perspective', mimeo, International Monetary Fund. Ghizari, E., 1992, 'Banking Reform in Romania' in Monetary and Banking Reform in Postcommunist Economies, ed. D. Kemme and A. Rudka, Westview Press. McKinnon, R., 1991, The Order of Economic Liberalization, Johns Hopkins University Press. McKinsey, Inc., 1993, 'Enhancing the Soundness of the Romanian Banking System', Final Presentation to the Financial Restructuring Committee, Bucharest, June. PlanEcon, 1993, PlanEcon Report, 10 November. Rostowski, J., 1993, 'Problems of Creating Stable Monetary Systems in Post-Communist Economies', Eumpe-Asia Studies, May. Scott, D.H., 1992, 'Revising Financial Sector Policy in Transitional Socialist Economies - Will Universal Banks Prove Viable?', Working Paper 1034, World Bank. Szekely, I.P., 1993, 'Economic Transformation and the Reform of the Financial System in Central and Eastern Europe', Discussion Paper 738, Centre for Economic Policy Research, London. Wijnbergen. S. van, 1992, 'Enterprise Reform in Eastern Europe', Discussion Paper 816, Centre for Economic Policy Research, London.

INDEX

Page references followed by 'n' indicate the note (s) on that page, those in italics indicate Figures/Tables. Agricultural Bank Estonia (now Union Bank of Estonia) 1 46, 158 Romania (formerly Bank for Agriculture and Food Industry) 2 1 4n, 2 1 6n Agrobank, Romania (Agricultural Bank) 2 1 6n Agroprombank Georgia 203, 204, 208 (Ukraine) 1 84 USSR 1 66 Antal! government, Hungary 71 Art-B case, Poland 1 20, 123-4, 1 25 Asset Management (US fund management company) 2 1 7n asset securitization 81 asymmetric key method (RSA) 1 30 ATMs 1 39 AvtoVAZbank 1 68 bad debts Czech Republic 27, 93, 94 Hungary 26 Poland 26, 93 policies to deal with, in PCEs 5-6 in post-stabilization PCEs 25-7 Romania 220-2 Russia 26, 1 70-1 solutions to 38 Ukraine 1 89-90 bad loan problems in Bulgaria 8� origin of, in Hungary and Poland 59-69 Balcerowicz programme, Poland 66 Baltic Independent 1 44

Bancarom (Romanian Bank) 223 Bankcoop, Romania 223 Bank for Agriculture and Food Industry, Romania (now Agricultural Bank) 2 1 4n, 2 1 6 Bank fo r Food Economy (BGZ) , Poland 5 , 91 Bank fo r Foreign Economic Relations, USSR see Vneshekonombank Bank for Industrial and Commercial Credit, Romania 223 Bank Handlowy, Poland 91 Banking Act 1 933 (Glass-Steagall Act) 45, 49, 1 1 0 banking conciliation project, Poland 76--80 banking laws Soviet (later Russian) 'On Banks and Banking Activity' 1 67 'On Enterprises and Entrepreneurial Activity' 1 67 'On Property' 1 67 'On the Central Bank of the Russian Federation' 1 67 Ukrainian (proposed) 'Banks, Banking Activity and Banking Supervision' 1 87, 1 88 'The National Bank of Ukraine ' 187 banking reform macroeconomic aspects of stabilization 20-7 microeconomic aspects of stabilization 1 7-20 regulatory structure in FSU 50-5 Romania 2 1 4 bad debt 220-2

