SUSTAINABLE BANKING .with the POOR Case Studies in Microfinance DOMINICAN REPUBLIC - ADEMI McDonald Benjamin Joanna Ledgerwood c,oRAL & \1 May 1999 RESi°0111:,rc: ri::I\\TER rEB 2 3 2000 The World M'rifNK . 1-- Sustainable Banking with the Poor (SBP) is a collaborative effort of the Gender and Poverty Team and Rural Finance at the World Bank, funded by the World Bank, the Royal Ministry of Foreign Affairs of Norway, the Swiss Agency for Development and Cooperation* (SOC), and the Ford Foundation. The study aims at improving the ability of donors, governments and practitioners to design and implement policies and programs to build sustainable financial institutions that effectively reach the poor. The SBP task managers are Lynn Bennett and Jacob Varon; the technical manager is Carlos Cueva~and the associate manager is Korotoumou Ouattara. The administrative assistant is Sangae Sherman. The World Bank SBP, J4-241, 1818 H Street, N.W. Washington, D.C. 20433 Phone (202) 458-0277 Fax (202) 477-2978 Internet: CCuevas@WORLDBANK.ORG or KOuattara@WORLDBANK.ORC: WEB PAGE: http://wwwesd.worldbankorg/htmVesd/agr/sbp *Note: In the back cover box, Swiss Development Corporation, should read Swiss Agency for Development and Cooperation (SDC). This study was p·repared by McDonald Benjamin, a financial economist at The World Bank, anc Joanna Ledgerwood, a consultantat Microfinance International. Both authors have followed ADEMl's wprk'for,:a f'!Umberofyears and prepared earlier reports on ADEMl's operations. '. . :. . •• : ~ The study would not have been possible without the warm welcome and support that the authors received from.all ,the staff at ADEMI. Particular thanks are due to Camilo Lluberes, President, and Pedro Jim~nez, Executive Director of ADEMI, who opened ADEMl's doors to the authors and unreservedly shared their information and insights on the institution. TABLE OF CONTENTS ABBREVIATIONS AND ACRONYMS ............................................................................. IV EXECUTIVE SUMMARY ............................................................................................... VII THE BIRTH OF ADEMI. ...................................................................................... VII THE FORMATIVE YEARS .................................................................................. VIII GROWTH AND CRISIS ........................................................................................ x CONSOLIDATION AND MATURITY .................................................................... x GRADUATION AND NEW DIRECTIONS ....... ... ................................................. XIII LOOKING TO THE FUTURE WITH BANCOADEMI ............................... ........... XIV I. THE DOMINICAN REPUBLIC: PAST AND PRESENT. ............................................. 1 SNAPSHOT OF THE DOMINICAN REPUBLIC .. ,................................................. 1 WHO ARE THE DOMINICANS? ........................................................................... 2 11. THE SHIFTING MACROECONOMIC AND FINANCIAL SECTOR CONTEXT. ......... 4 UPS AND DOWNS IN THE MACROECONOMY ................................................... 4 WHO'S WHO IN THE FINANCIAL SECTOR ........................................................ 6 FINANCIAL SECTOR CRISIS AND REFORM ..................................................... 8 ADEMl'S PLACE IN THE FINANCIAL SYSTEM................................................... 9 Ill. THE DOMINICAN "MICRO" ECONOMY ................................................................. 12 DEFINING "POOR", "INFORMAL" AND "MICRO" .............................................. 12 FEATURES OF POVERTY ................................................... .............................. 12 WHO BENEFITS FROM SOCIAL SPENDING? ................................................. 14 MICROENTERPRISES: SMALL BUT IMPORTANT........................................... 14 LENDING MICROENTERPRISES A HAND THROUGH MICROFINANCE ....................................................................... 17 IV. WHAT ISADEMl? ................................................................................................... 20 ADEMl'S LEGAL ORIGINS ................................................................................. 20 WHAT GAVE RISE TO ADEMl? ............................................................:............ 20 THE NEXT FIFTEEN YEARS ............................................................................. 21 ADEMI AT A GLANCE ........................................................................... ............. 22 ADEMl'S GUIDING PRINCIPLES ....................................................................... 22 WHAT ADEMI OFFERS ITS CLIENTS ............................................................... 23 V. WHOM DOES ADEMI SERVE? .............................. ............. :................................... 29 ADEMl'S MASSIVE MARKET NICHE. ................................................................ 29 CHARACTERISING THE CLIENTELE ............................................................... 29 HOW WELL DO ADEMl'S SERVICES FIT CLIENTS' NEEDS? ......................... 34 VI. HOW DOES ADEMI WORK? .................................................................................. 38 HE OVERALL STRUCTURE .............................................................................. 38 THE CRITICAL ROLE OF ADEMl'S HUMAN RESOURCES.............................. 42 FOLLOWING THE DOLLAR: THE LENDING PROCESS .................................. 45 FOLLOWING THE DOLL.AR: THE LOAN RE~OVERY PROCESS .................. .48 SUSTAINABLE BANKING WI1H THE POOR i ._ .. •'; •·- ... . VII. ADEMl'S PERFORMANCE .................................................................................... 53 THE PROFITABLE NON-PROFIT ...................................................................... 53 DECOMPOSING ADEMl'S EXCEPTIONAL RETURNS ....................... :............. 55 . RISK ................................................................................................................... 64 PROTECTING STAFF: ADEMl'S PENSION PLAN ............................................ 67 THE TRUE BOTTOM LINE: ADEMl'S IMPACT ON CLIENTS ........................... 68 VIII. LESSONS FRO_ M ADEMl'S SUCCESS, AND THOUGHTS ON ADEMl'S FUTURE .............................................................. 72 THE ESSENCE.OF THE ADEMI MODEL. .......................................................... 72 WHAT DID NOT WORK AND WHAT NEEDS MORE WORK ............................ 73 BANKING \ ON THE FUTURE .............................................................................. 77 BIBLIOGRAPHY ............................................................................................................ 82 BOXES 4.1: Why ADEMI No Longer Lends to Groups ...................................................... 27 8.1: Converting from Semi-Formal to Formal Microfinance - Five Recent Experiences ............................................................................... 79 FIGURES 2.1: Who's Who in the Financial System in the Dominican Republic ...................... 7 2.2: AD EM l's Net Worth is Comparable to That of Commercial Banks •in the Dominican Republic ......................................................................... 11 3.1: Gini Coefficient Measures for the Dominican Republic ................................. 13 3.2: Microenterprises Contribution to Total Value Added in the Dominican Republic in 1991 ..................................................................... 1.5 3.3: Microenterprises and the Job MarketAccounted for Almost One in Four Jobs in the Dominican Republic in 1991 ........................................... 16 5.1: Share of Women Entrepreneurs in ADEMl's Clientele of December 1996 ......................................................................32 5.2: Shift of ADEMI Lending Towards the Retail Sector ....................................... 33 (Number of Clients by Sector} ...................................................................... 33 5.3: A Look at the Growth and Dispersion of AD EM l's Client Base ...................... 36 6.1: The Structure of ADEMI, as of January 1997............ .. . .. . . .. ... ... .. . . ........ 39 6.2: ADEMI Enforces Collection on all Loans within 90 Days ... :........................... 52 7.1: ADEMl'sAssets and Net Worth (Figures in US$ Millions} ............................. 53 7.2: ADEMl's Subsidy Dependence Index Measures ........................................... 54 7.3: Analysing a Financial Institution's Return on Assets ..................................... 55 7.4: ADEMl's Funding Sources .. ;......................................................................... 60 (Expressed as Percent of Total Liabilities} ........................................60 7.5: Key Financial Indicators that Explain ADEMl's Profitability ........................... 63 7.6: Diversifying the Asset Base of ADEMl's Pension Fund ................................. 68 SUSTAINABLE BANKING wrm THE POOR ii 7.7: ADEMI is Having an Impact on Employment.. ........................... ;................... 70 8.1: ADEMI has Recovered All Initial Subsidies to Earn a.................................... 74 Cumulative Surplus .................................................................................. 74 8.2: Ademi's Maximum, Average and Minimum Loan Portfolios per Advisor in Millions of Dominican Pesos, as of September 1996................................. 76 TABLES 2.1 : Growth in Export Processing Zones .................................................................. 4 2.2: Movements in Macroeconomic Indicators ... ;..................................................... 5 2.3: Financial Sector Rationalisation Has Favoured Commercial Banks ............... 10 4.1: A Quick Look At ADEMl's Operations, As Of June 1997 ................................. 22 4.2: ADEMl's Loan Products, Prices and Terms by Type of Client, as of end-1997 .............................................................................. 25 5.1: ADEMI in the 1990s: Making Larger Loans, Serving Larger Enterprises ........ 30 5.2: ADEMl's Lending by Location, April-June 1997 .............................................. 33 6.1: How Microenterprise Loans are Made and Recovered at ADEMI ................... 46 6.2: Where the Buck Stops for Loan Approvals ............................................. .47 7.1: ADEMl's Financial Statements Yield Impressive Financial Ratios .................. 57 7.2: ADEMI Has Cut Unit Costs by Increasing Loan Size ...................................... 61 7.3: ADEMI Has a Healthy Portfq,Ho and Ample Loss Reserves ......................... 62 APPENDICES 1: Key Macroeconomic and Financial Sector Indicators ................................. 85 2: Poverty Indicators ................................................................................86 3: ADEMl's Incentive Scheme for Credit Advisors ......................................... 88 4: ADEMl's Financial Statements ............................................................... 89 5: Calculation of the Subsidy Dependence Index for ADEMI .................................... .'. .................................................. 91 6: ADEMl's Financial Performance Ratios ................................................... 92 7: Branch Level Income Statements and Portfolio Data in Dominican Pesos '000s ....................... : .......................................................... 94 8: ADEMl's Pension Plan ....................................... ;.................................................. 95 9: ADEMl's Impact on Clients ...................................................................96 10: Selected Indicators for ADEMI and for Members of the MicroFinance Netrwork........................................................... ; ....... 99 11 : • List of Case Studies .....................................................................................100 12: List of Discussionffechnical Papers ........................................ :............................................ 101 . SUSTAINABLE BANKING WITH THE POOR iii ABBREVIATIONS AND ACRONYMS ADEMI Associaci6n para el Desarrollo de Microempresas, Inc. (Association for the Development of Micro enterprises) ADOPEM Associaci6n Dominicana para el Desarrollo de la Mujer, Inc. (Dominican Association for Women's Development) BancoADEMI Banco de Desarrollo ADEMI, S.A. (ADEMI Development Bank) BNV Banco Nacional de la Vivienda (National Housing Bank) CA Credit Advisor CFI Corporaci6n de Fomento Indµstrial (Industrial Development Corporation) D.R Dominican Republic ECU European Currency Units EIB European Investment Bank EPZs Export Processing Zones FDD Fundaci6n Dominicana de Desarrollo (Dominican Development Foundation) FED Fundaci6n Economia y Desarrollo (Economics and Development Foundation) FondoMicro Fondo para la Promoci6n de la Microempresa (Foundation for the Promotion of Microenterprises) GAAP Generally Accepted Accounting Principles GDP Gross Domestic Product GNP Gross National Product GTZ Gesellschaft fiir-Technische Zusammenarbeit (Germany's Agency for Technical Cooperation) IDB Inter-American Development Bank IDECOOP Instituto de Desarrollo y Credito Cooperativo (Institute for Development and Cooperative Credit) ITC Inter-American Investment Corporation IMF International Monetary Fund . . INFOTEP Instituto de Formaci6n Tecnica y Profesional (Institute for Technical and Professional Training) INVI Instituto Nacional de la Vivienda (National Housing Institute) K-REP Kenya Rural Enterprise Program MIS Management Information System MUDE Mujeres en Desarrollo (Women In Development) NPLs Non-Performing Loans ODC Oficina de Desarrollo de la Comunidad (Community Development Office) PRODEM Fundaci6n para la Promoci6n y Desarrollo de la Microempresa (Foundation for the Promotion and Development of Microenterprises- Bolivia) PRO:MIPYME Programa para la Micro, Pequefia y Mediana Empresa (Programme for Micro-, Small- and Medium-Scale Businesses) SUSTAINABLE BANKING WITH THE POOR iv ABBREVIATIONS AND ACRONYMS CONT'D RCM Regional Credit Manager SBP Sustainable Banking with the Poor Project STP Secretariato Tecnico de la Presidencia (Technical Secretariat of the Presidency) USAID United States Agency for International Development Exchange Rates As of December 31 Dominican Republic Pesos per US$1 1995 US$1 = Pesos 20.0 1996 US$1 = Pesos 14, 1 1997 US$1 = Pesos 14.4 SUSTAINABLE BANKING WITH THE POOR V Country Profile Economic and Social Context Inflation GDP per Capita 1997 US$ 753 1995 12.5 Population (Million) 1997 8.1 1996 4.4 Population Density 1997 165 inhab./Km2 • 1997 8.3 SUSTAINABLE BANKING WITH THE POOR vi EXECUTIVE SUMMARY "The undersigned, citizens who are interested in strengthenmg the economic democracy of the country, have decided to contribute to addressing the problem ofunemployment . and underemployment via the formation ofan organisation that would promote the development ofmicroenterprises with the fundamental objective of incorporating marginalised groups into the economic life of the nation, an organisation that will be known as the "Association for the Development ofMicroenterprises, Inc. " 1 THE BIRTH OFADEMI The Association for the Development of Enterprises (ADEMI) is the brainchild of a Dominican businessman, Camilo Lluberes. Conceived in late 1982 as a way of addressing the plight of urban poor by "democratising" credit, ADEMI was legally constituted on December 14, 1982. It was blessed with the early support of a specialist in community development, a former Jesuit priest who was in the Dominican Republic working for the ACCION network. Stephen Gross brought together Llubere~ and Pedro Jimenez, a successful businessman who was managing a growing company in one of the Dominican Republic's export processing zones. With support from ACCION, Jimenez was brought on as Executive Director to assist Lluberes in launching this new venture. And so was born a leadership alliance that has been one of the key factors of ADEMI's exceptional success. • Right from the outset, ADEMI's philosophy was businesslike. This philosophy is captured in an aphorism often quoted by Jimenez, namely that "one does not have to become poor in order to assist the poor". This principle, which is easily misinterpreted, in practice underscores the importance of self-sustainability as a fundamental objective of ADEMI's operations-relying indefiiµtely on external subsidies to support lending operations was not an option for ADEMI. Therefore from day one the Association made sure it was not viewed as just another donor program, but rather as a business partner for low-income entrepreneurs. The decision to adopt a business approach to microcredit was the first, and arguably most important, key strategic decision taken by ADEMI, as it imbued all facets of ADEMI's life and operations, from the selection of its Board of Directors, to the location of its branches, to the incentive schemes it introduced for its staff, to the way it dealt with delinquent loans. A number of other key strategic decisions in ADEMI's life are documented below, but pursuing a social objective with a businesslike approach is clearly the dominant one, and the one that has most fundamentally contributed to ADEMI's success. ADEMI was incorporated under Executive Decree No. 745 of February 11, 1983, as one • of the first Dominican organisations dedicated exclusively to the development of By-Laws of the Association for the Development of Microenteiprises (ADEMI}, December 14, 1982. SUSTAINABLE BANKING WITH THE POOR vii microenterprises. 2 ADEMI's Constituent Assembly--consisting of distinguished members of the business and academic community--together with the Technical Secretariat of the Presidency contributed about RD$238,000 in grants for the first year (equivalent to US$238,000, given a RD$1:US$1 exchange rate at that time). Lluberes himself contdbuted RD$30,000 and offered a personal guarantee to Banco Popular, the largest private commercial bank in the Dominican Republic, for the bank to open a line of credit ofRD$50,000 for ADEMI. Free technical support was provided by ACCION, which advised ADEMI on implementing the ACCION model of group lending. 3 THE FORMATIVE YEARS I- The new institution quickly set about its first priority, namely designing the organisational structure, defining operating guidelines, and carefully selecting credit advisors. The constituent assembly elected a Board of Directors which formally appointed the Executive Director to manage day-to-day operations. Operations were initially confined to the capital city, Santo Domingo, starting in an office in the older part of town. ' ' In April 1983, ADEMI made its first loan. It began its lending in the same way as an informal lender would: the first client was well-known to the Executive Director, so the costs of obtaining information were low. Clearly with the growth of the institution came thousands of new dients (more than 40,000 as of June 1997), for whom information barriers had to be overcome. This was initially done via group lending schemes based on ACCION's model. However by 1984 this model had been rejected because it was felt that credit advisors were relying excessively on the peer pressure mechanisms and did not develop a thorough knowledge of their clients. This was a second key strategic decision for ADEMI. The screening mechanism for indivi_dual loans relied on a few key principles: first, ADEMI employed well-educated credit advisors who had themselves grown up in the barrios or low-income neighbourhoods that ADEMI was serving; second, credit advisors asked members of the community about potential clients in the same way that an informal moneylender might: Was the person viewed as reliable? Did he or she have a drinking problem or family problems that might impair his/her ability to repay? How long had the person been in business and what did clients think of the person's products or services? Third, ADEMI offered each borrower only a small initial loan, with a promise of-a larger one upon timely repayment. Fourth, ADEMI built 2 Previously the Dominican Development Foundation (FDD) had piloted microfinance programs as adjuncts to its small farmer activities-its experiences and those of ACCION elsewhere in Latin America proved valuable in the design of the ADEMI programs. In addition, the Dominican Association for Women's Development (ADOPEM) had just begun its operations in 1982. 