260 *z| World Bank Discussion Papers Africa Technical Department Series Financial Systems in Sub-Saharan A frica A Comparative Study Paul A. Popiel August I qLq Recent World Bank Discussion Papers No. 203 Poverty Reductiotn in East Asia: The Silent Revoluition. Ffida Johansen No. 204 Managing the Civil Service: The Lessons of Reforin Industrial Countries. Barbara Nunberg No. 205 Designring a System of Labor Market Statistics and In!fonnation. Robert S. Goldfarb and Arvil V. Adams No. 206 Infonnation Technology in World Bank Lending. Increasing the Developmental Impact. Nagy Hanna and Sandor Boyson No. 207 Proceedings of a Conference on Currenicy Substitution and Curreticy Boards. Edited by Nissan Liviatan No. 208 Developing Effective Employment Services. David FretweLI and Susan Goldberg No. 209 Evolving Legal Frameivorksfor Private Sector Developmetit in Central and Eastern Europe. Cheryl W. Gray and Associates No. 210 Marine Biotechnology and Developing Countries. Raymond A. Zilinskas and Carl Gustaf Lundin No. 211 Revitalizing Agricultural Researci in the Sahel: A Proposed Frameworkfor Action. Jan Weijenberg, Josue Dione, Michael Fuchs-Carsch, Adolphe Kere, andJacques Lefort No. 212 Institutional Optionsfor the Provision of Infiastnicture. Christine Kessides No. 213 The Contributionis of Itfrastructure to Economic Developmnent: A Review of Experience and Political Imnplications. Christine Kessides No. 214 From Macroeconomic Correction to Public Sector Reform: The Critical Role of Evaluation. Eduardo Wiesner D. No. 215 Chinia: Reform and Development in 1992-93. Peter Harrold and Rajiv Lall No. 216 The Reforn of Public Expendituresfor Agriculture. Bonni van Blarcom, Odin Knudsen, and John Nash No. 217 Managing Fishery Resources: Proceedings of a Symposium Co-Sponsored by the World Bank and Peruvian Ministry of Fisheries held in Lima, Peru,June 1992. Eduardo A. Loayza No. 218 Cooperatives and the Breakup of Large Mechanized Farns: Theoretical Perspectives and Empirical Evidence. Klaus W. Deininger No. 219 Development of Rural Financial Markets in Sub-Saharan Afnica. Sabapathy Thillairajah No. 220 The Maritime Transport Crisis. Hans J. Peters No. 221 Policy-Based Finance: The Experience of PostwarJapan. The Japanese Development Bank and The Japan Economic Research Insticute No. 222 Macroeconomic Management in China: Proceedings of a Conference in Dalian, June 1993. Edited by Peter Harrold, E. C. Hwa, and Lou Jiwei No. 223 Thie Development of the Private Sector in a Smnall Economy in Transition: The Case of Mongolia. Hongoo Hahm No. 224 Toward an Environmenital Strategyfor Asia. Carter Brandon and Ramesh Ramankutty No. 225 "Fortress Europe" and Other Myths about Trade: Policies touward Merchandise Imports in the EC and Other Major Induistrial Ecotiotnies (and What They Meanfor Developing Countries). Jean Baneth No. 226 Mongolia: Financing Education duritg Economic Transition. Kin Bing Wu No. 227 Cities without Land Mlarkets: Lessons of the Failed Socialist Experitnent. Alain Bertaud and Bertrand Renaud No. 228 Porfolio Investtnent in Developing Coduntries. Edited by Stijn Claessens and Sudarshan Gooptu No. 229 An Assessment of Vulnerable Groups in Mongolia: Strategiesfor Social Policy Platning. Caroline Harper No. 230 Raising the Productivity of Womnen Fanners in Sub-Saliaran Africa. Katfine Saito No. 231 Agricultiural Extension in Africa. Aruna Bagehee No. 232 Telecomnmunications Sector Refonn in Asia: Toward a Neew Pragmatism. Peter L. Smith and Gregory Staple (Continued on the inside back cover.) 260 = World Bank Discussion Papers Affica Technical Department Series Financial Systems in Sub-Saharan Africa A Comparative Study Paul A. Popiel The World Bank Washington, D.C. Copyright © 1994 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing August 1994 Discussion Papers present results of country analysis or research that are circulated to encourage discussion and comment within the development community. To present these results with the least possible delay, the typescript of this paper has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. Some sources cited in this paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Drive, Danvers, Massachusetts 01923, U.S.A. The complete backlist of publications from the World Bank is shown in the annual Index of Publications, which contains an alphabetical title list (with filU ordering information) and indexes of subjects, authors, and countries and regions. The latest edition is available free of charge from the Distribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66, avenue d'1ena, 75116 Paris, France. ISSN: 0259-210X Paul A. Popiel is senior financial economist in the Private Sector Development and Economics Division of the Africa Technical Department at the World Bank. Library of Congress Cataloging-in-Publication Data Popiel, Paul A., 1938- Financial systems in sub-Saharan Africa : a comparative study / Paul A. Popiel. p. cm. - (World Bank discussion paper ; 260. Africa Technical Department series) Includes bibliographical references. ISBN 0-8213-2996-0 1. Finance-Africa, Sub-Saharan. 2. Financial institutions- Africa, Sub-Saharan. I. Title. II. Series: World Bank discussion papers ; 260. III. Series: World Bank discussion papers. Africa Technical Department series. HG187.5.A357P67 1994 332'.0967-dc20 94-29423 CIP AFRICA TECHNICAL DEPARTMENT SERIES Technical Paper Series No. 122 Dessing, Supportfor Microenterprises: Lessonsfor Sub-Saharan Africa No. 130 Kiss, editor, Living with Wildlife: Wildife Resource Management with Local Participation in Africa No. 132 Murphy, Casley, and Curry, Farmers' Estimations as a Source of Production Data: Methodological Guidelinesfor Cereals in Africa No. 135 Walshe, Grindle, Nell, and Bachmann, Dairy Development in Sub-Saharan Africa: A Study of Issues and Options No. 141 Riverson, Gaviria, and Thriscutt, Rural Roads in Sub-Saharan Africa: Lessonsfro,m World Bank Experience No. 142 Kiss and Meerman, Integrated Pest Managemcnt and African Agriaclture No. 143 Grut, Gray, and Egli, Forest Pricing and Concession Policies: Managing the High Forests of West and Centrn Afiica No. 157 Critchley, Reij, and Seznec, Water Harvestingfor Plant Production, vol. 11: Case Stidies and Conclusionsfor Sub-Saharan Africa No. 161 Riverson and Carapetis, Intermediate Means of Transport in Sub-Saharan Africa: Its Potentialfor Improving Rural Travel and Transport No. 165 Kellaghan and Greaney, Using Examinations to Improve Education: A Study in Fourteen African Countries No. 179 Speirs and Olsen, Indigenous Integrated Farming Systems in the Sahel No. 181 Mining Unit, Industry and Energy Division, Strategyfor African Mining No. 188 Silverman, Public Sector Decentralization: &conomic Policy and Sector Investmetd Programs No. 194 Saint, Universities in Africa: Stabilization and Revitalization No. 196 Mabogunje, Perspective on Urban Land and Urban Management Policies in Sub-Saharan Afica No. 197 Zymelman, editor, Assessing Engineering Education in Sub-Saharan Africa No. 199 Hussi, Murphy, Lindberg, and Brenneman, The Development of Cooperatives and Other Rural Oranizations: The Role of the World Bank No. 203 Cleaver, A Strategy to Develop Agriculture in Sub-Saharan Africa and a Focusfor the World Bank No. 208 Bindlish and Evenson, Evaluation of the Performance of T&V Extension in Kenya No. 209 Keith, Property Tax: A Practical Manualfor Anglophone Africa No. 214 Bonfiglioli, Agro-pastoralism in Chad as a Strategyfor Survival: An Essay on the Relationship behteen Anthropology and Statistics No. 218 Mohan, editor, Bibliography of Publications: Technical Department, Africa Region-July 1987 to December 1992 No. 225 Dia, A Governance Approach to Civil Sevice Reform in Sub-Saharan Africa No. 226 Bindlish, Evenson, and Gbetibouo, Evaluation of T&V-Based Extension in Burkina Faso No. 227 Cook, editor, Involuntary Resettlement in Africa: Seleded Papersfrom a Conference on Environment and Settiment Issues in Afiica No. 232 Creightney, Transport and Economic Performance: A Survey of Deveioping Countries No. 238 Heath, Land Rights in Cite d'Ivoire: Survey and Prospectsfor Project Intervention No. 250 Rangeley, Thiam, Andersen, and Lyle, International River Basin Organizations in Sub-Saharan Africa No. 251 Sharma, Rietbergen, Claude R. Heimo, andJyoti Patel, A Strategyfor the Forest Sector in Sub-Saharan Africa No. 255 Mohan, editor, Bibliography of Publications: Technical Department, Africa Region,July 1987 to April 1994 Discussion Paper Series No. 82 Psacharopoulos, Why Educational Policies Can Fail: An Oveview of Selected African Experiences No. 83 Craig, Comparative African Experiences in Impkmenting Educational Policies No. 84 Kiros, Implementing Educational Policies in Ethiopia Discussion Paper Series (continued) No. 85 Eshiwani, Implementing Educational Policies in Kenya No. 86 Galabawa, Implementing Eduicational Policies in Tanzania No. 87 Thelejani, Implementing Educational Policies in Lesotho No. 88 Magalula, Implementitng Educational Policies in Swaziland No. 89 Odact, Implementitng Educational Policies in Uganda No. 90 Achola, Impletneting Educational Policies in Zambia No. 91 Maravanyika, Implementing Educational Policies in Zimbabwe No. 101 Russell, Jacobsen, and Stanley, International Migration and Developtnetit in Sub-Saharan Africa, vol. 1: Overview No. 102 Russell, Jacobsen, and Stanley, International Migration and Developmetnt in Sutb-Saharan Africa, vol. II: Country Analyses No. 132 Fuller and Habte, editors, Adjusting Educational Policies: Conserving Resources while Raising School Quality No. 147 Jaeger, The Effects of Economic Policies on African Agricailture: From Past Harm to Future Hope No. 175 Shanmugaratnam, Vedeld, Massige, and Bovin, Resource Management and Pastoral Institutiotn Buiilding in the West African Sahel No. 181 Lamboray and Elmendorf, Combatting AIDS and Other Sexuially Transtnitted Diseases in Africa: A Review of the World Bank's Agenda for Action No. 184 Spurling, Pee, Mkamanga, and Nkwanyana, Agricultural Research in Southern Africa: A Frameworkfor Actiotn No. 211 Weijenberg, Dione, Fuchs-Carsch, Kere, and Lefort, Revitalizing Agriciltural Research in the Sahel: A Proposed Framewvorkfor Action No. 219 Thillairajah, Development of Rural Financial Markets in Sub-Saharan Africa No. 230 Saito, Raising the Productivity of Women Farmers in Sub-Saharan Africa No. 231 Bagchee, Agricultaral Extension in Africa No. 234 Keck, Sharma, and Feder, Population Growth, Shifting Cultivation, and Unsustainable Agricultural Development: A Case Study in Madagascar No. 242 Biggs, Moody, van Leeuwen, and White, Africa Can Compete!: Export Opportunities and Challenges for Ganments and Home Products in the U.S. Market No. 251 Aryeetey, Baah-Nuakoh, Duggleby, Hettige, and Steel, Supply and Demandfor Finance of Smnall Enterprises in Ghana No. 252 Pinto and Mrope, Projectizing thle Goveratice Approach to Civil Service Reforn: An Institutiotial Environmetit Assessmentfor Preparing a Sectoral Adjustment Loan in the Gambia No. 258 Duncan and Muvandi, The Rate of Fertility Decline in Botswvana and Zimbabue No. 259 Scribner, Policies Affecting Fertility and Contraceptive Use: An Assessment of Tuelve Sub-Saharan Countries CONTENTS FOREWORD ............................................. ix ABSTRACT ............................................. xi ACKNOWLEDGEMENTS ........................................ xiii ABBREVIATIONS ............................................ xv EXECUTIVE SUMMARY AND AGENDA FOR ACTION ................... I 1. MACROECONOMIC AND FINANCIAL BACKGROUND .21 Overview .21 Introduction .21 The Macroeconomic Environment ........................... 21 Characteristics of Financial Systems in Africa .24 2. MACROECONOMIC AND FISCAL ENVIRONMENT .25 Overall Economic Performance ............................ 25 Domestic Saving ..................................... 28 Public Sector Finances .................................. 28 Effects of External and Internal Shocks ....................... 31 3. FINANCIAL DEPTH ..................................... 35 Factors of Financial and Monetary Depth. 35 Obstacles to Financial Deepening in the Study Countries .36 Monetary and Financial Depth in the Study Countries .39 4. FINANCIAL INSTITUTIONS AND RESTRUCTURING .43 Financial Institutions .43 The Lending Sector .43 The Commercial Banks .43 Development Finance Institutions .47 Housing Finance .47 Contractual Savings Institutions .48 Insurance Companies .49 Pension Schemes .49 Postal Institutions ...........0................ 50 Venture Capital .................................... 51 Leasing ......................................... 51 v Securities Markets .................................. 52 Informal Finance ................................... 54 Savings and Credit Potential ............................ 55 Formalizing Informality ............................... 56 Financial Restructuring ................................. 57 Techniques of Bank Restructuring ....... ................. 57 What Needs to be Done ................................ 60 5. MONETARY MANAGEMENT ................................ 63 Shortcomings of Direct Control Systems ....... ................ 64 Moving toward Indirect Management ........ ................. 65 Macrofinancial Environment . ............................. 66 Control of the Money Base ................................ 68 Coordination of Monetary and Fiscal Policy ...... .............. 68 6. PERFORMANCE IN THE STUDY COUNTRIES ..... .............. 69 Monetary Developments in the Study Countries ..... ............. 71 7. REFORM PERSPECTIVES .................................. 77 Reforms and Options . .................................. 78 Improving Existing Instruments .......... .................. 79 New Markets and Instruments ........... .................. 82 Monetary Programming and Decision Making ..... .............. 84 Clearing and Payment Systems ........... .................. 84 Interest Rates ...................................... 84 Strengthening Central Banks . ............................. 86 8. DEEPENING FINANCIAL SYSTEMS ........ .................. 89 Requirements for Successful Reform ........ ................. 90 Toward a Financial Sector Strategy ........ .................. 91 Promoting Securities and Capital Markets ...... ................ 94 REFERENCES .101 TABLES Table 2.1 Macroeconomic Indicators, 1981-90 .................... 26 Table 2.2 Gross Domestic Saving and Gross Domestic Investment, 1981-90 .............................. 29 Table 2.3 Budget Deficit Indicators, 1981-90 ..................... 30 Table 3.1 Monetary and Financial Currency Depth, 1981-90 .... ....... 40 Table 4.1 Share of Commercial Banks in the Financial Sector, Selected Countries, 1990 .44 vi Table 4.2 Costs of Intermediation, 1989 .......... . . .. . . .. . . .. . 46 Table 4.3 Banking Sector Analysis, 1990 ....................... 46 Table 4.4 Insurance Industry, 1989 ........ ................... 50 Table 5.1 Market Conformity of Monetary Instruments .... ........... 66 Table 5.2 Instruments of Monetary Control in Eight Developing Countries .67 Table 6.1 Source of Reserve Money Growth in Non-WAMU Countries .71 Table 6.2 Sources of Reserve Money Growth in WAMU Countries ..................................... 73 FIGURES Figure 1.1 Saving and Investment by Developing Region 1965 to 1987 ..... 23 Figure 2.1 External and Internal Shocks Indicators ..... ............. 32 Figure 4.1 Market Capitalization ......... .................... 53 Figure 6.1 Source of Reserve Money Growth Flow ..... ............ 70 vii FOREWORD Financial systems play a unique and essential role. They mobilize savings and allocate credit, and they limit, price and trade the risks resulting from these activities. Their contribution to the economy depends on the quantity and quality of their services and the efficiency with which they provide them. The 1980s were a difficult period for financial systems in sub-Saharan Africa. The economies of the region underwent steady economic deterioration as the result of an adverse economic environment, a fundamental worsening in the terms of trade, misguided economic policies, and falling investment. This situation adversely affected financial systems in the region and caused financial distress in a number of countries. To restore economic growth, most countries of sub-Saharan Africa initiated stabilization and adjustment programs. Most of these programs included a component aimed at stabilizing and reforming the financial system. This report analyzes and compares financial systems in a number of Anglophone and Francophone countries in sub-Saharan Africa. It follows their evolution during the 1980s and the outset of the 1990s. It compares their performance in monetary management. It examines the performance of banking sectors, the spread of financial distress, and the different methods of financial restructuring used to address this problem. It looks at the process of financial deepening and diversification and assesses the prospects for further financial development. Finally, it surveys the role of informal and semiformal financial mechanisms and institutions. The comparison yields a number of significant findings and a policy agenda for the future. It is evident that financial systems in sub-Saharan Africa need to develop and diversify further in order to support economic adjustment and growth. This will happen if the macro-financial framework remains steady, monetary management improves, legal, regulatory, and prudential frameworks are reformed, supervision is strengthened, property rights and contract laws are effectively enforced, and financial competition is increased. Some sub-Saharan countries have implemented such policy measures and have successfully reformed their financial system. More are far along on the reform path. These achievements suggest that for many financial systems of sub-Saharan Africa, the 1990s could well be a decade of recovery and renewed financial development. Kevin M. Cleaver Director Africa Technical Department ix ABSTRACT This report compares the performance of eleven financial systems of sub- Saharan Africa in the 1980s and early 1990s. It focuses on four areas: monetary management, banking sector performance, development of financial markets, and informal financial mechanisms and instruments. In a policy agenda for the future, the report draws conclusions and makes recommendations that are applicable to most sub- Saharan fimancial systems. xi ACKNOWLEDGEMENTS This study was sponsored by the World Bank's Private Sector Development and Economics Division, Africa Technical Department and the Ministry of Cooperation and Development of France. The Swedish International Development Agency and the International Finance Corporation participated in the financing. The study was conceived by Paul A. Popiel of the World Bank, Private Sector and Economics Division, Africa Technical Department and Paul Bayzelon, Ministry of Cooperation and Development, France. Paul A. Popiel guided the study and wrote the report along with Bernard Laurens, Central Bank of France, Francois Wohrer, Ministry of Cooperation and Development, France and Anand G. Chandarvakar, Consultant. We would like to thank a number of people who contributed to preparatory working papers or with detailed and substantive comments at different stages of the study: Teresa Barger, International Finance Corporation, Capital Markets, Sub- Saharan Department; Rudolf van der Biji, International Finance Corporation, Capital Markets Department; Michel Wormser, World Bank, Private Sector Development and Economics Division, Africa Technical Department; Tyler Biggs, World Bank, Private Sector Development and Economics Division, Africa Technical Department; Fernando Montes-Negret, World Bank, Financial Sector Development Department, Andre Ryba, World Bank, Industry and Energy Operations Division, Country Department I, Africa Region; Gerard Caprio, World Bank, Finance and Private Sector Development Division, Policy Research Department; Dinanath R. Khatkhate, World Bank, Public/Private Sector and Technology Development Division, Asia Technical Department; Per-Erik Berglund, Consultant, Conceptor AB; Professor Sergio Bortolani, Institute of Banking, University of Torino and Finafrica; Jacques Chavez, Caisse Francaise de Developpment; Stephane Cosse, Caisse Francaise de Developpement; Marie-France L'Heritau, Caisse Francaise de Developpement; Professor Michael Connolly, University of Miami; Professor Christian de Boissieu, University of Paris I, Pantheon-Sorbonne; Andre Hovine, Societe Franqaise de Conseil en Developpement; Lars-Ake Larsson, Sala Sparbank; Professor Alan Roe, University of Warwick; Professor Pradeep Srivastava, Harvard University; and Bernard Zimmermann, Ministere de la Cooperation et du Developpement. The paper was edited by Vincent McCullough and desktop publishing support was provided by Joan P. Pandit. xiii ABBREVIATIONS BCCI - Bank of Credit and Commerce International BCEAO - Banque Centrale des Etats de l'Afrique de l'Ouest CFAf - France de la Communaute Financiere en Afrique DFI - Development Finance Institution GDP - Gross Domestic Product MIS - Management Information System NBFIs - Non-Bank Financial Institutions NON-WAMU - Non-West African Monetary Union NPART - Non-Perforning Assets Recovery Trust ROSCAs - Rotating Saving and Credit Associations SME - Small and Medium Scale Enterprise SSA - Sub-Saharan Africa WAEMU - West Africa Economic and Monetary Union WAMU - West Africa Monetary Union xv EXECUTIVE SUMMARY AND AGENDA FOR ACTION If sub-Saharan Africa (SSA) is to provide its growing population with sufficient food, productive jobs and rising incomes, its economies must grow by at least 4-5 percent a year. To do this, the region must improve productivity, increase investment and dramatically raise levels of domestic saving L. Domestic savings in sub-Saharan Africa are strikingly low and have been declining in the 1980s-12.9 percent of Africa's aggregate GDP in 1981-85 and 12.1 percent in 1986-90. This is 28.9 percent below the saving level of South Asia, where average per- capita income is about 10 percent lower than in Africa. Indeed, the level of domestic savings in Africa remains below those of all other developing regions. To raise domestic savings, action must be taken on three key fronts. First, a significant effort is needed to generate public savings by both raising revenues and controlling expenditures. Second, mobilization of private savings must be improved and stepped up. Third, with a view to facilitating these processes and ensuring the efficient mobilization and allocation of financial resources, financial systems have to be reformed and deepened. Characteristics of Financial Systems in sub-Saharan Africa Financial systems in sub-Saharan Africa are shallow, narrow and undiversified. There is a set of general causes for this situation: 3 the early state of development of most economies in the region; e the low income of households and most economic agents; i lack of financial infrastructure; * a history of financial repression and governmental interference in the financial system; 3 lack of appropriate legal, regulatory and prudential frameworks; * instability of the macro-financial environment and lack of steadiness of macroeconomic policies; * weak monetary management; * widespread financial distress in the decade of the 1980s; and * acute scarcity of technical and managerial financial skills. Selected Study Countries This paper is principally based on a comparison of monetary management, financial deepening and of some features of informal and semi-formal financial systems in selected SSA countries. These countries are: Benin, Burkina Faso, Cote d'Ivoire, Ghana, Kenya, Mali, Madagascar, Niger, Senegal, Togo, and Uganda. They are divided in two 1/ World Bank. 1989. Sub-Saharan Africa - From Crisis to Sustainable Growth, A long Term Perspective Study. Washington, D.C.: World Bank. groups. The first group, referred to in this paper as the non-West African Monetary Union (non-WAMU) countries includes Ghana, Kenya, Madagascar and Uganda, countries which have individual Central Banks and currencies. The second group, referred to as the West Africa Monetary Union (WAMU) countries, includes Benin, Burkina Faso, Cote d'Ivoire, Mali, Niger, Senegal and Togo, countries that share a common Central Bank, the Banque Centrale des Etats de l'Afrique de l'Ouest (BCEAO) and a common currency, the Franc de la Communaute Financiere en Afrique (CFAf), which is fully convertible. However, most of the recommendations and conclusions included in this paper are relevant to all sub-Saharan financial systems. External and Internal Shocks: Fiscal and Monetary Implications Some common characteristics combine to render these countries' economies highly vulnerable to external and internal shocks. Per-capita income ranges from US$160 for Uganda to US$720 for Senegal with a median of US$340 for Kenya.Their economies remain very open, with the combined values of imports and exports ranging, in terms of GDP, from 28 percent for Ghana to 88 percent for Cote d'Ivoire and a median of 59 percent for Kenya 2L. This openness combines with the continuing dependence on a few primary agricultural or mineral commodities whose export earnings are subject to unstable world prices; and all countries are price-takers on the world export markets. Moreover, the economies of all study countries remain essentially agro-based which leaves them dependent upon unpredictable weather conditions. As a result, food output remains uncertain and food self-sufficiency, fragile. Their economies and financial systems are burdened by a legacy of inappropriate macroeconomic, fiscal and financial policies, and on top of that, the economies and budgets have little capacity to adjust in a timely fashion to shocks and their consequences. The shallowness and narrowness of the financial system is a factor contributing to this weakness. This vulnerability and its consequences have a specific impact on monetary as well as financial developments. External and internal shocks affect balance of payments and foreign reserves. They also affect budgetary outcomes. It is principally through these two channels that these impacts are transmitted to the monetary and financial system. These effects, in monetary terms, show up first as unforeseen changes in the net foreign assets position of the Central Bank and the banking system. Changes in net foreign asset levels require changes in monetary and fiscal stances. These stances affect in turn the level of credit and or reserve money. Changes in the level of credit and, or reserve money will reflect in movements of money supply. 2/ Based on the period 1981 to 1990. 2 Monetary and Financial Depth In the 1980s, there was virtually no financial deepening in the study countries. In fact, the shallowness of some financial systems increased. With some exceptions, the study countries experienced little or no economic growth, a necessary ingredient of financial development. Moreover, serious deficiencies in their financial infrastructures have been a crucial factor to financial shallowness. In financial intermediation, a strong infrastructure is characterized by an appropriate legal, regulatory and prudential framework; financial strength - in terms of the solvency and liquidity - of individual institutions; diversity of financial institutions; existence of an adequate amount of financial information - including accounting information of uniform and appropriate standards - and its integration into the financial intermediation process; and availability of relevant technologies and human skills. None of these factors are markedly present in the countries selected. In addition, the importance of informal financial markets and mechanisms in both groups of study countries is an obstacle to the deepening of financial systems. Informal mechanisms and institutions intermediate little. They rarely issue interest bearing financial instruments and by nature do not foster financial deepening. The share they occupy in financial operations in the study countries is an important constraint on the development of financial intermediation and hence on the deepening of the financial system. Structure of Financial Systems Like financial depth, the structure of financial systems in sub-Saharan Africa is marked by a history of financial repression and a decade of macrofinancial instability. These restricted financial intermediation to short-term maturities and kept it in the banking sector, preventing financial markets from deepening and diversifying. Monetary management through direct controls that was associated with financial repression has been an important determinant of this lack of deepening and diversification. Financial systems in the study countries are dominated by commercial banks. With the exception of Kenya, commercial banks represent 85-95 percent of the systems' total assets. In Kenya, non-bank financial institutions (NBFIs) developed rapidly in the 1980s to take advantage of loopholes in the Central Bank's monetary, credit and prudential regulations. These are, in fact, "near banks." The predominance of the banking sector does not translate into effective competition within the sector itself. In the study countries, on average more than 60 percent of the sector's assets is owned by, at most, four banks. Such a high degree of concentration blunts competition and represents a virtual oligopoly. At the same time, the banking sector is relatively small in terms of the GDP. Only nine banks from the sub-Saharan region are included in the group of the 2,000 largest banks in the world. There are several differences between the banking sectors of the two groups of countries. First, there are 7.76 branches per million inhabitants in the WAMU group of countries, as opposed to 14.4 per million in the non-WAMU group. Second and 3 consequently, the level of assets per employee is higher in the former. Third, operating costs in the WAMU group of countries are lower than in the non-WAMU group. Fourth, there is an important difference with regard to sources of funds. In the 1980s in the WAMU group, Central Bank refinancing provided about 35 percent of the funds the banks intermediated; in the non-WAMU group of countries, this percentage was about 6 percent. Financial Distress For all study countries, as for most SSA economies, the 1980s was a decade of financial distress, marked by unprecedent banking crises and restructuring. There were a number of causes to this situation. Economic deterioration, fiscal imbalances and macrofinancial instability with fluctuating inflation adversely affected the financial situation of banks and other financial institutions and eroded their capital base. From a microeconomic standpoint, mismanagement took its toll. Improper loan classification, poor monitoring of the loan portfolio and accrual of unpaid interest has been common problems in a number of financial intermediaries in study countries and the sub-Saharan region. Connected and politically motivated lending occurred in both groups of study countries. Fraud was also a contributing factor. Finally, legal, regulatory and prudential frameworks were most of the time inadequate, supervision weak and the "enforceability" of prudential regulation particularly weak. A common feature shared by both groups of countries in the authorities' late and belated recognition of financial distress. In many cases, bank supervisors were either technically unable or politically reluctant to take stock of the true extent of the losses of distressed banks and other financial institutions. Because of lack of political will, of lack of experience with financial restructuring or of the shortage of resources to pay for the losses, decisive action to deal with distressed instituins was postponed as long as possible. This allowed financial distress to spread and become sector wide in most study countries. Conditions for Successful Bank Restructuring For restructuring to be successful, financial fragility must be reduced. Some elements are beyond the scope of any policy (such as the evolution of the terms of trade). But some actions can be taken in specific fields. As illustrated by Ghana, it is possible to create a stable macroeconomic environment, with incentives favoring private-sector growth and competition. Macroeconomic and financial reforms in Ghana and Uganda also show that policy distortions can be reduced, interest rates liberalized, and competition within the financial system can be spurred. In Ghana, considerable progress was achieved in auditing and accounting. In Ghana and Kenya, among countries of the WAMU group, training in financial techniques and skills has been enhanced-but this is only a beginning and with the exception of Ghana, most of the tasks still lie ahead. In addition to these structural issues, some sequencing requirements exist for the re-structuring to be successful. Among them are three prerequisites: 4 relating bank restructuring to the situation of the public sector relating bank restructuring to enterprise restructuring linking bank restructuring to strengthening of bank supervision. In the study countries, too little attention has been paid to the public sector's financial situation in the face of restructuring requirements. If an insolvent institution cannot rescue another insolvent institution, an insolvent government cannot rescue an insolvent financial system. Similarly, a Central Bank which has been weakened by excessive lending to the public sector, with high interest liabilities and net foreign exchange liabilities, does not have the capacity to absorb bank failure losses. For largely indebted countries, such as the study countries, the design of the restructuring should avoid a cumulative deterioration of the debt-to-GNP ratio. Bank restructuring must also be closely linked with enterprise restructuring, as in Ghana. If not, the bulk non-performing loans will probably be never recovered and, moreover, the quality of the bank portfolio may deteriorate again after restructuring. It is useless to restructure the banking sector if prudential regulation and supervision is not also improved. Strengthening supervision is an absolute prerequisite. Both the technical skills and the enforceability of the recommendations of the supervisory body must be improved. A related prerequisite is the strengthening of the enforceability of contractual laws and obligations. Finally, study country economies, like most sub-Saharan economies, have to establish a permanent mechanism and institutional set-up to deal with institutions developing signs of insolvency. As financial systems are reformed and made more market based, the potential for failure will increase. In that context, three mechanisms or procedures should be established. First, an early warning system should be put in place in order to allow identification of "potentially distressed" financial institutions and deal with them before the cost of restructuring becomes unmanageable. Second, there must be a clear procedure, a sequence to deal with banks in difficulties. Third, the study countries and most sub-Saharan economies must establish a permanent mechanism and institutional set-up to deal with non-performing assets. More than one procedure could be involved and different mechanisms could be applied in different circumstances. The important point is that this mechanism, as flexible as it may be, should be institutionalized. Monetary Management As noted earlier, a prerequisite to higher domestic saving as well as to the soundness of financial institutions is macro-financial stability. To an important extent, this is brought about by effective monetary management. The movements of monetary aggregates can be managed by the Central Bank essentially in two different ways: (a) through its regulatory power; or (b) in its capacity of sole issuer of Central Bank money. Each way calls for different type of instruments, direct and indirect. 