228

Index

banking reform, Romania (cont. ) banking system 222-4 economy, real impact on 2 1 9--20 institutional design 2 1 5-1 7 policy, impact of 2 1 7-19 Ukraine bad debts 1 89--90 deposit insurance 193 experience of, to date 1 86-9 Central European experience, relevance to 193-6 main problems 183-5 payments system, structure of 1 90-3 banking regulation 'On Joint Stock Companies in Russia' 1 67 banking supervision Czech Republic 97-9, 1 09, 1 1 1 Estonia 1 59-60 Poland 97-9, 1 09, 1 1 1 Russia 1 76-8 Banking Supervision Department, NBG 205 Banking Supervision Unit, NBP 1 23-4 banking system in CE and FSU inter-bank market 1 3-14 non-cash payments system 9--1 1 Russian, problems in 1 70-8 1 universal banks (German-type) 1 1-1 3 USSR 1 66 Bank Inspection Department, BOE 1 59-60 bank management, Poland and Czech Republic internal controls 99--1 00 reform of l 09 Bank of Estonia see BOE Bank of Industry and Construction, Estonia (now Estonian Commercial Bank of Industry and Construction) 1 46, 1 58 bank privatization 6-9 Czech Republic 9 1 , 92-3, 99, 1 03-6, 1 1 1 Estonia 157-9 FSU countries 107 Hungary 71-3, 75 Poland 79, 9 1 , 92, 1 03, 1 06 Romania 224 Russia 1 07, 108 bankruptcy Estonia 1 43, 1 54 Hungary 66-9, 70, 75 Poland 77, 78 ' Banks, Banking Activity and Banking S}lpervision' 1 87, 1 88 Bank S4ski, Poland 92, 95, 101 Banque Franco-Roumaine SA 223 Basie Agreement on International Capital Standards 206

Basie requirements 96 Basie rules 93-4 BDE 158 Bekeski (Hungarian Finance Minister) 75 BGZ (Bank for Food Economy) , Poland 5, 91 BIS Standard 1 44 BNA network, Poland 1 30 BOE (Bank of Estonia) 1 42, 1 44, 1 46, 1 47, 1 48, 1 49, 1 50, 1 5 1 , 152, 1 53, 1 54, 1 56, 1 57, 158, 161 Bank Inspection Department 159-60 BPH (Kracow-based SOCB) 92 BPT Co 1 1 9 BRE (Export Development Bank) , Poland 90 Bucharest Bank 223 Budapest Bank 72, 73 Bulgarian National Bank 87 Cabinet of Ministers, Ukraine 1 88 Capital SA, Romania 2 1 7n CBR (Central Bank of Russia) 1 66, 1 68, 1 69, 1 70, 1 7 1 , 1 72, 1 73, 1 74, 1 75, 1 76, 1 77, 1 80, 201 , 2 1 6n CEC see Savings Bank, Romania Central Bank of Russia see CBR Central Clearing House, Russia 1 72 Central Computer Centre, Warsaw 1 30 Ceskoslovenska Obchodni Banka (CSOB) 92, 94 Chemical Bank (formerly Manufacturers Hanover Trust) , Romania 2 1 7, 223 Chernomyrdin government, monetary policy 1 76 CIS (Commonwealth of Independent States) 59 clearing and settlement systems, Poland ELIXIR 1 29-3 1 , 131 implementation of legal matters 1 25 SORB 1 3 1-2 SYBIR 1 26-9, 129 Clearing Centre, Moscow 1 72 COMECON 62, 63, 64, 70 commercial banks consolidated balance sheets 22, 23 establishment 42-3 Estonia 1 47-8 Poland 1 1 5-16 Russia 1 68 Committee for Social and Economic Information, Georgia 207 Computer Centres Central Computer Centre, Warsaw 1 30 NBP, Bydgoszcz 1 29 NCH, Warsaw 1 30