3 The ACCION model of group lending involves clients identifying three to five partners with whom to organise a group for the pmposes of obtaining loans from an ACCION affiliate. These typically uncollateralised, market rate loans are provided simultaneously to each group member, and must be repaid in weekly installments. Group members' loans are interlinked insofar as all members must repay their loans before any member can obtain a (larger) follow-on loan. ·susTAINABLE BANKING WITH THE POOR viii up a reputation of charging positive real interest rates and of enforcing loans, so that only serious candidates need apply. • In designing its loan disbursement and repayment processes, ADEMI took another key strategic decision, namely to involve.Banco Popular as a partner in these processes. Loans were disbursed in the form of cheques drawn on an account at Banco Popular and clients were required to cash the cheques at the bank. As ADEMI grew, it strategically locat~d its branches fairly closely to B~co Popular branches (or in a few instances to other banks' branches). Clients are given preprinted repayment coupons which they take to the bank with their payments. They then bring the bank receipts to ADEMX. The banks provide account statements to ADEMI, allowing for cross-checks. This procedure cuts down on administrative costs for ADEMI, sharply reduces the possibility of fraud or theft, (since ADEMI's offices handle only nominal amounts of cash), and creates a valuable paper trail for audits. It has also drawn many of ADEMI's clients more broadly into the formal fmancial system, since Banco Popular has sold to the best clients its own fmancial products, such as savings accounts and insurance services, or even lower interest rate loans to the very best ones. Banco Popular covers servicing costs on ADEMI's accounts with fee income and obtains valuable information about new clients at no cost in terms of risk exposure. An early problem for ADEMI was liquidity~ ADEMI had to grow quickly to reach an efficient scale of production and cover i~-- ~s~,. M~r~pver it had to make sure that it honoured its commitment to make available a new, larger loan promptly to clients who repaid on time . .Any delays could discourage borrowers and affect repayment performance. During its first year, ADEMI relied heavily on the Banco Popular line of credit,,.,which was increased with a further guarantee from Lluberes. Relief came in 1984 in the form of external assistance from the Interamerican Development Bank (IDB), which provided RD$230,000 in funding on concessional terms. The additional funding allowed ADEMI to open a second branch in Santo Domingo. At the end of 1984, Jimenez decided to resign as Executive Director to return to the inicrolenders, namely that of not private sector. This decision highlights a key difficulty for _ only attracting but also retaining qualified staff who can potentially command higher salaries in private for-profit enterprises. Jimenez's departure marked the end of the first phase of ADEMI's development. He left behind an institution with two branches serving about 1,000 clients, and with 20 employees, US$653,000 in assets, US$272,000 in equity, a net loss before grants of US$123,000, and an SDI of 151 percent at the end of its first full fiscal year of operations. 4 4 The SDI, or Subsidy Dependence Index, compares a financial institution's total subsidies in a given year from concessional loans, inadequately remW1.erated capital, grants, free technical support, etc, with the institution's interest income. The rationale is to show by how much interest revenue would have to increase in order to dispense with the need for subsidies. In this case, other things equal, it would have taken a 151 percent increase in interest income to be able to become subsidy-free, i.e. ADEMI was initially highly dependent on subsidies (see Yaron (1992), and Benjamin (1994) for an application to ADEMI). SUSTAINABLE BANKING WI1H THE POOR . ix GROWTH AND CRISIS A new Executive Director assumed the leadership of the Association in January 1985. Mirta Olivares quickly brought in new management, embarked on a programme of cooperation with USAID that yielded US$200,000 to support operating costs, drew on IDB funding totalling . US$420,000, and opened branches outside of the capital. The opening of branches in the southern region was a fourth key strategic decision for ADEMI, as it signalled the start of ADEMI as a national rather than merely a localized institution. During 1985-86, ADEMI' s total branch network increased quickly from 2 to 14 branches, covering most of the country .. Olivares also reintroduced the ACCION group lending methodology which had been abandoned by the earlier management. However, bureaucratic practices began to creep into ADEMI's procedures, leading to delays in lending. Accounting and record-keeping also fell behind as activities grew. With the pressure to lend, much of the funding was placed in new offices in the "interior'', i.e. ADEMI's offices outside Santo Domingo, which lacked the infrastructure to handle the volume of activities. Arrears began to rise significantly, reaching almost 30 percent of the loan portfolio (40 percent in parts of the interior). This conjuncture of events led to an internal crisis in August 1986, and the Board of Directors accepted the resignation of the entire senior management and several staff. The Board turned to Jimenez to resume the position of Executive Director (see Gomez [1989]). Thus ended the second phase of ADEMI's development. At the end of 1986, after almost two years under Olivares's leadership, ADEMI had a national network with 3,065 clients, 50 employees, US$878,000 in assets, US$342,000 in net worth, a reduced net loss before grants ofUS$36,000, and an SDI of 91 percent. A lesson to be drawn from this episode regards the challenges confronting institutions upon a transition in.management. This is particularly true if a charismatic leader such as Jimenez is followed by a more procedural manager, since the very same charismatic leadership that unifies an organisation into a team with a purpose can also generate expectations and allegiances that make the job of a subsequent manager more challenging. Since there is considerable loyalty to Jimenez among his staff, due to the exceptional way in which he has led the institution, this is an issue that could arise again for ADEMI when he eventually retires. The empirical literature on microfmance has often documented the fact that most microenterprises fail in their first four years of operations. Thereafter, the hazard rate drops significantly and the ·enterprises tend ·to stabilise. Much like the enterprises it serves, ADEMI itself experienced a very challenging first four years of life in which it almost collapsed more than once, due to both funding and managerial problems. However, once it survived the first four years, it embarked on a steady upward trend in terms of both outreach and self- sustainability. CONSOLIDATION AND MATURITY During the third phase of ADEMI' s development, beginning late in 1986, ADEMI began by putting its house in order, improving its service to clients and identifying new sources of funding. The German Agency for Technical Collaboration (GTZ), ACCION, Banco Popular SUSTAINABLE BANKING WITH THE POOR X and USAID (with funds channeled through FondoMicro) provided valuable technical and financial support during this period. On the operational side, accounting was updated via the introduction of a computerised data processing system, which provides immediate access to daily updated programme information for all offices. 5 Procedures were regularised and the credit advisor's manual was launched anew. _Arrangements between headquarters and interior branches were better defined and a new organisational structure with more decentralised decision-making was introduced. Nevertheless problems of communication and decision-making between the interior and managers based in headquarters have persisted. 6 With regard to loan1-delivery, group lending was permanently terminated as of 1987, with existing group members either becoming individual borrowers or being passed on to another institution. 7 The most important innovation was the introduction of an incentive system for credit advisors. This was the fifth key strategic decision made by ADEMI. The incentive system was designed above all to improve loan recoveries but also to balance two other important objectives, namely outreach and cost-efficiency. In particular, credit advisors' bonuses are linked to the recovery rate on their loans, the number of clients they serve and·the size of the portfolio they administer. In the early days of individual lending, each advisor had on average only 45 clients, and 60 was considered an ambitious target. Today each advisor has on average almost 140 clients! The incentive system was clearly used to gre~t effect to ratchet up the productivity of credit advisors. These individual incentife payments 'for credit advisors are in addition to annual bonuses for all staff that depend on the Association's overall performance. This system has contributed significantly to boosting staff morale, dedication and loyalty to the Association, as well as productivity and performance. To ensure the success of new branches, ADEMI decided in 1987 that it would not open offices in new areas unless it was specifically invited to do so by respected members of the community and felt that it could succeed. ADEMI's branches have generally been opened at the invitation of the local Chamber of Commerce, often in facilities provided by the chamber. This policy has been very successful in ensuring community support for loan collections by ADE:MI, with only one exception: ADE:MI opened a branch in the town of Barahona, but had to close it shortly thereafter when influential members of the community sided with delinquent borrowers, thereby creating an atmosphere that was not conducive to successful lending. In two other 'cases, s ADE.MI issues publicly available quarterly reports drawing on its information system. These reports contain data of interest to both donors (e.g. number of women clients), staff (e.g. credit advisors' portfolios) and financial data of interest to a broad range of readers (e.g. financial statements). 6 This problem has been addressed fairly successfully since i.996 by splitting the credit and administrative ftmctions of the overworked Director for the Interior and appointing a second Director for the Interior with a focus on credit issues. 7 Clients who were not ready for individual loans, in ADEMI's judgment, were passed on to the Dominican Institute for Integral Development, or IDDI, which bought their loans from ADEMI and/or provided follow-mi loans when the ADE.MI loans expired. SUSTAINABLE BANKING WI1H THE POOR xi branches proved to be financially unviable and, just as in the case of Barahona, ADEMI did not hesitate to close them. 8 The period 1987-88 was a period of consolidation. Most of the portfolio expansion came in the form of repeat loans to existing clients, rather than new loans. Indeed, the number of new loans declined by 21 percent in 1988. The main reason was a shortage of funding and the need to ensure that existing clients would have ready access to follow-on loans. In addition, the incentive system in place at that time encouraged credit advisors to play it safe and go for bigger loans once they reached the target of 90 clients, rather than going for new, smaller and more risky clients. Finally, the period of consolidation was needed following the rapid expansion of the network in the previous two years. Experience in other institutions has shown that uncontrolled growth can quickly undermine the institution. ADEMI was in danger of going the same way, but was able to make the necessary adjustments during this period before expanding further. This was particularly important in a period of macroeconomic stress with high inflation rates and (by 1990) declining real GDP. Indeed, even though ADEMI reported a positive net income before grants for the first time in 1988, that same year ADEMI experienced negative real growth in its equity for the first and only time in its existence, as growth in nominal equity failed to keep pace with rising inflation. The years 1989 to 1991 were years of macroeconomic crisis for the country. Eight out of 23 commercial banks collapsed, as did several development banks, mortgage banks and finance companies. However ADEMI weathered the storm and by 1991 it was subsidy-free, one of the only two organisations in the world dedicated exclusively to microfinance to be in that position at the time. 9 It did this, first, by continuing to charge positive real rates of interest on its loans even as inflation rose to 60 percent in 1990. 10 Second, ADEMI cut costs by slowing down the growth in the number of new branches, using up excess capacity, providing incentives for productivity increases and making larger loans. Indeed, while the average outstanding loan amount per client fell in real terms in 1990 to RD$2,580 (US$230), it rose by almost 60 percent in real peso terms in 1991 and by 20 percent per year thereafter (see Chapters V and VII). As of' 1991, after an exceptionally profitable year, 8 In the case of Barahona, ADEMI returned several years later (1996) and opened a successful branch. 9 The other institution was .Bank Rakyat Indonesia This assertion is bas'ed on data available on the leading microfinance providers outside the credit union movement (for which no full data are available, but no similar success is reported). Toe policy of positive real interest rates was essential, however, when inflation collapsed in 1992 as the result of an IMF-assisted program, ADEMI, like other financial institutions, allowed nominal rates to lag at higher levels. The exceptionally high real rates (over 40 percent) earned ADEMI considerable income, but also contn"buted to the largest number of non-returning clients in its history. ADEMI's real rates did fall thereafter, and it must be said that clients have very rarely complained about ADEMI' s interest rates (e.g. less than 6 . percent of borrowers in a sample by Gomez [1989)). However, organisations seeking to emulate ADEMI's experience cannot presume that the market will bear exorbitant rates to cover inefficient operations. SUSTAINABLE BANKING WTIH THE POOR xii ADEMI had 8,400 clients, 135 employees, US$5.2 million in assets, US$2.3 million in equity, a net income ofUS$1.6 million and an SDI of minus 16 percent. 11 GRADUATION AND NEW DIRECTIONS By 1991, ADEMI had embarked on the fourth phase of its development. This phase was driven by yet another key strategic decision. Many microfinance lenders have struggled with the issue of the "graduation" of their borrowers.. In some cases, e.g. Grameen Bank, graduation ceremonies were held for clients who had grown beyond the lender's target group, only to find that there was nowhere for these clients to graduate to. Successful repayment of small loans, and even the presence of a guarantee fund for graduates (as in ADEMI's case), did not tum them into attractive clients for commercial lenders. ADEMI therefore decided that it would not be the ciients who would graduate, but rather ADEMI itself. It est~blished a window for small-scale as opposed to micro loans so that it could grow with its clients. It also began to lend to new clients that joined ADEMI as small- to medium-scale firms. The small-scale lending grew rapidly after 1992, when ADEMI received an ECU 3 million (US$3 .45 million) loan from the European Investment Bank, which could only be utilised for small business loans in excess ofUS$50,000 each. This was followed by another EIB loan in 1995 ofECU 8 million (about US$10.5 million). J.'!\ ,. ')J!,o,,.·, ' i No restrictions were placed on the ·, use of repayments on these small businesl-loans, which:were therefor~ used to finance new lending at the micro level. In fact, the number of micro clients and the volume of lending to them continued to increase. However, in terms of portfolio_,expansion, small-scale lending began to dominate after 1992, e.g. accounting for four-fifths of the increase in portfoli.o in 1994. 12 This contributed to a sharp decline in administrative costs over outstanding loan portfolio, from 35 percent in 1990 to under 8 percent in 1996. 13 As of June 1997, small-scale lending accounted for 34 percent of the loan portfolio, although only 175 enterprises or about 1 percent of ADEMI's 16,200 active clients were classified as small- to medium-scale. The 1990s are marked by innovations in ADEMI. In 1992, ADEMI introduced a pension plan for its employees, a defmed benefit scheme that has been rated as actuarially sound by external actuaries. In 1993, it opened its small enterprise lending window at the central office in Santo Domingo. In 1994, ADEMI piloted home improvement loans to seasoned clients. These are medium term loans that enable clients to (re)build a part of their home at a time, e.g. a new roof or brick walls. Repeat loans effectively allow them to construct a full house. In 1996, 11 A negative SDI means that ADEMI coulci have reduced on-lending rates and still have remained fully viable- its lending rates were about 1.16 times higher than needed to yield a return on equity.equal to the opportunity cost of capital, with all costs adjusted to market or imputed opportunity values. • 12 While loans to small businesses accounted for only I% of new loans and 23% of pesos of lending approved in 1994, the fact that small business loan terms averaged around three to four years rather than ten months (see Table 4.2) meant that they grew far more rapidly as a share of the portfolio. This explains why they accounted for 80% of the increase in outstanding loan portfolio in 1994. . 