5 From Direct to Indirect Monetary Management Because of difficulties created by direct monetary control, greater use of indirect monetary instruments for monetary management is generally sought by Central Banks in sub-Saharan Africa and the study countries. Experience in other developing regions as well as in some countries of the region shows, however, that there are some prerequisites for successfully moving away towards implementing monetary management based on indirect methods and instruments. The macro-financial prerequisites are: * a stable macro-financial environment; * a public sector financial situation under control and sustainable in the long term; * no rampant financial distress in the financial system; and * close coordination of fiscal and monetary policies, since in the SSA region the budgetary situation is such a dominant factor to influence monetary management. In addition, in sub-Saharan Africa, transition requires the development of new short-term markets. There must be a minimum market base for supporting the implementation of indirect monetary management. At the very least, an interbank market and an embryonic form of money markets must be in place and competition must be present in these markets and the financial system at large. Another important prerequisite is sufficient autonomy and institutional capacity of the Central Bank. To efficiently manage money through indirect methods and instruments, the Central Bank must be allowed to control reserve money and influence the money base. That is, it must be free to manage its balance sheet in a way to influence the elements of the money base that can be controlled and the other elements whose movements can be offset. Moreover, the Central Bank has to have the appropriate legal and regulatory powers and the ability to enforce its directives. In relation to the functioning of money markets, the Central Bank has to have an information system which provides timely data on bank reserves and money conditions. The Central Bank's staff has to have the skills needed to implement its reserve management objectives through the day-to-day operation in the money markets. Finally, for these factors to play their role in reserve management, the mechanism by which changes in the monetary policy are transmitted to the real economy must exist and be clearly identified by policymakers. There must also be efficient monetary programming associated with an effective decision-making process to support and guide indirect monetary management. A shift to monetary management using predominantly indirect methods and instruments will affect the transmission mechanisms in several ways. It will imply a much greater role for interest rates in allocating (and in balancing the overall supply and demand for) credit. It should increase the importance of competitive forces in loan and 6 deposit markets. Interbank and short-term securities market play an important role in this competitive process. While in the long run, such a shift will benefit the efficiency of financial intermediation, and thereby of the mobilization and allocation of financial resources, there may be some short-term difficulties created by interest rate volatility and threats to bank solvency. Finally, the move to indirect methods and instruments of monetary management may also result in fundamental changes in the links between domestic and international markets: it will tend to link the whole domestic market more closely to international markets. With the lags that characterize monetary management through indirect methods and instruments, the diffusion of transmission channels, the building-up of experience with this type of management and the strength and abruptness of the external and internal shocks, Central Banks will need from time to time to have recourse, temporarily, to direct controls in order to ensure macro-financial stability. Therefore, they should maintain a range of instruments that includes some direct controls. Moral suasion is also important. Key Issue in Monetary Management Reserve money movements and changes in the underlying elements of the balance sheets of the Central Banks in study countries indicate that in the 1980s, they followed a passive monetary policy, dictated by the fiscal and budgetary requirements of the public sector, themselves a result of the lack of fiscal discipline and the desire of governments to offset the effects of internal and external shocks through fiscal relaxation. This fiscal and monetary stance varied from country to country in length and intensity. It was abandoned in Ghana in 1983, when the country embarked on a program of adjustment and financial reform. In general, it was discontinued earlier in the decade by countries of the non-WAMU group as compared to countries of the WAMU group. In the 1980s, however, in most of the study countries, the fact remains that Central Bank control of reserve money, base money and ultimately, money, remained weak. This weakness is a key issue related to monetary management. Close control of reserve money is needed to manage money movements so as to ensure monetary and price stability in the face of strong external and internal shocks and the consequences of their impact. The existing weakness inhibits Central Banks from responding appropriately to these shocks and their consequences. Control of reserve money is under- mined by factors related to the fiscal and institutional environments and by the inadequa- cy of monetary instruments and markets. Prospects for Financial Reform Since in the study countries as in most economies of sub-Saharan Africa, institutional shortcomings are such an important part of the issue and because they cannot be resolved in the short term, there is no first best solution to improve and reform monetary management. The challenge is to identify what best can be done on the basis of those existing elements that can be reformed in the short term and take appropriate 7 steps to strengthen the control of reserve and base money. A shift from direct methods of monetary control to indirect methods and instruments of monetary management is not necessarily the best option. In fact, such a shift in inappropriate circumstances can result in a loss of control of reserve and base money. A greater use of indirect methods and instruments in monetary management should be a medium-term goal for study countries as well as for most sub-Saharan economies. Experience with financial reform in other developing countries shows that a move towards the greater use of indirect instruments can have a reasonable probability of success only when several prerequisites are present. Even then, the critical path is by no means straightforward or smooth. Progress is to an extent a trial-and-error process. This is why an adequate decision-making process guiding the formulation and implementation of monetary policy together with strong and accurate monetary programming are key conditions for success. There are six prerequisites to the initiation of a reform leading to the greater use of indirect methods and instruments of monetary management. These are: * growing macro-financial stability and declining inflation; * a viable and improving fiscal and budgetary situation; * no pronounced or rampant financial distress in the financial system; X an operational and research capacity at the Central Bank allowing for the initiation of monetary programming; * an efficient decision-making process with respect to the formulation and implementation of monetary policy supported by an adequate management information system (MIS) in the Central Bank; and * a strong commitment on the part of the monetary authorities and the government to the reform (which would ensure the necessary coordination between the Central Bank, the Ministry of Finance and other government agencies). Development of Financial Markets Domestic saving mobilization benefits from the deepening and diversification of financial markets. Financial markets encompass money, bond and equity markets, including institutional investors and securities market intermediaries. Specifically, they consist of two main components: the money and capital market. The money market provides funds to meet short-term, fluctuating, financial requirements and principally serves the needs of the bulk users of funds, such as governments, banks, insurance companies and similar institutions. Instruments usually include government securities, public agencies securities, company commercial paper and repurchase agreements. The institutions involved are commercial banks (usually operating an interbank market) merchant banks, dealers, money market mutual funds and the Central Bank. The money market (including the interbank market) provides the 8 foundation for the implementation of monetary policy and the support to effective monetary management. The capital market provides long-term finance for creating means of production, that is, for the establishment and expansion of business enterprises. It consists of the "non-securities" segment and the "securities" segment. The first segment is designed to provide funds from financial institutions directly to enterprises and involves loans and quasi-loans. The second segment is designed to provide long-term finance not exclusively from institutions but also from the general public to business enterprises. The securities segment includes the new issues market and the (secondary) trading market. Obstacles to Financial Market Development Financial markets are undeveloped in sub-Saharan Africa. The financial systems remain heavily bank-centered. Deepening and diversification has not taken place. With the exception of Ghana and Zimbabwe, money markets are in their infancy; with the exception of Kenya, there is no significant development of leasing institutions, housing finance institutions, hire purchase, and retail credit companies. The long-term end of the market remains underdeveloped with small and weak contractual saving institutions and relatively small and inefficient stock exchanges. As a result, money and capital market intermediaries such as dealers, brokers, discount houses and merchant banks did not develop in sub-Saharan Africa. There are several reasons for this. Some are structural, some related to the incentive framework and some to past and present macroeconomic and financial policies. Among the structural reasons, the low level of economic development of sub- Saharan economies is a factor that contributed to the lack of development of financial markets. A second factor is the small size of most of these economies and their narrow resource base, which reflect on the range of markets and instruments. Severe impediments stem from financial repression and macroeconomic and macro-financial instability. Financial repression thwarts financial market development. By directing or manipulating credit flows and monetary parameters, it distorts the mobilization and allocation of financial resources, negating market signals and generally crowding out the private sector. A sound interest rate policy constitutes a key prerequisite for the successful development of financial markets. The very low or negative real interest rates that prevailed for many years in many economies of sub- Saharan Africa and the study countries have masked capital scarcity. In sub-Saharan Africa, a decade of macro-financial instability in the 1980s followed a decade of financial repression, the 1970s. This proved equally hostile to financial market development. Macro-financial instability, including high inflation, creates a high degree of uncertainty. High-risk premia need to be added to the basic remuneration of savings to attract and retain them. But when these risk premia exceed the expected return on productive investment, they curtail saving and financial investment. With inflation, market signals are distorted, and it becomes difficult for entrepreneurs to respond efficiently. Moreover, high inflation reduces the real value of financial holdings, discouraging savers and investors and inducing them to shift their 9 funds into real or physical assets which, at a time of inflation, are a better store of value. It also leads to financial market disintermediation. Among the policy factors, a major and determining one is the lack of an enabling environment for the development of the private sector and financial markets. An enabling environment is one that includes an adequate incentive framework. An adequate incentive framework, in turn, is one that ensures equity in the mobilization and management of funds. Ideally, the allocative influences stemming from the overall incentive system should be neutral, but such a situation is unlikely to prevail in the real world. A second-best and more realistic situation is one in which the incentive system is set to minimize distortions. The principal allocative influences and related distortions that have been identified stem mainly from tax and monetary policies. Fiscal policy affects the relative attractiveness of debt and equity instruments. Monetary policy impinges more on the relative yield advantages of deposit and non-deposit instruments, between bank and non-bank financial products. In most economies of sub-Saharan Africa, and in most of the study countries, the incentive system remains a patchwork of often unrelated or conflicting laws and regulations and its allocative influences are haphazard and distortionary. There are many distortions in the tax treatment of financial instruments. For example in most African countries, there are fiscal disadvantages related to equity investment, such as double taxation of income from corporations. Another common bias is the deductibility of interest, giving debt financing an edge over equity financing. There are similar distortions that affect the fiscal impact on private and government debt, where the latter is de jure or de facto subsidized and bears higher effective yields. Monetary and prudential policies exert their greatest effect on the relative advantages of deposit and non-deposit instruments of banks and non bank financial institutions (NBFIs). These influences are not all in one direction: some features favor deposits and banks while others favor non-deposit instruments and NBFIs. Banks enjoy explicit or implicit deposit insurance; NBFIs and their instruments do not always benefit from this advantage. There are reserve requirements on banks' deposits, but sometimes none on NBFIs' deposits and usually none on money market placements. In most of the study countries, NBFIs can issue paper with or without recourse whereas bank deposits necessarily are balance sheet items. As a matter of practice, bank loans are collateralized, while part of NBFIs' loans and money market advances are not. Banks are usually subject to interest rate ceilings, while NBFIs and money market placement and advances are not. Finally, capitalization requirements are higher for commercial banks than for merchant banks, NBFIs and market intermediaries. This last difference is not only unavoidable, given the fiduciary nature of bank deposits, but even actually desirable in most cases. All sub-Saharan countries suffer to different degrees from these biases. Requirements for Success For the deepening and diversification of financial markets to succeed, there are a few key prerequisites. These prerequisites are mostly related to the commitment of the 10 authorities to the reform, the existence of a enabling business and macro-financial environment as well as the existence of a logistical framework. Experience has demonstrated that no reform aimed at developing financial markets has the least chance of being carried out successfully if political commitment is relenting or vague. Political commitment must be solid, based on a thorough understanding of the long-term advantages of the reform, as well as of the hindrances and obstacles to the reform and their political cost. The successful development of financial markets requires an enabling environment. There are three key factors to an enabling environment that are prerequisites to the development of financial markets: * an incentive framework and a business climate supportive of entrepreneurship and private sector development; * a stable macro-financial environment resulting from stable macroeconomic policies, low inflation, and flexible interest rates; and * an all-out national training effort in managerial and technical skills related to financial operations and to introduce proper accounting procedures, adequate auditing and financial information dissemination. To be successful, financial reform must lead to a healthy competition between the public and private sectors for acquiring scarce financial resources. In particular, governments have to accept the abolition of fiscal biases favoring the public sector. But no matter how well designed is a reform of financial markets development, it stands a considerable risk of getting off-track or bogged down, if the implementation is not guided by bold and efficient decision-making with sufficient political clout and technical support. There is also need for highly-trained capable financial managers who can consistently follow up, evaluate and reorient the reform. Approaches to Financial Deepening Strategies for financial market development have followed two patterns- evolutionary or pro-active. The evolutionary approach consists of letting financial markets develop with the economy and, for the authorities, intervening ex post mainly through improvements or changes in laws and regulations, when major distortions or bottlenecks emerge or when the reaching of a further development stage warrants such interaction. Under this approach, the deepening and the development of the financial system is essentially market driven within an adaptable legal, regulatory and prudential framework. 11 The pro-active approach consist in providing a legal, regulatory and prudential framework that fosters and, when possible, accelerates financial market development by setting up mechanisms, institutions and instruments to that end. Experience has shown that for countries of sub-Saharan countries, the pro-active approach is more appropriate for three reasons. First, market forces are not strong enough to develop financial markets by themselves and a neutral incentive environment is not enough; active promotion and positive incentives are required. Second, a determining factor of the pace and strength of financial markets development is the institution-building capacity of the country concerned. Third, a pro-active approach allows for the use of the most efficient institutional set-up, mechanisms and for the adoption of a technology appropriate to the local conditions and level of development. Once the macro-financial environment is heading towards stability, a pro-active policy aimed at deepening and diversifying financial systems needs to address six issues. First, on which end of the market to focus development; second, what type of institution building path to adopt; third, how best to develop new markets and new instruments; fourth, how to ensure the liquidity of these instruments; fifth, how to organize financial information flows and dissemination; and sixth, what type of fiscal incentives, if any, to provide and on what basis. Development Priorities For sub-Saharan Africa and the study countries, a good part of the answer to the first issue lies in the macro-financial situation. If it has a history of stability that generated low inflation expectations and a sense of stability and continuity among the economic agents and investors, it does not matter too much at which end the process of developing financial markets begins. Country specifics, the institutional structure, and circumstances will be determining factors. In contrast, in a macro-financial environment that has only recently regained macro-financial stability and in which expectations are still unsettled, it appears better to start at the short term of the market. This is for two main reasons. First, economic agents and investors, lenders and borrowers, will more willingly commit or use funds with a short-term horizon; second, money markets strengthen monetary management and effective monetary management ensures macro- financial stability. As stability grows, maturities will lengthen naturally. Institution-Building Path Building the long-term end of the market requires a gradualistic approach. Broadly, this consists of developing first an appropriate legal, regulatory and prudential framework to govern and foster emerging market functions, then promoting institutions and anchoring them. The next step then becomes the further development of the legal, regulatory and prudential framework to adapt it to the progress made and foster the next stage of development. This sequence is then repeated. 12 Diversification of Instruments In narrow financial systems, the principal financial instrument used by economic agents is bank-intermediated loans. The diversification of financial instruments implies the emergence of other short- and long-term financial instruments such as bankers' acceptances, CDs, bills, bonds, other securities and, notably, equities. In particular, crop financing in the form of self-liquidating bills must be developed. Commercial bills and bankers' acceptance based on cooperative or "mutualistic" guarantees should be developed for, among other things, establishing a linkage between semi-formal and formal financial institutions. Extending the maturity of financial instruments is often initiated by extending the maturity of government debt instruments. Since longer maturities are associated with higher risk, in particular in macro-financial environments prone to instability and inflation, government debt, in principle riskless, alleviates to a degree this risk connection. In addition, the existence of a mechanism to ensure the liquidity of these longer term instruments on condition of positive returns is a strong supportive factor to the development of demand for these instruments. Leasing and housing finance have also proved to be good sources of longer term debt instruments owing to the peculiar maturity structure of their financial operations. Diversification of Institutions Developing financial markets requires fostering financial institutions that deal in short-term and longer term assets and liabilities as well as with equity financing. Establishing discount houses should be considered because they serve as a basis for efficient monetary management; can operate a collateralized interbank market (in particular where there is no trust between banks); are the locus for money markets; can introduce new financial mechanisms and instruments; can gradually extend their functions towards merchant banking; and are a nursery for talents and skills related to financial management and financial market operations. The supply of funds for housing should best be increased, not through administra- tive decisions, but through policies encouraging the growth of individual deposits in housing finance and other retail-saving institutions. The development of leasing raises a number of legal, accounting and fiscal issues. The legal issues include protection of leased assets' title, responsibility for maintenance, indemnity in favor of the owner, restrictions on the use of assets, creation of liens and other charges, insurance, rentals and default in payment and right to repossession. Venture capital will take time to develop. A strategy for the future development of contractual saving institutions hinges upon three developments. First, contractual saving demand is an increasing function of income. Second, contractual saving demand is a decreasing function of its price and third, there exists a positive relationship between the overall financial development of a country and the individual's ability and/or willingness to embark on contractual saving 13 contracts. Consequently, governments should rationalize insurance taxation, notably life insurance taxes, in order to stimulate greater financial saving. Information Flows The collection and dissemination of financial information is critical to the functioning of financial markets. In sub-Saharan Africa, this should be organized around three poles. First, in anglophone countries the collection and dissemination of financial information should be organized, starting from the registrars of public companies. Second, in virtually all sub-Saharan countries, the Central Bank stores data on the credit- worthiness of borrowing companies to monitor excessive borrowing exposure and credit- worthiness. Specific financial data from that source could be disseminated. Finally, audit companies have a double role to play. They should assist young enterprises and SMEs reaching a "formal" stage of development in organizing their accounting and establishing balance sheets and income accounts acceptable to financial institutions. Audit companies should also be assisting companies that envisage listing in organizing and disseminating relevant data and information. Strategy A strategy for the development of financial markets in sub-Saharan Africa should be two-pronged. First, it should aim at developing money markets and associated market functions and skills. It should then foster the transfer of some of these functions and skills to the budding securities and capital markets. Second, it should gradually promote the establishment of institutions that lead to the establishment and expansion of securities and capital markets. The strategy should rest initially on the establishment of broker and dealers institutions and guide them towards increasingly complex operations by means of a pro-active and evolving legal, regulatory and supervisory framework. The institution-building sequence should begin with promoting private placement, followed by investment companies and unit trusts. In terms of sequence, private placement, investment companies and unit trusts do not need a stock exchange to function and operate. The existence of a stock exchange facilitates the functioning of unit trusts in several ways, including the valuation of shares that the trust trades. The next step is the promotion of a flexible over-the-counter market with the proper legal, regulatory and prudential framework amongst existing brokers and dealers. Some self-supervision should be delegated to the association of brokers and dealers to prepare them for more formal self-supervisory mechanisms and institutions. Over-the- counter trading will eventually lead to the establishment of more formal trading mechanisms which then can be housed in a stock exchange. Increasing the Supply of Securities A major effort must be made to increase the supply of securities. Existing regulations limiting the type of securities that can be issued should be more flexible; 14 temporary or transient tax incentives for the public issuance of securities could be designed; the expenses of public offering could be made tax deductible. Finally, companies should be allowed to issue new equities at market price. Increasing the Demand for Securities Demand for securities may be expected to react favorably to the elimination of existing tax biases against equities and bonds. However, the reduction of tax biases and the implementation of tax incentives will only be successful if the tax system is respected, since no incentive can possibly compete with tax evasion. Another way to increase the demand for securities is to ensure that contractual saving institutions may invest in securities. Finally, one of the basic conditions for an increase in the demand for securities is the existence of a strong secondary market, or of mechanisms that ensure the liquidity of existing securities. For this reason, it is necessary to establish trading mechanisms when there is no formal secondary markets or to revitalize them where they exist. Improving Securities Intermediaries In accordance with the strategy outlined previously, it may become necessary at a later stage of financial market development to gradually upgrade the status of dealers and brokers into full securities intermediaries. Despite the fear of foreign domination of the market, foreign participation in the provision of these services should be encouraged. Foreign competition, even perhaps limited as to the number of foreign brokerage firms, could add vigor and legitimacy to the market, as well as introduce new market skills. Secondary Market Trading Mechanisms The core problem consists in choosing between two types of market: the order- driven or the price-driven market. For emerging markets, including those in Sub-Saharan Africa, order-driven markets are probably more suitable because of their low trading volume - it reduces the cost of intermediation. Moreover, market risks are much lower for brokers than for dealers. Consequently, a high capital requirement is not necessary for brokers, which makes it easier to develop this type of intermediary. Finally, careful attention should be paid to strengthen and develop clearance and settlement. In virtually all of sub-Saharan African countries, clearing systems are presently unable to face any substantial increase in the trading volumes. As often as possible, a central depository system should be established and payment against delivery should become the standard. Institutional Issues When trading volumes warrant it and the time comes to incorporate a stock exchange, this exchange can be owned and operated as a government agency or could be 15 a voluntary association or a corporation owned and operated by its members. At first sight, a privately-owned stock exchange is probably better suited to the development of a strong secondary market-private members are more willing to develop and promote the securities market than civil servants. However, given the low level of financial infrastructure in sub-Saharan Africa and the quasi-absence of securities firms which could become shareholders of the new exchange in most of the study countries, a public exchange could be considered as a temporary solution before an efficient network of financial intermediaries develop. Finally, very close attention should be paid to supervision, since secondary market development is closely related to this issue. The dynamics of a developing market will be best served by an institutional approach which would call for the concentration of monitoring, supervision and promotional functions in a single agency. Fragmentation of authority has proven to be detrimental to both, development goals and the necessary efficiency to deal with undesirable deviations from established market performance standards. To ensure the more effective enforcement of compliance with legal and regulatory provision, supervisory agencies should be granted considerable autonomy within government hierarchy. Informal and Semi-Formal Finance Informal finance refers to the operations of all lawful but unregulated entities such as rotating (and non-rotating) saving and credit associations (ROSCAS) such as the tontines, njangis and susus in West Africa and the chilembas in East Africa, moneylenders and money collectors and other providers of retail financial services. Its defining characteristic is that it is outside the purview of the legal, fiscal, regulatory and prudential framework of the monetary and financial authorities. A useful primary classification is between the spontaneous autonomous informal system comprising ROSCAS, moneylenders and similar mechanisms, which exists independently of the degree of official control of formal finance and the reactive informal sector which consists essentially of mechanisms to supplement shortcomings and gaps in the formal financial system or circumvent the repression of formal finance. Although there are no aggregative statistics on the scope and share of informal and semi-formal finance in the financial systems of countries of sub-Saharan Africa, the available sample survey and anecdotal evidence suggest that their share in financial activities is considerably higher than in other developing regions. Admittedly, informal and semi-formal finance have inherent drawbacks in economies of scope and scale, maturity-transformation, spatial transfer of savings, predominance of cash transactions, maturity-spectrum of instruments and operations, and shallowness of intermediation. Eventually, these deficiencies can only be remedied by an adequate and efficient formal system which therefore strongly underscores the need for an effective machinery of graduation to a formal status. The systems's potential for retail saving and decentralized lending is defined by its role and operating characteristics. The retail character of saving is evident from the fact that generally the average amount saved is very small and the average balances usually 16 below the minimum balances stipulated by the commercial banks. The evidence confirms the following broad features of informal and semi-formal finance in countries of sub- Saharan Africa. * Informal finance is much more extensive and diverse than formal finance and accounts for most of the financial services, other than term finance, provided to the rural sector. * Spontaneous well-established and successful informal group savings and credit systems exist in almost every SSA country (with the possible exception of Madagascar) based on voluntarism, autonomy, low transaction and information costs, convenience, flexibility, confidence and on non-tangible collateral (often an oral promise). * Informal finance concentrates on deposit mobilization and safekeeping services but the credit side, is based on information advantages and low transaction costs which is also a strong incentive for small savers despite the high interest rates on credit. In fact, the performance of institutions that offer both deposit and credit facilities is perceptibly better than that of those accepting only deposits. * The savings mobilized by rural banks, post office savings schemes and insurance corporations are generally preempted by government. e Semi-formal finance replicates the urban bias of formal finance insofar as it channels rural savings to urban credit. * Paradoxically, despite its proved advantages, there is "general denunciation of informal and semi-formal finance in sub-Saharan countries other than registered cooperatives and credit unions," which in fact demonstrates the challenge facing informal finance in sub-Saharan Africa. Official attitudes have to change drastically before constructive policies can be put in place. Central Banks and Informal and Semi-Formal Finance A conspicuous deficiency in African countries, as in other developing regions, is the lack of a central institutional agency to oversee and promote semi-formal finance and to foster viable linkages with the formal sector. The situation is a curious amalgam of benign neglect, hostility and prejudice because informal finance is associated (mistakenly) with the illegal or underground economy which therefore is regarded as being beyond the legitimate purview of the regulatory authorities. 17 It is imperative that Central Banks in sub-Saharan Africa accord high priority to the creation of properly-designed and staffed special units to research, monitor, promote and refinance the semi-formal financial sector. These cells would need to work in close collaboration with the legal and supervision department and credit information bureau of the Central Bank as well as the ministries of agriculture and cooperatives. For sub- Saharan countries who are members of the Central Bank of West African States (BCEAO), such cells will have to be necessarily located in the national agencies of the respective countries. The surveillance and supervision of semi-formal finance is a legitimate function insofar as an uncontrolled expansion of cooperative credit carries potential risks, as has been demonstrated by the Indian experience. But it is equally, if not more, important for the Central Bank to adopt a positive attitude and constructive policy approach towards semi-formal finance with a view to formulating and facilitating the operations and transactions within the sector as much as fostering market-oriented linkages with the formal sector based on an intermediate financial technology within an appropriate legal framework on the following lines: * Promotion of group borrowing and lending units for commercial banks, like the linkage being established between formal banks and stokvels in South Africa. * Creation of a panel of guarantee brokers from among, say, the tontinears on a salary cum-commission basis (like the compradores of commercial banks in Southeast Asia and stokvels in South Africa) to appraise and guarantee bank credit to ROSCAs. For this purpose, ROSCAs, money collectors and similar institutions and agents could be treated as borrowing entities empowered to borrow on the basis of group guarantee of credit. This formula is presently being tried in South Africa. * Offer of market-oriented rediscount and refinance facilities through the Central Bank for Promissory notes or other credit paper originating from the semi- formal financial sector which could be subject to a guarantee by commercial banks, authorized brokers or responsible guarantors. * Such rediscount and refinance facilities should also be extended to apex organizations of credit unions and cooperative societies. This would, concomitantly, involve appropriate Central Bank inspection and supervision procedures for semi-formal finance. Research Agenda In support of the foregoing policy agenda, the following research agenda could be considered: 18 * The legal and regulatory framework for informal and semi-formal finance, preferably as part of a comprehensive review of financial and bank legislation. * The positive role of the Central Bank in relation to informal and semi-formal finance and the scope for inter-Central Bank cooperation. * The actuarial and organization issues in promoting small-scale life insurance. * The nature, extent and possible correctives of urban bias in the financial system. * The scope for possible market-oriented linkages between the informal, semi- formal and formal sectors and for facilitating the graduation process. * Assessment of the technical assistance requirements for informal and semi- formal finance. * Design of proper attitudinal and investigative surveys. * The supportive payments and remittance system (for example giro). 