Index Consumer Price Index (CPI) , Estonia 1 42, 1 54 Coupon/US dollar exchange rate, April 1993--January 1 994 202 CPI (Consumer Price Index) , Estonia 1 42, 1 54 Credcom (Romanian Commercial Credit ) 223 Credit Bank, Romania 223 'credit crunch' hypothesis 20-5 CSOB (Ceskoslovenska Obchodni Banka) 92, 94 Dacia Felix Bank, Romania 223 debtor consolidation programme, Hungary 74, 76 deficit balance sheets causes and systematic solutions 2-4 deposit insurance (DI) 43, 47-50, 52-3, 54 Estonia 159 Romania 224 Russia 1 78--80 Ukraine 193 DIKs (Czech and Slovak voucher holders) 1 03-4, 1 05 Eastman Kodak 64n EBFRP (Enterprise and Bank Financial Restructuring Progamme) , Poland 5, 38, 39, 96 EBRD (Enterprise Restructuring and Private Sector Development) 92, 1 92, 2 1 7n, 22ln, 223n, 224 Economist, The 1 45 EDI (Electronic Data Interchange) , Poland 1 39 Eesti Pank Qy,arterly Review, Estonia 1 44 EFf, Poland 1 st inter-bank system (see also ELIXIR) 1 39 2nd inter-bank system (see also NBP Wire; SORB) 1 39 single bank system 1 39 Electronic Data Interchange (EDI) , Poland 1 39 Elektrobank, Russia 1 68 ELIXIR (electronic clearing system) Poland 1 27, 1 29-30, 131, 1 39 Enterprise and Bank Financial Restructuring Law 77 Enterprise and Bank Financial Restructuring Programme, Poland 5, 38, 39, 96 Enterprise Restructuring and Private Sector Development (EBRD) 92, 192, 2 1 7n, 22ln, 223n, 224

229

Estonia banking reforms bad debts 154--6 bank privatization 157-9 bank supervision 159-60 Central European experience, relevance to 1 60-1 deposit insurance 159 Estonian experience, relevance to others 1 61-3 initial steps 1 46-9 liquidity crisis, November 1992 1 49-54 main issues 1 43--6 payments system, reform of 157 real economy, effects on banking system 156 Finnish banks in 158 Estonian Commercial Bank of Industry and Construction (formerly Bank of Industry and Construction) 1 58 Estonian Social Bank (formerly Housing and Social Bank) 1 58 Estonian Supreme Council 1 48 EU PHARE 224 policy towards banking 105, 1 1 0, 1 45 'Europe '92' 46 ' evergreening' 1 8-19, 26 Eximbank Georgia 203, 204 Romania 223, 224 Ex-Im Bank, Ukraine 1 84 Export Development Bank (BRE) , Poland 90 Federal Reserve 1 20 Finance Minister (Russia) 2 1 6n financial settlements, Poland 1 1 7-18 financial system, role of in FSU 44--5 Financial Times 64n foreign banks, role of in Poland and Czech Republic 1 01-3 Foreign Trade Bank 71 Estonia (later Vneshekonombank) 1 46 Frankfurt Bucharest Bank AG, Romania 223 Free Democratic coalition, Hungary with Socialists (ex-communists) 75, 76 Fyodorov, B., Finance Minister (Russia) 216n GDP, CNG ratio Czech Republic 60 Estonia 1 54-5 Hungary 60 Poland 34, 35--6 General Credit Bank (VUB) , Slovakia 90, 94, 1 0 1 , 1 02, 1 04, 1 1 1

230

Index

General Savings Bank (PKO BP) , Poland 1 1 6 Georgian economy central bank, establishment of 1 99-203 monetary developments 199-203 present position 197-9 reform 203-209 Geraschenko, V., Governor, Central Bank of Russia 2 1 6n Glass-Steagall Act (Banking Act 1933) 45, 49, 1 10 Gosbank Estonia 1 46, 1 47 USSR 1 66 Hansapank, Estonia 1 58 Harvard Fund 1 04 HDF government 75 Herstatt Bank 1 62 Housing and Social Bank (now Estonian Social Bank) 1 46, 1 58 Hungarian banks financial position, government role in 7(µj financial restructuring 68 liquidation 68 Hungarian Investment and Development Corporation 72 IFC (Romania) 2 1 7n Ikarus, Hungary 63 IMF (International Monetary Fund) 1 20, 1 42, 143, 1 46, 1 5 1 , 1 52, 1 53, 1 83, 1 86, 1 87, 193 Imperial Bank, Russia 1 68 Industriyabank, Georgia 203, 204, 208 inflation Estonia 142 FSU 1 08 Hungary 75 ING (Dutch bank) 92 Inkombank, Russia 1 68, 1 72 Inter-Bank Settlement Department (ISD) 1 3 1 , 1 32 Interconfessional Bank, Romania 223 International Monetary Fund see IMF Investicna Banka, Slovakia 90, 102 Investicni Banka, Czechoslovakia 90, 1 1 1 Investment Bank (later Romanian Development Bank) 2 1 4n Investor Privatization Funds (IPF) , Czech and Slovak Republics 93, 1 03-6 Ion Tiriac Bank (Romania) 2 1 7n, 223n, 223 IPF see Investor Privatization Funds IS B/41 clearing and settlement system (Poland) 1 18--19, 119, 1 24, 1 26, 1 34, 1 35, 1 36