13 Interestingly, ADEMI reached self-sufficiency before embarldng on BIB-funded lending in 1992. SUSTAINABLE BANKING WITII THE POOR xiii ADEMI, in conjunction with Banco Popular, issued the ADEMI Mastercard. This innovative move is an important step towards fulfilling its mandate of improving access to credit for low- income entrepreneurs. It also provides a vehicle for greater future automation of its loan repayment and disbursement procedures. It again placed-ADEMI at the forefront of microfinance. Growth and success has bred competition. •In particular, in certain cities commercial banks have bid away some small- and medium-scale enterprises from ADEMI with better terms and a broader range of financial services. The USAID-supported Foundation for the Promotion ofMicroenterprises, FondoMicro, which was originally established a5i-an apex institution for microfinance, also opened a formal Bank for Small Enterprises that is engaging in direct lending to small-scale enterprises. ADEMI responded to the increasing competition late in 1996 by engaging local consultants to assist with strategic planning. The big question was whether to continue operating as a non-profit organisation or to formalise ADEMI by purchasing, converting into, combining· with, or establishing a formal bank. Several factors other than increasing competition drove ADEMI to review its status. First, in 1992 it had begun to mobilise credit from clients in the form of bond instruments. While these loans are permitted under the commercial code, a stricter interpretation by the monetary authorities might lead them to be viewed as deposits and therefore lead to a call for ADEMI to become regulated. These bonds currently account for about one- quarter of ADEMI's borrowings (or 11 percent of total resources). Second, the Dominican Republic's monetary code is being revised in a way that would allow the Banking Superintendency to regulate any institution whose primary activity is financial intermediation, even if it does not mobilise deposits. Third, reforms in the Dominican fiscal code have eliminated some of the fiscal advantages of non-profit status under Law No. 520. Fourth, ADEMI's unprecedented success in accumulating capital has led observers to question whom the capital of this non-profit truly belongs to, and led the President and Executive Director to look for ways of ensuring that it is properly managed when they eventually retire. Furthermore, from a regulatory perspective, there is a question as to whom creditors and regulators could tum to for ~dditional capital injections should the need arise-if the answer is "no one", ADEMI's risk rating could increase as ADEMI's creditors classify their assets in line with Basie guidelines. LOOKING TO THE FUTURE WITH BANCOADEMI The fifth phase of ADEMI' s development has just begun. In another key strategic decision (the seventh according to this overview), ADEMI had decided to continue operating as a non-profit but also to establish a new formal sector bank. On September 11, 1997, the Monetary Board approved a banking licence for BancoADEMI as a development bank. BancoADEMI opened its doors to the general public on January 2, 1998. It is introducing a full range of banking services for clients, except for checking accounts and certain foreign exchange- related services. ADEMI is the largest single shareholder of BancoADEMI (with 24 percent of shares), but other shareholders include staff and members of the Board of Directors, as well as the European Investment Bank (see Chapter VIII). While ADEMI originally planned to transfer its entire portfolio to BancoADEMI over a three-year transition period, it has greatly accelerated SUSTAINABLE BANKING WITH TIIEPOOR xiv that schedule, so that almost all qualifying ADEMI clients will have been transferred to the bank by the end of 1998. ADEMI is still deciding on the type of service to provide to clients who are too costly to serve under the prudential regu~ations applicable to formal banks. Much can be learned from ADEMI's exceptional success and also from its mistakes. Microfinance providers that have been fairly successful in their own right, such as the Ecuadorian Development Foundation (FED), the Alexandria Business Association (ABA) and the MultiCredit Bank of Panama, have drawn considerable insights from ADEMI's operations. Even before graduating into the banking sector, ADEMI had matured into a confident institution with a larger net worth than that of five of the 15 commercial banks in the Dominican Republic, and than all the development banks in the country (see Tecnicas de Planificaci6n [1997]). At the end of 1996, ADEMI had 200 employees serving over 16,200 clients, and US$39m in assets, US$15 million in net worth, and a net income ofUS$3.8 million for 1996, and continued to be self-sufficient with an SDI of less than 0.1 percent. In reaching the threshhold of the formal financial sector, ADEMI made a number of key strategic decisions, but none of them was more important than the original decision to adopt a businesslike approach in all its activities in order to fulfill its social objective of "democratising credit". The first part of this study places ADEMI's operations in the context of the history and culture of the Dominican Republic, the broader macroeconomy, and the "micro" economy that ADEMI serves. Thereafter, the study ex~µ:ies mor~ ~losely what ADEMI is about (including the products it offers), whom it serves (that is, key feattires of ADEMI's clientele), and how ADEMI operates (its structure, funding and lending procedures). Finally, the study considers how ADEMI has performed both in terms of its own financial results and of the results of its clients, and draws some conclusions on why ADEMI has succeeded, before closing with look a at new issues and constraints that (Banco)ADEMI faces as it looks to the future. SUSTAINABLE BANKING WITII THE POOR xv ' ' ••\'•;:N, , ,• CASE STUDIES IN MICROFINANCE . THE DOMINICAN REPUBLIC - ADEMI I. THE DOMINICAN REPUBLIC: PAST AND PRESENT 1-- The Sustainable Banking with the Poor (SBP) Project was launched by the World Bank in April 1994 as a joint effort between the Rural Finance Team in the Bank's Agriculture Department and the Gender and Poverty Team in the Bank's Asia Region. The purpose of the study is to improve the ability of donors, governments and practitioners to design and implement policies and programs to strengthen the financial sector and build sustainable financial institutions that effectively reach the poor. The core of the SBP study is a series of field-based case studies of institutions in Asia, Africa and Latin America that have pioneered innovative approaches for providing financial services to a significant number of low iilcome clients, The Association for the Development of Microenterprises (ADE:MI) was selected for a case study because of its exceptional success in providing financial services to micro- and small-scale enterprises in the Dominican Republic. This study provides a compreh~nsive overview of the context in which ADEMI operates, the methodology and performance of the Association, and future prospects for ADEMI as it stands on the threshold of becoming a formal financial institution. SNAPSHOT OF THE DOMINICAN REPUBLIC The Dominican Republic covers the eastern two-thirds of the Caribbean island of Hispaniola, which it shares with Haiti. Around 7.8 million Dominicans live in the country's 49,000 square kilometers of rugged highlands and fertile valleys, making the Dominican Republic one of the most densely populated countries in the Caribbean. With a GNP per capita of US$1,600, the Dominican Republic is considered a lower- middle-income country. Agriculture currently accounts for about 15 percent of economic activity, down from 22 percent in 1975. Industry's share of the economy has also shrunk to 22 percent as services, notably tourism, have boomed. Sugar and textiles, especially from export processing zones, are key exports, while fuel, energy an,d capital goods are the primary imports. The country's natural resources include nickel, bauxite, gold and silver. The country suffers from widespread poverty, notably in rural areas, where the State owns close to half of all arable land. Growth in agriculture averaged only 0.6 percent per year SUSTAINABLE BANKING WI'nI THE POOR 1 during 1985-95. This has fueiled rapid migration to urban areas (which now accou_nt for 55 percent of the population), especially to the capital city of Santo Domingo. About 70 percent of the urban population live in barrios or low-income neighbourhoods, Roughly 20 percent of the economically active population is openly unemployed and another 25 percent is estimated to be underemployed. World Bank data suggest that 21 percent percent of the population falls below the poverty line (see Chapter ill). The lack of formal employment opportunities has led to significant participation in the informal sector as well as to emigration to the United States, Venezuela and other countries. The economic stagnation of the 1980s (see Chapter II) arrested the long-term trend of improving social conditions. Life expectancy at birth is 71 years and the infant mortality rate is about 36 per 1000 live births, which is comparable to other lower-middle-income countries. Gross primary enrollment ratios are about 95-99 percent. However, almost 40 percent of the population lack ready access to safe water, resulting in a high prevalence of water-borne diseases. Moreover, maternal mortality rates are among the highest in the region, and 30 to 40 percent of childreµ under six suffer from moderate to severe malnutrition. Thus, there are two sides to this lush tropical country. On the one hand there are the colonial monuments, rainforests and beautiful beaches that attract more than a million tourists a year to the Dominican Republic. On the other, there is widespread poverty and a highly unequal distribution of income, largely as a result of political-economic factors including pervasive State interventions in goods, services and property markets, that have held back the economy (see Chapter IT). It is the marginalised populations that ADEMI has tried to reach with its financial services (see Chapters ill and V). WHO ARE THE DOMINICANS? Before the Spanish occupation in 1492, Arawaks inhabited the island, engaging principally in fishing and farming. 1' The Spanish introduced colonial sugar plantations that exploited native labour. Slaves were later imported from Africa to take the place of the Arawak laborers when fighting and disease led to the extinction of the native population. Spain later ceded Hispaniola to France in two parts, first the western portion (n_ow Haiti) in 1697, then the eastern portion (now the Dominican Republic) in 1795. Between 1809 and 1844, Dominicans rose in revolt against successive occupations by France, Spain and Haiti. On February 27, 1844, the Dominican Republic declared its independence from Haiti. Over the next several decades it subsisted largely on sugar exports, which it used to pay for imports from industrialising countries in Europe and the United States. Increasing indebtedness to several European nations led to a treaty with the United States in 1906, under which the United States administered the customs department and paid off the Dominican Republic's debts to external creditors. Internal disorder. led to the US Marines establishing a military government in 1916, but by 1924 a constitutional government had resumed control and the American occupation ended. In 1930, General Rafael Trujillo was 14 S~e Funk and Wagnall's Coiporation [1994). SUSTAINABLE BANKING WI1H THE POOR 2 elected president and established a ruthless dictatorship for the next 31 years. During this period, characterised by what the Organisation of American States termed "flagrant and . widespread violations of human rights," there was significant investment in public infrastructure, health facilities and housing projects. These were financed to an important extent with state revenues from public sugar estates, a colonial legacy that continued to drive the Dominican economy. That era ended with Trujillo's assassination in 1961. The presidential elections of 1962 were followed by further unrest, a military coup in 1963, a provisional government, and elections in 1966, which were won by Joaquin Balaguer. • High sugar prices, tourism and foreign investment boosted the economy in the late 1960s and 1970s, but a sharp decline in world sugar prices in the mid-1970s and a devastating hurricane in 1979 dealt major blows to the economy. Balaguer resumed the presidency in 1986 after terms by two other presidents (Guzman and Blanco) who had attempted to purge the military of right- wing elements and reform the economy. The protest riots of 1984 and 1985 ended the attempts at economic reform until 1990, when the Dominican Republic plunged into a deep economic crisis (see Chapter II). Following an election that was widely regarded as fraudulent in 1994, Balaguer bowed to popular protests and stepped down after only two years. In the elections of 1996, Leonel F emandez was elected president. Thus the Dominican Republic is a blend of cultures that, like most of Latin America, is heavily influenced by Spanish traditions . .The Domini~ans have had a long history of political unrest, of state control of the commandinlheights or"th~ economy, and of governments that have been largely controlled by the military. These political factors largely explain persistent economic distortions, the skewed distribution of wealth, and:the marginalisation of a large share of the population from the economic life of the country. These historical, cultural and political traditions also serve to explain ADEMI's mission of"democratising credit". In the words of Camilo Lluberes, ADEMI's founder and president: "Marginalisation and lack of opportunity give rise to economic and social disequilibrium which endanger the democratic stability of the country. We and all those who want. to ensure the democratic development of Dominican society have to contribute to reduce this disequilibrium." (Lewin [199lb]). SUSTA.INABLE BANKING WITH THE POOR 3 ' .-, -•;\ ·i~....... • ' -:, ) .r:~_:1 ~ . •~ 11. THE SHIFTING MACROECONOMIC AND FINANCIAL SECTOR CONTEXT UPS AND DOWNS IN THE MACROECONOMY The past fifteen years have been turbulent for the Dominican Republic. The poor have been particularly affected. During 1980-1990, real income per capita declined by 19 percent, while the average real minimum wage fell by 38 percent. An attempt to address the country's deepening economic crisis with IMF support failed in 1984 when the austerity measures, including a sharp devaluation and decreases in food and fuel subsidies, prompted rioting throughout the country. The government reverted to expansionary fiscal and monetary policies in 1985, financing rising public sector deficits by printing money. The foreign exchange market was controlled, leading to high shadow exchange rates of Dominican pesos per US dollar, and the Government allowed arrears to accumulate on its foreign debt servicing. Following a short-lived revival in the economy in 1986 and 1987, inflation soared to unprecedented levels, reaching 100 percent in 1990. That year saw capital flight rise to 6.5 percent of GDP, arrears on foreign debts amount to US$1.4 billion and foreign creditors suspend disbursements. Significant food and fuel shortages ensued, electricity production virtually collapsed and the financial system was thrown into crisis. Real GDP fell by over 5 percent and open unemployment reached 25 percent. The only sectors that performed well throughout the period of macroeconomic instability were tourism and export processing zones (EPZs). 15 For example, during 1983-88, value added in tourism grew by 23.5 percent while the overall economy grew at 2.5 percent. Even during 1989-93, with GDP growing on average at only 1.6 percent, tourism grew at 6.4 percent. The even more remarkable performance of the EPZs, which were insulated from the broader economy, is reflected in Table 2.1. While these sectors point to the Dominican Republic's competitiveness and have supported the economy in important ways (e.g. through labour- intensive production technologies and foreign exchange earnings), they cannot pull most Dominicans working in the broader economy out of poverty. To achieve that, ~e rest of the economy needs to reform. Export Processing Zones 23.5 16.4 19.0 6.8 Rest of Economy 1.8 3.1 2.2 -0.9 Source: The World Bank [1995] 15 The export processing zones are free trade zones that engage in the manufacture and assembly of non- traditional Dominican exports, notably textiles for the U.S. market. SUSTAINABLE BANKING WI1H THE POOR 4 A more durable reform process began in 1990 when the government launched stabilization measures as part ofihe New Economic Program. The program focused on fiscal and monetary restraint, exchange rate unification, elimination of large arrears c:·, foreign debts, accumulation of international reserves, liberalisation of trade and deregulation of financial markets (World Bank [1995]). This program was designed without external support and was a major success. Inflation fell from a peak rate of 100 percent during 1990 to 5 percent by late 1991, the government obtained a cash surplus as a result of lower investment expenditures and higher tax revenues, the exchange rate was unified, the Dominican Republic paid off its arrears on official aid loans, capital flight was reversed, and in 1992 the economy grew by 8 percent (see Table 2.2). In addition to the stabilisation measures, the government implemented structural reforms in the financial sector , trade reforms (including tariffication of non-tariff barriers and reduced effective rates of protection), rationalisation of prices ( e.g. of electricity), as well as reforms in the areas of taxes, education and the public sector. • Table 2.2: Movements in Macroeconomic Indicators Real Growth in GDP (in -22 2.2 4.4 -5.4 1.0 8.0 3.0 4.3 4;7 7.4 RD$,%) Inflation (%) 37.4 9.8 15.9 44.1 45.5 59.5 53.9 4.5 5.3 8.3 12.5 4.0 Exchange Rate (RD$/US$) 2.94 3,08 4.96 6.34 • 6.34 11.35 12.66 12.58 12.77 13.06 13.47 14.06 Current Account (Deficit) or ~~i:~: 1_:. • Smplus (US$ Millions) , (108) (183) (364) (19) (327) (2!10) (157) (708) (447) (68) (125) (517 Ratio of Reserves to hnports (months) 3.2 3.3 1.4 1.9 1.0 0.4 3.1 2.8 1.7 0.6 0.9 0.8 Ratio of Govt Expenditures 11.7 12.5 13.4 14.1 13.4 10.9 9.5 - 16.3 14.4 14. to GDP(%) _, ..... Ratio of Smplus or (Deficit) (1.3) 0.7 (0.3) (02) 0.3 0.6 1.1 - (0.5) 1.1 0.3 to GDP(%) GrowthinM2 16.2 74.0 16.7 50.7 30.5 39.1 37.6 27.9 26.3 9.6 17.5 21.2 Market Reference Interest 20.1 18.4 19.3 26.7 27.6 36.4 30.0 25.0 22.5 20.8 23.5 21.0 Rate(%) Source: Appendix 1 ,/,,,,,,/ After steady growth (3 percent) and low inflation (5 percent) in 1993, the reform program was suspended in 1994 due to the elections. Increased public spending, financed by printing money, led to renewed inflationary and exchange rate pressures. The central bank sold almost US$500 million in reserves to prevent a slide in the nominal exchange rate prior to the elections. However, a resumption of tighter macroeconomic policies in late 1994 improved the balance of payments and GNP growth accelerated. In 1995, the public surplus amounted to RD$1,400 million (US$104 million) or just over 1 percent of GDP, while in 1996, notwithstanding the national elections, growth surpassed 7 percent and inflation w~ held at 4 percent via a tight monetary policy. Interest rates remain high, with central bank certificates selling at 18 percent in 1996. These macroeconomic trends have had mixed effects on ADE.MI and on its clients. For example, the combination of inflation and unemployment in the late 1980s eroded real wages and depressed demand for ADEMI's clients' products, particularly in areas such as furniture SUSTAINABLE BANKING WITII THE POOR s sales and other larger expenditures. Inflation also eroded ADEMI's capital in 1988 and obliged it to raise nominal rates significantly in 1989 and 1990. On the other hand, ADEMI's nominal lending rates declined more slowly than inflation after 1990, yielding significant profits for the Association. These and other effects of the macroeconomic context are discussed in greater detail in Chapter VII (see Appendix 1 for further macroeconomic data). WHO'S WHO IN THE FINANCIAL SECTOR In 1947, the Dominican Republic established the Central Bank and the Banking Superintendency. 