19 1. MACROECONOMIC AND FINANCIAL BACKGROUND Overview This study examines and compares the situation and performance of the financial systems of eleven sub-Saharan African economies divided in two groups of the West African Monetary Union (WAMU) and the others (non-WAMU). The study is divided into four main sections. Chapters one and two provide basic background regarding the macroeconomic, saving and investment performance of the eleven economies especially during the period of the 1980s and early 1990s. Chapters three and four provide an overview of the existing financial sector structures in the study countries and some of the main roblems that affected them. Chapter five, six and seven are all concerned with different aspects of the problems of monetary management. Thus Chapter five provides some general discussion of alternative approaches to monetary management and especially of the advantages and disadvantages of indirect versus direct methods. Chapter six looks at the specific experiences of monetary control in the study countries and in particular at the practical problems involved in achieving control of reserve money. Chapter seven then provides a variety of suggestions for reforming present arrangements so as to achieve tighter monetary controls consistent with the practical constraints that most of the countries face. Finally chapter eight considers various aspects of the problem of achieving a deeper and more diversified financial system. The executive summary includes an agenda for action. Introduction Formal financial systems in sub-Saharan Africa are narrow, shallow, and undiversified, reflecting a decade of deteriorating economic conditions, macrofinancial instability, fiscal imbalances, and financial distress. Countries in the region are unable to mobilize and allocate financial resources efficiently-in particular, to foster domestic saving. Moreover, weak legal, regulatory, and prudential frameworks for financial intermediation, together with institutional shortcomings, have encouraged informal finance. The Macroeconomic Environment Sub-Saharan Africa (excluding Namibia and South Africa) comprises forty seven countries covering 23,000 square kilometers and with a population of about 500 million (1991). Despite a diversity of languages, religions, ethnic groups, and systems of government, virtually all countries of the region have gone through the same kind of economic evolution and difficulties. After independence, most countries in sub-Saharan Africa opted for a development pattern that relied on high levels of government intervention in the economy. The results, 21 distorted incentives and repressed financial systems, a constrained private sector, and high internal and external indebtedness. The economic history of the continent can be broadly divided into three periods: 1960-70, when economic growth outpaced population growth; 1973-80, when growth slowed and per-capita income growth was often negative; and 1981-90, which saw the stagnation or decline of most economies and the introduction of stabilization and adjustment programs. The economic deterioration of the 1980s was due not only to the perverse outcomes of the earlier development strategies but also to a series of external and internal shocks, including adverse weather conditions, chronic deterioration in terms of trade, increased international interest rates, and oil price hikes. Internal shocks included civil wars and unsustainable fiscal imbalances. There was also another, hidden shock. The principal economic activity of the region is agriculture, with more than 70 percent of the population depending on it for their livelihood. In 1975-90, however, the growth in population outstripped that in food production, which represents a continuing internal shock. Moreover, in the 1980s, the industrial sector in most countries stagnated or declined, worsening the economic situation. For most African economies, one or two commodities account for more than 50 percent of exports, making the economies sensitive to changes in terms of trade. For the whole region, terms of trade declined in the 1980s-by 4.5 percent a year for 1980-85 and by 1.1 percent for 1986-90. Export volumes also dropped in the 1980s. Among the major causes of the decline of exports was the government control over producer prices of major export crops, the indirect effects of overvalued exchange rates, and droughts. Because of both the deterioration in terms of trade and the decreases in export volumes, external and internal borrowings throughout sub-Saharan Africa have risen significantly. Total external debt of the region rose from about US$6 billion in 1970 to almost US$174 billion in 1991, with most of the increase occurring in the 1970s. In 1991 this debt was equivalent to 111 percent of the region's aggregate gross national product (GNP). This is high compared to other developing regions-for instance, 40.8 percent for Latin America. External borrowing per capita, too, is high-US$995 for Africa (per capita income of US$340) against US$350 for Latin America (per capita income of US$2,180). Gross investment peaked in sub-Saharan Africa in the mid-1970s at about 20 percent of the gross domestic product (GDP) and declined afterward (Figure 1.1). Why? The chronic deterioration in terms of trade and reductions in real income, growth in government deficits and reduced public domestic saving, the steady erosion of private domestic saving, and the tapering off of external borrowing in the 1980s. Foreign private capital flows to the region also dwindled in the late 1970s and early 1980s, mainly because of the lack of investor confidence, repressed financial systems, and the absence of financial markets. Foreign investment, US$794 million in 1980, fell to US$530 million in 1985; and in 1990, there was a net capital outflow of US$18 million. 22 Figure 1.1 Saving and Investment by Developing Region, 1965 to 1987 (percentage of GNP) E 1%5-73 Gross domreslc saving (S ) Gross domestic investmen (I ) [El 1974-82 S - I * 1983-87 201 2.1 [ -05 l$m//UXWz7Xg L~~~-0 . East Asia ...-::. - . . ...;- .- 0.4 . . v/X///Hg/X//Xi/t<7 -2.2 X /g / / South Asia -4.0 EuxLpe. I --04 the Middle East -[- - j 1 - - and rthAfrica 1.4 tatin America and - / 1 - the Carbbean L 3.7 30 20 10 0 0 lO 20 30 Note: Saving and investment are measured at current prices. Source: World Bank-, World Development Report 1989, New York: Oxford University Press. 1989. 23 Characteristics of Financial Systems in Africa When compared to financial systems in other developing regions (notably South Asia, Latin America, and East Asia), the financial systems of sub-Saharan Africa show some distinctive features. African financial systems are both shallower and narrower. In most African countries, there is considerable concentration in the banking sector, with little intrasectoral competition. Moreover, the capital base of banks and of relatively scarce non-bank financial institutions is low, rendering these institutions highly geared. On top of that, the systems are fragmented, and semi-formal and informal financial mechanisms and institutions are particularly important. The general causes for the unsatisfactory situation of the financial systems in Africa are: * The early stage of development of African economies * The low income of households and most economic agents * Lack of appropriate legal, regulatory, and prudential frameworks * Instability of the macrofinancial environment and the unsteadiness of macroeconomic policies * Weak monetary management * A history of financial repression and government interference in the financial system * Great scarcity of technical and managerial financial skills. Since the late 1980s a growing number of African countries have initiated stabilization and structural adjustment programs to correct imbalances affecting their economies and to lay the foundation for self-sustained growth. By 1993, thirty of the forty-five countries of the region were implementing programs of stabilization and adjustment, to stimulate growth and exports. Most programs include policy measures to reform monetary and financial management, restructure ailing banking sectors, and deepen financial systems. One major objective is to enhance of the efficiency of financial resources mobilization and allocation, with particular emphasis on strengthening and expanding domestic savings. 24 2. MACROECONOMIC AND FISCAL ENVIRONMENT This chapter surveys the macroeconomic and macrofinancial environments in eleven countries of sub-Saharan Africa-Benin, Burkina Faso, C6te d'Ivoire, Ghana, Kenya, Madagascar, Mali, Niger, Senegal, Togo, and Uganda-divided in two groups. The first group includes countries that do not belong to the West African Monetary Union (WAMU)-Ghana, Kenya, Madagascar, and Uganda-and which have individual central banks and currencies. The second group, the WAMU group, includes Benin, Burkina Faso, C6te d'Ivoire, Mali, Niger, Senegal, and Togo. These share a common central bank, the Banque Centrale des Etats de l'Afrique de l'Ouest (BCEAO), and a common currency, the franc de la Communaute Financiere en Afrique (CFAf). Notwithstanding these groupings, the countries have many features in common. The 1991 per capita income (GNP) of nine of the eleven countries ranges from US$160 for Uganda to US$410 for Togo. Only C6te d'Ivoire, at US$690, and Senegal, at US$720, have higher per capita incomes. In all countries agriculture is the predominant activity and by far the largest source of employment. In nine countries agriculture's contribution to GDP ranges from 32 percent to 49 percent. Only for Senegal is it below that range, at 21 percent, and for Uganda above the range, at 66 percent. All countries rely on a few commodities for export proceeds. The main export crops are cocoa, coffee, cotton, tea, and spices. In addition, C6te d'Ivoire and Ghana export timber. Several countries export gold, Niger exports uranium ore, and Togo exports phosphates. Industry's contribution to GDP is small. Services' contribution to GDP is, however, of the same magnitude as agriculture's. The services sector has numerous informal activities. All these countries share certain characteristics that condition their economic performance. First, their resource base is narrow. Second, their main economic sector, agriculture, is dependent on weather conditions, since most crops are rain-fed. From one year to the next, food output remains unpredictable and food self-sufficiency fragile. Third, their exports are undiversified, the bulk consisting of one or two agricultural and mineral commodities, and all are price-takers on world export markets. Fourth, the level of social and economic development is low. And fifth, these countries are burdened by a legacy of inappropriate macroeconomic, fiscal, and financial policies. All these characteristics combine to render these countries' economies vulnerable to external and internal shocks. Overall Economic Performance The 1980s have been difficult for these countries, marked by steady economic deterioration, substantial worsening in the terms of trade (Table 2.1), misguided economic policies, and falling investment. In the first part of the 1980s, most of these countries were affected by droughts, while civil strife, political turmoil, and transition toward multiparty politics have constrained or slowed economic activity in many countries. In addition, the macroeconomic and external environments have been difficult. 25 Table 2.1 Macroeconomic Indicators, 1981-90 (percent) Population GNP Group/Country growth per capita GNP growth Average inflation Terms of trade 1980-90 1991 (US$) 1980-91986-MR 1981-90 1986-MR 1981-85 1986-90 1981-90 WAMU Benin 3.2 380.0 2.1 -0.2 1.2 2.7 17.9 -12.5 3.7 Burkina Faso 2.6 350.0 4.0 2.1 3.0 5.5 2.5 -16.3 -14.3 Cote d'lvoire 4.1 690.0 0.3 -1.2 -1.6 2.0 17.5 -17.5 -0.3 Mali 2.5 280.0 2.5 4.2 1.4 2.8 5.9 2.8 8.8 Niger 3.4 290.0 -0.9 1.6 0.7 3.2 2.2 -36.0 -34.5 Senegal 3.0 720.0 2.9 3.3 2.7 6.0 -10.0 9.4 -0.2 Togo 3.5 410.0 1.8 3.1 2.8 4.9 -14.0 -0.5 -14.3 Nonweighted 3.2 446.0 1.8 1.7 1.4 3.9 3.1 -10.1 -7.3 Weighted 3.4 554.0 1.4 1.0 0.6 3.5 6.7 -11.2 -5.6 NON-WAMU Ghana 3.4 400.0 3.1 5.4 40.9 34.1 -2.5 -34.2 -35.9 Madagascar 3.9 340.0 4.1 5.2 8.9 9.5 12.6 -17.2 -6.8 Kenya 2.4 160.0 5.9 4.5 80.3 106.8 17.0 -16.8 0.3 Uganda 2.9 210.0 0.5 3.2 17.1 16.1 35.9 -8.2 24.9 Nonweighted 3.2 277.0 3.4 4.6 36.8 41.6 15.8 -19.1 -4.4 Weighted 3.4 299.0 3.5 4.8 33.7 37.9 24.7 -19.5 -6.6 ALL 11 COUNTRIES 3.2 384.0 2.4 2.8 14.3 17.6 7.7 -13.4 -6.2 Nonweighted 3.4 435.0 2.3 2.8 15.0 15.7 9.3 -14.9 -6.9 Weighted Sub-Saharan Africa 3.2 (339.0) 1.8 2.2 - - - - - (45 Countries) Source: World Bank, World Tables 1991, Trends in Developing Economies 1991, and World Bank Atlas 1991, Washington, D.C.; African Development Indicators, United Nations Development Program, New York; and World Bank staff estimates. As from 1982-83, terms of trade showed an improvement, peaking (for many countries) around 1985-86, only to decline sharply thereafter. For a few years, this blurred the severity of the underlying economic deterioration and dulled the urgency to tighten fiscal and monetary policies. During 1980-91, real GDP growth for the eleven countries averaged 2.3 percent a year, higher than for the whole sub-Saharan region, where real aggregate GDP grew at about 1.8 percent a year. For all eleven countries, however, real growth was below population growth, which averaged 3.2 percent a year and outpaced the 2.1 percent growth in food production (approximated by value added in agriculture). These numbers imply that income per capita declined in the 1980s and that food production per capita declined even faster. In general, the decline in agriculture's share in GDP was offset by the expansion of services, in the public sector in particular. 26 In the second half of the 1970s and in the early 1980s, most of the eleven countries were implementing a growth strategy based on public sector expansion and diversification. Expansionary fiscal and borrowing policies, however, eventually led to unsustainable internal and external imbalances. Favorable terms of trade masked the deleterious effects of this strategy, postponing the day of reckoning. It was only after 1986, when the terms of trade began a protracted decline, that the magnitude of the fiscal and external imbalances became visible. At the same time, the countries' access to foreign borrowing on commercial (or near-commercial) terms disappeared. With no recourse left but external borrowing to alleviate these imbalances and to restore a solid basis for self-sustained growth, the implementation of stabilization and adjustment programs became imperative. It was the protracted deterioration in the terms of trade in the second part of the 1980s that brought home the inevitability of severe adjustment programs. This general pattern of events is broadly applicable to all the countries, with differences between countries related mainly to timing. Uganda is an exception, however. It suffered from a long period of civil strife that extended into the early 1980s. Reconstruction started in earnest around 1986, and economic growth accelerated immediately, so that Uganda partly caught up with other countries in its group. The non-WAMU countries (Ghana, Kenya, Madagascar, and Uganda) implemented stabilization and adjustment programs earlier than most WAMU countries. Ghana initiated its Economic Recovery Program in 1983. By the mid-1980s, both Kenya and Madagascar were completing their first stabilization program. In these countries, along with Uganda, real growth picked up in the second half of the 1980s, a period during which they averaged 4.6 percent annual growth compared to 2.9 percent in the first half. Per capita incomes began to rise again, but growth was accompanied by significant inflation due to excess liquidity stemming from earlier recurrent monetization of deficits. Moreover, stabilization programs were disrupted by bursts of deficit monetization to make up unexpected shortfalls in revenues. Uganda experienced the worst inflation, averaging 80.3 percent in the first half of the 1980s and rising to 106.8 percent in the second half. For Ghana, inflation was higher in the first half, at 40.9 percent, before declining to 34.1 percent in the second half as the stabilization and adjustment programs unfolded and gradually alleviated the macroeconomic and fiscal imbalances. Indeed, inflation continued to decline, reaching about 15 percent at end-1992. Terms of trade for these countries improved temporarily in the first half of the 1980s, only to decline sharply in the second half (by 19.5 percent). Most WAMU countries began to implement stabilization and adjustment programs later in the 1980s than the non-WAMU group. Their real growth declined slightly from 1.9 percent a year in the first half of the 1980s to 1.7 percent in the second half. Several factors combined to delay implementation of stabilization and adjustment, including the WAMU monetary arrangements, which tempered expansionary policies; the convertibility of the CFAf; and extended budgetary support from external sources. Average inflation was lower than for non- WAMU countries but rising-0.6 percent for the whole of the 1980s but 3.5 percent for the second half of the decade. Terms of trade followed a pattern similar to that for non-WAMU countries, but with narrower fluctuations: improving in the first half of the 1980s and declining in the second half. 27 Domestic Saving During 1981-90, the savings performance of the eleven countries was markedly weaker than that of the whole region, where savings performance is already notably weaker than in other developing regions. Average gross domestic saving to GDP in the decade was 7.2 percent and gross domestic investment 15.7 percent, compared with 12.5 percent and 17.0 percent for the whole of sub-Saharan Africa. For the region, gross domestic saving to GDP declined from the first half of the 1980s to the second (12.9 percent to 12.1 percent), but for the eleven study countries the opposite is true-from 6.6 percent in the first half to 8.2 percent in the second (Table 2.2). Domestic saving has been affected by low income and sluggish growth, public sector dissaving, macrofinancial instability and financial repression, institutional weaknesses and lack of financial markets and instruments, and financial distress. The differences between the region and the study countries is due to the intensity and the extent of the constraints rather than to differences in the nature of these constraints. In the study countries, several phenomena have been more intense and widespread: macrofinancial instability, inflation, fiscal imbalances, and financial distress. Public Sector Finances In the 1980s, all study countries except Ghana experienced poor public finance situations, with budgetary outcomes unsustainable in the long term. Budgetary deficits were influenced by the terms of trade, which bottomed out in the early 1980s, peaked around 1985-86, and then began a steep decline. For many countries, particularly those in WAMU, this movement was not matched by corresponding fiscal and monetary adjustments. There were five causes of the lack of an appropriate fiscal response. First, there was a legacy of fiscal imbalances from the misguided policies of the 1970s. Second, revenue from taxes on international transactions, a major source of fiscal receipts, was adversely affected by the deterioration in the terms of trade. Third, for most countries, the 1980s were a time of institutional deterioration and poor governance, including lack of fiscal discipline, leaving countries illprepared and illequipped to stage an immediate and appropriate fiscal response to the abrupt deterioration in the terms of trade. Fourth, most governments initially sought to offset t:ie shock of this deterioration by increasing public expenditures; in a time of declining revenues and dwindling external borrowing opportunities, this widened deficits. And fifth, during the subsequent fiscal stabilization stage, many categories of expenditure proved to have little "compressibility," given the already low level of services provided, and could not be adjusted downward in time to match the decrease in fiscal revenues. Conventionally defined deficits are those of central governments (see table 2.3), but this measure generally underestimates the true public sector deficit. It does not take into account the quasi-fiscal deficit, which is defined as the total losses of public finance institutions, including the central bank. Because the financing of quasi-fiscal deficits (through either accounting accrual or monetization) is particularly relevant to liquidity and monetary management, the magnitude of this financing has been estimated for the study countries. For the non-WAMU countries, the 28 Table 2.2 Gross Domestic Saving and Gross Domestic Investment, 1981-90 (percentage of GDP) 1981-85 1986-90 1981-90 Group/Country Weighted Nonweighted Weighted Nonweighted Weighted Nonweighted Saving Invest- Saving Invest- Saving Invest- Saving Invest- Saving Invest- Saving Invest- ment ment ment ment ment ment WAMU Benin 1.8 17.8 4.4 13.8 3.1 15.8 Burkina Faso -0.5 20.4 6.2- 21.6 1.2 21.0 Cote d'lvoire 21.6 19.0 16.3 11.8 18.9 15.4 Mali -4.4 16.9 6.5 24.6 10.4 20.8 Niger 0.5 13.4 4.2 9.8 4.6 11.6 Senegal -1.5 11.3 6.7 11.7 2.6 11.5 Togo 20.2 25.7 11.9 21.7 16.0 23.7 Average 9.4 17.2 6.1) 17.8 10.5 14.3 8.0 16.4 10.3 15.8 6.9 17.1 NON-WAMU 5.1 8.8 Ghana 5.0 5.6 5.1 11.9 20.0 23.5 Kenya 20.5 23.4 19.6 23.9 4.0 10.8 Madagascar 1.4 9.1 6.6 12.3 3.3 9.9 Uganda 3.7 8.3 2.9 11.5 Average 10.4 13.7 7.7 11.6 10.6 16.6 8.6 14.9 10.2 15.2 8.1 13.2 All 11 Countries 9.5 15.6 6.6 15.4 10.5 15.4 8.2 15.9 10.2 15.5 7.2 15.7 Source: World Bank, World Tables 1992; and World Bank staff estimates. quasi-fiscal deficit has been estimated at about 4 percent of GDP for Madagascar (Le Hourerou and Sierral 1992) and at about 3.5 percent for Ghana (Islam and Wetzel 1991). There are preliminary indications that the Bank of Uganda also incurred losses. Thus, for non-WAMU countries, the aggregate quasi-fiscal deficit is equivalent to at least 1.8 percent of aggregate GDP for 1981-90. There are no indications that quasi-fiscal deficits have been incurred in WAMU countries. As a group, the eleven countries had an average central government deficit of about 7 percent of GDP for 1981-90, well below the 8.9 percent average for the forty- five countries in the region. The deficit is markedly different for the two groups however. For the non-WAMU countries, the deficit is below average and declining; for the WAMU group, it is above average and rising. Why? First, most non-WAMU countries embarked on programs of financial stabilization and fiscal adjustment earlier in the 1980s than did the WAMU countries. As a result, non-WAMU countries show a notable improvement in their budgetary situations in the second half of the 1980s. Second. there were marked differences in the patterns of deficit financing between the two groups, reflecting contrasting policy approaches and related monetary management stances. The methods of financing influenced in turn the levels of 29 deficits and were instrumental in stabilization performances. These differences related primarily to access to external financing, the existence of quasi-fiscal deficits, the magnitude of seigniorage and revenue, and the share of these different financing means in total financing. Because of WAMU's monetary arrangements, including full convertibility of the CFAf, and strong and permanent budgetary support from France, WAMU countries had for the whole of the 1980s relatively easy access to external financing. Non-WAMU countries financed a larger share of their deficit through money creation because of more restricted and irregular access to external financing and weak domestic financial markets. About 70 percent of the deficit of non-WAMU countries was financed domestically; for WAMU countries the share was about 20 percent (Table 2.3). Adding quasi-fiscal deficits boosts the deficits of non-WAMU countries 1.8 percentage points higher (in terms of GDP). Domestic financing is higher by the same amount, which brings the domestic share of total financing to 80 percent. According to available estimates, about 80 percent of the quasi-fiscal deficits of the central banks of non-WAMU countries was accrued as losses in revaluation accounts; 20 percent of these deficits was monetized. At the same time, about 80 percent of the "conventional" deficit was monetized (the remainder was financed through borrowing in a noninflationary way). Together, this accounted for considerable money creation: the result is an average of about 2.5 percent of the GDP for 1981-90. To the extent that this Table 2.3 Budget Deficit Indicators, 1981-90 (percentage of GDP) WEIGHTED NON-WEIGHTED Category 1981-90 1981-85 1986-90 1981-90 1981-85 1986-90 WAMU Countries Overall deficit -8.8 -7.4 -9.2 -7.6 -6.9 -7.7 Financing net 8.8 7.4 9.2 7.6 6.9 6.1 Foreign net 5.8 4.6 6.6 5.1 4.2 4.7 programmed 4.8 4.6 4.7 4.8 4.2 1.4 exceptional 1.0 0.0 1.9 2.6 0.0 0.6 Domestic net 1.5 1.7 1.2 1.2 1.6 1.2 Arrears 1.2 0.6 1.5 1.0 0.5 -0.2 Seingiorage 0.3 0.5 -0.1 0.3 0.6 Non-WAMU Countries Overall deficit -7.3 -8.3 -6.3 -6.9 -8.2 -5.8 Financing net 7.3 8.3 6.3 6.9 8.2 5.8 Foreign net 1.1 0.9 1.3 1.0 1.2 0.7 Domestic net 2.6 3.8 1.4 2.5 3.8 1.3 Seigniorage 1.8 2.0 1.6 1.6 1.6 1.8 Quasi-fiscal deficit 1.8 1.6 2.0 1.8 1.6 2.0 Note: Foreign finance includes foreign grants and net external borrowing. Source: World Bank and IMF data and staff estimates. 30 monetization led to an increase in money supply that surpassed the level of money balances that economic agents were willing to hold, such borrowing and monetization caused inflation and depreciation of the domestic currency. Monetization of deficits and quasi-fiscal deficits is an essential explanatory factor of the higher inflation for non- WAMU countries. Inflation and its consequences generated seigniorage revenue to the governments. (Seigniorage represents the amount of resources that the government gains from printing money). Money creation is an important source of revenue in many developing countries, including a majority of countries in sub-Saharan Africa. For non-WAMU countries, seigniorage revenue is estimated to have averaged about 1.8 percent of the GDP in 1981- 90. This revenue has proportionately reduced the need for financing the deficit from explicit and quasi-fiscal sources. It should therefore be added to both sides: overall deficits and their financing. Since seignorage represents domestic financing, its inclusion raises the share of domestic financing to 85 percent of total deficit financing. WAMU countries financed 72 percent of their deficit externally. According to available information, there was no quasi-fiscal deficit in this group-the WAMU monetary arrangements restrained monetization deficits. In addition, excess money supply would flow abroad because of the convertibility of the CFAf. As a result, inflation was distinctly lower in WAMU countries than in non-WAMU countries (table 2.1). Low inflation together with no reserve requirements at the BCEAO generated much lower seigniorage. Severely restrained in money creation and with low seigniorage revenues, WAMU countries accumulated sizable domestic arrears, equivalent to about 2 percent of their combined GDP. External arrears were of about the same magnitude. Because of CFAf convertibility, these arrears stemmed from a shortage of fiscal resources rather than of foreign exchange. Logically, therefore, they should be added to domestic arrears. Effects of External and Internal Shocks A key feature of the economies of the study countries is their vulnerability to external and internal shocks (Figure 2.1). One reason for the vulnerability is that the comparative advantage of these economies-the export of a few primary agricultural or mineral commodities requires relatively greater openess. As a percentage of GDP for 1981-90, the sum of exports plus imports (a measure of openness) ranges from 28 percent for Ghana to 88 percent for Cote d'Ivoire, with a median of 59 percent for Kenya. Moreover, the economies of all the study countries remain essentially agro-based, which leaves them exposed to unfavorable weather conditions. Because these economies rely principally on agricultural export commodities for their foreign exchange earnings, the serious long-term deterioration in the terms of trade for these commodities constitutes virtually a permanent shock. A second important cause of the vulnerability of these economies is their weak capacity to finance and adjust to shocks and their consequences. Adverse shocks are often financed in part by drawing down net international reserve holdings. These reserves 31 Figure 2.1: External and Internal Shocks Indicators Terms of trade indices, 1970-88 World interest rates, 1961-91 200 -1 16 180~~~~~~~~~~~~~1 iso ~ ~ ~ &14 80- -I, z E l E rl I,, X W 0 r 120~~~~~~~~~~~~~1 100~~~~~~~~~~~~~1 Note: Weighted by relative GDP Source: World Bank, World Tables, 1991 Source: IMF, International Financial Statistics Food production indices, 1970-90 Index of food Droduction ver capita, 1961-90 140 I .O 130 t - 13 - , 110 10. 6 100 0 61 67 770 80 85 88 61 65 71 75 80 es 91 Note: Weighted by relative GDP Note: Weighted by relative populationsize Source: World Bank data Source: IMF, International Financial Statistics Non-WAMU .................... .WAMU 32 provide a cushion against unforeseen shocks that would otherwise require a rapid downward adjustment in consumption. The low ratio of foreign exchange reserves to GDP and the frequently negative operations account position of the WAMU group of countries reduce the ability of foreign exchange reserves to cushion the adjustment to shocks. By the same token, the lack of a deep, well-functioning capital market makes it difficult for producers, public enterprises, and the government to obtain loans in difficult times; and the lack of access to the world capital market because of the aftereffects of the debt crisis makes it difficult to finance a current account deficit following adverse shocks. External and internal shocks such as changes in the terms of trade or changes in food production affect balance of payments and foreign reserves. A deterioration in the terms of trade, for instance, directly damages the current account of the balance of payments. External and internal shocks also affect budgetary outcomes since they may directly reduce or increase the proceeds from direct and indirect taxes. A fall in the world price of an export good, such as coffee or cocoa, directly reduces export tax receipts, for instance. Shortfalls in food production may call for food imports and their financing. Changes in terms of trade have an immediate impact on the foreign assets position of the central bank and commercial banks and other financial institutions and also on the budget. It is principally through these two channels that these impacts are transmitted to the monetary and financial systems. Empirical research into the effects of external and internal shocks on the fiscal and financial situation of the study countries established significant relationships between external and internal shocks and seigniorage (defined to include increases in the stock of base money and the inflation tax; Connolly and Popiel 1993). These results imply that authorities often seek to offset the effects of shocks through accommodating fiscal and monetary policies, which leads to inflation and, in turn, to seigniorage. For instance, a rise in food production leads to lower growth in seigniorage. This result is consistent with the reasoning that greater food production leads to greater revenues from taxes on production and exports and from income associated with food production. Consequently, there is less reliance upon central bank financing of fiscal deficits when food production rises. A change in terms of trade may or may not increase seigniorage. Seigniorage will increase if a decline in terms of trade leads to additional deficit and to central bank financing of this deficit. If the consequent decline in net foreign assets is not offset by deficit financing, seigniorage will decrease. This pattern of relationships between external and internal shocks and seigniorage emerges best for non-WAMU countries. 33 3. FINANCIAL DEPTH In the 1980s, there was virtually no financial deepening in the study countries; in some, financial systems became even shallower. There are a number of reasons for this situation. A lack of economic growth and an inadequate financial infrastructure inhibited financial deepening. Macrofinancial instability, inflation, and financial distress combined to cause relatively intensive disintermediation in a number of study countries and provoked a loss of confidence in the formal financial system. The study countries that launched their financial reforms earlier in the decade, such as Ghana, experienced a beginning of re-internediation toward the end of the 1980s. Factors of Financial and Monetary Depth Financial depth in developing countries is determined by four types of factors: macroeconomic, institutional, monetary, and financial. Financial deepening and increased savings are often but not necessarily two aspects of the same phenomenon. A greater use of interest-bearing financial instruments, which reflects an increase in domestic financial savings, leads to (and requires) financial deepening, since it calls for more sophisticated and diversified markets, a wider range of financial instruments, and a more developed institutional infrastructure. There is well-established empirical evidence relating financial depth in developing countries to the per capita income-', level of transactions in the economy, degree of development of the financial infrastructure, rate of return on financial assets, and exchange rate policy (see Singh 1992; Bhattacharyay 1988; and Craig 1990). Further empirical evidence suggests that in an environment of high inflation, real monetary balances contract. High inflation usually results in a flight of domestic financial assets into foreign financial assets or real assets. The level of financial depth is approximated by the ratio of M2 or M3 to GDP.!' The deepening of financial systems occurs largely through the expansion of the range and use of interest-bearing instruments, together with the development and diversification of financial markets and financial infrastructure. As financial systems and economies develop and per capita incomes increase, currency depth usually declines. This reflects a greater reliance on deposit money for funding transactions. The relationship between per capita income-considered as a proxy for wealth or level of development-and financial depth, mentioned previously, indicates that expansion of the formal financial system is part of the development process. Of course, causality can flow in either direction. Development entails wealth accumulation, and wealth can be held 1/ Empirical evidence shows that a I percent increase in per capita income is associated in developing countries with an approximately 1.5 percent increase in the indicators of financial depth. 2/ M2 is generally defined as Ml (that is, currency in the hands of the public + demand deposits) + time and savings deposits. M3 is defined as M2 + short-term certificates of deposit and similar financial instruments. 35 increasingly in financial form. And improving the management of the flow of financial resources in an economy can result in more efficient financial intermediation and investment selection, thus leading to faster growth and greater wealth. Obstacles to Financial Deepening in the Study Countries Deficiencies in financial infrastructure. Grave deficiencies in the financial infrastructures of the study countries have been a factor in their financial shallowness. A strong and diverse financial infrastructure is a catalyst of monetary and financial depth. A high and buoyant level of transactions in the economy, high per capita income, high propensity to save by governments, enterprises, and households, and high real returns on financial assets are necessary but not sufficient conditions. Unless there is an adequate financial infrastructure to support financial intermediation, these factors will not, by themselves, lead to financial deepening. The main reason that countries with income levels comparable to those of the study countries, such as China, India, or Pakistan have deeper monetary and financial systems is that they have a stronger and more developed financial infrastructure, including appropriate technologies and adequate human resources. In financial intermediation, a strong infrastructure is characterized by an appropriate legal, regulatory, and prudential framework; financial strength-in terms of solvency and liquidity-of individual institutions; diversity of the institutional structure; an adequate amount of financial information-including accounting information of uniform and appropriate standards-and its integration into the financial process; and availability of relevant technologies and human skills. None of these factors is noticeably present in the study countries. Moreover, the judicial system is weak and inefficient and the enforcement apparatus ineffective. Weakness in legal and regulatory framework. During the 1980s, the legal, regulatory, and prudential framework in all the study countries lacked the strength, consistency, and flexibility needed to withstand the turbulent changes in the macrofinancial environment, inhibiting financial deepening. The framework began to be reformed in some countries at the end of the 1980s, but problems remain and the reform process should be accelerated and expanded. In most of the study countries, contractual law is outdated and its force blunted by too many partial modifications; enforceability is weak as well. More specifically, laws governing loan recovery and collateralization are antiquated and their enforcement is time-consuming. Central banking and other banking laws were outdated for most of the 1980s, and slated for redrafting in several of the study countries at the end of the decade. None of the study countries features an adequate financial services act. With a few exceptions, prudential supervision was deficient in both coverage and enforceability during most of the 1980s. All of the study countries need new laws and regulations specifically governing new or changing financial activities. For instance, leasing is a financial activity well suited to countries in Africa, but it requires a supportive legal, regulatory, and prudential framework that none of the study countries has yet put in place. In addition, the enforceability of contractual laws must be made real and accessible to all economic agents at reasonable costs for judicial and other fees. 36 A similar situation prevails for prudential regulations and supervision. Of the non- WAMU countries, Ghana has considerably strengthened these elements and expanded the central bank to supervise financial institutions and enforce its recommendations when warranted. Kenya strengthened the central bank's supervision capacity, but the enforceability of its recommendations remains very weak. In Madagascar, little has yet been done to strengthen the central bank's supervision capacity. Uganda only recently set in motion a program with the aim of strengthening and expanding the supervision capacity of the central bank. In this area also, new supervision rules and procedures are needed in order to respond to emerging or deeply changing financial activities. The insurance industry is a case in point. The situation is comparable for WAMU countries. During most of the 1980s, the legal, regulatory, and prudential framework was outdated. For this reason, a new Regional Banking Act, common to the seven countries, was written in 1989 and became effective in 1990. A Regional Credit Unions Act is currently being drafted in order to support these types of institutions, which hitherto had not been formally recognized. A new regulation concerning insurance companies, common to the fourteen members of the CFAf zone, has been elaborated and was to come into effect in 1993. Prudential regulations and supervision have also been strengthened recently. A new prudential regulation for banks was adopted at the WAMU level and became effective in October 1991. A WAMU Banking Commission has been created to enforce the new prudential regulation and supervise financial institutions in the region. It is a regional body separated from the regional central bank. This Banking Commission is in charge of granting and withdrawing banking licenses. As a regional institution, it is less subject to political interference and can act more efficiently. But serious problems remain with regard to the effective enforceability of contractual laws and the access of all economic agents to justice at reasonable judicial fees. Widespread problems of solvency and liquidity. During the 1980s, virtually all the study countries suffered from intense disintermediation caused by the loss of confidence in the formal financial system, inflation, exchange rate devaluation (for non-WAMU countries), accumulation of domestic arrears by the states (for WAMU countries), and bank failures. These developments prompted the public to withdraw money from banks and financial institutions and keep it in domestic currency or invest it in foreign currency or in real assets. This situation contributed to financial distress. All the study countries suffered from more or less widespread financial distress (see chapter 5). This was observed in its most extreme form in Benin, where all the banks closed their doors, and in its mildest form in Togo. On the whole, financial distress deeply affected the financial systems of all countries included in this study and was a major impediment to efficient financial intermediation and financial deepening. With financial adjustment came the restructuring of banking systems in most of the study countries. Restructuring is continuing in most countries and has just been completed in some. But even in these countries, banks remain frail and reluctant to reassume the risks associated with lending to productive enterprises. Confidence in the formal financial system is returning only belatedly. 