benefits to central bank 1 22-3 to commercial banks 121-2 to customers 1 20--1 threat of liquidation 1 25 ISD (Inter-Bank Settlement Department) 1 3 1 , 132 Italo-Romena, Romania 223 joint-venture banks, Romania 223 Klaus, Vaclav 93 Know How Fund 95-6 Komercni Banka, Czech Republic 90, 92, 95, 1 0 1 , 1 04, 1 07 Konsolidacni Banka, Czech Republic 92 kroon (new Estonian currency, introduced June 1992) 1 42, 1 48 Kuchma, L. (President of Ukraine) 1 86 Law 7 (Romania) 221n Law No. 33/ 1991 (on banking, Romania) 224 Law of Insolvency (Bankruptcy) of Enterprises ( 1 March 1993) 1 77 Law of Banking ( 1 989, Estonia, superseded by 1993 Law) 1 48 Law on Banking Activity (Romania) 2 1 4n Law on Banks and Banking Activity, Georgia {adopted Aug. 1 99 1 ) 203 Law on Monetary and Credit Regulation (adopted Aug. 1991 ) , Georgia 203 Law on the Bank of Estonia 1993 (superseded 1989 Law) 1 48, 158 Law on the National Bank of Georgia (Aug. 1 99 1 ) 199, 203 Law Regarding the Status of the National Bank of Romania 214n LIBOR (London Inter-Bank Offered Rate) 87 loan consolidation, alternatives to 80--2 loan consolidation programme, Hungary 72, 73, 74, 76 Lombard credit 1 37, 1 40 M2, Estonia 1 48, 150 McKinsey report (Romania) 222 macroeconomic stabilization 20--7, 1 09 Manufacturers Hanover Trust, Romania (now Chemical Bank) 2 1 4n Menatepbank, Russia 1 68, 1 72 Message Authentication Code 1 30 Metallgesellschaft 1 06 Metallinvestbank, Russia 1 68 microeconomic stabilization 1 7-20 Mindbank, Romania 223

Index Minister of Finance Russia 1 77, 1 80 Ministry of Finance Federal Czechoslovakia 92 Hungary 63, 65, 67, 70, 7 1 , 73 Poland 76, 77, 79, 95, 97, 98 Ministry ofJustice, Estonia 158 MISR Romanian Bank SAE, Romania 223 MoF see Ministry of Finance Mosbusinessbank, Russia 1 68, 1 72 Moscow Clearing Centre 172 Narva Commercial Bank, Estonia 154 National Bank of Poland Act 1 25 National Banks Czechoslovakia 90 Georgia see NBG Poland see NBP Romania 2 1 4, 215, 2 1 5-16, 2 1 7, 224 Ukraine (NBU) 1 84, 1 87, 1 88, 1 93, 1 94 National Clearing House Company 1 1 5, 1 32, 1 35, 1 37, 1 38 clearing systems 1 26--7 ELIXIR 1 29, 1 39 establishment of 1 25-6 RCHs 1 26 SORB 1 3 1 , 1 32 SYBIR 1 27-9 NBG (National Bank of Georgia, formerly local branch of Gosbank, USSR) 199-203, 204, 206, 209 Banking Supervision Department 205 NBP (National Bank of Poland) 90, 95, 98, 1 1 4, 1 1 5-16, 1 26, 1 28, 1 34, 1 35, 1 36, 1 39, 140 Art-B case 1 23-4, 1 25 Banking Supervision Unit 1 23 IS B/ 41 1 1 8-23 monetary policy and payments system 137-8 payments system, impact on inter-bank money market 1 38 SORB 132 NBP Act, January 1989 1 1 6, 1 25 NBP Computer Centre, Bydgoszcz 1 29 NBP Wire (2nd clearing system) 1 20, 1 3 1 , 1 39 NBU (National Bank of Ukraine) 1 84, 1 87, 1 88, 193, 194 NCH ( see also National Clearing House Company) banking payment services 1 36 Computer Centre, Warsaw 1 30 effect on float 1 35-6 head office 1 29 implementation of clearing systems 1 26--7 Key Management Centre 1 30