16 Subsequently, the Insurance Superintendency and the National Housing Bank were also established. These four agencies are responsible for regulating the Dominican financial sector (see Figure 2.1 ). In particular, the Monetary Board of the Central Bank determines monetary, credit and exchange rate policy to be implemented by the Central Bank. The Banking Superintendency ensures adherence to banking laws by banking institutions, while the Insurance Superintendency does the same for regulated insurance providers. Finally, the National Housing Bank (BNV) acts as an apex institution for the country's savings and loans associations. Financial inte~ediaries in the Dominican Republic can broadly be divided into • monetary or non-monetary institutions, where the former can mobilize sight deposits and therefore create money. Monetary institutions are limited to commercial banks, including the state-owned Reserve Bank. Together with the Central Bank they make up the banking system. Commercial banks are regulated under the General Banking Law of 1965 (No. 70~). Non-monetary institutions can in tum be -classified into banking institutions or non-bank financial intermediaries. Banking institutions can mobilise term deposits and are regulated by the Banking Superintendency. They include savings and loan associations, mortgage banks and development banks. 17 By contrast, non-~ank financial intermediaries are allowed to intermediate financial resources but cannot mobilise deposits, are not governed by banking laws, and are not regulated by the Banking Superintendency. They must, however, adhere to the broader commercial codes applicable in the Dominican Republic. 16 It also created the national currency, the Dominican Peso, and passed commercial banking laws. 17 Savings and loan associations are private non-profit mutual institutions and thus do not have authorized capital. Mortgage banks, governed by Law No. 171 of 1971, are designed to support the construction industry by providing financing for industrial par.ks, offices, and other commercial developments. With rapid industrial growth in the 1960s, development banks were established under Law No. 299 of 1966, with a view to providing medium to long term loans for production, mining, transport and related activities. These used to be referred to as Finance Companies but their renaming as banks was authorised in 1982 by the Monetary Board. Other institutions in this grouping include the Irufostrial Development Corporation (CFI), the Agricultural Bank and the Institute for Development and Cooperative Credit (IDECOOP), which assists cooperative societies . •SUSTAINABLE BANKING WITH THE POOR 6 Figure 2.1: Who's Who in the Financial System in the Dominican Republic Regulatory Institutions Central Bank Banking Superintendency Insurance Superintendency • National Housing Bank Financial Intermediaries Monetary Institutions Non-monetary.Institutions Central Commercial Banking Non-bank Financial Bank Banks Institutions Intermediaries •Mortgage Banks •ADEMI •Development Banks •INVI •Savings and Loan •FDD Associations •ODC •Others •Instirance companies · •Others Source: Jimenez {1997] ADEMI, together with a series of government-owned financial institutions, NGOs and insurance companies, fall into the category of non-monetary, non-bank financial intermediaries. 18 • 18 Other agencies in this category include the National Housing Institute (INVI); the Community Development Office (ODC), which is the product of an agreement with USAID and is used to channel financial resources for capital investments to farme,s, small industries and other groups with limited access to the banking system; and the Dominican Development F01mdation (FDD), a private non-profit organisation designed to establish associations of small farmers and offer them technical and financial support. .Note that these institutions are not regulated, although clearly in the case of insurance companies, whichfall in the same category, more specific insurance laws apply and the companies are regulated by the lnsmance Superintendency. SUSTAINABLE BANKING WITII THE POOR 7 FINANCIAL SECTOR CRISIS AND. REFORM . During the late 1980s and early 1990s, there was a major financial crisis that led to the collapse of a number of financial institutions (Fis), including commercial, development and mortgage banks as well as scores of finance companies. Eight of the 23 commercial banks in the Dominican Republic, accounting for 20 percent of coinmercial hanking assets, collapsed during 1988-91. The rising inflation rate, exchange rate instability and capital flight led to significant demand for liquidity which could not be met by a range of financial intermediaries with funds concentrated in unproductive assets such as speculative investments and connected lending. Whereas the Banking Superintendency _had previously seen its role as being limited to ensuring that the reserve requirements on deposits were duly adhered to, the collapses underscored the absence of genuine preventive mechanisms to preempt these types of crises. The illiquidity of the institutions quickly led to their ceasing to make payments and to insolvency and closure. 19 Beginning in the early 1990s, reforms were introduced with a view to increasing the solvency and capital adequacy offmancial institutions, achieving greater transparency, strengthening supervision of Fis and increasing openness and competition in the sector. For example, in January 1991 interest rates were liberalised and a single reserve requirement (of • 20%) was established for all institutions. In June 1993; a Monetary Board ruling introduced new banking norms for classification of assets, (modified slightly in a September 1994 ruling). In October 1994, the Board established a standard format for the fmancial statements of all intermediaries. Other rulings established operating and accounting instructions for the valuation of collateral, treatment of contingent liabilities, provisioning and the determination of connected lending. ·1n September 1995, issuers of credit cards became subject to banking norms, including with regard to provisioning and disclosure. Notwithstanding these reforms, further strengthening of the Banking Superintendency is required, a proposed monetary law codifying these rulings has yet to be approved by Parliament (see below), and the Central Bank's handling of the failed Bancomercio raises questions about the monetary authorities' willingness to implement prompt corrective actions when rulings are not adhered to by key players. 20 The most important element of the reforms is a refocusing of banking regulations and the attention of the Banking Superintendency from adherence to reserve requirements on deposits, to maintaining the quality ofloan portfolios. As of 1993, banks were required to classify all assets by risk categories and to make adequate provisions for loan losses for each class of assets, 19 In the case of development banks, the termination of fiscal incentives that had made their deposits attractive to clients worsened the situation by reducing demand for their deposit services. 20 Notwithstanding serious liquidity problems for several years, as well as evidence of connected lending and questionable disclosure practices, Bancomercio (formerly the second largest private bank in the D.R.) was kept afloat with massive nm.ding from the Central Bank (estimated at over RD$1. l billion, or one-third of Bancomercio 's liabilities) until shortly before the 1996 elections, when the Central Bank took over the institution. To date, it app.ears that the State, rather than shareholders, have taken the major losses (see Calvo [1996]). SUSTAINABLE BANKING WITH THE POOR 8 • beginning with the largest clients and covering the entire portfolio within three years. Based on the new norms, an initial assessment in 1994, covering 99 percent of assets of banking institutions, found loss risk ratios averaging 9.4 percent (Tecnicas de Planificaci6n [1997]). In addition, in a departure from generally accepted accounting principles (GAAP), banks have been required since 1992 to recognise interest income on a cash basis, rath,er than an accrual basis. The reasoning has been that improperly applied accrual accounting has reduced transparency and allowed banks to overstate their returns. Both these rulings (on asset classification and income recognition) have had major implications for ADEMI. First, ADEMI continued to record interest income on an accrual basis, in accordance with international GAAP. When it changed its status from a non-bank financial intermediary to a regulated institution it had to conform with Dominican banking norms. Second, the classification of commercial loan assets places significant weight on collateral and client quality, as reflected in audited financial statements. Loans for which there is inadequate collateral or documentation may be classified as class C loans of real or significant risk, requiring loan loss provisions of 10 to 20 percent, regardless of the client's repayment record. While asset classification in the case of personal loans places greater emphasis on repayment records than on collateral and client quality, the Superintendency has to date not recognised repayment records built up outside the banking system. This has been a major blow for ADEMI's smaller clients (see Chapter VIII). Finally, a new monetary code under debate in the Parliament would expand the potential regulatory auth6'rity of the Banking Superintendency to include institutions such as ADEMI, whose primary function is financial intermediation (even if they do not mobilise deposits). These and other issues related to further proposed financial reforms, as well as ADEMI' s reactions to the reforms, are taken up again in Chapter VIII on ADEMI's future. ADEMl'S PLACE IN THE FINANCIAL SYSTEM As a result of the reforms, the scope of lending for commercial banks has been broadened to include categories previously reserved for specialised banks (i.e. mortgage banks and development banks). This has led to several mergers and rationalisation of the industry. Indeed, between December 1994 and June 1996, the number of intermediaries declined from 273 to 240. The fifteen commercial banks have increased their share of total assets in the financial system ( excluding insurance companies), whereas mortgage banks and development banks have declined in importance (see Table 2.3). SUSTAINABLE BANKING WI1H THE POOR 9 Table 2.3: Total Assets of Financial Intermediaries in RD$ millions Commercial Banks 36,231 40,609 44,681 52,224 Development Banks 3,108 2,970 2,168 1,951 6% 5% 3% 3o/c Mortgage Banks 770 662 679 140 1% 1% 1% 0o/c Savings and Loans 7,175 9,147 10,029 11,727 13% 15% 15% 16o/c Finance Companies 3,999 4,296 4,081 4,244 7% 7% 6% 6o/c Other private lenders 157 145 141 189 0% 0% 0% 0o/c Subtotal 51,440 57,829 61,779 70,475 93% 93% 93% 98o/c Public: BNV & CFI 3,928 4,616 4,724 1,600 7% 7% 7% 2o/c Total 55,368 62,445 66,503 72,075 100% 100% 100% l00o/c Source: Tecnicas de Planificacion [1997] ADEMI's assets (which are not included in Table 2.3) as of 1997 exceeded those of two of the 15 commercial banks in the Dominican Republic (see Tecnicas de Planificaci6n [1997]). Moreover, ADEMI's loan portfolio, which grew by an average of 46 percent per year during the 1990s, has surpassed that of all 16 development banks, both mortgage banks, 14 out of 19 savings and loans associations and two of the 15 commercial banks. As of September 1996, ADEMI's capital was RD$191 million (US$14 million). By June 1997, ADEMI's net worth had grown by over RD$50 million, comfortably placing ADEMI among the ten largest financial intermediaries in the country, ranked by net worth (see Figure 2.2). 21 21 With the transfer of assets to BancoADEMI in 1998, ADEMI's assets clearly shrunk while the new bank's assets grew rapidly. As a financial group, ADEMI and BancoADEMI have retained their preeminent position among development banks, although a clearer picture of the group's financial position will only emerge with a consolidated financial accounting for 1998. SUSTAINABLE BANKING WI1H THE POOR Figure 2.2: ADEMI's Net Worth is Comparable-to That of; ommercial Banks in the Dominican Republic (Figures in Millions of Pesos as of September 1996) 1,200 1,000 Ill 0 Ill - CII Cl. 800 0 Ill C ~ 600 SJ .5 .c 400 ,:: 0 :: - CII z 200 (/) :a 0 ::E ... [!I 3 0 :r: (/) c w .g ~ cu O ~ Cl) 0::: c. 0 Ill a.. m e a.. o._ Cl) ::E c( z u:: CP Jj Source: Tecnicas de Planificacion [1997] Clearly ADEMI has become an important player in the Dominican financial systiem, and its review of its status as a non-bank financial intermediary is timely (see Chapters VII and VIII, respectively, on ADEMI's performance and future prospects). • SUSTAINABLE BANKING WITH THE POOR 11 111. THE DOMINICAN "MICRO" ECONOMY DEFINING "POOR", "INFORMAL" AND "MICRO" In 1995 the World Bank carried out a poverty assessment of the Dominican Republic. Drawing on an income-expenditure survey conducted by the Fundacion &onomia y Desarrollo (FED) in 1992 (see Dauhajre et al. [1994]), the World Bank defined a consumption basket that would provide a minimum caloric requirement as well as basic non-food needs. The cost of the food basket and additional basic needs was used to define the poverty line, while the cost of the food basket alone was used to determine the extreme poverty line. In 1992, these two cut-offs were, respectively, US$30 per month and US$20 per month per person at the national level, with figures that were somewhat higher (lower) than these cut-offs in urban (rural) areas. The assessment found that one in five Dominicans lived in poverty and one in ten Dominicans lived in extreme poverty in 1992. A large proportion of the urban poor in the Dominican Republic are employed in the informal or semi-formal sector, i.e. in enterprises that may or may not be legally registered and that do not adhere to all the requirements of the commercial code, labour laws, environmental laws, the fiscal code, etc. Many of these informal enterprises are microenterprises, i.e. enterprises engaged in non-crop production or trade activities that employ between one and ten paid or unpaid, employees, including the proprietor. Not all microenterprises are informal (e.g. small consulting firms) and not all informal firms are microenterprises (e.g. informal small-scale enterprises with 10-50 employees). 22 Finally, as we shall see in Chapter VII, many informal enterprises are successful and raise their owners and employees out of poverty, so that not all these entrepreneurs are poor. • FEATURES OF POVERTY Poverty in the Dominican Republic is characterised by the following key features (see Appendix 2): • It is more prevalent in rural areas (30 percent) than in urban areas (11 percent). • It is disproportionately high for children (27 percent compared to 21 percent for the total population), because the poor tend to have larger families. Indeed, almost one in five children suffered from chronic malnutrition in the Dominican Republic in 1991. • There was a significant gender imbalance in the mid-1980s-23 percent of women were poor, compared to 15 percent of men. This imbalance was removed and even slightly reversed by 1992 (16 percent for women and 17 percent for men), largely due to increased 22 Based on a survey of Dominican households flmded by FondoMicro, Dauhajre [1992] suggests that most microenterprises in the Dominican Republic are informal, but that only about two-thirds of informal sector activity in the D .R is accounted for by microenterprises. Overall. the informal sector is estimated to account for one-third of economic activity in the Dominican Republic. SUSTAINABLE BANKING WI1H THE POOR 12 female participation in the labour force in the export processing zones. However, almost one in ten mothers still faced chronic caloric deficiencies in 1991. The overall prevalence of poverty in the-Dominican Republic depends heavily on movements in real wages. The poorest 20 percent of the population derive almost 60 percent of their income from wages, compared to only 31 percent of income for the richest 20 percent of the population. According to FED surveys, when real wages decreased by about 16 percent in 1986-89 (as inflation outpaced nominal wage growth), poverty increased by 34 percent and extreme poverty by 31 percent. During 1990-92, inflation was brought back under control, real wages increased by 29 percent and poverty fell by 16 percent, with extreme poverty falling by 34 percent. These movements in real wages are also reflected in the trends in the distribution of income in the Dominican Republic between 1986 and 1992, as measured by the Gini coefficient. 23 The coefficient rose from an already relatively high value of 0.47 in 1986 to 0.55 by 1989, indicating an even higher concentration of income among the relatively wealthy, before falling back to 0.49 in 1992 (see the movements in the distribution of income in Figure 3 .1 ). Figure 3.1: Gini Coefficient Measures for the Dominican Republic 100.0 CII 80.0 ~ 0 .5 60.0 _ _ _ 1986 ii C i z 40.0 _____ 1989 _. _. _ . ___ 1992 'o 20.0 c CII e CII 0.0 D. -20.0 Percent of Population (Poorest to Richest) Source: Appendix 2 Relative differences in expenditure patterns between rich and poor also explain changes in relative welfare between the rich and poor (see Appendix 2). I<' or example, the poorest 20 percent spend 61 percent of their incomes on food (notably rice, edible oils, beans, sugar and milk), while the richest 20 percent spend only 41 percent of incomes on food. The rich spend larger shares of their incomes than the poor on transport, clothing, schooling, electricity and water. The poor are therefore disproportionately affected by the high levels of protection on rice (with domestic prices about 2.5 times world prices), red beans and sugar, and reduced but still 23 The Gini coefficient measures the degree of inequality in the distribution of incomes or wealth for a population. A coefficient of 0 indicates perfect equality, whereas a figure of 1 ~dicates perfect inequality. Actual Gini coefficients for income distributions have ranged from about 0.28 in China (1971) to 0.64 in Ecuador (1970- 75). The most recent figure for the United States is around 0.35. SUSTAINABLE BANKING wrm THE POOR 13 high protection on milk. On the other hand, the rich benefit disproportionately from subsidies on cooking gas and from both explicit subsidies and low cost recovery on public utilities (e.g. water supply and sewerage). WHO BENEFITS FROM SOCIAL SPENDING? The structure of public expenditures has been highly regressive. Public spending on education and health, which potentially has significant benefits for the poor, has been among the lowest in the region ( The World Bank [ 1992], [ 1995]). Indeed it amounted to less than 3 percent of GDP in the early 1990s. Moreover the quality of public education has also been poor, so that low-income entrepreneurs frequently make private schooling for their children one of their priority expenditures once their enterprises begin to earn higher incomes with ADEMI's support (see Chapter VII and Lewin [1991]). Expenditure on social assistance has also been minimal (0.3 percent of GDP in 1992, see IDB [1994]). Instead of recurrent expenditures on social seivices, the Government has stressed infrastructure and housing projects with high visibility. Moreover, it has focused infrastructure expenditures on urban areas, although poverty is largely a rural phenomenon. The benefits of such projects, e.g. for infrastructure and public housing, have accrued disproportionately to the middle and upper classes. MICROENTERPRISES: SMALL BUT IMPORTANT About 75 percent of Santo Domingo's 3 million inhabitants live in barrios. They are largely immigrants from rural areas who have come to the capital in search of work. With the unemployment rate in Santo Domingo at about 20 percent and underemployment affecting another 25 percent of the labour force, a large proportion of this population has sought work in the informal sector at least on a part-time basis. There are an estimated 330,0Q0 microenterprises in the Dominican Republic (according to the findings and projections of a 1992 survey described in Dauhajre [1992]). These enterprises generate over 760,000 jobs, or about 23 percent of the jobs in the economy, and produce the equivalent of about 31 percent of the official gross domestic product, or just under one-quarter of total official plus informal value added (see Figure 3 .2). SUSTAINABLE BANKING WI1H THE POOR 14 Figure 3.2: Microenterprises Contribution to Total Value Added in the Dominican Republic in 1991 Mero Government Forrrel fi,ance 8% corm-erce Mero industry 13% Forrrel services 5% 8% 7% Mero services 2% Mero other 1% Forrrel convrerce 11% ·Forrrel Forrrel industry infrastructure Forrrel 12% 19% agriculture '14% Source: Dauhajre [1992] The value added by microenterprises, which is conservatively estimated in the smvey at RD$29 billion in 1991, is three times larger than the contribution to GDP of the public sector. The importance ofmicroenterprises in varibus subsectbts is even higher. For example, in the area of commerce, sales by microenterprises exceed the value of formal commerce that is tallied in the national accounts (see Figure 3 .2). Average monthly sales for microenterprises were estimated at RD$14,626 (US$1, 160) in 1991, with a standard deviation ofRD$4,900 (Dauhajre [-1992]). Over half of the value added by microenterprises came from commerce, one-third from manufacturing and just under 10 percent from services. Thirty-seven percent of Dominican homes (or over half a million households) receive income as a result of a family member's participation in a microenterprise. 