37 Lack of institutional diversification. Financial systems in all the study countries are characterized by lack of diversification. The mainstay of the system remains the banking sector. Kenya is an exception. C6te d'Ivoire had the most diversified system among the WAMU countries but lost much of its financial fabric in the 1980s, when the system was financially distressed. In all study countries, the structure of the financial system is lopsided, in that it concentrates on urban areas, leaving many rural regions financially unserviced or underserviced. Deficient financial information. The quality, depth, and extension of accounting are an important factor for growth of the financial system and the economy in general. Accounting quality varies by group and individual country (United Nations 1991). But even in countries where accounting is at its best, it suffers from grave shortcomings related to the specific legal and regulatory framework; to the quality, uniformity, and enforceability of standards; to the supply of qualified accountants and accounting technicians; and to education and training. Most of the non-WAMU countries have outdated legislation for organizing and regulating accounting. Of the four, only Ghana and Kenya have bodies that set accounting standards. The WAMU countries, although members of a regional monetary union, have disparate and often outdated legal and regulatory frameworks. They follow, however, a standard accounting format related to national accounts. This format is somewhat outdated and too complex for use by medium-size and small enterprises, since it is designed primarily for large, formal enterprises. The dissemination of financial information is only beginning in the study countries. Some countries have centralized information on bank lending that is kept by the central bank or a specialized institution. But this information is internal to the banking sector and is not disseminated. In WAMU countries, the BCEAO is supposed to centralize financial information on companies at the regional level, but the system is not yet operational. In both groups of countries, registrars of companies have valuable financial information in their files, but most of the time these agencies are underfunded, understaffed, and unable to exploit and circulate the information they hold. Scarcity of human resources and skills. In all the study countries, there is a scarcity of human resources and skills related to financial operations. There is need for high-quality financial education, including practical training at three levels: financial management, complex financial operations, and basic bank training. Basic bank training is provided in- house by some international banks and in banking institutes. Kenya recently established a banking institute. Ghana is in the process of establishing one. No operative training center exists in Madagascar or Uganda. In WAMU countries, there are two other types of training in addition to the in-house training organized by international banks with subsidiaries in the area. Basic bank training is organized at the national level for each of the seven countries by the professional banking associations, which deliver national diplomas. Financial management and complex financial operations are taught at the regional level by the banking institute of the central bank. But much more urgently needs to be done in this field. 38 The informal financial sector. Informal financial systems are important in Sub-Saharan Africa and for both groups of study countries. But informal mechanisms and institutions are weak intermediaries. They rarely issue interest-bearing financial instruments and by nature do not foster financial deepening. The share they occupy in financial operations in the study countries is, by its sheer size, an important constraint on the development of financial intermediation and, hence, on the deepening of the financial system. Monetary and Financial Depth in the Study Countries There is evidence that in the study countries, as in most African countries, currency is used not only for payment but also as a means of short-term saving, despite inflation. This use of currency stems from several causes. Deficiencies in the financial infrastructure are one. Either financial institutions do not reach into certain areas or the use of their services is too complex or too costly for small savers; hence, many households and other economic agents do not have access to deposit money and term savings. Another cause is the persistent lack of confidence in formal banking and high potential for fraud which inhibit the use of checks. Payment systems are inefficient and slow, further discouraging their use. One appeal of informal mechanisms is that they use currency. For instance, "susus" in Ghana and "traveling bankers" in Benin operate almost entirely with currency. Because of the lack of appropriate financial instruments and markets, the best guarantee of liquidity is the holding of currency. The propensity to hold currency is compounded by the importance in study countries of trade in the small enterprise sector where liquidity is an operational requirement. As a group, the study countries show a financial depth (M2/GDP) of 23.4 percent for 1981-90, below the regional average of 27 percent (Table 3.1). The ratio has been stable throughout the period, reflecting the lack of economic growth and the absence of financial deepening. Monetary depth (M1/GDP), at 16.4 percent, is slightly higher than the regional average of 16 percent. As a result, the share of quasi-money in M2 is lower for the study countries than for the region. There are notable differences between the two groups of study countries. Non- WAMU countries experienced financial deepening from the first to the second half of the 1980s. In contrast, Ml remained roughly stable and below the average for the region. In countries of this group, there appears to be easy substitutability between currency and foreign exchange and real assets. Hence, financial deepening took place through the expansion of quasi-money, which grew from 5.2 percent of GDP to 7.4 percent. The principal reasons for this would be the adjustment process initiated by these countries early in the 1980s, which stabilized the macrofinancial environment, and the restructuring of their financial systems, in particular, the relaxation or liberalization of interest rates. The period 1981-85, however, was a time of disintermediation. The average ratios of financial and monetary depth were significantly higher in the 1970s, higher even than in the second half of the 1980s. Hence, non-WAMU countries experienced disintermediation in the first part of the 1980s and some re-intermediation in the second part of the 1980s, though without reaching the pre-1980 level. 39 Table 3.1 Monetary and Financial Currency Depth, 1981-90 (percentage of GDP) Sub-Saharan Eleven study Seven WAMU Four non-WAMU Countries countries countries countries Average (1981-85) Currency/GDP - 7.7 9.1 5.7 Ml/GDP 17.2 16.7 18.9 13.7 Quasi- 10.1 5.9 6.9 5.2 money/GDP 27.3 22.6 25.8 18.9 M2/GDP Average (1986-90) Currency/GDP - 7.3 8.5 6.0 Ml/GDP 14.6 16.1 18.4 13.4 Quasi- 12.0 8.0 8.8 7.4 money/GDP 26.6 24.1 20.4 20.8 M2/GDP Average (1981-90) Currency/GDP - 7.5 8.7 5.8 Ml/GDP 15.9 16.4 18.7 13.5 Quasi- 11.1 7.0 7.8 6.3 money/GDP 27.0 23.4 23.3 19.8 M2/GDP Source: IMF, International Financial Statistics, IMF data, and staff estimates. WAMU countries appeared to experience less fluctuation in indicators of financial depth from the first to the second period. But this picture is misleading. When "frozen" liabilities are taken into consideration-liabilities of banks closed or being restructured- financial depth in the WAMU countries falls, according to the best estimates available, to about 20.4 percent of GDP for the second half of the 1980s. But it is likely that the major part of "frozen" liabilities affects quasi-money. Thus, WAMU countries, in contrast to non-WAMU countries, would have experienced notable disintermediation in the second half of the 1980s but for the freezing of liabilities that took place. If Ml and M2 were adjusted for frozen liabilities, they would likely stand (by best estimates) at about 15.5 percent and 20.5 percent of GDP, as against the 19.8 percent and 26.5 percent presently shown. There are a number of causes of the greater financial shallowness of the non- WAMU group of countries. The first is much higher inflation, which averaged 33.7 percent a year in the 1980s, against 0.6 percent for WAMU countries. Inflation is a disincentive to conserve currency and other financial assets. As inflation rises, economic agents increasingly seek to move out of currency and financial assets into real assets or foreign exchange. The second cause is the convertibility of the CFAf, the currency common to the WAMU countries, which permits unconstrained purchase of foreign exchange at a fixed rate. In non-WAMU countries, with the exception of Ghana and to some extent Uganda, there exist active parallel markets in foreign currencies. A third cause is definitional. It is a fact that the financial disintermediation that occurred in the 40 mid-1980s in the non-WAMU countries is reflected in the monetary data, whereas the financial disintermediation that took place later in the decade in the WAMU countries does not show up in the monetary data. If the data were adjusted, as illustrated in the preceding paragraph, the picture would show the beginning of re-intermediation in the countries of the non-WAMU group and the full impact of disintermediation in the countries of the WAMU group. 41 4. FINANCIAL INSTITUTIONS AND RESTRUCTURING FINANCIAL INSTITUTIONS The Lending Sector Financial systems in the study countries are dominated by commercial banks, and the nonbank financial institutions are underdeveloped. In all countries but Kenya, the share of commercial banks' assets in the total assets of the financial systems ranges from 85 percent to 95 percent (Table 4.1). Kenya, where this share is 45 percent, is an exception for two reasons. First, the Kenyan financial system is private-sector- driven and dynamic and thus prone to diversify in response to the demand for varied financial services. Second, the most important nonbank financial institutions are near-bank financial institutions, which organically could be considered commercial banks. Their number increased rapidly in the 1980s, when commercial banks set them up to circumvent the Bank of Kenya's credit ceilings and reserve requirements, regulations that do not apply to nonbank financial institutions. Although these deposit-taking institutions can neither engage in foreign exchange transactions nor offer checking accounts, they are deeply involved in consumer finance. As a result, between 1980 and 1985, the assets of these near-bank financial institutions tripled while commercial banks' assets only doubled. The Commercial Banks The predominance of the banking sector does not translate into effective competition among banks. In the study countries, more than 60 percent of the assets of the banking system is owned by, at most, four banks. In some countries, such as Ghana, Mali, and Uganda, one commercial bank alone accounts for more than 50 percent of assets. Such high concentration reduces competition and results in oligopoly. Introducing competition within the banking sector is an intricate policy issue. Prudential requirements call for solid and solvent banks. By the standards of the study countries and the sub-Saharan African economies, that means that banks must be relatively large. Given the difficult and risky lending environment, a large number of small banks with a weak capital base might present a potential systemic risk to the sector. However, a banking sector composed of large banks offers opportunities for oligopolistic behavior. Large, strong banks have a tendency to absorb small newcomers, preempting added competition. The way out of this dilemma is to promote competition from outside the banking sector by fostering nonbank financial institutions. Such institutions initially require a smaller capital base (which makes them more accessible to local entrepre- neurship) and, in a healthy and growing financial system, are likely to find some specific comparative advantage, providing needed competition to the banking sector. This channel 43 Table 4.1 Share of Commercial Banks in the Financial Sector, Selected Countries, 1990 (local currency, billions) Commercial bank Total assets of share of total the commercial Total assets of financial sector Group/country banks financial sector assets (%) WAMU Benin 219 232 94 Cote d'Ivoire 1,532 1,718 89 Madagascar 1,157 1,286 90 Senegal 666 740 90 Non-WAMU Ghana 394 462 85 Kenya 65 155 42 Uganda 49 57 86 Source: World Bank and IMF data and staff estimates. for introducing competition can only work, however, when prudential supervision of nonbank financial institutions is adequate. In the study countries, banking accounts for a relatively small share of GDP and, compared with banking in other developing regions, does not seem to have expanded strongly. Only nine sub-Saharan banks are included in the 2,000 largest banks in the world (table 11, statistical annex). The development of banking has been largely limited to cities and commercial centers, although in Anglophone African countries some expansion of bank networks did occur in rural areas. Growth of banking has also been hampered by competition from semiformal and informal financial institutions. In particular, rotating savings and credit associations have made it more difficult for banks to attract retail savings. At the time of independence in the 1960s, banks in the region and in the study countries were privately owned. But the 1970s were marked by the growth of government-owned banks. By the end of that decade, governments held, directly or indirectly, a majority interest in more than half of the region's banking institutions and a 44 minority interest in 40 percent of the other banking institutions in the region (tables 12 and 13, statistical annex; Tenconi 1992). By the early 1990s, the dominance of government in the banking sector began to subside. At the end of 1990, the number of banks in the region in which governments held a controlling interest had been reduced from 140 to 118, and private interests had assumed a stronger role. By end-1990, the private sector held a controlling interest in 40 percent of the banks in the region. In sub-Saharan Africa and the study countries, banks only lend short term, mainly on overdrafts that are rolled over, which is normal in an unstable macrofinancial environment. For the study countries, domestic credit extended by commercial banks averages 25.4 percent of GDP, ranging from 4.8 percent for Uganda to 40.1 percent for C6te d'Ivoire, with Madagascar at the median of 22 percent. In comparison, the figures are 96 percent for China, 47.6 percent for India, and 39.2 percent for Pakistan. For all the study countries as a group, demand deposits average 10.7 percent of GDP and time deposits, 7.9 percent. The relatively low level of time deposits renders the banks' liabilities prone to volatility and constrains the intermediation process. It also explains banks' preference for overdraft lending. In the study countries, commercial banks have three main sources of funds: time deposits, demand deposits, and central bank refinancing. Despite the apparent lack of competition, the cost of intermediation (calculated as the operating income or gross earning margin divided by average total assets) remains within acceptable limits (Table 4.2). With the exception of Uganda, all study countries have costs of intermediation between 3.13 percent and 7.21 percent. The two most important factors behind the relatively low cost are administered interest rates and the nature of commercial banking. For most of the 1980s, interest rates were administered de jure or de facto in all the study countries. In the WAMU countries, the banks' margin is also fixed at 5 percentage points-as administrative controls substitute for competition. Banks in the study countries, as in most African economies, cater mainly to an industrial and commercial clientele. With the exception of Ghana, they do little retail business, which keeps costs down. Finally, the 1980s were a time of economic stagnation, which affected operating income. Despite reasonable operating costs (from 1.94 percent of total assets for Benin to 7.21 percent for Kenya, with a median of 3.46 percent for Niger), banks make losses in three of the eleven countries. Except in C6te d'Ivoire and Togo, the number of banks per capita is lower in WAMU countries, at 7.8 branches per million inhabitants, than in non-WAMU countries, at 14.4 branches per million inhabitants (Table 4.3). Moreover, because of relatively more extended networks, assets per employee are lower in non-WAMU than in WAMU countries. Consequently, operating costs are higher in non-WAMU countries. 45 Table 4.2 Costs of Intermediation, 1989 (as a percentage of average total assets) Group/Country Operating income Operating costs Other net costs Profit before tax WAMU countries Benin 3.13 1.94 -0.25 1.43 Burkina Faso 5.48 3.69 2.89 -1.10 Cote d'lvoire 5.11 3.43 1.64 0.04 Mali 4.26 2.67 1.19 0.39 Niger 4.34 3.46 1.42 -0.54 Senegal 6.84 3.86 2.36 0.61 Togo 5.55 3.97 1.69 -0.11 Average 5.25 3.43 1.69 0.12 Non-WAMU countries Ghana 6.97 5.83 1.70 -0.56 Kenya 10.65 7.21 1.45 1.99 Madagascar 6.93 3.07 1.57 2.29 Uganda 23.64 19.4 1.12 3.12 Average 7.45 3.81 1.57 1.96 Source: BCEAO; central banks of Ghana, Kenya, Madagascar, and Uganda; World Bank and IMF documents and staff estimates. Table 4.3 Banking Sector Analysis, 1990 Branches per Commercial Assets per Commercial million bank employee Group/country Banks Branches Staff inhabitants assets (US$ (US$ millions) thousands) WAMU/countries Benin 4Y 11 241 2.47 686 28.46 Burkina Faso 7b/ 66 1,475 7.72 824 558 C6te d'lvoire 141" 175 1,493 15.11 4,802 3,219 Mali 6 40 862 5.00 551 639 Niger 7 11 649 1.57 655 1,009 Senegal 10 46 1,627 6.43 2,087 1,282 Togo 8_d 40 1,187 11.90 761 641 Average 57 389 9,741 7.76 10,369 1,064 Non-WAMU/Countries Ghana 9 261 9,413 18.00 1,456 154 Kenya 28 362 6,100 15.40 3,140 514 Madagascar 4 118 4,300 10.56 721 167 Uganda 12 223 5,566 13.29 220 39 Total 53 964 25379 14.4 5,347 211 a. Excluding CAI. b. Excluding SCPCE. c. Excluding CAA and BNDA. d. Excluding SNI. Source: BCEAO; Commission Bancaire; World Bank and IMF documents and staff estimates. 46 Another important difference concerns sources of funds. In 1981-90, central bank refinancing provided about 35 percent of intermediated funds in the WAMU group and about 6 percent in the non-WAMU group. Development Finance Institutions Development finance institutions (DFIs) were set up to provide enterprises and projects with long-term finance that commercial banks were unable or unwilling to supply. In the 1970s, their mandate was broadened to include the promotion of priority sectors. Using government or donor funds, development finance institutions frequently extended credits to activities judged too risky by other lenders. In practice, these institutions found it difficult to remain financially viable while financing projects with high economic returns but low financial returns. Monitoring and control of borrowers also proved problematic because long-term lending does not involve the day-to-day contact with customers that commercial banks maintain. The narrow specialization of development finance institutions made it difficult for them to diversify risks. They have been particularly vulnerable to the risk concentration that characterizes financial asset portfolios in the region. They also have been vulnerable to cycles in commodity prices and business activity. Frequent government interference in their operations also undermined the soundness of portfolios. As a result, in the 1980s, many development fmance institutions were liquidated in the study countries-C6te d'Ivoire liquidated all of them (Yaron 1992). Nevertheless, there was no reason to shut down development finance institutions that could mobilize funds, expand services, and remain profitable. In Ghana, the three former development finance institutions have now introduced commercial banking services. They represent about 23 percent of assets and 12 percent of deposits in the country. In Senegal, one of the two remaining development finance institutions, Banque de I'Habitat du Senegal, remains involved in housing finance. Its liabilities are mainly time and savings deposits. Housing Finance Housing finance, while widespread in Latin America, is underdeveloped in sub- Saharan Africa. The mechanisms and instruments for funding mortgages vary from country to country. Some countries-Benin, Burkina Faso, Madagascar, Mali, Niger, Togo-have no specialized system, with commercial banks providing most housing finance. In C6te d'Ivoire, the specialized housing lender has been liquidated, and no new system has taken its place. In other countries, one or more institutions specialize in housing financing, with commercial banks involved as well. Contractual savings institutions also contribute in the financing or refinancing of these transactions. In Anglophone countries and the study countries, there were initially well- developed networks of building societies. But because of the structure of their assets and liabilities, few of these institutions survived the financial repression of the 1970s and the macrofinancial instability of the 1980s. Those that survived did so by expanding into 47 short-term financial intermediation. The virtual disappearance of building societies meant the disappearance of a retail financial service serving individual savers, a majority of whom did not have access to the formal banking system-an example of financial de- institutionalization and disintermediation caused by financial repression and macrofinancial instability. Three housing finance systems (in Kenya, Senegal, and Uganda) are based on the collection of deposits from private persons and institutions. Ghana's is financed by the National Pension Fund. None has recourse to the financial market, not even those that formally use market instruments. The systems in Kenya and Senegal appear to be vulnerable to a rapid contraction of deposits, since they entail a high level of maturity transformation. Ghana's system appears to be wellbalanced, with uses backed by funds of similar duration. In Uganda, building societies offer only short- to medium-term loans (no more than three years). Moreover, they hold a large part of their portfolio in liquid assets (notably deposits in commercial banks). In countries that have had high inflation rates (except Uganda), mortgage rates are indexed to maintain competitiveness in attracting funds and to be able to take a position on the market. The selection of customers mainly from the middle class and careful management of repayments help keep the system healthy. The key challenge in developing housing finance institutions in most African economies is to design a mechanism that supports a high degree of maturity transformation in an unstable macrofinancial environment prone to inflation. One promising model has been developed in Colombia, based on the unit of constant purchasing power, or UPAC. Savings kept for the purpose of securing a mortgage earn slightly positive real rates of interest. There are no heavy penalties for withdrawing the savings except the loss of ranking for obtaining the mortgage. Colombia's central bank stands ready to assist the housing institutions in case of a run on deposits. Thanks to this arrangement UPAC-affiliated housing finance institutions have a stable core of deposits available for long-term lending. In its twenty years of operation, the system has grown rapidly and has never required massive central bank intervention (Rosas and Lauchlin 1986). Some such mechanism should be considered for the study countries, taking into account the specificities of their monetary and financial environment. Contractual Savings Institutions Contractual savings institutions, such as life insurance companies, funded pension schemes, and national provident funds have generally predictable liabilitiesY*' They are potentially good sources of finance for investment in corporate bonds and equities. In high-income countries, they are the main suppliers of long-term finance. They provide savers with opportunities to diversify and with the benefits of investing in portfolios 1/ Though unfunded pension schemes and social security systems are not usually considered contractual savings institutions, they have been included in the analysis for convenience. 48 selected by professional investors. Contractual savings institutions still make up only a small part of the financial systems in the study countries. Their growth has been impeded by three major obstacles. First, macrofinancial instability and financial repression have together discouraged long-term saving by keeping real rates of return negative on financial products offered to the public. Second, preemptive use of the funds by governments have interfered with their development as a source of long-term corporate fiance. Third, the absence of an appropriate legal, regulatory, and prudential framework has stifled the development of contractual saving institutions. Insurance Companies Insurance companies accumulate substantial reserves to meet their future claims and obligations, playing an important role as financial intermediaries in the mobilization and allocation of resources. Development of the insurance industry depends on many factors, from level and distribution of income and wealth to social and cultural structures. Macrofinancial stability is also important since economic agents are less likely to embark on long-term contracts in a volatile environment. The regulatory framework and the tax treatment of insurance premiums are important considerations as well. Several factors have held back growth of the insurance industry in the study countries. Most important have been the poor financial condition of many insurance companies, large losses related to fraud and inadequate supervision, and inadequate regulatory framework and tax treatment. In Cote d'Ivoire, Ghana, Kenya, and Senegal, the number of insurance companies is high and the average premium issued per company is low, at less than 0.05 percent of GDP (Table 4.4). Weak supervision and fragmented markets allow financially nonviable insurance companies to remain in operation. In Benin, insurance remains a legal monopoly. In other countries, the insurance market is wellbalanced, with competition but no excess fragmentation. Pension Schemes Pension systems are quite different in the francophone and anglophone study countries. In the francophone countries, pension systems are unfunded, are headed by a national agency, and have participation by employers and employees. In principle, such pay-as-you-go systems do not accumulate reserve funds, since current contributions are expected to meet current payouts. However, these agencies establish small back-up reserves to weather economic downturns and to offset growth over time of the ratio of retired to active workers. Thus, some funds are available for investmnent. The anglophone study countries have generally adopted funded schemes in which workers accumulate retirement savings in individual accounts. The macroeconomic impacts of the two systems are well documented.2' Funded schemes are more likely to 2/ See, for instance, the summary of the Diamond Consumption/loans model in Annex 4 (Diamond, National Debt in a Neoclassical Growth Model, AER. 1965). 49 Table 4.4 Insurance Industry, 1989 (percentage of GDP) Number of Issued Life insurance companies premiums premiums WAMU Countries Benin 1 0.50 0.00 Burkina Faso 3 1.10 0.00 Cote d'Ivoire 38 2.60 0.30 Mali 5 0.53 0.00 Niger Senegal 22 1.20 0.30 Togo 5 1.25 0.11 Uganda Non-WAMU Countries Ghana 21 0.70 0.05 Kenya 39 2.30 0.20 Madagascar 41 0.85 0.05 Source: CICA, compilation of staff report. spur development of the financial market than unfunded schemes, which typically lead to lower capital intensity, lower output per capita, lower average wages, and higher interest rates. Private pension funds, usually established by companies and directed by private managers, also have a significant potential for accumulating capital. Among the study countries, only Kenya and Ghana have private pension funds. In Kenya, the total capital stock drawn on by private pension funds is estimated at about the same level as in the public retirement system. Some investment companies manage up to fifty pension funds. By extension, they also begin to handle the management of private wealth. This joining of private pension funds and management companies, an original feature of the Kenyan financial system, opens up important possibilities for the deepening of the financial system. Postal Institutions Most economic agents in the study countries are small-scale, in terms of both the volume of savings and their financial needs-farmers, owners of microenterprises, artisans, migrant workers, and so on. Together, however, they represent the largest group of potential users of financial services, and a potentially significant source of savings. Yet many have little or no access to financial institutions. Postal institutions (postal checking systems and postal savings banks), with their extensive networks, could 50 bridge this gap and provide basic financial services across the study countries. The costs for the customer would be small, and services would be available to a large section of the public. That potential isn't being realized in most of the study countries-Kenya may be the exception. Postal systems are in distress, and many are under major restructuring. The reasons include preemptive use of the funds by governments, weak separation between postal institutions and treasuries, and lack of dynamism, a consequence of bureaucratic behavior. Venture Capital Venture capital is temporary financing in equity capital or loans, with returns linked to profits and carrying some measure of managerial control (see Sagari and Guidotti 1992). Venture capitalists expect losses on some ventures to be greater than with traditional financing, but they invest because they think that greater-than-normal returns on other ventures will more than make up for those losses. Venture capital is ideally suited to projects involving a lack of collateral, making it an alternative to finance from development finance institutions. While there are no venture capital enterprises in the study countries, there are financial companies that could provide risk capital under the appropriate financial arrangements. In the study countries, venture capital is embryonic, and its development runs up against the problem of promoting new projects in growing niches and bringing together entrepreneurs having the appropriate managerial capacity and financial partners interested in supporting them. The development of venture capital enterprises would require an environment conducive to private sector initiatives, a tax system that fosters rather than penalizes venture capital operations, an appropriate mechanism to absorb losses, and an "exit channel," such as a stock exchange. Neither of the first two prerequisites exists in the study countries, and only four of the countries have fledgling stock exchanges. Social aspects may also impede the development of venture capital since entrepreneurs in the study countries do not seem to be willing to open up the capital of their companies to people or entities they do not know personally. Leasing Leasing offers a number of advantages as a form of finance for small and medium-size companies. Financing conditions are based on cash flow generated by the equipment. The guarantee is represented by the equipment, which removes the need for tying credit to unrelated collateral. Responsibility for equipment maintenance can be assumed by the lessor or the lessee. Leasing institutions are usually among the first private institutions in developing financial systems to issue medium-term bonds, thereby strengthening the bond market. For all these reasons, leasing grows fast in developing countries. The share of leasing in capital formation (excluding building and construction) is 20-28 percent in developing countries. 51 Despite these advantages, leasing is not welldeveloped in the study countries. First, there is no appropriate legislative or regulatory framework. Although most countries specify that leased equipment remains the property of the lessor, there are no provisions specifically governing disposition of the property if the lessee stops payment. Second, interest rates on leasing are higher than those applied by banks on overdrafts and short-term loans because of a dearth of markets for the medium-term financial instruments that leasing companies use to fund their operations. And third, refinancing of leases can cause problems: with the amount paid monthly established at the start, the lease must be attached to a medium-term loan, at a fixed rate. If the leasing company is successful, it may face difficulties finding the appropriate resources. Securities Markets Securities markets are essential to development since they provide long-termi debt and equity finance for the corporate sector and the government. By making long-term investment liquid, securities markets mediate between different maturities and maturity preferences of lenders and borrowers. Furthermore, securities markets facilitate the spread of business ownership and the reallocation of financial resources among corporations and industries. Several developing countries have made great strides in recent years in establishing and expanding equity markets, and such markets now exist in about forty countries. Sub-Saharan Africa has the least developed capital markets and perhaps the fewest in relation to other developing regions. Only about nine countries have established stock exchanges, but a few more are doing so. Of the study countries only C6te d'Ivoire, Ghana, Kenya, and Uganda have stock exchanges, though trading has not begun in Uganda's. The ratio of market capitalization to GDP is dramatically low in the three countries: 6.15 percent in C6te d'Ivoire, 1.2 percent in Ghana, and 5.17 percent in Kenya (Figure 4.1). Ratios are 50 percent in Chile and 114 percent in Malaysia. Activity is sluggish in the three markets even in Kenya's, which is the most developed. The turnover ratio is 1.3 in C6te d'Ivoire, 0.5 in Ghana, and 2.2 in Kenya, two to three times lower than in the stock exchanges of Bangkok or Djakarta. The markets are even narrower than the number of listed companies or turnover ratios would suggest. Of the thirteen companies listed in Accra, three represent two-thirds of the market capitalization (four, if only securities sold to the public are taken into account). One of these companies alone represents 40 percent. Of the fifty-four companies listed on the Nairobi Exchange, some thirty are dealt in, and major management companies limit their participation to six or seven "blue chips." In Abidjan, five companies account for 75 percent of transactions. The same litany of factors explain the weak development of capital markets in the study countries: the financial repression of the 1970s and the deteriorating economic environment of the 1980s, administered interest rates, unfavorable tax systems, weak and unclear legal and regulatory frameworks, preferential treatment for banks, and inefficiencies in stock exchange operations. 52 Figure 4.1 Market Capitalization (in termns of GDP) s a g .. .. . . . . . . . . . . . ............ 