231

operational risk 132-4 Organizational Committee 1 25 packages transferred 133 settlement risk 1 34-5 statistical data 1 32 NEB (North Estonian Bank, formed by integration of UBB with NEJSB) 154, 1 58, 1 63 NEB and UBB, rescue package 1 63-5 Nederlandsche Bank 187 Neftekhimbank, Russia 1 68 NEJSB (North Estonian Joint-Stock Bank) 1 47, 1 50, 1 5 1 , 1 54, 155, 1 56, 1 58 NMP, Georgia 197, 201 North Estonian Bank see NEB North Estonian Joint-Stock Bank see NEJSB Novell LAN (PC installed in RCHs, Poland) 1 29, 1 3 1 OECD (Organization fo r Economic Cooperation and Development) 56 'On Banks and Banking Activity' 1 67, 1 75, 1 76 Patronbank, Romania 223 PCEs (post-communist economies) , German-style universal banks in 28-9, 39, 93, 1 1 1 effect of inflation 34 suitability in early state of transition 29-32 in intermediate stage of transition 32-5 PHARE programme 95--6 PKO BP (General Savings Bank) , Poland 91, 1 1 6 PKO SA, Poland 9 1 Polish Banking Association 1 25 Polish banking conciliation project 76--80 Polish banks payments system development of 1 39-40 initiatives for modernization 1 1 9-20 NCH role 1 1 5 requirements for 1 1 3-14 ' twinning' with Western banks 95, 101 Polish Development Bank 106 Ponzi schemes, Estonia 151 POSs (Points of Sale) 1 39 Postbank, Romania 223 Private Property Funds, Romania 2 1 7 Prominvest Bank, Ukraine 1 84 Promstroibank, USSR 1 66, 1 68, 1 74 Raba, Hungary 63 Rating Information Centre 1 73

Index

232

RCH (Regional Clearing Houses) , Poland 1 26, 1 30, 1 3 1 SYBIR 1 27-9, 1 34 recapitalization of banks Hungary 73 Poland 77, 79 Regulation Q 45 Restructuring Agency, Romania (set up Dec. 1993) 221 Retail Price Index (RPI ) , Georgia 1991-3,

1 98

Revalia Bank, Estonia 154 Richter Gedeon 64n Romanian banking system 1994, 223 Romexterra, Romania 223 Rosselkhozbank, Russia 1 74 RPI (Retail Price Index) Georgia, 1991-3

198

in PCEs at time of transition 1 6-- 1 7 privatization 3 3 SORB (clearing system) 1 31-2, 1 34, 1 35, 1 38, 1 39 SRQR (sales related quick ratios) 35 State Holding Company, Hungary 73 Statni Banka, Czechoslovakia 90 STF agreement 1 46 Stock Exchange Prague 92 Warsaw 1 38 Stolichny Bank, Russia 1 68, 1 72 SWIFT 1 29 SYBIR (paper-based clearing system, Poland) 1 26--9, 1 30, 1 3 1 , 1 32, 1 34, 1 35, 1 36, 1 37, 1 38