24 Indeed, informal microenterprises are the most important source of employment in the economy, exceeding not only the formal commercial and industrial sectors and the government, but also the important agricultural sector (see Figure 3 .3). Altogether, microenterprises are estimated to account for almost one in four jobs in the Dominican economy. 24 These figures consider only the direct incomes, thereby ignoring important multiplier effects on incomes resulting from the economic activity of microenteiprises. SUSTAINABLE BANKING WITH THE POOR 15 • .,1·, • . , - . ,. . . . , . . • I, ., • ,; ~ :;.~ • • • Figure 3.3: Microenterprises and the Job Markets in the Dominican Republic in 1991 Mero enterprises Agricuture Forrrel comrrerce Government Forrrel industry Free trade zones Tourism Other Forrrel sector 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% Percent of Total Employment Source: Dauhajre [1992} Microenterprises are born and disappear very rapidly. A two-part survey by the Foundation for the Promotion of Microenterprises (FondoMicro) of almost 4,000 Dominican households, found 622 microenterprises operating in March 1992 and 571 operating in March 1993 (see Cabal [1993]). However, the true change in activity was considerably larger, with a 25 percent increment in (new) enterprises being more than offset by the closure of30 percent of the microenterprises. The reason for this high closure rate is that microenterprises are highly susceptible to changes in economic conditions (e.g. a shortage of an input, or a decline in demand), generally lack capital (even though they are typically not highly leveraged) and sometimes lack managerial and commercial skills. Most enterprises are wound up within less than four years.25 In many cases, the informal sector is a secondary source of income that is abandoned as soon as formal sector employment opportunities arise. For example, during the early 1990s there was a boom in employment in the export processing zones, particularly for women, and at the same time, women headed 61 percent of the enterprises surveyed that were closed during the 1992-93 period. In other cases, low-income entrepreneurs operate microenterprises as their primary source of income. It is these enterprises that tend to yield more durable jobs and incomes. That is why ADEMI has refused to finance start-up operations (see Chapter V), requiring a minimum of one year of op~rations to qualify for a loan. Cabal's [1993] research also reveals interesting gender differences in the operation of low-income enterprises. Overall, women head slightly less than half the microenterprises. The enterprises that they operate tend to be smaller than those for men: owners account for over half the employees in enterprises headed by women (the comparable figure for male-headed enterprises is 32 percent). Female-headed microenterprises also tend to employ far more women 25 For similar findings regarding African microenteq,rises, see Liedholm and Mead [1993), McPherson (1992). SUSTAINABLE BANKING WI1H THE POOR 16 than jointly- or male-headed enterprises and to employ far fewer paid employees (12.5 percent compared to 46.9 percent for male-headed enterprises). Female-led enterprises tend to concentrate more on commerce (45 percent) and services (46 percent) than on manufacturing. This gender difference has an important bearing on women's demand for credit. The FondoMicro survey indicates that both female-led enterprises (83 percent) and male-led enterprises (78 percent) said they did not use formal credit during 1992-93. However, female entrepreneurs were also more likely than their male counterparts to believe that they did not need formal credit (72 percent compared to 64 percent). One reason is that women headed few manufacturing firms. Such firms were more likely to demand credi~ than commercial ·or service enterprises, not least because they were on average larger and had more paid employees. To obtain credit, manufacturing firms were also more likely to tum to suppliers, banks and finance companies, than to families, friends or to the enterpreneurs' own resources. 26 Thus, gender, line of production and scale of production all have an important bearing on the demand for financial services. LENDING MICROENTERPRISES A HAND THROUGH MICROFINANCE Most microenterprises depend primarily on the entrepreneur's own capital injections. For those that do receive credit, informal fµiance is an µnportant source of working capital. In general, family members and friends (incltiding through'rotating savings and credit associations) account for 35 percent of credit sources and are the leading source of capital other than the O'_Vner. Suppliers, moneylenders and other informal sources account for about 28 percent of credit sources (see Cabal [1993]). Suppliers' credit is particularly favored by female-led enterprises, which value ease of access and reliability of finance. Funds from moneylenders cost up to 10 percent per day, 20 percent per week, 75 percent per month or 240 percent per year. These sources are used to finance working capital and emergencies, but rarely for fixed capital investments. Only one in 16 enterprises surveyed by FondoMicro were found to have received credit from banks or finance companies and only about 2 percent had received credit from an NGO prior to 1992. While 40 percent of the micro entrepreneurs declared they had savings accounts, only one in six had a checking account at a bank. Thus, even though the microehterprise sector is highly dynamic, it remains at the margins of the financial system. Providers o/Microjinance. There are several governmental and NGO providers of microfinance and other business development services for microenterprises in the Dominican Republic (see Goldmark [1996], Paxton and Cuevas [1996]). They include: ADEMI; the Association for the Development of the Province of Espaillat (ADEPE); the Dominican Association for Women's Development (ADOPEM); the Association for Investment and Employment (ASPIRE); the Dominican Development Centre (CDD); the Dominican Development Foundation (FDD); the Development Fund (FONDESA); the Agroforestry Rotating Fund (FORESTA); the Enterprise Foundation for Development and Services 26 On the other hand, as noted earlier, there are fewer manufacturing microenterprises than commercial ones. SUSTAINABLE BANKING WITII THE POOR 17 (FUNDESER); the Dominican Institute for Industrial Technology (INDOTEC); the Institute for Technical and Professional Training (INFOTEP), and Dominican Women in Development (MUDE). In addition, FondoMicro, an apex institution funded by USAID that provides market- ·rate loans to microfinance providers, established a formal bank in 1997 (Banco para la Pequefia Empresa, see Chapter VIII). This bank provides credit directly to micro and small-scale enterprises (focusing on the latter). Whether by size of portfolio or number of clients, ADE:MI has been the dominant provider of financial services to microenterprises in the Dominican Republic. Indeed, until 1994, ADEMI accounted for up to 80 percent ofFondoMicro's portfolio. The two most important other non-bank providers of micro-credit are: • ADOPEM: established in 1982, it has 6,400 clients (90 percent of whom are female) with lower average incomes than ADEMI's clients. In fact, ADOPEM estimates that 70 percent of its clients are in the poorest quintile of the income distribution. ADOPEM provides loans for service, manufacturing and commercial enterprises for terms of 4 to 24 months at an average nominal interest rate of33 percent per year. It provides both individual and group loans, with the former dominating the portfolio. • The Dominican Development Fo~dation (FDD): established in 1966, it has approximately 13,500 clients and a portfolio ofUS$1.7 million. It provides technical assistance and training, as well as loans for service, manufacturing, commercial and agricultural enterprises for terms of 4 to 18 months at an average nominai interest rate of 29 percent. Around 45 percent of its clients are female and about 20 percent are rural. FDD estimates that 85 · percent of its clients are in the second poprest q1.Jintile. There has long been competition for staff, particularly for field level agents, between ADOPEM, ADEMI and FondoMicro--ADOPEM has been particularly affected, due to its lower income target group and consequently lower staff salaries. However, until recently, there has been little competition for clients between these providers, because of the relatively low penetration of the microenterprise market, and to a certain degree of specialisation by income levels and by location (rural versus urban). The situation is changing rapidly in Santo Domingo, where FondoMicro's bank has established small-business loan operations in areas that ADEMI has served for some time. A New, Subsidized, Government Lending Program for Low-Income Borrowers. One important area of concern is the Government's decision to engage in the direct provision of micro-credit as a result of the success of microfinance providers such as ADEMI. Early in 1997, it launched a Credit Programme for Micro-, Small- and Medium-scale Enterprises (PROMIPYME). The programme is publicly funded, housed in the Ministry of Commerce and Industry, and channeled through the State-owned Banco de Reservas, (the country's largest bank). Over RD$200 million in credit (or two-thirds of the programme's initial capital) was disbursed in the first five months of operations to 1,500 qualifying borrowers, and it has successfully secured further injections of resources. PROMIPYME's loans, which range from RD$50,000 to RD$1 million, bear a substantial interest rate subsidy (14.5 percent, which is below the cost of borrowing for the Government). Banco de Reservas receives a spread of just 4 SUSTAINABLE BANKING WITII TIIE POOR 18 percent but assumes the credit risk. ADOPEM, MUD~ and certain cooperatives have also received funds from PROMIPYME for micro-credit to very low income beneficiaries (i.e. loans in the RD$ l, 000-100, 000 range), to be on-lent at no more than 28 percent per year. This programme, which is heralded by its sponsors as the fulfillment of an electoral promise, runs counter to the lessons of experience regarding subsidized credit programmes, (see Yaron, Benjamin and Piprek [1997]). Notwithstanding efforts by its dynamic executive director to ensure transparency and asset quality, there are significant political risks associated with PROMIPYME. Thus it could easily undermine ¢.e very financial market that it is purportedly designed to develop. Nevertheless, barring a truly substantial increase in public funding for PROMIPYME, for the next few years ADEMI/BancoADEMI will almost certainly continue to be the single largest provider of microenterprise finance in the Dominican Republic. SUSTAINABLE BA :}KING WI1H THE POOR 19 IV. WHAT IS ADEMI? ADEMl'S LEGAL ORIGINS The Association for the Development ofMicroenterprises, Inc. (ADEMI), is a non profit organisation that was constituted on December 14, 1982, when its founding members signed and agreed to the By-Laws of the Association (ADEMI [1982]). It was formally established by virtue of Presidential Decree 745 on February 11, 1983, under Law No. 520 of July 26, 1920, which governs public interest organisations. Under this law, it is exeippt from tax on its net income, and is entitled to mobilize grants and loans both domestically and abroad, and to make loans to clients. While it can perform the functions of a non-bank financial intermediary, ADEMI cannot mobilize sight or term deposits. After operating as a non-profit intermediary for 14 years, ADEMI established a new development bank, called Banco de Desarrollo ADEMI, S.A. (or BancoADEMI). The By-Laws of the bank were approved by ADEMI's Board in June 1997 and on September 11, 1997, BancoADEMI was granted a banking licence by the Monetary Board. As such, it is supervised by the Banking Superintendency and falls under Law No. 299 of 1966, governing development banks, rather than the law for non-profit organisations. On January 2, 1998, BancoADEMI .opened its doors to the ~eneral public. ADEMI has begun a transition process in which all clients will eventually be transferred to BancoADEMI. •As soon as the end of 1998, it is expected that most clients will have made the transition to BancoADEMI. Once the transition is completed, ADEMI is currently planning to operate simply as a holding company for the bank (see Chapter VIIl for a discussion of this transition process). WHAT GAVE RISE TO ADEMI? ADEMI was founded by a Dominican businessman, Camilo Lluberes, who was concerned that political democracy in the Dominican Republic would be jeopardized unless Dominicans made a personal effort to promote greater economic democracy. With technical support from Stephen Gross, and in partnership with Pedro Jimenez, the future Executive Director of ADEMI, Lluberes began to outline the purpose, structure and operating procedures of ADEMI. He also brought on board partners to become founding members and to add to his own financial commitment to the Association's initial capital·. Finally, the team obtained a line of credit for ADEMI from Banco Popular and a grant from the Technical Secretariat of the Presidency. ADEMl's founders regarded unemployment and underemployment as the key problem to be addressed for the urban poor (see Chapters II and III). They noted that low income businesspersons were often extremely entrepreneurial and that microenterprises offered excellent jobs prospects for the growing population in Santo Domingo's barrios or low income SUSTAINABLE BANKING WITii THE POOR 20 neighbourhoods. However these enterprises lacked access to adequate credit. Therefore, ADEMI set out to do its part in reducing poverty, promoting employment and integrating marginalised groups into the economic life of the nation, particularly by " democratising" credit, i.e., making credit available to creditworthy low-income entrepreneurs. In order to expand opportunities for low-income entrepreneurs, ADEMI also set out to provide demand-driven technical support to its clients through ADEMI's field officers or "credit advisors". However, the .Association did not see its primary role as a technical assistance agency-as "credit advisors" suggests, the credit was to come first and the advice second. More broadly, as an apolitical, non-religious and not-for-profit association, ADEMI also saw itself as a point of contact and communication between the business community, municipal authorities, the Central Government and lower-income groups in the country. THE NEXT FIFTEEN YEARS ADEMI began lending in April 1983. Credit was provided using both individual lending and solidarity group lending methods and was targeted at microenterprises. In addition to credit services, field officers provided business advice to clients, notably in such areas as financial record-keeping and management of assets and liabilities. Right from the outset,·ADEMI sought to bring its clients into the formal system by relying on Banco Popular to cash loan cheques and accept loan payments (and initially by literally accomp~ying the clients to the bank). ADEMI soon began to face liquidity problems as its client base grew more rapidly than the resources at its disposal. However, by 1985 it was fac~g the opposite problem, namely a surplus of liquidity, after it had successfully mobilised external resources from the Inter- American Development Bank. ADEMI rapidly establis~ed a branch network in towns and cities around the country, but by 1986 the pace of expansion was leading to institutional shortcomings that were contributing to poor record-keeping and high arrears. Beginning in 1986-87, ADEMI focused attention on improving its operating efficiency by implementing a professional administrative system and by formulating policies for personnel, loan monitoring, and communication between ADEMI offices. An incentive scheme was also introduced for ADEMI's staff to encourage low delinquency and a minimum size of portfolio (see Chapter VI and Appendix 3). Internal audit offices were created to increase controls, and the services of lawyers were retained as a measure to strengthen loan enforcement. Additional funding was sought from various agencies including the Germany's Technical Cooperation Agency (GTZ), Pueblito Canada, USA.ID and IDB. In 1989 ADEMI reached operational self-sufficiency and by 1991, it had grown to become a financially self-sufficient organization (see Chapter VII). As its clients' businesses continued to expand, ADEMI increased its loan sizes to accommodate demand, extending the range of services offered. It also began to lend directly to small-scale enterprises, rather than just growing microenterprises, after it secured a loan from the European Investment Bank (EIB) in 1992. Four years later, in partnership with Banco Popular, ADEMI launched a Mastercard for its clients, furthering its objective of''democratising credit". Finally, in 1997, ADEMI established a form.al bank for micro-, small- and medium scale enterprises, BancoADEMI, that is SUSTAINABLE BANKING WI1H THE POOR 21 offering not only credit but also savings instruments (time deposits and CDs) to its clients - see Chapter VITI. ADEMI AT A GLANCE As of June 1997, ADEMI had a network of25 branches serving over 16,000 clients. ADEMI stood on the threshold of entering the ranks ofthe formal financial sector with RD$606 million (US$43 million) in assets and RD$242 million (US$17 million) in net worth. It has served over 40,000 micro and small scale enterprises during the past 15 years and had a direct impact on the jobs of over 200,000 owners and employees of micro and small scale enterprises. Chapter VII provides further details on ADEMI's performance and on that of its clients. Table 4.1: A Quick Look At ADEMl's Operations, As Of June 1997 Number of active clients 16, 185 Outstanding loan portfolio RD$377,204,234 (US$26.9 million) Average loan size Micro: RD$26,290 (US$1,875) Small: RD$738,468 (US$52,700) Total assets RD$606,221,870 (US$43.2 million) Net worth RD$242,565,858 (US$17.3 million) Number of staff 202 Number of credit advisors 120 Number of branches 25 Source: ADFlvfl [1997} ADEMl'S GUIDING PRINCIPLES ADEMI operates in accordance with several important guiding principles which largely explain its success (see Ledgerwood and Burnett [1995], and Chapter VITI): • Business-like Operations: ADEMI has a social objective, but its operating approach is businesslike. This first principle imbues all facets of ADEMI's operations and underlies virtually all other guiding principles for ADEMI. • Pursuit of Self-Reliance: Far from viewing itself as a transitory donor program, ADEMI aims to establish itself as a permanent source of financial services to entrepreneurs who are unable to access credit through formal channels. To achieve permanence, an institution has to be self-sustaining. Therefore ADEMl's choice of products, branch locations, operating SUSTAINABLE BANKING WITII TIIE POOR 22 , ' procedures, etc. is guided by whether they meet the clients' needs and contribute to self- sustainability. • Independence from Donors: While subsidies are welcomed, ADEMI is not dependent on donor assistance. Nor is its Board of Directors controlled by donor ag·encies (but rather by private sector professionals). The aim is to be self-sufficient so as to be able to maintain a client orientation rather than a donor orientation. • Partnership with Clients: ADEMI operates on the principle that low-income entrepreneurs are experienced in their areas of work, know how to use credit, are capable of paying positive real rates of interest on their loans, want to be treated as business partners rather than as beneficiaries of paternalistic assistance, and will seek training when they feel they need it, rather than when the lender thinks they need it. • Responsible Lending: ADEMI carefully screens its clients and actively pursues delinquent borrowers. In the words of Pedro Jimenez, "There are no bad clients. There are only bad loans." The idea is therefore not to blame the clientele for poor performance, but rather to focus on internal accountability and responsibility.27 • Focused Services: Like many successful microlenders that seive poor (as opposed to extremely poor) households, ADEMI follows a minimalist approach. It does not engage in the provision of social or human develqpment seivices, marketing or formal technical training courses (for which it connects its clients with the government Institute for Technical and Professional Training, INFOTEP). 28 Rather it specialises in providing easily accessible credit and technical advice related to the enterprises that are financed. • Broad Outreach: Although ADEMI has expanded the range of loan sizes offered and, implicitly, its target markets to include small and even medium size businesses, it is still dedicated to the "full democratisation of credit", and thus aims to reach even the smallest of viable enterprises. For example, initial loans can be as little as $100 or less. Moreover, ADEMI gears its processes and procedures towards enabling the Association to reach the largest possible number of clients in a viable manner. 29 WHAT ADEMI OFFERS ITS CLIENTS ADEMI provides a range of lending seivices to micro- and small scale enterprises, including loans for working capital, investment credit and Mastercards, as well as pilot programs 27 A corollary to this point is that ADEMI does not engage in activities it believes it cannot handle successfully. For example, in the late 1980s, ADEMI declined an invitation by the President of the Dominican Republic to expand its operations into lending for agriculture, due to the high risks of lending to agriculture, the lack of skills in ADEMI for this type of venture, and the Government's policy of mandating subsidised rates of interest on agricultural loans. 28 INFOTEP is a government operated training institute funded by all registered businesses in the Dominican Republic. Each Dominican business is required to provide one-half of one percent of its sales revenue as a "donation" to INFO'IEP to fund its operations. 29 This is proving more challenging with the establishment of BancoADEMI (see Chapter VIII). SUSTAINABLE BANKING WITII TIIB POOR 23 for home improvement loans and for loans to schools and hospitals operating in low-income neighbourhoods. On the savings side, ADEMI has, as a non-bank financial intermediary, accepted loans from clients in lieu of deposits, and paid a market rate (14-18 percent) on these deposit substitutes (which have now been taken over by BancoADEMI). Finally, ADEMI offers technical assistance to its clients through its credit advisors. Credit for.Mero-and Small Scale Enterprises. Loans to microenterprises are ADEMI's most important product, accounting for 66 percent of the volume of loans outstanding as of 1997. The remainder is largely credit to small scale enterprises. Since 1993, ADEMI has begun to increase the range of loans sizes that it offers to its clients, both in order to grow with its clients and to serve a perceived "gap" in the credit market for small enterprises that cannot obtain credit from the formal banking sector (see Chapters V and VII on increasing loan sizes). In order to serve the small business market, ADEMI established a small lending window in the head office for loans in excess of US$36,000, (which has now been taken over by BancoADEMI) and has applied more traditional credit analysis and underwriting practices to small business loans (see Table 4.2). Loan Terms and Amounts. The size and terms of each loan provided by ADEMI are determined by the credit advisor and approved by management. There are currently no minimum or maximum limits to loan sizes. However, ADEMI does classify its credit into microenterprise versus small business lending, and draws a further distinction between small, emerging • microenterprises and stronger ones that are growing into small businesses (Table 4.2). Smaller microenterprises tend tp receive smaller, shorter-term working capital loans (ofup to one year), whereas growing microenterprises and small businesses make somewhat greater use of longer term, investment credit (of about two years). Interest Rates. Contrary to the practices of most state-owned development finance intermediaries, higher interest rates and up-front commissions are charged on smaller loans than on larger ones. The effective annual interest rate for smaller microenterprise loans approximates 66 percent depending on the loan term. (A three-month loan term has an effective rate of 84 percent; an 18-month loan term has an effective rate of 61 percent), whereas it is closer to 28 percent for small businesses. The main reason is that the average cost of making smaller loans is higher, (even though approval processes for the smaller loans are more streamlined) and that ADEMI wants to minimise the cross-subsidy from larger to smaller loans (see Chapter VII). The scope for such a cross-subsidy is in any case constrained to a certain degree by greater competition in the small-scale lending market. SUSTAINABLE BANKING WITH TIIE POOR 24 Table 4.2: ADEMl's Loan Products, Prices and Terms by Type of Client, as of end-1997 Typical features of entelJ)rise assets under assets of Assets of more US$10,000; US$10-20,000; than US$20,000; 1-10 employees 10-20 employees 10-50 employees Size and Term of Loans Average loan size (04-06/97) US$1,100 US$7,300 US$58,300 Typical range ofloan sizes US$36,000 Average term ofloans IO months 20months 34months (BIB-financed loans: 48 months) Repayment schedule Monthly monthly monthly Price of Loans Monthly interest rate (variable, 3% 2.5% 2.2% and market driven) Basis for interest calculation d~clining balance Declining balance Up-front commission t P~:~ "· 2% 2% Daily p~nalty interest on arrears 0.2% 0.2% 0.2% Other Requirements Savings requirements . None none none CoHateral requirements None required, none required, but Mandatory, with guarantor often partial collateral deed, 1.4- encouraged collateralisation of 2.1 times loan larger loans value Other requirements 1 year in business; 1 year in business; 1 year in business business pUIJ>ose business pUIJ>ose business plan and proper accounting Sources: ADFMI. {1997}, Jimenez {1997], LedgerwoodandBurnett{l995) Payment Frequency. Loans are amortised in monthly, rather than weekly, installments, in order to enable clients to "capitalize" from use of borrowed funds. There are also no forced savings requirements or counteivailing balances, since forced savings proved unpopular in ADEMl's early years (due to high inflation, frequent devaluations of the Dominican peso and low interest rates on savings accounts), and clients prefer to reinvest all surpluses in their enterprises. Loan Eligibility. To qualify for an ADEMI loan the microenterprise must be a minimum of one year old and demonstrate the potential to create new jobs. Clients must be over 18 years of age, show initiative and a minimum level of business aptitude to assure their success and growth. For both small arid microenterprises, loans must be for working capital purposes or for acquisition of equipment and machinery (fixed assets). ADEMI does not make consumption SUSTAINABLE BANKING WITH THE POOR 25 .. \ . i-.:•- _·:;.-'_; : . ~. r :: 1 1 _~::-.·,.{Ji•-J•.-v'.• ' . loans. 30 In the case of small-scale enterprises, the loan application must be supported by a business plan and adequate accounting records of the financial condition of the enterprise, as well as documented authorization from the company's Board of Directors for the signatory to receive the loan from ADEMI. There are no formal collateral requirements on loans to microenterprises, although ADEMI does encourage the use of guarantors and does reposess equipment or seize other assets of clients on whom it has to foreclose. All borrowers are required to sign a legal contract and promissory note, stating their obligation to repay funds at specified terms. When feasible, borrowers are required to sign deeds for property or machinery as loau guarantees, particularly in the case of growing enterprises and of investment credits. 31 In some cases, guarantors may be asked to pledge security on behalf of the borrower. For small business loans, collateral is mandatory and must exceed the value of the loan by a safe margin (generally 1.4-2.1 times loan value). While the cost of enforcing very small loans may appear high, the benefits accrue in terms of higher overall repayment rates, since clients who witness legal enforcement of loans against other entrepreneurs are more likely to repay their own loans. However, collateral is ultimately mainly a substitute for good information, and ADEMI prefers to rely on good information to pre-screen clients rather than on collateral to enforce bad loans. Therefore ADEMI's advisors investigate the reputation of the applicant in much the same way as an informal moneylender might before making loans, and develop a working relationship with their borrowers through frequent visits to detect any problems as early as possible. 32 ADEMI used to provide loans to groups of low-income entrepreneurs, relying on peer pressure rather than the threat of foreclosure on collateral to encourage repayment. Howe~er, ADEMI terminated its group lending in 1986 (for reasons that are described in Box 4.1), and all loans are now made only to eligible individuals or companies. 30 However, clients can use the ADEMI Mastercard, discussed below, for consumption pmposes. 31 ADEMI estimates that most deeds (80 percent) cover only 20-30 percent of the value of the loan. 32 With the transfer of clients to BancoADEMI, legal requirements are becoming more stringent for loans to qualify in the two lowest risk.loan categories and thus entail minimal loan loss provisioning. SUSTAINABLE BANKING WITH THE POOR 26 Box 4.1: Why ADEMI No Longer Lends to Groups In 1987, ADEMI terminated group lending and began to lend exclusively to individual physical or legal persons. The rationale for this deci&ion was: • • Group members' enterprises experienced diverse growth rates, which eroded the cohesiveness of the groups and led to high turnover rates for clients. • Efforts to standardise loan terms in the face of diverse levels ·o f demand for credit resulted in lower quality services and discouraged the growth of individual businesses. • The depth of knowledge that credit advisers acquired about their clients was limited. They relied more on mutual guarantees than on loan quality to evaluate applications. • The solidarity group lending methodology was viewed as better suited to poverty alleviation programmes than to ADEMI's "democratisation of credit" objectives. The shift to an individualised product meant that existing clients had to be re-evaluated under new eligibility criteria. Clients were screened not on the basis of their ability to collateralise loans, but on their repayment capacity, the growth potential of their businesses, and their reputation. A large share of ADEMI's group clients were granted individual loans. Others were "spun off' to other NGOs providing group credit. - Source: Ledgerwood and Burnett [1995) Technical Assistance. In order to strengthen the portfolio further, ADEMI supports its clients by offering technical assistance through its credit adv~sors. The technical assistance is not mandatory, and focuses particularly on constructing financial statements and managing various aspects of the business, e.g. inventories, cash flow and marketing. Credit advisors may suggest additional technical assistance to clients, based on their individual needs, and point them towards courses offered by the State-owned Institute for Technical and Professional Training (INFOTEP). ADFMI MasterCards. In May 1996, ADEMijoined with Banco Popular to launch the ADEMI MasterCard. In the first few months, over 3,000 credit cards were issued to ADEMI's good clients and ADEMI now has over 5;000 clients with Mastercards. The advertising logo of "A great card for the micro enterprise" captures the spirit of ADEMI's attempts to" democratise credit" by making credit cards available to low-income entrepreneurs. Clients can use the cards in the Dominican Republic to withdraw cash at 45 Banco Popular branches and 60 automatic teller machines, or to pay light, telephone and other bills. To qualify for the cards, ADEMI's clients must have monthly incomes of RD$3,000, have borrowed at least RD$3,000 from ADEMI, and have a good repayment record. ADEMI receives a payment of about US$ l Ofor each client it recommends to Banco Popular, plus a small share of the fees charged by Banco Popular to vendors who accept the ADEMI-Banco Popular Mastercards in their stores. The credit risk is borne entirely by Banco Popular, which charges its normal credit card rate on the ADEMI MasterCard (6 percent per month). While the credit cards will probably replace some of ADEMI's loans to its clients, due to the added convenience of a credit card, ADEMI hopes to use the card in future to automate its disbursement and repayment processes. SUSTAINABLE BANKING WI1H THE POOR 27 Home Improvement Loans. In 1994, ADEMI began to pilot home improvement loans at the request of one of its Board members. The loans are made available to clients in good standing fqr home improvements (ADEMI 11:1akes no mortgage loans or home construction loans). Often clients build one wall with a loan, repay, and then build another wall with a second loan, and so on. The loans average US$6,000 and are repayable in three years. Interest rates on the home improvement loans are positive but below market rates. They have gradually been increased from 18 percent to 22 percent without affecting repayment rates. While BancoADEMI had planned to take over the programme and expand it beyond the pilot area in Neyba, to cover the entire country, regulatory requirements for clients to hold formal titles to the properties may preclude it. Loans to Schools and Hospitals. In 1989, ADEMI began a loan programme for fixed assets for private primary schools and medical dispensaries. The programme was designed to enable schools to expand and medical dispensaries in marginal areas to purchase of laboratory or dental equipment. This programme has remained small, relative to ADEMI's other lending activities. Loans From Clients. Since ADEMI is not a formal financial institution, it is not allowed to accept deposits on a fiduciary basis. However, the law does allow ADEMI to receive loans from individuals, including its clients. In 1993, ADEMI began to issue bonds to clients including non-borrowers, at 20 percent per year, or about 2 percent above the commercial deposit rates. The cost of these funds has declined by four to five percentage points since 1993, in line with declining market interest rates. ADEMI backs these loans, which are effectively deposit-substitutes, by its line of credit from Banco Popular. Currently, loans from clients account for 23 percent of ADEMI' s borrowings. Due to the cost of mobilizing funds, ADEMI has preferred to issue larger denomination bonds. However, BancoADEMI has begun to offer both savings accounts and certificates of deposit to the public (see Chapter VIII). SUSTAINABLE BANKING WI1H THE POOR 28 I V. WHOM DOES ADEMI SERVE? ADEMl'S MASSIVE MARKET NICHE With an estimated 330,000 microenterprises in the Dominican Republic and little competition, there is no shortage of potential clients for ADEMI. Clearly not all enterprises need loans-according to Cabal [1993], 70 percent of entrepreneurs say they do not-but there are still an estimated 100,000 enterprises that are interested in formal credit. With such a large, underserved market segment and with its well-designed lending servi£_es, ADEMI has been able to make substantial profits by lending to micro- and small scale enterprises (RD$54 million in 1996, on 13,300 loans totalling RD$395 million). ADEMI aims to reach the largest number of low-income entrepreneurs possible in every area that it serves. Its branches are located throughout the Dominican Republic, with six in Santo Domingo and 19 in other urban and semi-urban areas throughout the country. ADEMI · sets no specific quotas with regard to, e.g., outreach to women, specific income groups, etc. Its portfolio is highly diversified, both geographically and by type and size of enterprise. However, its overall target is to serve two broad categories of enterprises: micro and small-scale. Microenterprises are served at all 25 branches, whereas a separate department serves the small- scale enterprises directly from the headqu~ers . . . in Santo '·, Domingo. _ Microenterprises. Within the microenterprise category, ADEMI distinguishes between subsistence enterprises with one operator and less than US$1,000 in assets, expansion enterprises with two to ten workers and assets ofup to US$10,000, and transforming enterprises, with 10-20 workers and up to US$20,000 in assets. In virtually all cases, the operations and administration are concentrated in the hands of one or two people (the owners), the activities are labour- intensive, and the enterprises depend almost exclusively on local resources and markets. Small scale Enterprises. The small-scale lending category includes enterprises with 10 to 50 employees, monthly sales ofRD$250,000 or more, assets exceeding RD$500,000 and loan amounts of no less than RD$50,000, and usually more than RD$500,000. Small business generally have senior employees at the management level other than just the owner. These businesses are frequently more capital-intensive than are microenterprises. CHARACTERISING THE CLIENTELE General Features. ADEMI's clients generally live in urban or semi-urban barrios which have minimal infrastructure, poor housing, few community facilities, high population density, low family incomes, high unemployment and underemployment, uncertain rights to land, high rates of malnutrition and high mortality rates (see Chapter ill). Most ADEMI clients have five to ten years of experience in their area of work and on average three years of experience running their enterprises. ADEMI' s larger loan clients have attended junior high school and are between 30 to 40 years old. Their smaller loan clients have generally completed the third or fourth grade of primary school, .and about 3 0 percent are illiterate. These entrepreneurs' businesses have little organization in terms of record-keeping, documentation and accounting, no prior formal credit SUSTAINABLE BANKING WITH THE POOR 29 history, limited collateral and low capacity to absorb shocks. (Ledgerwood and Burnett [1995]). ADEMI also lends to professionals who cannot get jobs in the formal sector and want to start their own businesses. They include engineers, architects, economists, administrators and chemists who may be "start-ups", but have demonstrated expertise in their field. Finally, the small scale enterprises that approach ADEMI for loans are typically informal and cannot access commercial bank credit. 