40%. . .. . . ... CN- E NOIRE . . .. .. NA = NIGERA JOR =JORDAN IND INDONESIA 10% THA THkJLA1D 0% GHA KEN CP/ NGAZWE JOR IND 7HA Source: IFC, Emerging Stock Markets Factbook, 1992 and other sources 53 Until the end of the 1980s, central banks regulated interest rates in C6te d'Ivoire, Ghana, and Kenya. Many countries kept the lending rate low-sometimes highly negative in real terms-to encourage investment under conditions in which project opportunities were limited. Such easy access to subsidized credit encouraged suboptimal investments and induced enterprises to rely on bank financing rather than to float equity or to issue bonds on the market, stifling the development of financial markets. Stock market operations have also been inefficient. In C6te d'Ivoire, for instance, the costs of issuing equities in 1990 reached 5.5 percent of the issue, compared with 0.5 percent to borrow from banks. Moreover, the trading system has not permitted prices to reflect changes in demand and supply. In Kenya, it can take up to 60 days to deliver securities. The stock exchanges further suffer from overstaffing, a shortage of skilled and trained personnel, little or no computerization, and, until recently, a lack of dynamism. Until the end of the 1980s, securities on the three markets received unfavorable tax treatment. Kenya introduced a capital gains tax in the mid-1970s. In C6te d'Ivoire, one of the fundamental problems has been multiple taxation of dividend income prior to its receipt by the shareholder. Typically, corporate profits are subject to corporate income taxes and then, if paid as dividends, to withholding tax and personal income tax at the rate applicable to individual shareholders. The legal and regulatory framework for securities markets remains weak. Responsibility for regulation and supervision of markets for securities, public companies, and institutional investors is not defined; incentives for companies to go public are lacking; requirements for information flows and disclosure are weak and ambiguous; and accounting and auditing standards remain inadequately specified (UN 1991). In C6te d'Ivoire, for instance, there is almost no provision against insider trading. Kenya is working to introduce a comprehensive legal framework that includes rules on trading, intermediation, information disclosure, regulation of takeovers, and professional codes of conduct for brokers and underwriters. But despite these encouraging developments, company law is still in its infancy: disclosure requirements are not clear, and protection clauses for shareholders are not included. Informal Finance Informal finance covers all lawful but unregulated activities, such as rotating (and non-rotating) savings and credit associations (ROSCAS), like the tontines, njangis, and susus in West Africa and the chillembas in East Africa, moneylenders and money collectors, and other providers of retail financial services. All are outside the purview of the legal, fiscal, regulatory, and prudential framework of the monetary and financial authorities. Since, typically, these informal and semiformal operations are small-scale in terms of membership and scale of operations, their activities are of special relevance to mobilization of retail savings. The strength and persistence of informal finance reflects its relatively lower information and transaction costs and, above all, its ease of access to low-income groups who do not have links with formal finance. It demonstrates that availability of basic financial services is of prime importance to low-income economic agents and that availability of credit is more important than its price. The absence of the 54 formal legal and regulatory and prudential framework and, in particular, supervision is compensated for by custom-based peer supervision. Although there are no aggregative statistics on the scope and share of informal and semiformal finance in SSA and study countries' financial systems, their share in financial activities is considerably higher than in other developing regions. Informal and semiformal finance has some inherent drawbacks in economies of scope and scale, maturity transformation, spatial transfer of savings, predominance of cash transactions, maturity spectrum of instruments and operations, and shallowness of intermediation. These deficiencies can only be remedied by an adequate and efficient formal system, which strongly underscores the need for an effective machinery of graduation to formal status. It is nevertheless imperative to examine and tap the full savings and credit potential of informal and semiformal finance at the retail end of the clientele spectrum comprising such small-scale target groups as women, landless labor, marginal farmers, artisans, petty traders, and small-scale and medium-enterprises. Savings and Credit Potential Informal finance is much more extensive and diverse than formal finance and accounts for most of the financial services, other than term finance, provided to the rural sector. Formal rural credit accounts for less than 10 percent of total credit in most SSA countries and it is the more affluent who have the greatest access to formal credit. Informal finance also accounts for a big share of financial services in urban areas. Spontaneous, well-established, and successful informal group savings and credit systems exist in most of Africa (except, maybe, Madagascar), based on voluntarism, autonomy, low transaction and information costs, convenience, flexibility, confidence and on nontangible collateral (often an oral promise). ROSCAs and moneykeepers are more prevalent in Africa than in Latin America and Asia; credit unions too have been remarkably successful in Africa, but in contrast, have generally failed in Latin America. Informal finance concentrates on deposit mobilization and safekeeping services on the credit side, information advantages and low transaction costs which are strong incentives for small savers despite high interest rates on credit. In fact, the performance of institutions that offer both deposit and credit facilities is perceptibly better than that of those that accept only deposits. Successive studies have established that there is "strong correlation between savings and loans, and a financial institution cannot function without both. But its advantage of being small scale and segmented also imposed distinct limits to the expansion of informal finance. Moreover, the savings mobilized by rural banks, post office savings schemes, and insurance corporations are generally preempted by government amd semiformal finance replicates the urban bias of formal finance insofar as it channels rural savings to urban credit. Paradoxically, despite its proved advantages, there is "general denunciation of informal and semiformal finance in SSA countries other than registered cooperatives and credit unions (RFM Main Report). " This is the challenge facing informal finance in 55 Africa. Official attitudes have to change drastically before constructive policies can be put in place. Formalizing Informality Central bank surveillance of informal and semiformal finance is a legitimate function insofar as uncontrolled expansion of cooperative credit carries potential risks. It is as (if not more) important for the central bank to adopt a positive attitude and constructive policy approach toward informal and semiformal finance. The aim should be to foster market-oriented link with the formal sector based on an intermediate financial technology within an appropriate legal framework. This would include * Linkage of informal entities like ROSCAs, money collectors, and so on, to the central bank, as done, for an example, in the registration of susu collectors to the Bank of Ghana. * Promotion of group borrowing and lending units for commercial banks. - Creation of a panel of guarantee brokers from among, say, the tontiners on a salary cum commission basis (like the compradores of commercial banks in Southeast Asia) to appraise and guarantee bank credit to ROSCAs in the informal sector. For this purpose, ROSCAs, money collectors, and similar institutions could be treated as borrowing entities empowered to borrow with group guarantee of credit. The allocation of such credit to individuals within the ROSCA should be left to consensus. * Offer of market-oriented rediscount and refinance (that is without any subsidy) facilities through the central bank for promissory notes or other credit paper originating from the informal sector (like the "Kalata" in Malawi). These could be guaranteed by commercial banks, authorized brokers, or responsible guarantors like the 'Mboni' in Malawi or the tontiners in West Africa. * Such rediscount and refinance facilities should also be extended to apex organizations of credit unions and cooperative societies. This would, concomitantly, involve appropriate central bank inspection and supervision procedures for semiformal finance. Such activities would not involve any direct financing and management by central banks, whose primary functions (controllers of money supply and prudential supervisors) should always remain unimpaired. Indeed, it is incumbent on the central bank to ensure that informal and semiformal finance does not pose problems for monetary management or create prudential risks, without unduly regulating these institutions. 56 FINANCIAL RESTRUCTURING As in many other regions of the world, the 1980s were a decade of financial distress for all the study countries. Perhaps the most visible case, however, was Benin, which saw a complete collapse of its banking sector in 1987-89. Banks in the WAMU group felt the effect of the crisis earlier in the 1980s, because of the structure of their balance sheet. Their loans traditionally have exceeded their deposits, while the opposite has been true in the non-WAMU and other sub-Saharan Africa countries. Problems with banks normally come to light only when a liquidity crunch develops. Before that, banks are generally able to hide their insolvency through various accounting manipulations, such as delaying the classification of nonperforming loans, capitalizing interest on those loans, and overvaluing dubious loan securities and guarantees. Because of their high loan-to-deposit ratio, banks in the WAMU group of countries were much more rapidly dependent on liquidity assistance from parent institutions or central banks. In contrast, countries with a low credit-to-deposit ratio had more room for maneuver before resorting to central bank credit and could delay the liquidity problem by drawing on reserves. Techniques of Bank Restructuring Recognition of insolvency. In the study countries, there was a belated recognition of financial distress. In many cases, governments and bank supervisors were either technically unable or politically reluctant to admit the extent of insolvency. Decisive action was avoided as long as possible, and the situation, of course, deteriorated. Usually, after drawing on reserves, banks attempted to shore up liquidity, by borrowing more from the central bank, increasing government deposits, reducing foreign assets, or increasing foreign liabilities. The increased reliance on government deposits which with the economic crisis became increasingly volatile, generally aggravated the situation, especially when governments eventually had to withdraw deposits to meet their own liquidity needs. This pervasive lack of will to tackle the problem left the most seriously affected to fall into such a state of illiquidity that, when the problem was finally tackled, the costs of the restructuring had rocketed. Governments were forced to deal with the problem at the worst moment, when the bank runs threatened systemic bank failure. By then, losses were so high that it was necessary to seek assistance from external donors. Case-by-case approach. When they did finally begin dealing with the problem, most African and study countries chose to do so on a case-by-case basis, attempting to first solve the most urgent problems. When the crisis was limited to a few institutions, individual measures were tailored to the special circumstances of the banks in trouble and isolated cases of liquidation, restructuring, and merger took place. This piecemeal approach, however, had serious shortcomings. It did not provide the authorities with an 57 overall view of the problem-that is, did not give a measure of the magnitude of the crisis nor the total cost of restructuring. Restructuring or liquidation. While Benin and Guinea did not have much choice but wholesale liquidation, some African countries and some study countries (like Ghana and Madagascar) chose to rehabilitate their banking systems through restructuring; others (most WAMU countries) through a mixture of restructuring and liquidation. Restructuring is less disruptive of financial internediation, may be less costly as the banks continue to operate and the costs may be spread over a longer period or even hidden. In several African countries and some study countries such as Uganda, the domestic payment system was operated solely by financially distressed banks; restructuring those did not disrupt the payment system. Restructuring, however, often throws good money after bad; it often only prolongs the agony of the terrninally ill institutions. For instance, the restructuring of the agricultural development bank BNDA (C6te d'Ivoire) began in 1987, but did not succeed, so that a liquidator had to be appointed in 1991. Restructuring mechanisms. Financial restructuring in the study countries has been marked by heavy governmental involvement and, in this context, two types of mechanisms are available: the stock and the flow techniques. Flow solutions. Some of the flow solutions were and still are used in the study countries. They usually have four aspects: (i) central bank or treasury liquidity support at subsidized rates for problem banks; (ii) allowance of high intermediation spread; (iii) use of seigniorage, including the inflation tax; and (iv) deregulating to allow new income sources. In the WAMU countries, banks under restructuring received strong support from the regional central bank, notably through the consolidation of claims at a subsidized rate of 3 percent over a period of 15 years. In Madagascar, the treasury granted interest-free loans to the three banks under restructuring. These loans are to be repaid out of recoveries from the delinquent borrowers. Allowance of high spread and using seigniorage including the inflation tax, is more a characteristic of non-WAMU countries. Inflation in Ghana, Kenya, Madagascar, and, above all, Uganda has reached high levels. Even when lending rates in Ghana and Kenya were controlled, deposit rates were allowed to remain relatively low to allow a more generous spread to the banking sector. The last flow solution is to deregulate banks and allow them to do business outside traditional areas, such as securities trading, investment banking, travel services, or insurance. This technique was not widely used in WAMU, although since 1976 commercial banks in C6te d'Ivoire have been allowed to act as brokers on the stock exchange and some Togolese banks sell insurance. Deregulation was used on the fringe, largely because there was no adequate supervision. Flow solutions are only likely to help alleviate financial distress when insolvency is confined to a few in the financial system. Moreover, the real sector problems must be the result of temporary shocks, and management of banks, enterprises, and the public 58 sector must be fairly effective. When it is not the case, a more radical approach must be used. Stock solutions. Basically, stock solutions can take four aspects: (i) liquidation; (ii) merger; (iii) capital injection by private or public sector; and (iv) carving out of bad debts to the central bank or a specialized government's agency. Liquidation was widely used in WAMU countries. In Benin, the three existing banks were liquidated. Liquidated banks reached five (out of nineteen) in C6te d'Ivoire, four (out of ten) in Niger, six (out of sixteen) in Senegal, and one (out of nine) in Togo. This was not the case in non-WAMU countries. In Ghana, Madagascar, and Uganda, no banks were closed. In Kenya, merger was preferred to liquidation; this was true also of Burkina Faso. It was tried in C6te d'Ivoire and Senegal but failed because mergers took place between insolvent and quasi-insolvent institutions, which is not much of a solution. Some recapitalization and cleaning of the balance sheet must take place beforehand. C6te d'Ivoire and Senegal have not followed this rule, which partly explains the final liquidations that took place. In Kenya, merger was one step of a comprehensive process including recapitalization, and the new institution, the Consolidated Bank of Kenya, has some chance of viability. Capital injection by the private or public sector was broadly used in all study countries. In the WAMU countries, foreign shareholders, traditionally international French banks, as well as the governments were asked to recapitalize the restructured banks. In Ghana, the same thing happened. In Kenya, the new bank was recapitalized, notably by the conversion of parastatal deposits into equity. In Madagascar, new international banks took a share of the capital of the previous problem state-owned banks, permitting the privatization of the banking sector. The problem of loan recovery. The final cost of restructuring is difficult to estimate since it depends on loan recovery. In some African countries, responsibility for recovery is left to the restructured financial institution or to the liquidator; in most (and virtually all study countries), it has been entrusted to a special recovery unit such as the Societe National de Recouvrement in Senegal or the Non-Performing Assets Recovery Trust in Ghana. Institutions to handle nonperforming assets are of two types. In Benin, C6te d'Ivoire, and Senegal, the recovery unit centralizes the disposal of assets to reduce administrative costs and speed up recovery of nonperforming loans. With the exception of Ghana, this turned out to be excessively slow in study countries because of legal inefficiencies. The purpose of agencies of this first type does not consist in supporting banks under a restructuring process. In Ghana, and to a lesser extent in Burkina Faso, the central unit has a broader mandate. In Ghana, for instance, it received the validated nonperforming assets, which were replaced by interest-rate-bearing Bank of Ghana's bonds in the balance sheet of the problem banks. The unit can liquidate the assets, restructure debts, or help indebted enterprises back to health. The main problem with these institutions, however, is that they can become bottlenecks to rapid change, although 59 they fulfilled an essential function in study countries without which financial distress would have not been tackled effectively. In African and study countries, recovery of nonperforming assets has been disappointing (except for Ghana), mainly because of half-hearted commitment of the authorities and legal and administrative obstacles. Also, after the most obvious bad assets have been recovered or sold, costs as a proportion of the recovery unit cash flow mount rapidly, as recoveries gradually decrease. What Needs to be Done? In the long term, financial fragility must be reduced. But how? Ghana has shown that it is possible to create a stable macroeconomic environment, to bring the fiscal and budget situation under control, and to establish market-based incentives that favor private sector growth and competition. Ghana and Uganda also show that policy distortions can be reduced and interest rates liberalized, and that competition within the financial system can be spurred. All the study countries have begun to overhaul their legal, regulatory, and prudential framework. Some progress has been made in transparency and auditing and accounting. Training in financial techniques and skills has been enhanced in some countries. These ongoing and planned reforms will eventually strengthen the financial systems and make them better able to respond quickly to future problems. Given the higher risks inherent in the economic environment and the poor quality of management and information systems, it may be wise in a number of countries to require bank capital adequacy ratios in excess of the 8 percent of risk-weighted capital recommended by the Bank of International Settlement. In addition to these structural issues, there are some basic short-term requirements for financial restructuring to be successful. Among them are three prerequisites relating bank restructuring to the fiscal and financial situation of the public sector, enterprise restructuring, and prudential regulations and supervision. In the study countries, too little attention was paid to the financial situation of the public sector, which was asked to finance a major part of the restructuring process. If an insolvent institution cannot rescue another insolvent institution, an insolvent government cannot rescue an insolvent financial system. A treasury cannot bear the cost of bank restructuring if public finances are in a sorry state and the government has reached its borrowing limits. For heavily indebted countries, such as the study countries, restructuring should avoid a cumulative deterioration of the debt-to-GNP ratio. Of course, the financing of the restructuring through the issue of bonds increases the stock of debt and worsens the debt- GNP ratio. But the restructuring should not be designed in such a way if it implies a further cumulative deterioration of the ratio. The solvency condition indicates that a government can meet its interest bill by issuing more bonds, but only if the economy's real growth in output and income equals or exceeds the real interest rate. If that is not the case, the government must run a sufficient surplus to pay that part of its interest bill that cannot be met by printing bonds, to avoid a deterioration of the debt-to-GNP ratio. This surplus must equal the current government debt times the excess of the real interest rate 60 over economic growth. Unfortunately, this was not the situation prevailing in most of the study countries when restructuring plans were implemented. With the exception of Ghana, none "grew out" of financial distress. Similarly, a central bank that has been weakened by excessive lending to the public sector, with high interest liabilities and net foreign exchange liabilities, does not have the capacity to absorb losses due to bank failure. A central bank in that position may end up monetizing its losses through the growth of monetary liabilities, generating inflation. Only a central bank with non-interest-bearing liabilities, high capital, and reserves represented mainly by foreign exchange assets would be in a position to absorb losses due to bank failures. The central bank angle should also be taken into account when designing bank restructuring plans. Bank restructuring must also be closely related to enterprise restructuring, as in Ghana. If not, most nonperforming loans will never be recovered and the quality of bank portfolios may deteriorate again after the restructuring. This aspect has been taken into account in Ghana. For potentially viable enterprises, the Non-Performing Assets Recovery Trust (NPART) can grant a debt moratorium, debt rescheduling, and conversion into subordinated debt. Moreover, the government encourages the establishment of venture capital companies to increase equity finance. Such a model, where the central recovery unit may play an important role in enterprise restructuring, could inspire other African countries. Finally, it is futile to restructure the banking sector if supervision is not also im- proved. Both the technical skills and the enforceability of the recommendations of the supervisory body must be improved. In Kenya, the technical supervision ability of the central bank was strengthened but enforcement remained weak. As a result, two years after restructuring, some financial institutions again face difficulties. On the other hand, the WAMU Banking Commission created in 1990 has both the technical skills and the power to enforce its recommendations. As such, it constitutes quite an improvement on the previous situation. The supervisory body must have an early warning system to control damage and to head off problems quickly before financial distress becomes widespread. Basic indicators of difficulties can be a rapid growth in assets, deposit rates higher than market rates (such as, for instance, the BCCI in Togo), the late submission of bank reports, rapid staff turnover, and so on. Off-site and on-site inspections are both necessary for bank supervision to be efficient. As financial systems are reformed and become more market-based, the potential for failure will increase. Three procedures should be established. First, an early warning system should be put in place to allow the identification of "potentially distressed" financial institutions and to deal with them before systemic hazards develop or the costs of restructuring get out of hand. Second, there must be a clear and systematic procedure to deal with financial institutions in difficulties. Third, countries must establish a permanent mechanism and institutional set-up to deal with nonperforming assets. The choice of options is wide, and different mechanisms could be applied in different circum- stances. Distressed banks may be asked to recover bad assets. An agency could do that. 61 The central bank may or may not be involved. Debts and enterprises may or may not be restructured in the process. More generally, entry and exit policies must take into account the growth potential of financial systems, the ability of the central bank to supervise a larger number of institutions, and the need for competition. The strategy for new entry should be tailored to country-specific features and market conditions. In countries that already have a reasonable number of private banks, unrestricted free entry tends to reduce profits and increases the likelihood of failure. The solvency of banks in general can be enhanced by limiting entry to sound, well-managed, and well-capitalized institutions. 62 5. MONETARY MANAGEMENT Experience has shown and empirical research has confirmed that the behavior of monetary aggregates, particularly their rate of expansion, has an important bearing on the level of prices and on the balance of payments, as well as on other factors of activity and growth of the "real" economy, such as investment (Fry 1988; Brunner 1989; and Mosser 1989). Movements of monetary aggregates can be managed by the central bank in essentially two ways: through its regulatory power or in its capacity as sole issuer of money-that is, through financial institution balances at the central bank and the amount of currency in circulation. The aim of direct monetary control is to influence the behavior of financial institutions principally by imposing limits on the price or quantity of credit and money. The aim of monetary management through indirect methods and instruments is to influence their behavior through market forces. To achieve this, the central bank uses its power to create and destroy central bank money and allows variations in money market interest rates to be freely transmitted through the economy. Direct instruments act on financial institution credit, while indirect monetary management influences financial institution liquidity and thereby the potential of these institutions to grant credit as well as this credit's cost. Direct instruments ration the net domestic assets of the financial system, while indirect instruments limit the net domestic assets of the central bank. Direct instruments administer either prices (interest rates) or quantities (credit aggregates), while monetary management through indirect methods and instruments operates by influencing demand and supply. When monetary management is based on direct control of credit, macroeconomic policy and credit allocation are part of the same process. Since base money consists of currency in circulation and financial institution balances at the central bank, indirect monetary management requires the central bank to be capable of managing its balance sheet. Under direct monetary management, banks and other financial institutions allocate loans within direct limits established by the monetary authorities: interest rates may be regulated, the overall capacity to extend credit may be limited by credit ceilings, and the composition of the portfolio of banks may also be regulated. Under direct monetary management, banks and other financial institutions have automatic access to the central bank to refinance credit, sometimes within a global limit. Under monetary management based on indirect methods and instruments, banks and other financial institutions are free to allocate credit, but access to the central bank is no longer automatic. They have to compete for funds in the interbank market and money markets and for the deposits of economic agents. Interest rates in these markets are an important factor of competition. Because the central bank is a participant, it has the capacity to influence rates. Banks and other financial institutions then extend credit in accordance with the quantity and cost of the loanable funds available to them, as well as the maturity and risk related to specific lending operations. The actions of banks and other financial institutions influence other 63 institutions, markets, and the value of assets. All those changes (be they the results of direct or indirect monetary management) affect in turn the expenditure decisions of households, businesses, and the government. How then can monetary aggregates and their components best be regulated in pursuit of policy objectives in economies of the sub-Saharan region, including the study countries? Theory asserts that macroeconomic and macrofinancial stability is maintained when the supply of money expands at approximately the rate at which demand for money is growing. This assertion raises three questions. Can the demand for money be influ- enced by the monetary authorities? Is the demand stable and predictable? How can the supply of money then be managed in relation to demand? Shortcomings of Direct Control Systems Direct instruments of monetary control consist, most commonly, of credit ceilings, directed credit, fixed and administered interest rates, and subsidized interest rates. When properly implemented and monitored, direct instruments can be useful in achieving narrowly defined (often temporary or transitory) targets, such as keeping a bank's overall credit expansion under a certain ceiling. They can also be effective when used in combination with indirect instruments. However, the macroeconomic impact of direct controls may be unpredictable because of the scope for evasion and avoidance. Moreover, direct controls are difficult and costly to design and enforce over a prolonged period of time, and they are administratively cumbersome and prone to moral hazard. Finally, direct controls have several perverse side effects (Caprio and Honahan 1991; Johnston and Per Brekk 1989; World Bank 1989; and Gillis and others 1987). In the study countries and throughout the region generally, direct controls have created numerous difficulties in the mobilization and allocation of financial resources, financial intermediation, economic growth, and macroeconomic policy. Some of these effects were inherent in the controls; others resulted from the interaction of the controls, external conditions, and other government policies, such as those related to the fiscal deficit, trade, or the exchange rate. Studies also show that direct controls have reduced savings and distorted the allocation of resources. Compliance and regulation costs have increased, financial competition and intermediation have been undermined, and financial innovation has been stifled (Fry 1988; World Bank 1989; Johnston and Per Brekk 1990). Intermediation spreads have widened. Existing lines of business have been favored at the expense of new enterprises. Direct controls have often included administered interest rates. Under such regimes, nominal interest rates for both deposits and loans have been kept low and real rates have often been negative. The most frequently cited rationales for administered credit allocation and interest rates have been the need to offset "market failures," the need to promote economically or socially justifiable projects, or, more plainly, the need to finance the activities of public enterprises and the expansion of the public sector. These and other arguments failed to recognize that demand for credit is derived from opportunities to engage in productive investment. Allocating credit does not create these opportunities (Brunner 1989). Like- wise, these rationales failed to recognize that interest rates eventually reflect the economic 64 demand for credit is determined by the pattern of economic activities and to presume that capital and loanable funds are not scarce. Moving toward Indirect Management These difficulties and perverse side effects have induced the monetary authorities of many developing countries, including some countries in sub-Saharan Africa, to move toward indirect monetary management, mostly market-based. Base money (or the monetary base) consists of currency in circulation and reserves of banks and other financial institutions at the central bank. In practice, the central bank has little influence in the short term on the volume of currency in circulation, which is determined by payment practices, transactions in the economy, behavioral patterns, and the depth and diversification of the financial system. Movements in the reserve money, the noncurrency part of the monetary base, are determined mainly by movements in other elements of the central bank's balance sheet. In African countries, movements in reserve money are determined mainly by movements in the net foreign assets and net domestic assets of the monetary authorities. These result from various external and internal factors, including changes in terms of trade, the exchange rate, capital movements, and domestic credit. Because African countries are heavily dependent on trade (and because the prices of primary commodities on which they depend for exports are volatile), changes in terms of trade have had important impacts on their economies. They have had direct impacts on net foreign assets. These external shocks are obviously beyond the control of monetary authorities. With changes in net foreign assets almost completely outside the monetary authorities' direct control, their control over base money can be achieved by influencing net domestic assets. These principally represent net claims on the central government, public enterprises, and the private sector. But central banks from sub-Saharan Africa are mainly answerable to ministries of finance. By the standards of industrial countries and of other developing regions of the world, central banks in sub-Saharan Africa have little independence in setting monetary policies. They have little leeway in restricting credit to the government, and the absence of markets for bills, bonds and notes further limits their scope of action, leaving little alternative to deficit monetization. The policy variables left to manipulate are credit to the private sector and to public enterprises, to the extent that lending to public enterprises is not also politically determined. In the non-WAMU countries, credit to public enterprises was generally politically determined during most of the 1980s. In the WAMU zone and the BCEAO, credit to public enterprises was determined by the national government, not the central bank. In effect, that left credit to the private sector as the major instrument available to the central banks to influence movements in reserve money. By its nature monetary management through indirect methods and instruments reduces government interference (Table 5.1). Experience has shown that in a suitable macrofinancial environment, this type of management is more efficient than the use of direct controls and enhances the efficiency of the financial system (Killick and Matthew 65 Table 5.1 - Market Conformity of Monetary Instruments Low High Direct instruments Interest rate regulations Net credit ceilings Sectoral credit ceilings Credit ceiling with trading Gross credit ceilings in unused margins Indirect instruments Reserve requirements Open market operations Rediscount policies Auctioning money market liquidity Swaps Note: Market conformity can be defined as the degree to which the use of an instrument leaves room for market forces to determine the outcome of the process of money creation. 1990; Duesenberry and McPherson 1991; Bank Negara Malaysia 1989). There are, however, some prerequisites for successfully moving away from monetary management based on direct controls toward monetary management based on indirect methods and instruments. (Table 5.2 describes some of the instruments used for monetary control in a sample of eight developing countries outside Africa.) Macrofinancial Enviromnent Indirect monetary management requires a stable macrofinancial environment. The process of transition itself is likely to increase the variability of monetary parameters. Interest rates will become more volatile, and new, indirect ways of monetary management are likely to lead initially to undershooting or overshooting of new intermediate and operational targets, with possibly wider fluctuations in credit, monetary aggregates, or interest rates (Meek 1991). Macrofinancial instability at a time of transition would magnify the variability of monetary parameters, endangering transition. Conversely, macrofinancial instability complicates the conduct of indirect monetary management. Moreover, the public sector financial situation must be under control and sufficiently stable to be sustainable in the long term. If not, the reserve money supply is likely to go offtrack sooner or later. That is particularly likely to happen in countries in sub-Saharan Africa, where central banks are subordinated to ministries of finance and where markets for government bills and bonds are thin or nonexistent. Some recent studies have shown that in sub-Saharan Africa, the budgetary situation is a dominant factor in monetary management (Killick and Mwega 1990; Duesenberry and McPherson 1991). 