RSA (asymmetry key method) 1 30 Russian banks banking system problems bad debts 1 70-1 bank supervision 1 76-8 Central and East European experience, relevance to 1 8 1 deposit insurance 1 78-80 payments system 1 71-3 structure 1 73--6 clearing systems, creation of 1 72 commercial banking sector, current state 1 67-70 inter-bank clearing houses 1 72 ownership structure 169 Russian Economic Barometer 1 69, 1 70

Takeda Uapan) 64n Tartu Commercial Bank see TCB taxation issues, Poland and Czech republic 97 Tbilisi Inter-Bank Currency Exchange (founded Spring 1993) 202 TCB (Tartu Commercial Bank) , Estonia 1 47, 1 50, 1 5 1 , 1 52, 1 54, 155, 156 Telbank (data transmission network, Poland) 1 1 9, 1 26, 1 30 TELBANK-P (X-25 protocol) 1 30 TELBANK-T (public telephone network, Poland) 1 28 Tokobank, Russia 1 68 Turkish-Romanian Bank, Romania 223 Tveruniversal Bank, Russia 1 72 TVK 64n

Savings Bank Estonia 1 46, 1 49, 1 58, 159 Georgia 199, 200, 201, 203, 204 PKO BP, Poland 9 1 , 1 1 6 Romania (CEC) 2 1 4, 2 1 6, 2 1 7, 224 Slovakia 102 Ukraine (Sberbank) 1 84, 193 USSR (Sberbank) 1 66, 1 74, 1 75 Siberian Trade Bank 1 72 SOCB (state-owned commercial banks) 3 1 , 32, 38, 39, 93, 95 Bank Sh1ski, Poland, 92, 95, 1 0 1 BPH (Krakow) 9 2 Poland 90- 1 , 9 6 , 99 Socialist ( ex-communist) coalition, Hungary with Free Democrats 75, 76 Societe Generale, Romania, 2 1 4n, 2 1 7, 223 SOE (state-owned enterprises) 1 7, 19, 24, 25, 27, 30, 31 credit allocation of 27-8

UBB see Union Baltic Bank Ukraina Bank (Agroprom) 1 84 Ukrainian banks, payments system 190-3 Ukrsots Bank, Ukraine 1 84 'underbanking', Romania 2 1 6-- 1 7 Unikombank, Russia 1 68 Union Baltic Bank (UBB) , Estonia 1 47, 1 49, 150, 1 5 1 , 1 55, 156 integration into NEB 1 54, 158 liquidation 1 54 rescue package 1 63-5 Union Bank of Estonia (formerly Agricultural Bank) 158 UNISYS 131 UNISYS All 1 30 UNISYS micro A 1 3 1 United Clearing House, Russia 1 72 UPS (installed in all NCH units) 1 29 US Internal Revenue Service 97 US regulatory system 45 USSR, banking system 1 66

Index VEB 1 47, 1 5 1 , 152, 153, 1 55, 156, 1 63, 1 64, 1 65 VEB-Fund 1 53, 1 54, 1 64, 1 65 Vneshekonombank (see also VEB) Estonia (previously Foreign Trade Bank) 1 46, 1 47 USSR (Bank for Foreign Economic Relations) 1 66, 204 voucher privatization Czech Republic 60, 103-4 first wave 94, 1 0 1-2 second wave 92 V-SAT (back-up telecommunication network) 1 3 1 VUB see General Credit Bank, Slovakia Warsaw Central Computer Centre 1 30

233

Stock Exchange 92 Wasserstein Perella (US merchant bank) 2 1 7n WBK (SOCB based in Poznan) 92, 95, 97 Western twin 1 0 1 Western banking evaluation of 45--6 World Bank 76, 79, 95, 1 1 0, 1 83, 221n, 224 loan to Poland 99 X-25 protocol (TELBANK-P) 1 30 Zhilsotsbank Georgia 203, 204 USSR 1 66 Zhivnostenska Banka, Czechoslovakia 90, 95 Zloty Stabilization Fund 96