33 Size ofF.nterprises. While fairly small microenterprises still constitute the vast majority of ADEMI's clients, there has been a clear shift towards lending larger amounts and to larger clients (Table 5.1). Microenterprises with less than RD$140,000 (US$10,000) in assets and loan amounts under RD$50,000 (US$3,500) accounted for 89 percent of ADEMI's 13,300 loans during 1996, but for only 43 percent of the RD$395 million in new lending. Small-scale loans averaging RD$816,000 (US$58,000) now account for 21 percent of the portfolio, compared to 6 percent in 1992. Number of New Loans 14,152 13,621 13,184 14,193 13,289 Microenterprises 99.6% 99.5% 99.0% 88.7% 88.6% Stronger Microenterprises 10.4% 10.6% Small-scale Enterprises 0.4% 0.5% 1.0% 0.9% 0.8% Volume of New Lending (RD$m.) 174.1 251.6 295.3 3~6.4 395.1 Microenterprises 94.4% 85.8% 77.1% 40.8% 42.9% Stronger Microenterprises 35.4% 36.2% Small-scale Enterprises 5.6% 14.2% .22.9% 23.8% 20.9% Average Loan Size (RD$) Microenterprises 11,655 15,946 17,430 12,534 14,403 Stronger Microenterprises 92,889 100,999 Small-scale Enterprises 191,705 475,159 521,043 706,840 816,144 Source: ADE!v1I Quarterly Reports (See ADE!v1I [J 997]) There are three main reasons for the shift in portfolio composition towards larger lo~: 33 Businesses at all levels in Dominican Republic use a system known as double accounting in which two separate sets of books are kept. One set reflects the real financial condition of the business, and is used internally for management of the business. The second set, which understates the perfonnance of the business, is presented as the official version, is "audited" and is used to keep the business from being heavily taxed. This practice is important since high taxes can mean the difference between the survival and failure of small businesses. The problem arises that, while their true financial condition may be favourable, small businesses cannot obtain credit from banks using their official records, since prudential regulations would place them in a high risk category (entailing prohibitive rates of interest). Thus small enterprises either self-finance orrely on highly priced credit from.financieras or loan sharks. ADE:MI recognizes their problem and contends that until taxes are refonned to assist small businesses, ADEMI can make an important contnbution to the cmmtry's economic development by helping "informal" businesses to grow. SUSTAINABLE BANKING WTIH TIIE POOR 30 ( • Inflation: since loan sizes are expressed in current pesos, nominal increases are required to keep pace with inflation. However, inflation rates have generally been below 10 percent in recent years (see Chapter II and Appendix 1). -• Success ofADElvfl's clients: The scale of operations, retained earnings and borrowing .capacity of most of ADEMI's clients have grown with loans from ADEMI (see Chapter VII). As the clients have grown, so have ADEMI's loan sizes increased to meet their requirements. Thus in 1995, ADEMI began to distinguish between smaller microenterprises and strengthened ones that are transforming into small businesses . .• New lending to small businesses: In 1992, ADEMI began to borrow from the European Investment Bank (EIB), which required that its loans to ADEMI be on-lent to small businesses in the first instance in amounts of no less than US$60,000 per loan, although no restrictions were placed on the use of the repayments-and ADEMI relies heavily on the repayments to fund its microfinance portfolio. This explains the dramatic increase in the average loan size for small enterprises after 1993. Whereas this category had consisted almost entirely of graduated microenterprises before 1992, after 1992, new small-scale enterprise clients were sought out. Nonetheless, graduated microenterprises still accounted for almost 60 percent of ADEMI's "small business" category in 1997.34 Did ADEMI's expansion irito small-scale lending entail a loss of its original "vision" as a bank for microenterprises? As of 1997, tb:einswer seemed to be no. First of all, as noted in Chapter Ill, ADEMI's microenterprise target market has always been slightly above the subsistence enterprises served by ADOPEM, FDD and others. Thus over half of ADEMI's clients take out loans in the US$400-1,300 range, compared to only 1 in 12 clients who take out loans under US$400. Nevertheless, low-income entrepreneurs can and do still take out initial loans of as little as US$70 (see ADEMI [1997]). There is moreover no evidence that the number of new micro-businesses receiving loans was affected by the influx of EIB funds (Table 5.1). The increase in loan sizes can largely be explained by the successful functioning of the incentive system for borrowers, which involves access to a larger follow-on loan upon, timely repayment of the current loan. Finally, the incentive system for credit advisors still encourages them to seek out microenterprise clients, since they must have a minimum of 130 clients to earn bonuses, and, until the opening ofBancoADEMI in January 1998, advisors lost the strongest microenterprises to the head office when the enterprises graduated into small businesses. However, the opening ofBancoADEMI has .inevitably re-opened the question of ADEMI's client focus. Successfully launching BancoADEMI has inevitably commanded . management focus to a far greater ex.tent than has developing the microenterprise client base at ADEMI. Moreover, several factors have contributed to an acceleration of the transfer of clients from ADEMi to BancoADEMI, even though the integration of the smallest enterprises into the BancoADEMI is proving extremely challenging (see Chapter VIII). There is considerable internal debate on how best to accommodate clients who are very costly to serve under the prudential regulations governing banks. This is a priority issue for management, which has 34 Under the most recent BIB loan, approved in November 1997, the BIB agreed to a substantially lower initial loan to small businesses, namely ar01md US$35,000. SUSTAINABLE BANKING wrm THE POOR . 1·.1 31 I .. . ,,. , . attempted to draw microenterprises into the formal financial sector and offer them a broader range of financial products. This is evinced by the fact that, even though larger loans are more profitable than smaller ones (see Chapter VII), over three-quarters of the loans issued to date by BancoADEMI have been for less than RD$50,000 (US$3,500). To date, ADEMI and BancoADEMI remain the main source of finance for microenterprise in the Dominican Republic, and unless they are partly displaced by PRO:MIPYME (see Chapter ID), they are likely to continue serving an increasing number of the country's low-income entrepreneurs. Nevertheless, as a regulated financial institution, a shift in BancoADEMI's lending focus seems ·inevitable. Gender ofADFMI's clients. During ADEMI's first year of operations, women accounted for one in nine borrowers. That number increased to one in three by 1988, and reached almost one in two at the end of 1996 (46 percent of the 15,370 active clients). More than half of ADEMI's active clients in the service sector in 1996 were women (55 percent), whereas men accounted for two out of three ADEMI clients in the manufacturing sector (Figure 5. 1). Figure 5.1: Share of Women Entrepreneurs in ADEMI's Clientele as of December 1996 Manufacturing, Manufacturing, SeNices, Female Male Female· 5% 11% Services, Male 6% 5% 34% 39% Source: ADFMI Quarterly Reports ( See ADFMI [1997}) Location ofClients. Sixty percent of ADEMI's services are provided in the Santo Domingo area (see Table 5.2). This reflects the concentration of the country's low-income (urban) enterprises in the barrios of Santo Domingo. As a result of the presence of the small enterprise window at the head office, the largest loans on average are also made in the capital city. The Eastern region, which accounts for the next largest share of ADEMI's clients (17 percent), has the smallest average loan sizes, while the northern city of Santiago and the Southern region have important concentrations of strengthened microenterprises. SUSTAINABLE BANKING WI1H THE POOR 32 i Table 5.2: ADEMl's Lending by Location, April-June 1997 Total 4,016 123,022,678 Small Enterprises 0.6% 15.9% Santo Domingo 58.8% 48.3% ·santiago 4.2% 4.2% South 12.6% 11.4% East 17.2% 14.8% North 6.6% 5.4% Source: ADFlv11 {1997) Types ofActivities. Merchants, shop-keepers, confectioners, mechanics, beauty salon operators and restauranteurs accounted for the vast majority of ADEMI's clients (86 percent). In December 1995, 32 percent of ADEMI clients were active in manufacturing, 22 percent in services and 46 percent in retail. . These proportions, which had been fairly consistent for a number of years, changed dramatically in 1996 and 1997, with a sharp swing away from manufacturing and services towards retail activities. In particular, over 70 percent of ADEMI's active loans in the past two years have been for retail, rather than manufacturing or service activities. This does not appear to be an intentional or planhed strategy. Nor does it appear to be simply the product ofa change in the methods of classification. Figure 5.2: Shift of ADEMI Lending Towards the Retail Sector (Number of Clients by Sector) 12000 10000 8000 6000 4000 2000 1993 1994 1995 1996 11/1997 11 senAces ■ Manufacturing 0 Retail I Source: ADF.MI Quarterly Reports (see ADF.MI {1997)) SUSTAINABLE BANKING W11H TIIE POOR 33 The implications for the structure of ADEMI's lending are potentially enormous. For example, food vendors have a high turnover of inventories and need a rapid succession of working capital loans. With regard to liquidity risks, vendors' exposure to customers is very low, since they are almost always paid in cash and offer credit to only their best and most regular customers. With regard to market risks, there is an inelastic demand for their product, so that inflation is not likely to depress demand for their services too much. On the other hand, furniture manufacturers require more equipment and have a longer production cycle, their products remain in inventory for longer periods unless they are made to order, they often make sales on credit rather than for cash, and the credit tends to be to customers who are not "regulars" and need to .borrow on_ longer terms, e.g. one to two months or more. They may have their own borrowings on shorter terms and therefore must handle the term transformation problem. Finally they have a more elastic demand for their product, so that in periods of inflation or depressed growth, buyers postpone their purchases of furniture (or if they have already bought, are-likely to struggle with servicing of their credits). Therefore the risks, required loan terms, and debt carrying capacity are vastly different for manufacturers than for retailers. Arguably the potential implications for ADEMI's cost structure and risks of such a dramatic shift in the composition of the portfolio have not been evaluated sufficiently, and point to the need for greater strategic planning. This is not to say that ADEMI has not considered these factors at all. For example, in 1994, there were only two types of activities with arrears in excess of9 percent, namely ceramics (20 percent) and jewelry (12 percent). By the following year, support for both activities had been cut in half-in each case to less than one-half of I percent of ADEMI's lending. However, the risk implications of sharp changes in the nature of loans demanded need to be analysed thoroughly. In addition, from a broader development perspective, ADEMI may want to consider how its portfolio allocation, not only by type of activity but also by location, gender and scale of lending, can be of greatest benefit to the country in terms of employment and skills development. HOW WELL DO ADEMl'S SERVICES FIT CLIENTS' NEEDS? Cabal [ 1993] found that trust in the lender and ease of access to funds are the main reasons that low-income entrepreneurs choose particular sources of credit. These factors are much more important than the cost of funds (3 8 percent considered trust the most important factor, compared to 17 percent for the cost of funds). This finding holds for both men and women, although male-led enterprises and/or larger enterprises attach greater importance to the cost of funds than do female-led and/or smaller enterprises. Female-led enterprises place particular emphasis on_ ease of access to credit. Overall, ADEMI's operating strategies take into account these preferences, and clients consequently appear to be extremely satisfied with ADEMI's services. The most important source of satisfaction for clients is that ADEMI' s credit advisors cultivate a relationship with them that typically goes beyond that of a convenient financial exchange to one of genuine partnership .. Interviews with both credit advisors and clients suggest that the relationship between the two is a leading factor in promoting strong repayment performance. Many clients SUSTAINABLE BANKING wrrn THE POOR 34 f say they feel a strong sense of obligation to their advisor, as well as to ADEMI as a whole (see Ledgerwood and Burnett [1995]). There are a host of other reasons for the high degree of client satisfaction: loan decisions are made quickly and the transaction costs to the client of applying for ADEMI credit are low. Applicants need not visit ADEMI's offices. They can simply call the office and have a credit advisor visit their home or business at their convenience. Potential clients are not required to attend training courses prior to receiving a loan. Neither are they required to accumulate savings for a specified period oftime. 35 Group lending, which was initially adopted to increase outreach and keep costs down, was discontinued when it was found to increase transaction costs for clients, result in less individualised services and discourage the rapid growth of individual businesses (see Chapter IV). Finally, timely repayments almost automatically ensure access to larger follow-on loans, and the loan terms and interest rates are better than those provided by moneylenders (see Chapter IV). Areas in which services to clients could be improved include: • Repayment period. While clients are generally satisfied with the repayment periods on their loans, there are some exceptions (see Gomez [1988], Ledgerwood andBumett [1995]). This is because the loan terms have tended to depend more on the size of the loan than on the cash flows of the borrower. For example, whereas fixed capital loans are usually larger and are amortised over longer periods than wdrking capital loans (and even include a short grace period), there is little differentiation between the terms on fixed or working capital loans for different lines of activity for the borrower. A vendor with a high turnover of its working capital may receive the same two-month loan term on a working capital loan of a given size as a producer with a slower rotation of its capital. As a result, on occasion, clients have had to borrow from other sources to make timely repayments to ADEMI. • Preclusion of concu"ent loans. Clients at times come across opportunities that require a quick additional loan before they have paid off an existing one, however ADEMI requires that existing loans be paid off before new ones are issued. ADEMI might explore occasional use of concurrent loans for more seasoned clients if clients can easily service both loans as well as other known liabilities. • • Repayment process. In order to repay ADEMI loans, clients are required to visit the nearest branch of Banco Popular, make their monthly payments, and bring the stamped receipts to ADEMI. This two-stage payment process is extremely effective from ADEMI's point of view (see Chapter VI), but it does result in higher transaction costs to clients, who often have to wait in long queues at the bank. It also results iii costs to ADEMI when clients neglect to submit their receipts to the ADEMI office. Nevertheless, this is changing with the move to BancoADEMI, which plans to issue debit cards that would facilitate payments by clients and loan monitoring by the bank. 35 While, the screening and application processes for small business loans are similar.to those ofbanks, ADEMI is less bureaucratic, and will accept a broader range of collateral as well as a lower ratio of collateral to loan value (1.4:1, compared to 2:1 for banks). SUSTAINABLE BANKING WITH THE POOR 35 • Credit card applications. Although the approval process is supposed to be automatic once ADEMI approves a client for an ADEMI-Banco Popular Mastercard, approved clients in areas outside the capital have had their applications delayed or rejected by the local Banco Popular branch pending additional guarantees or income levels. ADEMI is aware of this problem and has been working with Banco Popular to fix it. Overall, though, the products and services provided by ADEMI appear to be extremely well-suited to clients' needs, as is evinced by ADEMI's sustained profitability, clients' response in interviews, data on the impact on clients (see Chapter VII) and ADEMI's high client retention ratio of 40 percent (see Figure 5.3). 