66 Table 5.2 - Instruments of Monetary Control in Eight Developing Countries Instruments of Other Primary issues daily money market-based Refinance Re6erve Country of securities market instruments facilities Direct controls requirements management Argentina Weekly auctions of Repurchase/reverse Several, Interest rate Actively particip- repurchases, including regulations. varied. ations in government foreign exchange subsidized Quotas on papers swaps. facilities. nonregulated in the form of central rate deposits, bank acceptances bills. and swaps. Brazil Weekly auctions of Repurchase/reverse Several, Selective credit Not actively central bank bills (and repurchases, and inicuding controls. varied. treasury bills, but mainly outright sales. subsidized for budgetary finance). facilities. Indonesia Auctions of central bank Repurchase Several, Interest rate and Not actively bills. The frequency of agreements and including credit controls varied. auctions has varied daily auctions of subsidized lifted on part of between daily and central bank bills. facilities. financial reform. weekly. Malaysia Weekly auctions of Discount window, Several, Interest rates Not actively treasury bills solely for outright sales and including liberalized, varied. government finance purchases of subsidized selective credit purposes. government facilities. controls securities, re- remain in effect. purchase/reverse repurchases, foreign exchange swaps, re- cycling of government deposits. Mexico Weekly auctions of Outright sales and Weekly Credit ceilings. Actively treasury purchases and auctions of varied. certificates. repurchase/reverse fixed deposits repurchases. and placements from the central bank. Philippines Weekly auctions of Repurchase/reverse Single facility Interest rates Actively treasury repurchases, covering an have been varied. bills. foreign exchange array of liberalized. swaps. purposes. Interest rate charged is linked to market interest rates. Sri Lanka Weekly or more frequent Repurchase/reverse Several Interest rates Actively auctions of treasury bills. repurchases, including liberalized. varied. foreign exchange subsidized Some credit swaps. Second- facilities, controls ary window for sale remain but are and purchase not strictly of treasury bills. enforced. Thailand Weekly auctions of Repurchase/reverse Central bank Several, Interest rate Not actively treasury repurchases, bond issues. including ceilings, direct varied. bills.(Mainly to satisfy discount window. subsidized controls on demands for liquid assets; facilities. credit of limited importance for last used in monetary management.) 1984. Source: Barry R. Johnston and Odd Per Brekk, "Monetary Control Procedures and Financial Reform,' Washington, D.C., IMF, 1989. 67 The financial system must also be in good health. If a substantial part of the system is in financial distress, that is likely to undermine efforts to move toward indirect monetary management. Because most indirect instruments are market-related, markets need to work efficiently. Markets such as interbank and money markets do not operate properly under conditions of major financial distress. Moreover, because of the added difficulties of predicting the level of money supply and keeping it under control, the transition is more difficult. In sub-Saharan Africa, transition from direct to indirect controls also requires the development of new financial markets. There must be a minimum market base-at least an interbank market and embryonic money markets-and competition must be present in these markets and in the financial system at large. Control of the Money Base The central bank must be free to manage its balance sheet in a way that allows it to influence the elements of the money base that can be controlled and the other elements whose movements can be offset. Moreover, it has to have the appropriate legal, regulatory, and enforcement powers and an information system that provides timely data on bank reserves and money conditions. Finally, its staff has to have the skills needed to implement its reserve management objectives through day-to-day operations in the money markets. Coordination of Monetary and Fiscal Policy Especially in sub-Saharan Africa, where the government's budgetary situation is a dominant factor influencing monetary management, indirect management requires the close coordination of fiscal and monetary policy so that the public sector's deficit financing requirements are consistent with the objectives of monetary policy. Further, there must be close monitoring of fiscal out-turns, in particular if their predictability is weak. Staff from the central bank and the ministry of finance must be able to work effectively together. For all these factors to play their role in reserve management, there must be a mechanism through which changes in monetary policy are transmitted to the real economy, and the mechanism must be clearly identified by policymakers and monetary managers. Competition in financial markets tends to produce the most efficient channels. Flexible interest rates that respond to market signals are in turn essential to this competition. Moreover, to ensure that the promotion of competition does not subject the financial system to unwarranted systemic risk, financial supervision must create the prospects that major financial failures will be minimized. Finally, there must be efficient monetary programming associated with an effective decision making process to support and guide indirect monetary management. 68 6. PERFORMANCE IN THE STUDY COUNTRIES In the 1980s, monetary developments in the study countries were shaped mainly by four factors: the pace of economic growth; the impact of external and internal shocks, especially on the balance of payments and the budget; the balance sheet of the central bank; and the outcome of monetary management related to these factors. With few exceptions, the study countries experienced slow or stagnant real growth, deteriorating food production, and declining per capita income. Hence, real money balances did not expand. Concomitantly, little seigniorage was generated from incremental increases in real money stock. Severe deterioration in the terms of trade depressed foreign exchange earnings and constrained foreign trade. Financial distress and financial restructuring affected central bank balance sheets and imposed new claims on government budgets and central banks. All of these factors, combined with poor governance and rigid expenditure structures, deepened public sector deficits and led to their increased monetization. The non-WAMU countries that did not apply fiscal adjustment measures, such as Ghana and Kenya, monetized much of their public sector deficit or quasi-fiscal deficit. In all countries, central banks had to support distressed financial institutions. For the WAMU countries, where monetization of deficits is statutorily limited, governments resorted to roundabout monetization through credit directed to public enterprises and government-owned banks. Moreover, the regional central bank was called on to maintain liquidity in the distressed financial systems of some countries. In the second part of the 1980s, there were huge accumulations of arrears in public sector domestic debt payment. These developments reflected the evolution of base money and reserve money in the study countries. Changes in three balance sheet items-net claims on government, net claims on financial institutions, and net foreign assets-accounted for most of the movements in base money in both groups of study countries. Other factors were only occasionally important-restricted deposits, for instance, exerted a contradictory influence in some of the study countries in the early 1980s, and the category "other items net" (mainly claims on governments and public sectors) exerted an expansionary influence in the late 1980s. In sub-Saharan Africa, external and internal shocks are among the main causes of exogenous movements in the money base, particularly in reserve money (Figure 6.1 and Tables 6.1 and 6.2). The effects of these shocks may be offset or magnified by the response of the central bank or other fiscal and financial authorities to the shocks. Contractionary or expansionary movements in net foreign assets produce corresponding changes in reserve money unless they are offset by changes in other categories of the central bank balance sheet. 69 Figure 6.1 Source of Reserve Money Growth Flow (billions of UMOA Francs) /: - . - . .. .. . .. . . . - .: . .. . ..... . .. ... .. ... 0 - -.. - - - c ...n.O..n. . . - ..wew: : :: . ::: : ::: . -... .C ..... -, .et. / . ozei.:t -:: ''- '-: .: ' .-. ... :-.. . . ..-i.. 4200 } __ /~~~~~~~~~~~~~~~~3 -:: m f V . -WMU --:: EJOth.r fromli Not OIRaetrioced Deposit 100- i- - m.-: nI. -u-., :. - -1 00 -300) ....NOIV-WA MLJ - -1 co 11~~~~ . .. . Q.t . __. . . ..eveMn T.* -.AnY -200 T; -50 7 1 -7 o 985 71985- 7 .9907 _300~~~~~~~~~7 Table 6.1 Source of Reserve Money Growth in Non-WAMU Countries (billions of CFA francs) 1981 1982 1983 1984 1985 1980-85 1986 1987 1988 1989 1990 1985-90 Flow Flow Change in reserve money 243.5 196.9 (329.1) (196.0) (10.2) (94.9) (23.1) (23.3) 9.9 8.6 (23.8) (51.7) Change in: Net claims on govermment 280.2 102.1 (4.2) (328.4) (15.9) 33.7 (48.5) 18.2 (51.5) (67.4) (11.0) (160.2) Claims on nonfin pub. ent. 61.6 205.5 (319.9) 19.4 33.5 0.1 (22.9) (17.6) (15.1) 15.2 (13.0) (53.4) Claims on other financial institutions 2.2 3.0 (3.6) 1.9 5.9 9.3 (4.4) (2.8) (0.2) 0.9 (1.8) (8.3) Claims on dep. money banks 0.0 0.1 (0.1) 0.0 0.1 0.2 (0.1) (0.1) (0.0) 0.0 0.0 (0.1) Restricted deposits (18.2) (78.1) 34.8 44.5 0.4 (16.7) 10.1 10.3 11.8 3.8 0.9 36.9 Other items net (43.9) (24.3) 113.4 30.6 7.1 82.9 (9.0) (7.7) 43.9 37.1 (26.8) 37.5 Counterpart funds (0.8) (0.3) 1.9 0.6 (0.4) 1.1 0.1 (0.1) 0.7 0.4 0.5 1.5 Capital account (0.8) (0.2) (3.6) (0.7) (0.6) (5.9) 1.3 (0. ) (0.1) (0.5) 0.5 1.0 Change in net foreign assets (36.9) (10.9) (147.8) 36.2 (40.2) (199.6) 50.4 (23.4) 20.4 19.1 26.9 93.3 Memorandum items Reserve money (end-year) 524.3 721.2 392.2 196.2 186.0 162.9 139.6 149.5 158.1 134.2 Change in reserve - 37.6 (45.6) (50.0) (5.2) (12.4) (14.3) 7.1 5.7 (15.1) money (%) Source: IMF. International Financial Statistics, IMF data. Monetary Developments in the Study Countries Non-WAMU group. In the first half of the 1980s, the evolution of reserve money in non-WAMU countries was influenced mainly by movements in net foreign assets, net claims on government, and other items net (see Table 6.1). Net foreign assets declined, exerting a contractionary influence. Net claims on government and other items net expanded modestly, yielding lower expansionary influences. Contractionary influences exceeded expansionary ones for the period, resulting in significant shrinkage in reserve money. In the second half of the 1980s, reserve money shrank as well, though the forces moved in opposite directions. Net foreign assets increased, yielding an expansionary influence, and net claims on government and on official entities decreased, yielding an even greater contractionary influence. These patterns in reserve money movements reflect the general change in macroeconomic and fiscal policies that took place in most non-WAMU countries in the 71 1980s. Expansionary policies dominated in the first half of the decade, depleting foreign assets and swelling credit to the public sector. Macrofinancial stabilization and adjustment programs and greater fiscal discipline characterized the second part of the decade. Net claims on government declined, reflecting net repayments of outstanding claims to the central bank by the public sector. Ghana began the adjustment process early in the decade, whereas Uganda was a later starter. Kenya went through a stabilization program in the first half of the 1980s and an adjustment program in the second half, but with less continuity than Ghana. Madagascar followed a sequence of adjustment measures broadly comparable in timing and intensity to Kenya's. The expansionary or contractionary impact of net credit to the public sector has, however, to be adjusted for public sector financing through quasi-fiscal deficits run by central banks. This deficit financing had an expansionary influence on reserve money in non-WAMU countries. In some cases, other items net represents claims on the public sector. That happens when governments, following prolonged periods of fiscal austerity unrelieved by revenue increases, turn to various modes of off-budget financing. Frequently observed in sub-Saharan Africa is deficit financing through the quasi- fiscal activities of the central bank. In some countries of the group, part of this financing has been channelled through central bank revaluation accounts. All banks in the non- WAMU countries show a substantial revaluation account on their balance sheets, originating from the need to account for changes in the value of net foreign assets following devaluations. Since net foreign assets shown on central bank balance sheets in the four countries became increasingly negative in the 1980s, their increased value in domestic currency following devaluations had to be reflected on the asset side of the balance sheet by an offsetting entry called the revaluation account. No monetization was involved since the transaction was purely an accounting measure. Over time, however, revaluation accounts began to record other foreign exchange operations and transactions, becoming a mix of accrued and cash operations and producing losses. In addition, in several countries, the revaluation account represented a substantial proportion of central bank assets and that share did not yield any revenue. The imbalance between non-interest- bearing assets and interest-bearing liabilities resulted in losses for the central bank, with direct expansionary monetary impact. The expansionary or contractionary impact of net credit to the public sector also has to be adjusted for central bank advances to distressed public financial institutions. The effect of this financing of public entities often appears in claims on financial institutions, a balance sheet item different from credit to government. In all countries of this group, central banks came to the rescue of distressed financial institutions and, with the exception of the Bank of Ghana, the traces still show on their balance sheets, though in relatively modest amounts. WAMU group. For the first half of the 1980s, the evolution of reserve money in the WAMU group of countries was influenced mainly by movements in net foreign assets, net claims on governments, and claims on financial institutions (see Table 6.2). Net foreign assets 72 Table 6.2 Sources of Reserve Money Growth in WAMU Countries (billions of CFA Francs) 1981 1982 1983 1984 1985 1980-85 1986 1987 1988 1989 1990 1985-90 Flow Flow Change in reserve money 95.0 26.6 31.9 114.1 110.0 377.5 76.7 18.0 (50.9) (15.2) 57.2 85.8 Change in: Net claims on govermnent 116.5 75.9 160.9 60.9 22.5 436.7 12.5 (46.4) 28.4 (43.6) (12.2) (61.4) on financial institutions 137.6 90.9 74.7 (103.6) 13.2 186.4 31.0 84.6 43.4 (71.8) 64.7 151.9 Other items net 11.9 (14.3) 17.2 37.0 (48.5) 3.2 (26.5) (18.8) (5.8) (27.2) (12.8) (91.1) Change in net foreign assets 171.1) (125.9) (220.9) 119.7 149.3 (248.8) 56.7 1.6 (116.8) 127.4 17.5 86.3 Memorandum items Reserve money (end-year) 536.8 563.3 595.2 709.3 819.3 896.0 914.0 863.1 847.9 905.1 - Chne in reserve money(%) 21.5 5.0 5.7 19.2 15.5 _ 9.4 2.0 (5.6) (1.8) 6.7 - Source: IMF, International Financial Statistics, IMF data. declined, yielding a contractionary influence. Net claims on government and claims on fmancial institutions expanded by far more than the decline in net foreign assets, for an expansionary influence that brought about a huge net increase in reserve money. A reversal of this pattern of reserve money movements occurred in the second half of the 1980s. Net claims on government and other items net, apparently containing significant claims on government, exerted contractionary influences while net foreign assets and claims on financial institutions had expansionary impacts. Expansionary influences outweighed contractionary influences by a moderate margin, leaving a small net increase in reserve money. The regional central bank, BCEAO, is restricted by statute from lending to a government more than 20 percent of the government's past year's revenue. The 20 percent is calculated on a cumulative basis so that if a government's revenue does not increase from one year to the next, the BCEAO is not allowed to extend incremental credit to the government. This limit has been circumvented, however. Some banks, including publicly-owned ones, were induced to lend to public enterprises, and a large part of this lending was "refinanced" by the BCEAO. This amounted to additional credit to the public sector-hence the huge increase in net claims on government. In the second half of the 1980s, several developments constrained central bank direct and indirect lending to public sectors. First, the 20 percent rule was enforced more strictly. Second, 73 financial distress, particularly the insolvency and liquidity of publicly owned financial institutions, curtailed lending to public enterprises. Finally, toward the end of the decade, the BCEAO strictly limited its refinancing activities in order to stem the hemorrhaging of liquidity resulting from the worsening financial distress. These developments led, however, to a huge increase in public sector payment arrears, which peaked at 2 percent of aggregate GDP. By the end of the decade, the BCEAO had consolidated on its balance sheet all the credit and refinancing to insolvent banks in the second half of the decade, the majority of them government-owned. Almost all the banks were eventually closed. BCEAO's assets representing this refinancing could-and possibly should-be considered as claims on government since the insolvent banks that were closed belonged to governments. But as of today, these BCEAO assets are recorded under claims on financial institutions. Comparison of the two groups. For both groups, claims on financial institutions increased over the decade and exerted an expansionary influence on reserve money. This movement in reserve money reflects the financial problems of many financial institutions in the study countries and the consequent central bank support. This movement was much less pronounced in the non-WAMU group of countries than in the WAMU group of countries, however. In the first half of the 1980s, a striking similarity between the two groups of countries was the fall in net foreign assets. This drop reflected, among other things, the sharp deterioration in terms of trade between 1977 and 1983-by 60 percent for the WAMU group of countries and about 100 percent for the non-WAMU group. A second similarity was the increase in net claims on government, which in the case of the WAMU group more than offset the relatively smaller decrease in net foreign assets. Where the two groups of countries diverged was in how they achieved additional financing of the public sector: in the WAMU countries by BCEAO refinancing of bank credit to the public sector, and in the non-WAMU countries by monetizing the fiscal deficit and quasi- fiscal deficit. In the second half of the 1980s, the divergence widened. In both groups net foreign assets increased and net claims on governments decreased, but in the WAMU countries there was a marked increase in claims on financial institutions, reflecting BCEAO refinancing of insolvent financial institutions and the eventual consolidation of these claims. For the non-WAMU countries, net foreign assets exerted the only expansionary influence. As a result, there was a net increase in reserve money in the WAMU group and a net decrease in the non-WAMU group. There was also a strong build-up in domestic payment arrears in the WAMU group. This analysis highlights the close relationship between the fiscal and the monetary situation in sub-Saharan African countries. With narrow financial systems and no markets for government debt, excessive deficits are bound to be monetized sooner or later, unless they are "financed" through the accumulation of public sector arrears. Also emerging from the analysis is the importance of fiscal adjustment as a prerequisite to financial reform. Countries of the non-WAMU group implemented fiscal adjustment programs earlier and more thoroughly. As a result, in the second part of the decade reserve money 74 was under better control. The WAMU countries began fiscal adjustment in earnest only in the last years of the 1980s, and the second half of the decade showed notable increases in reserve money in a stagnant economic environment. It can be postulated that without the build-up of domestic arrears, the increase in reserve money would have been much greater. The analysis also reflects the different positions of the central banks relative to financial restructuring. In the non-WAMU countries, central banks engaged in little permanent financing of bank restructuring, and they kept liquidity flowing into distressed banks. In Ghana, the nonperforming assets were only a transitory item on the Bank of Ghana's balance sheet. Nonperforming assets resulting from credit extended to public enterprises were absorbed by the budget, and nonperforming assets resulting from credit extended to private sector entities were absorbed by the Non-Performing Assets Recovery Trust (NPART). In Kenya, Madagascar, and Uganda, however, financial problems remain, though to different degrees. In Kenya and Madagascar, financial restructuring has not yet reached the stage of allocation of losses. In Uganda, losses may end up being absorbed by the institutions in distress before they are recapitalized. In Kenya and Madagascar, the issue remains open. In the WAMU countries, the BCEAO injected liquidity into failing banks through "refinancing" and later consolidated these short-term claims into long-termn assets, burdening its balance sheet with a substantial amount of frozen assets. The impact on the monetary situation was, however, similar in both groups: by keeping failing banks liquid, central banks kept increasing the supply of reserve money. 75 7. REFORM PERSPECTIVES Weakness in the control of reserve and base money and, ultimately, in the control of the money supply is a key reason for the generally poor monetary performance in the study countries (except Ghana) and in most sub-Saharan African economies. In the short term, it leads to inflation, currency depreciation, capital flight, and macrofinancial instability. Macrofinancial instability in turn complicates monetary management and undermines its effectiveness, creating a vicious circle. In the longer term, it not only undermines the effectiveness of macroeconomic management, but also hinders adjustment and prevents the diversification and deepening of the financial system, thereby enhancing the potential for macrofinancial instability. Thus a second vicious circle is created, more extended in time. The causality between weak monetary management and the inadequacy of instruments and markets is reciprocal. Monetary management involving extensive interest rate and credit controls and subsidies is typically associated with underdeveloped and uncompetitive markets. Direct monetary controls hinder the emergence of new financial instruments, markets, and institutions. Lack of those leads monetary authorities to rely on direct controls. More generally, direct monetary control creates numerous difficulties with respect to the efficient mobilization of savings and the efficient allocation of financial resources, and hence with the efficiency of financial intermediation, thereby hindering financial deepening (Johnston and Per Brekk 1990; Duesenberry and McPherson 1991). These detrimental effects make a strong case for strengthening and improving the control of reserve and base money and for reforming the ways and means of monetary management. The reform is intended to move monetary management away from direct controls toward the greater use of indirect methods and instruments, increasing the scope of monetary policy for stabilization purposes and introducing more flexible arrangements for setting interest rates and allocating credit. These are all important elements in improving savings mobilization, resource allocation, and financial intermediation, which in turn can help attain adequate rates of economic growth. An early stage in the reform may involve the improvement of existing instruments of direct control. The WAMU zone is presently involved in a broad reflection on economic integration. Taking as a starting point the reality of the monetary union, the authorities are considering consolidating the advantages conferred by a single currency through an effective integration of the economies of member countries. This project would eventually lead to the transformation of the West African Monetary Union into the West African Economic and Monetary Union (WAEMU). In the field of economic management directly related to monetary policy implementation, the studies undertaken focus on multilateral monitoring of fiscal and budgetary policies and harmonization of taxes and tariffs. The establishment of the WAEMU will probably take many years. In the immediate future, this project offers a new means for the conduct of monetary policy. A truly regional 77 monetary policy, through the promotion of regional money markets and instruments and the implementation of region wide liquidity management by the BCEAO, would be a reachable goal for the WAMU in the short term. In fact, given the existence of a commnon currency, all that is needed in the field of monetary management is some adaptation of the WAMU arrangements. There is also potential for developing a regional stock and bond market. A greater use of indirect methods and instruments in monetary management is often viewed not only as desirable, but also as inevitable. Considering the specific features of sub-Saharan African economies and the study countries, what are the prerequisites to such reform? And what approaches should be considered and sequences and critical paths tried? These topics are discussed in the second part of this chapter. There are seven prerequisites to monetary management reform: * Growing macrofinancial stability and declining inflation * A viable and improving fiscal and budgetary situation * No pronounced financial distress in the financial system * A sufficient core of human resources and skills in central banking and financial operations and management at the central banks * An operational and research capacity at the central banks allowing for the initiation of monetary programming, including a timely information base * An efficient decision making process with respect to the formulation and implementation of monetary policy, supported by an adequate management information system at the central banks * A strong commitment from the monetary authorities and the govermient to the reform (ensuring the necessary coordination among the central bank, the ministry of finance, and other government agencies). Reforms and Options In the study countries as well as in most countries of sub-Saharan Africa, the transition to a system of monetary management that relies on greater use of indirect methods and instruments is particularly delicate because of the lack or weakness of short- term financial markets. However, reform cannot wait for these markets to develop or strengthen. Instead, reform must begin even if the markets remain underdeveloped, an approach that requires gradualism, close monitoring, and flexibility. It calls for a progressive reform, starting with the improvement of existing instruments and procedures. A second stage would include the establishment (or strengthening) of an interbank market, the introduction of an adequate discount policy, and the creation (or strengthening) of an auction for treasury bills. The third stage of the reform would promote money markets and secondary trading, thus establishing a strong base for monetary management relying on a greater use of indirect methods and instruments. 78 Improving Existing Instruments Credit ceilings. Ideally, the use of overall and individual credit ceilings should be phased out for the day-to-day management of liquidity and monetary aggregates. However, this can happen only when an alternative method of monetary management, relying on greater use of indirect methods and instruments, is operational and effective. In the meantime, the credit ceilings in use can be improved; individual credit ceilings can be made more flexible and market-responsive. A major limitation of credit ceilings is that they discourage savings mobilization and lead ultimately to financial disintermediation. Traditionally, individual credit ceilings are based on the share of financial institutions in total credit. This should be modified to induce financial institutions to make greater efforts at mobilizing savings and to stimulate competition among them. Individual credit shares should be partly linked to changes in deposit levels. Alternatively, loan-to-deposit ratios can take on increasing importance in the formulation of credit ceilings. For new banks, initial ceilings should be based on their business plan. Such reforms took place in Ghana and Madagascar and could be introduced in all the study countries. Individual credit ceilings can be linked to other variables, such as the level of gross or net foreign assets, the share of domestic credit or credit to the private sector, or performance in recovering past-due loans. Such approaches, which make the credit ceiling more responsive to performance, were used successfully in the 1970s by Ghana and in the WAMU zone. The basis for the establishment of credit ceilings by monetary authorities would vary from one country to another, according to specific circumstances. But the study countries should make every effort to see that the ceilings are flexible and less stifling of competition among financial institutions. As a transition measure, this approach should be considered in Kenya and Madagascar. In the WAMU zone, BCEAO is considering abolishing the ceiling in the wake of its recent reforms. Finally, penalties should be enforced when ceilings are breached (Guash and Gleassner 1992). Liquidity ratios. Liquidity ratios are essentially prudential regulations. But when they are used to finance the public sector, often at subsidized interest rates, as in Kenya and Madagascar, they create distortions in resource allocation and the cost of funds. To alleviate this problem, interest rates paid on bills and bonds captive under liquidity ratios should be raised to approximate a market-based interest rate. This approach was about to be introduced in Madagascar before the political turmoil broke out. In addition, when open-market-type operations are introduced, part or all of these bills and bonds should be made marketable, creditworthiness permitting. This provides a ready-made stock of instruments for the central bank's auction. Reserve requirements. These are a discretionary instrument that contributes to the management of reserve money in a flexible way. They are a powerful tool of monetary policy that, through reserve money and the multiplier, constrains money supply without the disadvantages of credit ceilings. But this instrument is more suitable for counteracting structural changes and trends in reserve money levels than for short-term liquidity management. They are also costly to financial institutions because they complicate liquidity management and require portfolio adjustments. That is why changes 79 are usually accompanied by a grace period-which reduces their usefulness when sudden intervention is required. Reserve requirements are not a good instrument for absorbing excess liquidity that stems from the inadequacy of existing instruments of monetary management. In such a situation, there is a danger of increasing ratios stepwise, with no way to absorb liquidity should the ratios need to be reduced. It is a sort of trap that leads to high spreads and disintermediation. Ghana was coming close to that at one point in its reform, but fortunately had by then an operational treasury bill auction system that allowed it to correct the reform course in time and escape the trap. The instrument's flexibility is also limited because it imposes an implicit tax on the financial institutions subject to it unless interest is paid on the required deposits. Reserve requirements that are unremunerated or remunerated at below-market rates raise the cost of financial intermediation and may distort competition to the extent that some categories of financial institutions escape the requirement. Ultimately, the implicit tax is borne by the customers of the financial institutions. Reserve requirements that are too high may therefore lead to avoidance practices, such as the creation of financial instruments that are not subject to these requirements or to the establishment of affiliated nonbank financial institutions that generate deposits indirectly, as in Kenya. Finally, there must be no overall excess liquidity in the financial system that would lead financial institutions to maintain excess reserves at the central bank. But during transition and reform, required reserves may have to be used more frequently than otherwise appropriate, until open-market-type operations are functioning effectively and the right discount policy is in place. Among the study countries, only the non-WAMU group uses reserve requirements, and there is room for some improvements in their implementation, particularly in their definition, the calculation of their level and basis, the definition of their scope, and their remuneration. As is the practice in most countries, the definition of reserve requirements should be broadened to include cash in the vault. Such an improvement would require that central banks be satisfied that the financial institutions are reporting such cash accurately and in a timely fashion and that these cash positions can be adequately monitored. Calculation of reserve requirements should specify a minimum holding to be met on average over a given period of time, leaving financial institutions with some day-to-day flexibility. Such a formula reduces the volatility of short-term rates, since institutions running a shortfall (relative to their required average holding) at times when liquidity is tight and rates high can add to the pressure on rates if they lack this flexibility. At the same time, the implementation of reserve requirements should be as "contemporaneous" as the reserve accounting and management information systems permit. To that end, two or three reporting periods per month can be determined by the central bank. The period for which liabilities at financial institutions are averaged to calculate the amount of reserves required should match the reserve period itself. Long gaps between reporting periods leave important lags, while too-frequent reporting is difficult to manage. A balance must be found, suitable to every country's conditions. All deposit-taking financial institutions should be covered by reserve requirements, and the ratio should be uniform. As long as some categories of institutions are not 80 included or different ratios are applied, deposits will be shifted from institutions included to institutions that are not included or that benefit from lower ratios, destabilizing the money multiplier and complicating monetary management, as in Kenya. As of October 1993, the BCEAO has instituted a system of reserve requirements on all banks and finance companies, calculated as a function of sight deposits and short term credit, excluding crop credit. Penalties are imposed for institutions failing to hold the reserve level. Since there are separated targets for domestic credit in each country and variations in reserve requirements by country are explicitly permitted, the system amy become inconsistent with a regional market. Refinance policy. Various reforms in the study countries have already simplified the refinance system. Special windows have been closed, preferential rates abolished, and rates unified. This is probably the policy area where the most progress has been made, though there is still room for further rationalization of discount policy. Refinancing is a powerful monetary instrument for managing reserve money. Although it can be used independently, it is more effective when linked to reserve requirements and an efficient interbank market. At present, non-WAMU central banks, with the exception of Ghana's, use a double approach to refinancing, with one window for collateralized refinancing and one for overdrafts or advances. In Madagascar, the overdraft facility is combined with the "money market" facility run by the central bank. When there is a penalty rate, it is more readily applied at the discount window than through the overdraft facility. As a result, financial institutions prefer to access the advance or overdraft facility. While earlier reforms abolished specifically designed refinance systems, a quasi-preferential system has replaced them in practice. Indeed, in all non-WAMU countries except Ghana, the liquidity needs of distressed financial institutions and of institutions that have extended credit to the public sector have been accommodated without much restraint under this advance facility. It is this area of monetary policy that brings out most clearly the passive side of the monetary stance of central banks in the study countries. Like the central banks in the non-WAMU countries (except Ghana's), the BCEAO followed until recently a two-track refinance policy, with a refinance window and a "money market" facility. Until October 1993, the BCEAO operated a "money market" which was in fact a mechanism to transfer liquidity from countries which had a surplus of it which had a shortage of it, at rates determined by the central bank. The new money market scheme introduced in October operates on a regional basis and involves weekly bids from a pre-announced duration from participating banks from all countries of the region. The BCEAO then matches supply and demand at the bids rate. The BCEAO guarantees money market transactions. Under this new system banks have also the option to transact in the interbank market without a guarantee. This market is still very limited, especially as concerns cross-border activity within the zone. Reform in this area of refinance policy should aim at establishing procedures that distinguish clearly between refinancing policy as a monetary instrument for injections of reserve money into the financial system and refinancing policy as a mechanism for temporary assistance to financial institutions running out of liquidity. Moreover, it should also aim at enhancing the active use of market-based monetary instruments and 81 facilitating money market development. This objective relates to both the availability and the cost of refinance credit. Central banks should consider-mainly in their capacity as lender of last resort-introducing or activating through the discount window a regular refinance facility to provide collateralized short-term credit for, say, up to three months. Access to this facility should not be automatic, and the amount available should be at the discretion of the central bank, following the policy principle of uncertainty of refinancing. The overall volume of refinance available every period (month or quarter) through this facility should be based on the monetary policy stance of the day and on a monetary programming exercise. The basic discount rate should be related to the interbank market rate, the auction rate, or a rate approximating them if these markets do not exist or if they function poorly. In addition, a graduated system should ensure that the discount rate increases with the volume and frequency of access to the facility. Repeated access to the facility by an institution should trigger increased scrutiny or even inspection of the institu- tion. Liquidity support to financially distressed institutions or to institutions needing to re-finance claims on the public sector has frequently been a source of uncontrolled money creation in the study countries. Kenya, Madagascar, Uganda, and the WAMU countries are the most glaring examples. It is essential that the central banks establish procedures to regain control of the supply of reserve money in such circumstances, starting by abolishing any open-ended facilities. In some of the study countries-Kenya or Madagascar, for example-this should be done gradually, to avoid causing the collapse of part of the financial system. In fact, in 1989, when BCEAO had to curtail refinancing aimed at injecting liquidity into distressed financial systems, Benin's system collapsed and several financial institutions in other member countries had to be closed down. As a transitional arrangement, liquidity should be allocated to individual financial institutions on the basis of specific criteria linked, when warranted, to progress in financial restructuring. Also, to avoid violent swings, the maximum amount that can be accessed al any period should be limited and preferably announced in advance. Because of the nature of this temporary overdraft or advance facility, the interest rate should not be punitive. But a time limit should be set after which financial institutions having continuous recourse to the facility should be closed, merged with a healthy financial institution, or taken over by a public agency established for the purpose of financial restructuring. This time limit time-together with a financial restructuring plan-should be established as soon as it becomes clear to the central bank that the liquidity injection required is not going to be temporary. In addition, an institutional mechanism should be established to handle distressed financial institutions. New Markets and Instruments Interbank market. Except in Ghana, interbank markets are absent or function poorly in the non-WAMU countries, and there is no formal national or regional interbank market in the WAMU zone. A well-functioning interbank market is one of the two key markets that facilitate the move toward indirect monetary management. There are three principal factors that prevent interbank markets from emerging or functioning well in the 82 non-WAMU study countries: lack of trust among financial institutions in a context of financial distress or restructuring, administered interest rates, and lack of market-makers. The Bank of Ghana faced these three problems and addressed them in reverse order. It promoted a market-maker and, in the context of a gradual liberalization of rates, helped the market-maker launch the interbank market on a collateralized basis. A discount house was created, followed by a second one to introduce competition; a third is to start functioning soon. This sequencing worked well and seems to be a good means of starting an interbank market in a time of financial restructuring, when trust among institutions is lacking. A second-best solution would be for the central bank to establish and run the interbank market on a collateralized basis until it is well established, providing at the outset the needed infrastructure and confidence. The central bank would act as a broker, bringing the parties together. To deal with the problem of confidence, a lending financial institution could request the borrowing bank to deposit collateral with the central bank, but the central bank should not under any circumstances act as a guarantor for interbank transactions. This is the option apparently chosen by BCEAO under the new money market formula. Alternatively, it is common practice in many countries for financial institutions to establish credit lines in relation to other participants in the interbank market, based on such elements as their ratings and the intensity of their relationship. Open-market-type operations. There are many ways to organize and conduct a treasury bill (T-bill) auction. Moreover, arrangements and procedures keep evolving with experience, circumstances, monetary management stance, and changes in instruments and markets. At every stage, improvements are possible. A question all countries address at the outset is whether the T-bill should be issued by the ministry of finance (or the treasury) or the central bank. Only Ghana chose the central bank. Technically, the provenance of the instrument makes no difference, as long as it is auctioned by only one institution. In practice, there are advantages and drawbacks to both options. Interest on a T-bill issued by the ministry of finance is paid by that institution and there may already be, as in Kenya, a stock of T-bills in the markets (built under liquidity ratios), with financial institutions having grown used to the practice. Such a stock facilitates the initiation of open-market operations. The drawback is that when T- bills are issued by the ministry of finance, the size, timing, and yields of the issue are usually decided on the basis of fiscal and budgetary requirements rather than monetary considerations, which really ought to be the central element of such decisions. Auctions should be integrated into the monetary management framework as an open-market-type operation and should become the fulcrum of monetary management by the central banks. Equally important, these decisions should be based on the monetary programming exercise. The timing of the auctions should be harmonized with the dates and periods governing reserve requirements. The question of integrating the T-bill auction into the monetary management framework is one that the Bank of Uganda is now addressing. Should the BCEAO decide to establish a T-bill auction, it would have to issue its own bills since most of the treasuries of the WAMU zone have lost their creditworthiness. In such a case, fiscal coordination becomes very important. In all cases, 83 close coordination between the central bank and the ministry of finance or the treasury is necessary in the preparation and the operating of auctions. At the beginning, there should be only one T-bill maturity and it should be short. Experience in several developing countries shows that secondary market trading is better fostered by the creation of a deep market in one maturity than by the development of several shallow markets. The larger the stock, the easier secondary trading becomes. In Ghana, the ministry of finance issues its own instruments independent of the central bank. The ministry's sales on tap are intended for other financial institutions and the general public whereas the central bank's auction involves discount houses and banks. But the Bank of Ghana rediscounts all instruments so issued, its own and those of the ministry of finance at a unique rate, irrespective of their nominal yields, ensuring consistency of the interest rate structure. Monetary Programming and Decision Making Establishing or strengthening monetary programming is particularly important in the study countries. Programming gathers and synthesizes available data and information, translates them into monetary policy implementation scenarios, and feeds them into the decision making process. The decision making process needs to be flexible, rapid, and bold, given the sensitivity to external and internal shocks of the monetary and financial systems in the study countries and in most sub-Saharan economies. The effectiveness of the monetary programming and decision making process is especially important because it has to compensate for the information gaps that exist in the study countries. These countries should place high priority on strengthening and developing monetary programming and decision making. Clearing and Payment Systems In all the study countries, clearing and payment systems can and should be improved. Weaknesses in these systems contribute to macrofinancial instability and are one of the main causes of the narrowness of the reserve money core. Efforts should be made to improve and expand the clearing system, speed up clearing in more remote areas, promote computerization and the use of checks and electronic transfers, and strengthen the accounting framework for the clearing process. Moreover, efficient clearing and payment systems are a prerequisite to the development of interbank and money markets and T-bill auctions. Interest Rates A flexible interest rate environment is a prerequisite for the development of an effective monetary policy and a monetary management framework that relies on indirect methods and instruments and operates through market-based instruments. But, as in monetary management reform, the transition from administered to flexible rates is delicate because of the weakness or absence of short-term markets. 84 Only one of the study countries-Ghana-has a fully flexible market-based interest rate structure. Kenya and Madagascar have fully liberalized rates, but their central banks continue to guide the structure. In Madagascar, the interest rate structure is anchored to the basic discount rate that is determined by the central bank. In Kenya, the central bank influences tender rates at the auction through moral suasion. In Uganda, the central bank sets a minimum bank deposit rate and a maximum bank lending rate. As of October 1993, the BCEAO had liberalized the regional interest rate structure. It abolished the fixed interest rates margin that banks had to observe in the region. the structure is supported by 4.5 percent minimum interest rate on savings accounts. The only other administrative constraints on interest rates are the floor on short terms deposits which is set 200 basis points below the money market rate and a "usury cap" on lending rates which is set at twice the rediscount rate. Liberalization has been a gradual process for all the study countries. Given the weakness or absence of short-term markets, gradual liberalization was the only viable policy alternative and, for sub-Saharan Africa, it worked well. Three main lessons can be drawn from the way interest rates were liberalized in the study countries, two of them illustrated by Ghana's reform experience. The first is that inflation has to be on the decline for the central bank to keep interest rates positive in real terms. The second is that it is more effective to create short-term markets-at least an interbank market and an auction-before fully liberalizing rates. There can be no market-based rates without markets. A third lesson is that it becomes increasingly difficult to maintain a fully administered rate structure when the financial system is private-sector-driven and fairly diversified. On their way to liberalizing interest rates, all the non-WAMU countries struggled-or are still struggling-to keep rates positive in real terms. When inflation is high, it is difficult for the central bank to keep rates positive all the time. It faces two problems. One is to estimate the duration of the inflation burst. In Uganda, for instance, some periods of inflation have been short, and there was no need to adjust rates to keep them positive in real terms at the inflation peak, but only to keep them positive in real terms on average. The other problem is that high real lending rates are detrimental to economic and financial activity. So the central bank faces a dilemma-a dilemma made worse if part of the financial system is in distress (there is a danger of increasing distress) or is being restructured. In Ghana, when the central bank fully liberalized rates, they became negative in real terms. A major reason for that development was that banks were reluctant to pass rate increases on to the few viable borrowers they had kept after cleaning up their portfolios. Banks were fearful of jeopardizing their borrowers' viability and thereby their own. The best way out of this dilemma, clearly, is to bring inflation down as quickly as possible, which comes back to the issue of control of reserve money, base money, and money supply and the fiscal budgetary situation. Establishing an interbank market and at least an auction system, if not open- market operations and the core of a market in short-term securities, is a powerful means to facilitate the liberalization of interest rates. That is the reform path Ghana followed. The reasons are obvious. It provides a market-based, short-term anchor rate. It provides the central bank with a framework for managing rates indirectly anid at the margin. It 85 provides mechanisms that allow financial institutions to arbitrage rates and maturities freely and according to market signals. And finally, it preempts the need for the authorities to interfere with market mechanisms in order to set the rates. Interest rate developments in all the WAMU countries illustrate these points. It is increasingly difficult to maintain an administrated rate structure when the financial sector is fairly diversified and private-sector-driven. During the period that rates were administered in Kenya, the effective rate structure differed increasingly from the administered framework mainly because of the existence of fees and charges and of a wholesale market for funds. Effective interest rates were notably higher than the administered ones on about 50 percent of the banks' funding liabilities and assets. Since yields on government securities were set according to the interest rate structure, they were not competitive. To overcome this distortion and fund at least part of the public sector deficit in a noninflationary way, the authorities took recourse to liquidity ratios and moral suasion. This brought in further distortions and accentuated the disintermediation of the part of the financial system affected by the authorities' actions. These developments indicate that when interest rates are liberalized, fees and commissions should be liberalized along with them. With the new interbank and money markets in the WAMU zone, two prominent, market-related rates are the money market rate and the interbank rate. Under the new system, the banks in need of liquidity would borrow on the money market, then on the interbank market before approaching BCEAO's rediscount facility. This systems allows the central bank to apply the principle of uncertainty of refinancing. Strengthening Central Banks Central banks in the study countries, as in all African economies, have been widely exposed to external and internal shocks. In the process of absorbing the impact of these shocks and their fiscal and monetary consequences, they have accumulated huge claims on governments yielding little or no return. At the same time, they have had to shoulder a number of quasi-fiscal and developmental tasks that distracted them from their main function of money management. All this weakened them financially. In addition, the pervasive pressures of an inimical economic and financial environment left them organically weak and undermined their quest for greater autonomy. All financial reforms in the non-WAMU study countries include a component to strengthen the central bank, financially and institutionally, beginning with a redefinition of the institution's role to focus it primarily on monetary management. Reforms are already under way in Ghana and Kenya and are being pursued in Madagascar and Uganda. The new acts recognize that the central bank must play a decisive role in controlling the money supply in order to promote and maintain the stability of the currency; stimulating the mobilization and efficient allocation of financial resources; and regulating, supervising, and directing banks and other financial institutions in order to protect the stability of the financial system without affecting the banks' autonomy and efficiency. Under the new acts, the explicit or implicit quasi-fiscal and developmental 86 roles are being terminated. The new acts also explicitly recognize the autonomy of the central banks. However, the legal status of the central bank is only one of the several elements that determine its autonomy, including less-structured factors such as arrangements and coordination between the central bank and ministries and other agencies of the govern- ment and the quality of the central bank's research department (Cukierman, Webb, and Nyeapti 1992). To that end, institutional mechanisms are needed that allow the central bank to state explicitly and officially its monetary policy position-in particular, when it differs from that of the ministry of finance or other government agencies. If there is a substantive disagreement between the central bank and the ministry of finance or another government agency, arbitration may eventually be required and the central bank may have to yield. However, experience in several developing countries shows that when there is an institutional mechanism that allows the central bank to state its position, that position is considered more seriously by the authorities. Malaysia permits disagreement that persists after a first attempt at compromise to be aired in the Parliament. Other developing countries have different procedures for reconciling divergent positions, but most provide the central bank with a forum to state and explain its position. Strengthening the central bank financially requires relieving the bank of claims on the government and distressed financial institutions that it has accumulated and that burden its balance sheet. Often, the assets representing these claims are the revaluation account, credit extended to financial institutions, and the central bank's losses. Ghana shows one way of shedding this burden. The central bank's revaluation account was replaced by a bond on the ministry of finance, with the understanding that all additions to the revaluation account would be settled in cash by the treasury. The bond carried a below-market interest rate that would provide an income flow sufficient to ensure the financial autonomy of the Bank of Ghana. Bad assets that the Bank of Ghana had taken over from financial institutions being restructured were treated in two ways. Assets representing nonperforming loans on public enterprises were taken over by the ministry of finance; those representing nonperforming loans on private enterprises were transferred to the Non-Performing Assets Recovery Trust. The government of Ghana is now preparing the recapitalization of the central bank. While there are a number of other ways to financially strengthen the central bank, there is no escape from three facts. First, the government must ultimately take over the losses related to the central bank's quasi-fiscal operations, including foreign exchange and foreign-debt-servicing operations. Second, a formula has to be designed to relieve the central bank of the losses it had to take over from ailing banks. In Chile, a long-term agreement was made with the banks to repurchase the bad assets according to an established schedule; in Spain, a special fund was constituted to refinance these assets. In several countries, the ministry of finance funded the losses. The formula chosen depends on the country's circumstances and political preferences. Third, the central bank will have to be recapitalized sooner or later. The case of the BCEAO is a complex one. It carries on its balance sheet a considerable amount of assets representing the bad loans of financially distressed institutions that it refinanced in order to keep the institutions afloat. Later, that short-term 87 refinancing was consolidated into medium-term assets paying an interest rate well below the reference "market" rate. The total of these assets is equal to at least the net worth of the regional central bank. Some distressed institutions assisted by BCEAO have been closed in the wake of the restructuring of the financial systems in member countries. Most of these institutions were government-owned. Logically, most of these assets are claims on governments. But the BCEAO cannot record them as such because it is constrained by the 20 percent limit on credit to governments. This problem, together with BCEAO's weak capitalization, will have to be addressed. 88 8. DEEPENING FINANCIAL SYSTEMS There are many reasons for diversifying and deepening the financial systems in sub-Saharan Africa through the development of financial markets. The first is that it increases the resilience of economies and financial systems to external and internal shocks. The second is that it is necessary to meet the need for investment and related long-term finance. In the next decade, the net flow of foreign capital to most sub-Saharan African countries is likely to be less than in the past, which has important implications for financial sector development. Countries will be forced to rely primarily on domestic savings to finance investment. Because of heavy reliance on foreign financing in the past, external debt exceeds domestic debt in virtually all sub-Saharan African countries. Moreover, practically all long-term credit was provided through foreign loans. The decline in funding from abroad will make living with a shallow domestic financial market and little long-term finance difficult and will constrain investment and economic growth (World Bank 1989). Because of the unsettled macrofinancial conditions and the shallowness and narrowness of financial systems in most sub-Saharan African countries, investors tend to rely on short-term credit to fund investment projects. Banks offer lines of credit and short-term loans that they roll over regularly as long as the borrower is in good standing. This type of finance is used extensively in place of long-term finance, for which it is only an imperfect substitute because of the risk of nonrenewal and, from the viewpoint of the enterprise, because higher leveraging becomes riskier the more earnings fluctuate. The enterprises that can best afford to be highly leveraged are large capital-intensive concerns with highly predictable earnings. Young enterprises and other firms with less predictable income streams need to rely more on equity finance and retained earnings. Finally, if a majority of enterprises in a country are excessively leveraged, the potential for adding to macrofinancial instability rises. This is particularly likely in the study countries and other sub-Saharan African countries that, owing to their narrow resource base, show a highly concentrated risk pattern in terms of financial assets and their management. There is also potential for creating financial distress as the risk pattern reflects on the portfolios of banks and other financial institutions. A different but equally important reason for deepening and diversifying financial markets is the enhancement of competition within the financial system. To achieve its general objective, which is to contribute to faster, sustainable, and socially equitable economic growth, the financial system must be efficient and solvent. Efficiency in financial resource mobilization and allocation is best promoted by competition, with as few imperfections as possible. This efficiency depends on heeding market signals, and for these signals to be valid, the competitive system must function properly. 89 The characteristics of a perfectly competitive market are well known from economic theory (Legarda 1986): * A large number of buyers and sellers * Freedom of entry and exit * Perfect knowledge (transparency). Dilemmas arise when actual conditions fall so far short of these conditions as to throw doubt on the competitiveness of the market. They also arise when competition cannot be promoted to the same degree in all segments of the financial markets. In the bank-centered financial systems of the study countries and other sub- Saharan African economies the commercial or deposit-taking banks, which make up most of the financial system, have exhibited strong oligopolistic features in the interest of institutional stability and solvency. Because of the vulnerability of the economies to external and internal shocks, capital requirements should be stiff and high. To promote strength, monetary authorities may have to encourage mergers, resulting in larger and fewer banks. There is thus a contradiction in promoting bank-centered financial systems. Operational considerations tilt policy toward fewer and larger units, which diminishes real competition and therefore eases the pressures for efficiency. What is likely to emerge is a system with marked oligopolistic features. Large institutions become powerful vested interests and typically resist pressures for increased competition. There is another problem in achieving long-term competitive conditions in a bank-centered financial system. The process of competition is not necessarily a stable equilibrium. After a while, the strong institutions eliminate or absorb the weak ones, which results in market concentration and oligopoly. There is therefore need to look for other sources of competitive pressure. That is where money and capital markets can play a crucial role. The lower capital requirements and the greater importance of entrepreneurship relative to capital in the nonbank financial institutions mean that two characteristics of the competitive market are potentially present: a greater number of participants and greater ease of entry. This should lead to a higher degree of competition and therefore efficiency. Moreover, various recent studies have cast doubt on the assertion that larger unit size and fewer numbers lead to economies of scale or scope (Federal Reserve Bank of Kansas 1988). Requirements for Successful Reform Commitment to reforms. Experience has demonstrated that no reform aimed at developing financial markets can succeed when political commitment is vague. There has to be an appreciation of the implications of the market-oriented philosophy behind the functioning of the reformed system. Political commitment is important because financial reforms involve difficult decisions regarding cost of credit, the inefficiency of some financial institutions, and the removal of distortions favoring well-entrenched financial constituencies. Closure or rehabilitation of troubled financial institutions also results in employment cutbacks. 90 A successful financial reform must lead to healthy competition between the public and private sectors for scarce financial resources. The potential for budget deficits will likely rise because of the liberalization of the financial sector, which entails extra cost in the form of higher interest rate payments on domestic public debt or possibly a reduced share of domestic credit to the public sector. Nevertheless, governments should abolish fiscal biases favoring the public sector, such as the tax exemption of government bonds in Kenya and Togo. An enabling environment. Removing obstacles to the deepening and diversification of financial markets creates an enabling environment. The reform of the environment can be organized around three major objectives: * Creating and maintaining a stable macrofinancial environment, based on stable macroeconomic policies, low inflation, and flexible interest rates * Establishing an incentive framework and a business climate supportive of entrepreneurship and private sector development * Fostering an all-out national training effort in managerial and technical skills related to financial operations and the use of proper accounting procedures, and adequate auditing and financial information dissemination. The logistical framework. The capacity for carrying out effective financial reform also rests on the underlying regulatory framework and the competence of those entrusted with its implementation. The central banks, in order to efficiently discharge their supervisory and regulatory functions, need reinforcement of their information systems, improved and articulated regulations, and enhanced supervision of credit. At the level of banks and nonbank financial institutions, upgraded procedures in relation to credit policies, loan reviews, and management systems must accompany the changes undertaken by the central bank. No matter how well designed a reform of financial markets might be, it stands a considerable chance of getting bogged down if the implementation is not guided by bold and efficient decision making backed up by sufficient political pressure and technical sup- port. There is a need for highly trained, capable financial managers who can consistently follow up and evaluate the reforms. Most African countries are in dire need of financial experts, analysts, and economists who can be entrusted with the task of building local data banks and other information systems vital for studying the impact of financial reforms. Technical assistance is often necessary. Toward a Financial Sector Strategy The deepening and diversification of financial markets calls for a proactive approach by the authorities. The pace of this development is ultimately determined by each country's institutional capacity. The strategy to follow is one aimed at establishing market locuses, such as brokers, dealers, and foreign exchange bureaus, and promoting a 91 consistent sequence of institution building through an adaptable and evolving legal, regulatory, and prudential framework. Planned or regulatory approach? Strategies for financial market development have been either evolutionary or based on a proactive policy. The evolutionary approach consists in letting financial markets develop as the economy develops, intervening through changes in laws and regulations only when major distortions or bottlenecks emerge or when the reaching of a further development stage warrants it. The pattern and pace of development of financial markets are determined by the pace of overall economic development and existing incentives. In situations where the pace and pattern of economic development are mainly market-driven, this approach may produce satisfactory results to the extent that changes in laws and regulations are timely and do not counteract market forces. The deepening and development of the financial system is then essentially market- driven within an adaptable legal, regulatory, and prudential framework. This approach requires a neutral incentive system with regard to markets and instruments. The proactive approach consists of providing a legal, regulatory, and prudential framework that fosters and, when possible, accelerates financial market development. This framework supports the setting up of mechanisms, institutions, and instruments that promote and facilitate this development as the economy grows and market functions expand. This approach may involve, temporarily, a less-than-neutral incentive system with regard to markets and instruments to support financial market development. For sub-Saharan Africa in general and the study countries in particular, the second approach is more appropriate. Experience has shown that the evolutionary approach is not the best way to achieve strategic objectives. In these countries, regulatory bodies are run by officials and civil servants who know much about regulation but little about how a financial system works. Clearly, in Cote d'Ivoire, Kenya, Nigeria, and Zimbabwe, attempts to pursue strategies of financial market development and to implement related policies through legal and regulatory bodies have failed. Too many detailed regulations at the outset have impeded the development of money and securities markets and rendered stock exchanges inefficient and costly. Three arguments favor the adoption of a proactive approach. First, market forces are not strong enough for financial markets to develop by themselves. Years of financial repression have prevented market innovation, limited financial diversification, and introduced biases against the private sector. A neutral incentive environment is not enough; active promotion and positive incentives are required. Second, a determining factor of the pace and strength of financial market development is the country's institution-building capacity. The private sector and authorities have to join hands to establish or foster key market institutions and skills that determine and condition financial market development. Witness the way in which discount houses in Ghana have been set up to become the locus for money market development. Third, a proactive approach encourages the use of the most efficient institutional set-up and the adoption of a technology appropriate to local conditions and level of development. It also allows for organizing locally the required training infrastructure. 92 Once the concept of a proactive approach has been adopted, the first thing a government needs to do is establish a working committee on financial market development, which would guide the efforts toward financial sector diversification and development and would focus on issues that emerge as markets develop. Private sector representatives should predominate on this committee. Efforts should focus on specific measures to remove the obstacles to financial market development and to accelerate the pace. Attention should be steered away from restrictive laws and regulations. Though suggestions could be made concerning an appropriate regulatory regime, clear strategic decisions need to be made first. The committee should be given a broad mandate and, if possible, the power to implement the proposed changes. A good example of the positive impact that such a committee can have is the Capital Market Authority (CMA) of Kenya, established in 1989 to develop the securities market and to supervise the regulation of the country's capital market. Despite some weaknesses, the CMA has made considerable progress in its four years. It proposed and implemented the incorporation of the Nairobi Stock Exchange, reform of the tax system, and establishment of a formal trading floor. Diversification of instruments. In narrow financial systems, the principal financial instrument used by economic agents is bank-intermediated loans. The diversification of financial instruments implies the emergence of other short- and long-term financial instruments, such as bankers' acceptances, CDs, bills, bonds, other securities, and, notably, equities. In most sub-Saharan African countries, the short-term end of the market offers more immediate prospects for development. In particular, crop financing in the form of self-liquidating bills must be developed. Commercial bills and bankers' acceptances based on cooperative or "mutualistic" guarantees should be developed to establish a link between semiformal and formal financial institutions. Development of such short-term financial instruments, responding to real domestic needs in the sub- Saharan region, would go a long way in developing money markets. The first step in extending the maturity of financial instruments is often the extension of the maturity of government debt instruments. Since longer maturities are associated with higher risk, especially in macrofinancial environments prone to instability and inflation, this risk is lessened when government debt instruments are involved. In addition, the existence of a mechanism to ensure the liquidity of these longer-term instruments on conditions of positive returns is a strong supportive factor in the development of demand for these instruments. Leasing and housing finance have stimulated the development of longer-term debt instruments, owing to the peculiar maturity structure of their financial operations. Universal or specialized banking? Equity finance can be provided by securities markets or by a universal banking system, a system that combines commercial banking, involving collecting deposits and extending loans, and investment banking, involving issuing, underwriting, placing, and trading securities. In industrial countries such as France and Germany, the success of universal banking appears to reflect not only some economies of scope and scale enjoyed by large and diversified financial institutions but also the 93 importance of universal banks in monitoring corporate performance and controlling the behavior of corporate managers. In principle, the development of financial systems could follow two different paths of evolution-one toward universal banking, the other toward a securities-market-oriented financial system. However, a closer look at the situation in the study countries raises further questions. The commercial banking system strongly dominates the financial system-and a few commercial banks dominate the sector. Competition is weak, so reform should aim to increase competition both within and from outside the commercial banking system. Implementation of universal banking would reinforce the oligopolistic position of a few commercial banks. Even if offered the opportunity, commercial banks in sub-Saharan Africa are ill suited to providing higher-risk financing since they attract primarily short-term deposits and are limited in their ability to transform short-term resources into long-term assets. In a world of imperfect information, those banks are conservative. Moreover, they are often unwilling to promote new services that are not part of their core business. Witness the functioning of the Abidjan Stock Exchange. In 1974, it was decided that only commercial banks would be allowed to act as brokers. Nineteen years later, it is clear that commer- cial banks have done very little to promote the emerging securities market. As a general rule, countries with universal banking normally exhibit a higher debt-equity ratio. Universal banking calls for a strong tradition of self-regulation and self- supervision, because potential conflicts of interest are more numerous than in a specialized financial system. It also calls for a strong and sophisticated legal, regulatory, and prudential framework and extensive and effective supervision. Until prudential regulation and supervision reach those levels of technique and efficiency, the authorities should exercise great prudence in encouraging any moves toward universal banking. More generally, financial market development requires fostering financial institutions that deal in short-and longer-term assets and liabilities, as well as equity financing-such as discount houses, mortgage finance houses, savings institutions, and leasing and venture capital companies. Promoting Securities and Capital Markets In sub-Saharan Africa, development of financial markets should begin at the short- term end. The pace of development depends on institutional capacity. Since there is little possibility of accelerating financial market development beyond institutional capacity, a strategy of securities and capital market development in Africa should be two-pronged. First, it should aim at developing money markets and associated market functions and skills. It should then foster the transfer of some of these functions and skills to the budding securities and capital markets. Second, it should gradually promote the establish- ment of institutions needed for the expansion of securities and capital markets. Functions that are essential to the operations of money and other short-term markets can be transferred without too much modification to securities and capital markets. These involve, among other things, financial operations in short-term instruments, management of portfolios based on short-term instruments and associated 94 treasury functions, term transformation, matching supply and demand for various financial instruments, and market-making. The financial management and trading functions can gradually be extended toward private placement, investment funds, and unit trusts and greater dealing in medium- and long-term instruments. An institution-building strategy aimed at developing securities and capital markets should have a focus and follow a sequence. Ghana's financial reforms show the effectiveness of establishing market locuses that crystallize functions and attract a critical mass of human resources and skills. In a generic sense, this locus takes the form mainly of trading and market-making; in a specific sense, it takes the form of a brokerage or dealership function and, as the market develops, of discount houses and merchant banks. When the foreign exchange market is liberalized, this locus also takes the form of foreign exchange bureaus. Ghana and Uganda are good illustrations of how this process works. Institution building should therefore rest initially on establishing broker and dealer institutions and guiding them toward increasingly complex operations through a proactive and evolving legal, regulatory, and supervisory framework. The first stage of this process should be the development of private securities placement-that is, the sale of shares to pre-identified purchasers, such as institutional investors. A legal, regulatory, and prudential framework for private securities placement should be established concurrently with the formulation of a legal, regulatory, and prudential framework for brokerages and dealerships. Most countries have used private placement for privatizing public enterprises-for instance, Ghana, C6te d'Ivoire, Kenya, Senegal, and Uganda-but none of these countries first established an overall framework. Doing so would provide a good basis for investment and facilitate further privatization. A second stage is the promotion of investment companies, privately held companies (usually limited liability) that invest in a broad range of securities. The investment company could be a passive investor or could take an active role in the management of the companies in which it invests. A legal, regulatory, and prudential framework should be formulated to govern and promote investment companies. The advantages and disadvantages of an investment company are broadly similar to those of private placement, except that an investment company represents a more advanced legal formalization of securities operation and trading. Clearly, both institutions should relate to the functions of dealership and brokerage, since it is around these functions that they are most likely to crystallize and grow. The third stage in institution building is the promotion of unit trusts. Neither private placement nor investment companies contribute directly to wider share ownership, nor are they geared to liquidating holdings rapidly when the need arises. A unit trust is an investment mechanism that provides many investors with a means of participating in a diversified portfolio of investments. As for the other institutions, a legal, regulatory, and prudential framework should be developed to govern and supervise unit trusts. In fact, these three mechanisms could benefit from a single, unified legal, regulatory, and prudential framework (Popiel, 1990). In terms of sequence, private placement, investment companies, and unit trusts do not need a stock exchange to function and operate, although the existence of a stock exchange facilitates the functioning of unit trusts in several ways, including the valuation 95 of shares that the trust trades. However, Botswana established a successful unit trust-its shares valued in an ingenious and efficient way-before the establishment of a stock exchange. The next step is the promotion of flexible over-the-counter markets among existing brokers and dealers, again with the proper legal, regulatory, and prudential framework. Some self-supervisory capacity should be delegated to the association of brokers and dealers, to prepare them for more formal self-supervisory mechanisms and institutions. Merchant banks may or may not be associated, depending upon each country's circumstances. Over-the-counter trading will eventually lead to more formal trading mechanisms, which will eventually be housed in a stock exchange. At the same time, more sophisticated institutions may be promoted, such as closed-end investment trusts. As the stock exchange emerges with its supervisory and governing bodies, attention will have to be paid to keeping the clearing and settlement mechanism at the forefront of new developments. The evolving legal, regulatory, and prudential framework should focus on the attainment of one more fundamental objective: ensuring optimal transparency in an "in- formed" market. Achieving this objective involves timely disclosure of both corporate and market information. Little progress has been made in terms of establishing acceptable accounting standards and norms for external auditing. Accounting and auditing require- ments, financial statements, and forms of presentation still need to be upgraded. Professional associations should be encouraged and supported in their efforts to correct existing inadequacies. Transparency is a fundamental prerequisite for the attainment of an informed securities market. Increasing the supply of securities. A basic problem facing all sub-Saharan African countries is the paucity of securities to trade-only thirteen are listed on the board in Accra's stock exchange and fifty-five in Nairobi. Several measures, in addition to the liberalization of interest rates, can be introduced to increase the supply of securities. First, regulations limiting the type of securities that can be issued should be made flexible, to offer companies a broad choice, such as nonvoting preferred shares that can match the preferences of local entrepreneurs. Second, temporary tax incentives could be designed for the public issuance of securities, such as lower tax rates, accelerated tax depreciation, and investment credit for companies that issue securities to the public. None of these tax incentives is presently available in the study countries. Also, the expenses of public offering should be tax- deductible. Levels of supply are also determined in a fundamental way by issue price considerations. Many companies are reluctant to issue shares to the public at a below-the-market price, so companies should be allowed to issue new equities at market price. The supply of securities can also be expanded significantly through privatizations or divestment of the numerous public sector minority interests in private or mixed ownership companies, some of which may even be traded on the secondary market (as, for instance, in C6te d'Ivoire). Given the intensive level of active public sector involvement in the economy in the study countries, privatization programs should fuel 96 stock markets for some time to come (there are more than twenty public enterprises in Niger for instance; in C6te d'Ivoire, eighteen companies should be privatized in the near future; and in Kenya, one 139 companies are expected to be privatized in the long run). However, careful planning and selection are needed since not all state enterprises are suitable for broad ownership. Privatization of public enterprises should take place within a legal, regulatory, and prudential framework that governs all operations and trading of that type, whether related to securities of privately owned or state-owned companies. Increasing the demand for securities. Even if it is not a problem usually faced by emerging securities markets, some African countries may eventually suffer from limited demand for securities (such as in C6te d'Ivoire), partly because of low levels of financial saving. In addition to ensuring a transparent and informed market, which will strengthen savers' confidence, other actions can be taken. Demand for securities may be expected to react favorably to elimination of tax biases against equity and even bonds. Often, dividends are subject to taxes (sometimes to triple taxation), but interest on bank deposits is not. Capital gains may be subject to taxes, discouraging savers from investing. However, the reduction of tax biases and implementation of tax incentives will be successful only if the tax system is respected, since no incentive can compete with tax evasion. Another way to increase the demand for securities is to ensure that the investment regulations of contractual savings institutions permit this type of investment. To develop, securities markets need these institutional investors. Actually, they are often more important than individual investors in the initial stages of market development (in Botswana, for instance, they represent two-thirds of the market business). For small savers, adapted products could be developed. Medium-term bonds are one option that matches the short-term horizon of most small investors. Unit trusts and other types of mutual funds can also be established to offer small investors diversification and some professional management. It must be kept in mind, however, that sophisticated products can quickly become too complicated for small savers. Finally, one of the basic conditions for increased demand for securities is the existence of a strong secondary market or of mechanisms that ensure the liquidity of existing securities. Savers need to know that they will be able to sell their assets when they want (high marketability) and at a highly predictable price (liquidity proper). For this reason, it is necessary to establish trading mechanisms when there is no formal secondary market or to revitalize them where they exist. Improving securities intermediaries. At a later stage of financial market development, it may become necessary to gradually upgrade the status of dealers and brokers into full securities intermediaries, possibly by providing opportunities to graduate from brokerage or dealership into formal discount houses or merchant banks, investment banks, or securities firms. Africa has no underwriters in the primary market to encourage companies to raise finance through public offerings of securities. In some countries, regulations limiting selling prices and underwriting commissions have prevented the development of new-issue institutions. In others, it is difficult to obtain an investment 97 bank license or to operate under regulations designed for commercial banks. Brokers are often underfinanced and inadequately trained individuals, rather than properly capitalized securities firms. In C6te d'Ivoire, brokers are commercial banks not keen on developing a securities market. In none of the studied markets do brokers make available for potential investors published research on particular companies or on specific issues. Because of these inadequacies, steps have to be taken to encourage the development of securities firms. Impediments to the formation and operation of securities companies should be removed. An appropriate legal, regulatory, and prudential framework for merchant, investment, and securities banking should be established. Brokers should be allowed and encouraged to incorporate, bringing more capital into the securities business. Despite the fear of foreign domination of the market, foreign participation should be encouraged. Foreign competition, even if the number of foreign brokerage firms is limited, could add vigor and legitimacy to the market, as well as introduce new market skills. Stock exchanges. When trading volumes warrant establishing a formal stock exchange, both public ownership and operation (as in C6te d'Ivoire) and private ownership and operation by stock exchange members (as in Ghana and Kenya) are possible. Private ownership is probably better suited to the development of a strong secondary market, since private members are more willing to develop and promote the securities market than civil servants. Witness the poor results of the publicly owned Abidjan Stock Exchange. However, given the low level of financial infrastructure in sub-Saharan Africa and the dearth of securities firms that could become shareholders of the new exchange in most of the study countries, a public exchange could be considered as a temporary solution while an efficient network of financial intermediaries develops. In Botswana, which is often cited for its successful small securities market, the only stock brokerage house is owned largely by the public development finance institutions. Membership should be limited to individuals and companies that have a commitment to the growth of the securities business. In addition, members should engage in a modicum of public business during the year. To retain membership, a certain minimum level of business should be required. A situation such as that in Cote d'Ivoire, where one formal broker has engaged in no business for years, is not acceptable. Moreover, it is probably desirable to block direct membership for commercial banks, though they could be permitted to have affiliates in the securities business, which would avoid conflicts of interest and promote the development of the market. Finally, close attention should be paid to supervision, since secondary market development is closely related to this issue. The dynamics of a developing market are best served by an institutional approach that concentrates monitoring, supervision, and promotional functions in a single agency. Fragmentation of authority has proven to be detrimental to the attainment of development goals and to efficiency in dealing with undesirable deviations from established market performance standards. In Kenya, for instance, to be listed on the stock exchange a company needs the approval of the Capital Issue Committee of the Treasury, the Exchange Control Department of the central bank, the Committee of the Nairobi Stock Exchange, and the Capital Market Authority. 98 Interagency fragmentation of authority over market activities, issuers, investor interests, and intermediaries impairs focused market supervision and the promotion of sound market development. To ensure more effective enforcement of legal and regulatory provisions, supervisory agencies should be granted considerable autonomy within the government hierarchy. 99 REFERENCES Adams, Dale W. 1990. Taking a Fresh Look at Informal Finance. EDI. Washington, D.C.: The World Bank. and Delbert A. Fotchett. 1992. Informal Finance in Low-Income Countries. Westview Press, Boulder, Colorado, U.S.A. African Development Indicators. 1992. New York, N.Y., Washington D.C.: United Nations Development Program, The World Bank. Aidoo, Judith, F. 1989. "Report on the Feasibility of a Stock Exchange in Ghana." Judith F. Aidoo. Washington, D.C.: World Bank. Akaah, I.P., K.Q. Dadzie and B. Dunson. 1987. "Formal Financial Institutions as Savings Mobilizing Conduits in Rural LDCs." Savings and Development 3, XI. Milan, Italy. Amoako, Adu and K. Menyah. 1987. "Inflation, Devaluation and the Insurance Industry in a Less Developed Economy: The Case of Ghana." Savings and Development 3, XI. Milan, Italy. Aryeetey, E., and F. Gockel. 1989. Mobilizing Domestic Resources for Capital Formation: The Role of Informal Financial Markets in Ghana. Legon, Ghana: University of Ghana. Bank Negara Malaysia. 1989. Money and Banking in Malaysia. Kuala Lumpur, Malaysia: Bank Negara. Berglund, Per-Erik, Sergio Bortolani, Marie-France L'Heriteau, Lars-Ake Larsson. August 1992. Study Report on Retail Saving by Semi-formal Financial Institutions. Internal report prepared by Africa Technical Department, Africa Region. World Bank, Washington, D.C., U.S.A. Bhattacharyay, Biswa N. 1988. "Development of Financial Infrastructure: An International Comparison." Saving and Development 4 XII. Milan, Italy. Bhatia, Rattan, J. May. 1985. "The West African Monetary Union--An Analytical Review." Occasional Paper 35. Washington, D.C.: International Monetary Fund. Blejer, M.I., and C. Ke-Young. "Measurement of Fiscal Impact--Methodological Issues." Occasional 59. Washington D.C.: International Monetary Fund. 101 Braverman, A., and J.L. Guash. 1986. "Rural Credit Markets and Institutions in Developing Countries." World Development, Vol. 14. Brimmer, A. 1992. "Central Banking and Economic Development." Development Digest 19(2). Brunner, Karl. 1989. "The Role of Money and Monetary Policy." The Federal Reserve Bank of St. Louis Review 71(5). Caprio, Gerard, Jr., and Patrick Honohan. 1991. "Monetary Policy Instruments for Developing Countries." Paper Presented at a World Bank Symposium. Washington, D.C.: World Bank. Cole, David C.. and Betty F. Slade. September 1991. "Recent Developments in Indonesia Monetary Management." Consulting Assistance on Economic Reforn (CAER), HIID, Cambridge, Massachusetts, U.S.A. Mimneo. . September 1991. "Monetary Policy Regimes in East and Southeast Asia." Working Paper. Overseas Development Institute, London, U.K.: Regent's College. Connolly, Michael, and Paul A. Popiel. 1993. "Monetary Adjustment to Real Shocks in WAMU and Other Selected African Economies. " Internal Paper. Africa Technical Department, Africa Region. Washington, D.C., U.S.A.: The World Bank. Unpublished. Corrigan, G.E. 1990. "The Role of Central Banks and the Financial System in Emerging Market Economies." Federal Reserve Bank of New York, Vol. 15., #2. Craig, R. Neal. 1990. "Macrofinancial Indicators for 117 Developing and Industrial Countries." Policy Research Working Paper 58. World Bank, World Development Report Office, Washington, D.C. Duesenberry, James and Malcom F. McPherson. June 1991. "Monetary Management in Sub-Saharan Africa - A Comparative Analysis." Consulting Assistance on Economic Reform. CAER Discussion Paper No.7. HIID, Cambridge, Massachusetts, U.S.A. Mimeo. Giri, J. 1990. "Formal and Informal Small Enterprises in the Long-Term Future of Sub-Saharan Africa." Working Paper #9. Geneva, Switzerland: International Labor Organization. 102 Fisher, Bernhard et. al. 1986. "Sparkkapitalbildung in Entwicklungslanderen - Engpasse und Reformasatze." Forshungsbenchte fur das Bundesministenium fur Wirtschaftlich Zusammenarbeit, Vol 78. Munchen, Germany. Fry, Maxwell J. 1988. Money, Interest and Banking in Economic Development. Baltimore, Maryland, U.S.A.: Johns Hopkins University Press. Havrilesky, M.T. and J. T. Boorman. 1982. Money Supply, Money Demand and Macroeconomics Models. Arlington Heights, Illinois, U.S.A.: Harlan Davidson Inc. Huppi, Monika, and Gershon Feder. 1989. "The Role of Groups and Credit Cooperatives in Rural Lending." Policy, Planning and Research Working Paper No. 284. Washington, D.C., U.S.A.: The World Bank. Ibanez, Fernan. 1989. "Venture Capital in Entrepreneurial Development." Policy, Planning and Research Working Paper No. 53. Washington D.C.: The World Bank. Islam, Roumeen, and Deborah L. Wetzel. 1991. "The Macroeconomics of Public Sector Deficits: The Case of Ghana." Policy Research Working Paper 672. World Bank, Country Economics Department, Washington, D.C. Johnston, Barry R., and Odd Per Brekk. 1990. "Monetary Control Procedures and Financial Reform: Approaches, Issues, and Recent Experience in Developing Countries." In Philippe Callier, ed., Financial Systems and Development in Africa. EDI Seminar Series. Washington, D.C.: World Bank. Khan, Moshin S., and D. Knight Malcolm. 1981. "Stabilization Programs in Developing Countries: a Formal Framework, IMF Staff Papers vol. 28, pp 1-53. Washington, D.C.: IMF. Killick, Tony, and F. M. Mwega. 1991. "Monetary Policy in Kenya - 1967 - 88." Working Paper. Overseas Development Institute. London, U.K.: Regent's College and Martin Matthew. 1991. "Financial Policies in the Adaptative Economy." London: Working Paper. Overseas Development Institute. London, U.K.: Regent's College. Lane, Christopher. 1991. "Exchange Rates and Effectiveness of Monetary Policy." Working Paper. Overseas Development Institute. London, U.K.: Regent's College. . 1991. "The Effectiveness of Monetary Policy in Cote d'Ivoire." Working Paper. Overseas Development Institute. London, U.K.: Regent's College. Legarda, B. 1986. "A Philosophy of Financial Markets Building." Discussion Paper, Washington, D.C., U.S.A.: International Finance Corporation. Unpublished. 103 Le Houerou, Philippe, and Hector Sierra. 1992. "Estimating Quasi-Fiscal Deficits in a Consistency Framework: The Case of Madagascar." Internal Paper. Washington, D.C., U.S.A.: World Bank. Unpublished. Long, Millard. 1987. "Crisis in the Financial Sector." State Bank of India Monthly Review. Bombay, India. Masuoka, Toshiya. 1990. "Assets and Liabilities Management in the Development Coun- tries." Policy, Planning and Research Working Paper No. 454. Washington, D.C., U.S.A.: World Bank. Manchester, J. February 1989. "How Money Affects Real Output." Journal of Money, Credit and Banking, Vol. 21(1). Meek, Paul. April 26, 1991. "Central Banking Liquidity Management and the Money Market." Internal Paper. Africa Technical Department, Africa Region. Washington, D.C.,. U.S.A.: World Bank. Mimeo. Mosser, Patricia C. 1991. "Changes in Monetary Policy Effectiveness: Evidence from Large Macroeconomic Models." Federal Reserve Bank of New York Quarterly Review 17(1). Mwega, F.M. April 1989. "An Econometric Study of Selected Monetary Issues in Kenya", Working Paper # 42. Overseas Development Institute. London, U.K.: Regent's College. , S.M. Ngola, and N. Mwangi. 1990. "Real Interest Rates and the Mobilization of Private Savings in Africa." Research Paper No. 2. African Economic Research Consortium., Neal, Craig R. March 1990. "Macrofinancial Indicators for 117 Developing and Industrial Countries." Development Economics. Washington D.C.: The World Bank. Nelson, Eric R. April 1991. "Monetary Management in Sub-Saharan Africa." AID Consult- ing Assistance on Economic Reform Project. HIID, Cambridge, Massachusetts, U.S.A. Ogunbiyi, A., and A.O. Orafidiya. 1988. "Prospects of Livestock Insurance in Nigeria." Savings and Development, 4, Xli. Milan, Italy. Page, Sheila. September 1991. "The Relationship Between Formal and Informal Finance", Improving Management in Africa and Asia. Working Paper. Overseas Development Institute. London, U.K.: Regent's College. Popiel, Paul A. 1990. "Developing Financial Markets in Sub-Saharan Africa." EDI Working Papers. Washington, D.C., U.S.A.: World Bank. _ 1991. "Developpement des Marches Financiers - Presentation Generale du Probleme." Document intere. Institut de Developpement Economique. Washington,D.C., U.S.A.: Banque Mondiale. Non publie. 104 Prywess, Menahem. 1990. "Risk Facing U.S. Commercial Banks." Policy, Planning, and Research Working Paper No. 534. Washington, D.C., U.S.A.: World Bank. Roe, Alan. 1990. "Financial Systems and Development in Africa." EDI Working Papers. Washington, D.C., U.S.A.: The World Bank. , and Paul A. Popiel. 1987. "Managing Financial Adjustment in Middle-Income Countries." EDI Working Paper No. 11. Washington, D.C., U.S.A.: World Bank. __ 1990. "The Restructuring of Financial Systems in Latin America." EDI Working Paper No. 25. Washington, D.C., U.S.A.: World Bank. Rosas, Luis Edouardo, and Currie Lauchlin. 1986. UPAC - A Theory Converted into a Successful Reality - Colombian System of Savings and Housing. Bogota, Colombia: Instituto Colombiano de Ahorro y Vivienda. Ryding, J. 1990. "Housing Finance and the Transmission of Monetary Policy." Federal Reserve Bank of New York, Quarterly Review, Vol. 15, #2. Sagari, Silvia with Gabriela Guidotti. 1992. Venture Capital: Lessons From the Developed Worldfor the Developing Markets. Discussion Paper #13. Washington, D.C.: International Finance Corporation. Santorum, Anita. May 1990. "The Control of Money Supply in Developing Countries: China 1949-1988." Working Paper No. 35. Overseas Development Institute. London, U.K.: Regent's College. Sheng, Andrew. "Financial Fragility and the Contractual Nature of Banking - Hard Contracts and Firm Regulations." Unpublished. CECFP. Washington, D.C., U.S.A.: World Bank. . "The Contractual Nature of Banking and Financial Crises: Theory and Experience." Unpublished. CECFP. Washington, D.C., U.S.A.: The World Bank. _ "Bank Restructuring, Techniques and Experience." September 1992. Draft. CECFP. Washington, D.C., U.S.A.: World Bank. Sheng, Andrew. 1989. "Bank Restructuring in Malaysia -- 1985-1989." Policy, Planning and Research Working Paper No. 54. Washington, D.C., U.S.A.: World Bank. Singh, J. P. 1992. Draft Handbook on Financial Sector Operations. Central Operations Department, Operations Policy Unit. Washington, D.C., U.S.A.: World Bank. Sowa, Nii Kwaku. May 1991. "Monetary Control in Ghana: 1957 - 1988." Working Paper No. 45. Overseas Development Institute. London, U.K.: Regent's College. Tally, Samuel, and Ignacio Mas. 1990. "Deposit Insurance in Developing Countries. Policy, Planning, and Research Working Paper No. 548. Washington, D.C., U.S.A.: World Bank. 105 Tenconi, Roland. 1992. "The African Banking System in the 1980s and Early 1990s." Africa Region. Washington, D.C., U.S.A.: World Bank. Thomas, J.J. September 1991. "The Financial Needs of the Rural and Informal Sector: Commercial Banks and Informal Institutions." Working Paper. Overseas Development Institute. London, U.K.: Regent's College. Uganda Stock Exchange. August 17, 1990. "Recommendations of the Workshop." Kampala. "Trends in Developing Economies." 1991. Washington D.C., U.S.A.: The World Bank. United Nations. 1991. Accountancy Development in Africa, Challenges of the 1990s. New York: United Nations. Vaez-Zadeh, R. and S.P. Leite. 1986. "Effectiveness of Selective Credit Controls: An Empirical Test Applied to India." The Journal of Development Studies, Vol. 22, #1. Vinay, Bernard. 1988. "Zone Franc et Cooperation Monetaire." 1988. Focal Coop, Ministere de la Cooperation et du Developpement. Vittas, Dimitri and Michael Skully. 1991. "Overview of Contractual Savings Institutions." Policy, Planning, and Research Working Paper No. 605. Washington, D.C., U.S.A.: World Bank. World Development Report 1989. New York: Oxford University Press. Yabile, K.R. 1987. "Provision of Liquidity Management: An Incentive to Curb Default in the Ivory Coast's Agricultural Credit Programs." Savings and Development, 4, XI. Yaron, Jacob. 1992. Assessing Development Finance Institutions: A Public Interest Analysis. World Bank Discussion Papers No. 174. Washington, D.C, U.S.A.: World Bank. Younger, Stephen D. March 1991. "Monetary Management in Ghana." Development Alternatives. Mimeo. 106 Distributors of World Bank Publications ARGENTINA EGYPT, ARAB REPUBLIC OF KENYA SAUDI ARABIA, QATAR Carlos Hirsch, SRL Al Ahramn Africa Book Servioe (E.A.) Ltd. jarir Book Store Galeria Guemes Al Galaa Street Quaran House, Mfangano Street P.O. Box 3196 Florida 165,4th Floor-Ofc. 453/465 Cairo P.O. Box 45245 Riyadh 11471 1333 Buenos Aires Nairobi The Middle East Observer SINGAPORE, TAIWAN, Ofidna del Libro Internacional 41, Sherif Street KOREA, REPUBLIC OF MYANMAR,BRUNEI Alberti 40 Cairo Pan Korea Book Corporation Gower Asia Pacific Pte Ltd. 1082 Buenos Aires P.O. Box 101, Kwangwhamun Golden Wheel Building FINLAND Seoul 41, Kallang Pudding, #04-03 AUSTRALIA, PAPUA NEW GUINEA, Akateeminen Kirjakauppa Singapore 1334 FIJI, SOLOMON ISLANDS, P.O. Box 128 Korean Stock Book Centre VANUATU, AND WESTERN SAMOA SF-00101 Helsinki 10 P.O. Box 34 SOUTH AFRICA, BOTSWANA D.A. Information Services Yeoeido For singe titles: 648 Whitehorse Road FRANCE Seoul Oxford University Press Mitcham 3132 World Bank Publications Southern Africa Victoria 66, avenue dIlena MALAYSIA P.O. Box 1141 75116 Paris University of Malaya Cooperative Cape Town 8000 AUSTRIA Bookshop, Limited Gerold and Co GERMANY P.O. Box 1127, Jalan Pantai Baru For subscription orders: Graben 31 UNO-Verlag 59700 Kuala Lumnpur Intemational Subscription Service A-1011 Wien Poppelsdorfer ADeeSS P.O. Box 41095 53115 Bonn MEXICO Craighall BANGLADESH INFOTEC Johannesburg 2024 Micro Industries Development GREECE Apartado Postal 22-860 Assistance Society (MIDAS) Papasotinou S.A. 14060 Tlalpan, Mexico D.F. SPAIN Hpuse 5, Road 16 35, Stournara Str. Mundi-Prensa Libros, S.A. Dhanmondi R/Area 106 82 Athens NETHERLANDS Castello 37 Dhaka 1209 De Undeboom/InOr-Publikaties 28001 Madnd HONG KONG, MACAO P.O. Box 202 BELGIUM Asia 2000 Ltd. 7480 AE Haaksbergen Libreria Internadonal AEDOS jean De Lannoy 46-48 Wyndham Street Consell de Cent, 391 Av. du Roi 202 Winning Centre NEW ZEALAND 08009 Barcelona 1060 Brussels 7th Floor EBSCO NZ Ltd. Central Hong Kong Private Mail Bag 99914 SRI LANKA AND THE MALDIVES BRAZIL New Market Lake House Bookshop Publicacoes Tecnicas Intemacionais Ltda. HUNGARY Auckland P.O. Box 244 Rua Peixoto Gomide, 209 Foundation for Market Economy 100, Sir Chittampalam A. 01409 Sao Paulo, SP Dombovari Ut 17-19 NIGERIA Gardiner Mawatha H-1117 Budapest University Press Limited Colombo 2 CANADA Three Crowns Building Jericho Le DiffLuseur INDIA Private Mail Bag 5095 SWEDEN 151A Boul. de Mortagne Allied Publishers Private Ltd. Ibadan For single titles: BoucherviDe, Quebec 751 Mount Road Fritzes Fackboksforetaget 14B 5E6 Madras - 600 002 NORWAY Regeringsgatan 12, Box 16356 Narvesen Information Center S-103 27 Stockholm Renouf Publishing Co. INDONESIA Book Department 1294 Algoma Road Pt. Indira Limited P.O. Box 6125 Etterstad For subscription orders: Ottawa, Ontario Jalan Borobudur 20 N-0602 Oslo 6 Wennergren-Williams AB KIB 3W8 P 0. Box 181 P 0. Box 1305 Jakarta 10320 PAKISTAN S-171 25 Solna CHINA Mirza Book Agency China Financial & Economic IRAN 65, Shahrah-e-Quaid-e-Azam SWITZERLAND Publishng House Kowkab Pubhshers P.O. Box No. 729 For single titles: 8, Da Fo Si Dong Jie P.O. Box 19575-511 Lahore 54000 Librairie Payot BeLjing Tehran Case postale 3212 PERU CH 1002 Lausanne COLOMBIA IRELAND Editorial Desarfotlo SA Infoenlace Ltda. Government Supplies Agency Apartado 3824 For subscnption orders: Apartado Aereo 34270 4-5 Harcourt Road Lima I Librairie Payot Bogota D.E. Dublin 2 Service des Abonnements PHILIPPINES Case postale 3312 COTE D'IVOIRE ISRAEL Intemational Book Center CH 1002 Lausanne Centre d'Edition et de Diffusion Yozmot Literature Ltd. Suite 1703, Cityland 10 Africaines (CEDA) P.O. Box 56055 Condominium Tower I THAILAND 04 B.P. 541 Tel Aviv 61560 Ayala Avenue, H.V. dela Central Department Store Abidjan 04 Plateau Costa Extension 306 Silom Road ITALY Makati, Metro Manila Bangkok CYPRUS Licosa Commissionaria Sansoni SPA Center of Applied Research Via Duca Di Calabria, 1/1 POLAND TRINIDAD & TOBAGO Cyprus College Casella Postale 552 International Publishing Service Systematics Studies Unit 6, Diogenes Street, Engomi 50125 Firenze Ul. Piekna 31/37 #9 Watts Street P.O. Box 2006 00-677 Warszawa Curepe Nicosia JAMAICA Trinidad, West Indies Ian Randle Publishers Ltd. For subscription orders: DENMARK 206 Old Hope Road IPS Journals UNITED KINGDOM SamnfundsLitteratur Kingston 6 UlI. Okrezna 3 Microinfo Ltd. Rosenoerns Alit 11 02-916 Warszawa P.O. Box 3 DK-1970 Fredenksberg C JAPAN Alton, Hampshire GU34 2PG Eastem Book Service PORTUGAL England DOMINICAN REPUBLIC Hongo 3-Chome, Bunkyo-ku 113 Livrana Portugal Editora Taler, C. por A. Tokyo Rua Do Carmo 70-74 ZIMBABWE Restauraci6n e Isabel Ia Cat6ica 309 1200 Lisbon Longman Zimbabwe (Pvt.) Ltd. Apartado de Correos 2190 Z-1 Tourle Road, Ardbennie Santo Domingo P.O. Box ST 125 Southerton Harare Recent World Bank Discussion Papers (continued) No. 233 Land Reforn and Farm Restructuring in Ruissia. Karen Brooks and Zvi Lerman No. 234 Population Growth, Shifting Cultivation, and Unsustainable Agricultural Development: A Case Study in Madagascar. Andrew Keck, Narendra P. Shamma, and Gershon Feder No. 235 The Design and Administration of Intergovernmental Transfers: Fiscal Decentralization in Latin America. Donald R. Winkler No. 236 Public and Private Agricultural Extension: Beyond Traditional Frontiers. Dina L. Umali and Lisa Schwartz No. 237 Indonesian Experience with Financial Sector Reforn. Donald P. Hanna No. 238 Pesticide Policies in Developing Counitries: Do They Encourage Excessive Use?Jumanah Farah No. 239 Intergovernmental Fiscal Relations in Indonesia: Issues and Refonn Options. Anwar Shah and Zia Qureshi, with Amaresh Bagchi, Brian Binder, and Heng-fu Zou No. 240 Managing Reduntdancy in Overexploited Fisheries. Joshua John No. 241 Institutional Change and the Public Sector in Transitional Economies. Salvatore Schiavo-Campo No. 242 Africa Can Compete!: Export Opportunities and Challenges for Garnents and Home Products in the U.S. Market. Tyler Biggs, Gail R. Moody, Jan-Hendrik van Leeuwen, and E. Diane White No. 243 Liberalizing Trade in Services. Bernard Hoekman and Pierre Sauve No. 244 Women in Higher Education: Progress, Constraints, and Promising Initiatives. K. Subbarao, Laura Raney, Ha]il Dundar, andJennifer Haworth No. 245 "hat We Know About Acquisitiotn of Adult Literacy: Is There Hope? Helen Abadzi No. 246 Fornulating a National Strategy on Itrfonnation Technology: A Case Study of India. Nagy Hanna No. 247 Improving the Transfer and Use of Agricultural Information: A Guide to Infonnatiotn Technology. Willem Zijp No. 248 Outreach and Sustainability of Six Rural Finance Institutiotis in Sub-Saharan Africa. Marc Gurgand, Glenn Pederson, and Jacob Yaron No. 249 Population and Iticome Change: Recent Evidence. Allen C. Kelley and Robert M. Schmidt No. 250 Submissiotn and Evaluation of Proposalsfor Private Power Generatiotn Projects in Developing Countries. Edited by Peter A. Cordukes No. 251 Supply and Denandfor Finance of Small Enterprises in Ghana. Ernest Aryeetey, Amoah Baah-Nuakoh, Tamara Duggleby, Hemamala Hettige, and William F. Steel No. 252 Projectizitig the Governatce Approach to Civil Service Refonn: An Instituitional Environment Assessmentfor Preparing a Sectoral Adjusttnetnt Loan in the Gambia. Rogerio F. Pinto with assistance from Angelous J. Mrope No. 253 Small Finns Itformally Financed: Studiesfrotn Bangladesh. Edited by Reazul Islam, J. D. Von Pischke, and J. M. de Waard No. 254 Indicatorsfor Monitoritig Poverty Reductiorn. Soniya Carvalho and Howard White No. 255 Violence Against Women: The Hidden Health Burdetn. Lori L. Heise with Jacqueline Pitanguy and Adrienne Germain No. 256 Women's Health and Nutrition: AIaking a Difference. Anne Tinker, Patricia Daly, Cynthia Green, Helen Saxenian, Rama Lakshminarayanan, and Kirin Gill No. 257 Imnprovitig the Quality of Primary Education in Latin America: Towards the 21st Century. Lawrence Wolff, Ernesto Schiefelbein, andJorge Valenzuela No. 258 How Fast is Fertility Declinintg in Botswana and Zimbabwe? Duncan Thomas and Ityai Muvandi No. 259 Policies Affectitig Fertility and Contraceptive Use: An Assessment of Twelve Sub-Saliarant Countries. Susan Scribner The World Bank Headquarters European Office Tokyo Office 1818 H Street, N.W. 66, avenue d'lena Kokusai Building Washington, D.C. 20433, U.S.A. 75116 Paris, France 1-1 Marunouchi 3-chome Chiyoda-ku, Tokyo 100, Japan Telephone: (202) 477-1234 Telephone: (1) 40.69.30.00 Facsimile: (202) 477-6391 Facsimile: (1) 40.69.30.66 Telephone: (3) 3214-5001 Telex: Msci 64145 WORLDBANK Telex: 640651 Facsimile: (3) 3214-3657 MCI 248423 WORLDBANK Telex: 26838 Cable Address: INTBAFRAD WASHINGTONDC 12996 FID 100 0-8213-29'96-0 FINANCIAL SYSTEM SUB/AFR 11w I I I ~1 10o111 I11 I12111 400000015125 $10 .95 ISBN 0-8213-2996-0