36 Over time, clients will leave a lender for several reasons, e.g. they do not need a loan at the time, or they graduate to or turn to other sources of credit, or they discontinue their work in their enterprises (due to retirement, emigration, or alternative employment). Some of these reasons are a sign of success for the lender, others are either neutral or are implicit criticisms of the services of the lender. There have been no studies to track the reasons that some of ADEMI's clients have become inactive----although it appears that the highest departure rates for clients have occurred one year after an economic downturn, i.e. in 1991, 1992 and 1995 (see Figure 5.3). Nor have there been thorough analyses of retention ratios for other microlenders. However, ADEMI's 40 percent retention ratio appears to be higher than the ratio of active clients to total clients served by the successful microfinance programmes of PRODEM/Bancosol in Bolivia (under 25 percent, see for example PRODEM [1996]) and of the many group lending programs that report high turnover in group membership. Figure 5.3: A Look at the Growth and Dispersion of ADEMl's Client Base 20000 15000 10000 .l!l 5000 C .!! u 0 0 t -5000 .a §-10000 z -15000 -20000 -25000 Source: Authors' Calculations based on Quarterly Reports (see ADFMI {1997}) 36 This ratio measures the percentage of all earlier clients who are still active as of a given year. SUSTAINABLE BANKING WITH THE POOR 36 While a fairly high retention ratio might be an indicator of dependency or of inability to graduate, ADEMI's case suggests that its client retention ratio is a sign of success: most clients have developed a capacity to handle larger volumes of credit and service them in a timely manner, as well as to employ more workers and accumulate more assets (see Table 5.1 and Chapter VII). In this sense, clients have "graduated" but have chosen to remain with a lender they trust that has grown with them. As ADEMI itself "graduates" with the establishment of BancoADEMI, the Association will' be in a position to broaden its range of financial services and thus to match its clients' requirements even more fully_31 37 The client retention ratio will be a useful indicator of BancoADEMI's success, once it is established, since BancoADEMI will face significantly more competition than did ADEMI. SUSTAINABLE BANKING WI1H THE POOR 37 VI. HOW DOES ADEMI WORK? THE OVERALL STRUCTURE Institutional Structure. The supreme body in ADEMI's institutional structure is the general Constituent Assembly comprising all its associates. 38 It meets once a year and nominates a board of trustees, i.e., a Board of Directors, made up of7 to 21 members (currently 15), headed by a president. The Board has full powers of representation of ADEMI before third parties and of entering into contracts on behalf of the Association. Board members are elected for a two-year period but there are no term limits. They are not compensated for their activities or expenses. The Board members are Dominicans drawn from professional and academic fields, such as banking, commerce and industry. They are nominated on the basis of their ability to provide practical and political support to ADEMI within the local environment, instead of as contributors or fund-raisers. In addition to the general assembly meeting, the full Board of Directors meets once to discuss the budget and another time to review the audit. 39 At other times, executive powers are delegated to a seven- member Executive Committee of Board members-including the Executive Director-that it appoints (see Figure 6.1). The Executive Committee convenes approximately every six weeks. Its members· are more actively involved with the organisation, e.g., they are represented on the Credit Committee, which approves all the largest loans. However, the Executive Committee also delegates broad powers to the President who in tum meets with the Executive Director for most decision- making. The Executive Director implements decisions taken by the Board. He or she also directs the day-to-day operations of the Association, including supervising and directing the credit programme and overseeing all fmancial operations. He or she maintains relations with national and international agencies on behalf of ADEMI. In staff-related matters, the Executive Director is assisted by an internal Executive Committee comprising the Executive Director, the heads of ADEMI's departments, the internal auditor and the head of accounting. ADEMI has nine departments: planning and programming, administration and human resources, legal, fmance, treasury, small enterprises, microenterprises in Santo Domingo, microenterprises in the interior, and credit (see Figure 6. I). Around 75 percent of ADEMI's staff work directly with the clients, mainly as credit advisors and administrative assistants. The remaining 25 percent work on support functions, such as administration, treasury and computing/data management. 38 This section descnoes the structure of ADEMI. BancoADE.MI has adopted a very similar structure to that of ADE.Ml, and staff responsibilities, procedures and operating principles have carried over to the new bank, with only a few changes as discussed in Chapter vm. 39 External audits are con.ducted by a "Big Six'' firm and the results are made available to ADE.Mi's funders and to its Constituent Assembly. SUSTAINABLE BANKING WITH THE POOR 38 Figure 6.1: The Structure of ADEMI, as of January 1997 Executive Committee l External Audit ------I-~◄-----------. External Advisors___.. Executive Director l .__ _ _~-i....1--------!I . - - - - - - - - - - - .- - ,1 Internal Audit I IntemalExecuttve Commt1tee I - I I I l Dir. Plan. Director Dir. Small Dir. Micro- Credit Dir. Micro- & Program Finance Enterprises enterprises Director enterprises Sto. Domingo Interior Manager Legal Office of I I Admin& Office Treasurer I I Human RCMI RCMII RCMID Resources -Central '-- -Eastern Small Busi- . Central South East ness Credit -Payroll -Interior Mgr. VII _J _J -Purchasing -L. Mina ~es 1-14 -Herrera -San. Crist. -M. Plata -Filing ..... RCMV -San Pedro -Services ~ - -, -Bani i.',. S.D, East -H. Mayor Mgr. Planning WAccounting Officers 1-3 -Azua -San Juan -Romana & Program ~es 1-14 -Neyba -Higuey -Barahona -Sab. de Mar I ..... Data RCMVID ..... RCMIX -El Seybo Processing L- S.D. West S.D. East -Boca Chica H Statistics I I ~es 1-14 ~es 1-14 RCMIV RCMVI -- Public Relations RCMXI S.D. West -- RCMX S.D. East North Santiago [z;es 1-14 ~es 1-14 -San. Fco. 1-zon~s A-H -La Vega -Dajabon -Mao -Bonao Note: RCM = Regional Credit Manager SOURCE: JIMENEZ [1997] SUSTAINABLE BANKING WI1H THE POOR 39 Management Structure. The organisational structure of ADEMI provides for six senior management positions, in addition to the executive director. The Director of Planning and Programmes is responsible for monitoring and evaluating ADEMI's ability to meet its goals and objectives and coordinates most programmes, studies and agreements to this end. The current Director of Planning and Programming acts as Deputy Director in the Executive Director's absence. The Director of Finance conducts financial analyses of ADEMI's activities; supervises ADEMI's accounting and technology systems; manages the use of donor funds; authorizes petty cash, disbursement of funds and transfers among bank accounts; reviews the payroll; and is responsible for the preparation of financial statements and reports. The Director of Small Enterprises is responsible for all loans to small enterprises, i.e. for about 20 percent of the value of the portfolio, and currently supervises a manager and three loan officers. The Directors ofMicroenterprises, Santo Domingo, and ofMicroenterprises, Interior of the Country, oversee all field work conducted by the Regional Credit Managers (RCMs) and credit advisors (CAs ), and are responsible for overseeing the loan application process-including approving medium-sized loans-tracking loans in arrears, conducting staff evaluations and ensuring that the staffs training needs are met. The Director ofMicroenterprises, Santo Domingo, has five Regional Credit Managers, each of whom is responsible for 14 Zones of . lending activity-there is one credit advisor per zone. The Director ofMicroenterprises, Interior of the Country, has four Regional Credit Managers, each of whom is responsible for various branches and zones. He or she recommends new areas for ADEMI to engage in and liaises with community organisations. Given the nationwide scope of his or her work and the consequent potential for bottlenecks at the manag~rial level, his or her work is focused on operational issues and is supported by the newly created position of Director of Credit, who concentrates on credit issues. In addition to the six directors and the Executive Director, senior management comprises the Manager of Administration and Human Resources, who inter alia maintains the payroll and oversees record-keeping; the Legal Counsel, who supervises ADEMI's lawyers, perfects ·collateral and guarantees, and enforces loans· in arrears beyond 60 days; the Treasurer, who ensures that funds are available for planned disbursements and oversees the issuance of loan cheques; and the Internal Auditor, who reviews accounting and financial operations to verify their authentidty, accuracy and conformity with established policies and procedures. All senior managers report to the Executive Director, except for the Internal Auditor, who reports both to the Executive Director and to the Executive Committee of the Board. In a recent innovation, all the senior managers were assigned branches which they visit around 5-6 times per year to ensure regular contact with front-line activities and provide a direct channel for credit advisors and administrative assistants to voice their ideas and concerns to the senior management. The senior managers appear to enjoy the opportunity to change routines, visit the field offices and keep in touch with events "on the ground". Middle management also plays a key role in ADEMI's structure. It comprises Regional Credit Managers, who organize and supervise credit advisors' activities. They monitor CA's SUSTAINABLE BANKING WI1H TIIE POOR 40 visits to clients, supervise the quality of interactions with clients, participate in the evaluation of loan applications and technical assistance requirements and oversee staff training. Regional Credit Managers also approve all loan rene~als under RD$30,000. RCMs report to the Directors ofMicroenterprises and to the Credit Director. Branch Structure. ADEMI operates 25 branch offices throughout the country. All of them are within walking distance of a commercial bank branch in order to make loan payment as easy as possible for the clients. Four of the five offices located in Santo Domingo are owned by ADEMI, while the fifth one is rented. A typf cal Santo Domingo branch consists of a Regional Credit Manager, up to 14 credit advisors, a lawyer, and administrative assistants (generally one assistant for every three advisors). The branches are organized into blocks that reflect the potential market. Each credit advisor is responsible for one block. All reporting and monitoring of loans and expenses in the Santo Domingo branches is done through the Head Office. Branches in the interior generally have a manager, two to five credit advisors, one or two administrative assistants, and a lawyer on a retainer. Reporting and monitoring of loans and expenses is done at the branch with Head Office compiling the final reports and financial statements. ADEMI only opens branches in a given community if the Association has been invited to do so by a local organisation. The idea is to ensure a high degree of local support and cooperation. Many of the organisations thafinvite ADEMI also provide office space for its branches. These organisations include focal Chambers of Commerce (9), Development Associations (4), Foundations (2), the Catholic Church (I), a cooperative, a saving~ and loans association, a medical centre, and the youth club in Boca Chica. • In addition to the five branches in Santo Domingo, ADEMI operates in eight zones in the second largest city, Santiago, and has five northern branches in San Francisco de Macoris, La Vega, Mao, Bonao and Dajabon; six southern branches in San Cristobal, Bani, Azua, San Juan de la Maguana Neyba and Barahoma, and eight eastern branches in monte Plata, Hato Mayor, Sabana de la Mar, E Seybo, Higiiey, La Romana, San Pedro de Macoris and Boca Chica. Communications with headquarters are not always timely. On-line connections are expensive since the local phone industry is dominated by an inefficient public utility. O~y a few branches have modem connections to head office. The remainder rely on couriers for the delivery of cheques, diskettes and other financial records between the branches and the head office. This can lead to errors, and makes "real time" accounting impossible: for example, reports produced at the end of each month by the head office often reflect arrears on clien~' accounts in the Interior simply because the information has not yet reached the head office and been entered by the computing staff. Therefore management also places little weight on information regarding arrears of less than five days. Full autom,ation of the branches, further connection of the larger branches to headquarters, transmission of information from smaller branches via nearby larger ones, and the future use of Mastercards (or other automated banking cards) to make payments should reduce these problems. SUSTAINABLE BANKING WITH THE POOR 41 The key lessons to be learned from ADEMI's overall structure are that: · • it is quite straightforward with clearly defined responsibilities and chains of command. It is also fairly decentralised and bureaucracy is kept at a minimum; • the Board is made up of business-oriented nationals, and delegates most powers while retaining oversight over the budget, auditing and the largest single loan risk exposures; • the branch locations are determined by business prospects, convenience for clients and receptivity of the community; • the Association's design and functions are fully oriented towards its basic objective: provision of high-quality credit services to microenterprises; • Credit advisors are given considerable autonomy in decision-making, and are held accountable for their decisions through an effective monitoring and incentive system. Since the CAs' interaction with the clients is crucial, ADEMI's structure is designed in large measure to facilitate and support the work of the CAs. THE CRITICAL ROLE OF ADEMl'S HUMAN RESOURCES Strong management, attention to staff motivat!on and compensation, and insistence on quality service and respect for clients at all levels of personnel are cornerstones of ADEMI's success as an organisation. Particular emphasis is placed on selecting, training and motivating credit advisors to provide excellent service to clients, since CAs are the first members of ADEMI's staff that clients meet, and the ones with whom they interact most frequently. Credit advisors screen, recommend, monitor and advise clients. They enjoy a high degree of freedom in their work, together with clear accountability for the size and quality of their portfolios. The CAs typically arrive at their desks in the morning, receive a daily update on which of their clients has paid or not paid, or is due to make a payment that day or in the next few days. They then make several phone calls to clients and confer with other agents and with the manager, after which they spend most of the day visiting between four and ten existing or prospective clients by motorcycle. 40 Advisors typically cover a radius of two to five kilometres and are responsible for 120 to 160 individual clients. By spending up to three-quarters of their time meeting clients in their work-places, CAs become known in their zones, get to know their clients intimately, listen to their concerns and provide advice as needed. CAs can also easily be approached by new clients and can identify new opportunities and challenges for the Association. Staff selection and training. Credit advisors are selected informally, usually through recommendations from existing advisors, although the human resources department does receive 40 These motorcycles are leased by ADEMI to the CAs and become their property within four years. CAs are responsible for maintaining them and ADE:MI pays the CAs for the use of their motorcycles. SUSTAINABLE BANKING WITH THE POOR 42 applications from outside of the ADEMI network. Character, skills and enthusiasm are th~ key variables in determining appointments. However, potential credit advisors must also have an undergraduate degree, preferably but not necessarily in accounting, business administration, or economics. These qualifications are important insofar as they enable CAs to back up ADEMI's credit services with technical assistance. Most advisors in Santo Domingo are from the State University which draws its students from neighbourhoods similar to those in which ADEMI operates. In the interior regions, advisors must be from the town in which they wiH work. All credit advisors are male. 41 After interviews by every level of management, new CAs spend the first week reading through the operations manual for credit advisors and learning about the institution from the manager for human resources. Thereafter, they are given two weeks of hands-on training, accompanying existing advisors on their rounds, ·interviewing and screening potential clients and conducting all the work of a credit advisor. ADEMI has found that most candidates who last the • first two weeks stay on as advisors. Those that meet the requirements are assigned to their districts for another ten-week test period in which they inherit clients from credit advisors who are overstretched, and begin to build up a new client base. Once they enter into full service, advisors are encouraged to maintain their professional skills and are provided with ongoing training and professional development of a min4num of five days per year. Recently some of the more seasoned CAs have complained that the induction process has been shortened and that this has resulted fu i~ss inculcation of the ADEMI mystique. While the c~uld not confirm the accuracy or impact of this suggestion, it does highlight the authors _ importance that CAs attach to being part of a family with a v,ision. The annual meeting of all ADEMI' s staff provides an interesting opportunity to renew that vision and bring the ADEMI family together under one roof. At these "Encounters", ADEMI's staff exchange experiences, build collegiality, review performance and establish targets for the next year. Staff are also exposed to new techniques in other countries and .: motivated to look for new ways of improving interchanges with clients. They also discuss innovations such as the market for leasing or discounting, the ADEMI Mastercard, and the establishment of BancoADEMI. Sta.ff incentives. Staff incentives consist of merit-based promotions; competitive salaries·; a performance-related incentive scheme; and attractive insurance benefits: • Promotions. Most senior positions in ADEMI are filled from within, especially by successful credit advisors. Promotion incentives were particularly important during the rapid 41 Early on in its operations, ADEMI came up against an important outreach issue: should women be employed as credit advisors? While women's participation in ADEMI as credit advisors is not proscribed, it ended relatively early in ADEMI's operations, apparently because there is a cultural disposition in the Dominican Republic against women riding motorcycles (which·are the vehicles of choice for field agents). However, given the effective role women have played as field officers in other microfinance organisations, more could probably be done in this area. Women do play an important role in the office operations of ADEMI, such as accounting, auditing and treasmy, and there are women on ADEMI's Executive Committee and Board. Finally, BancoADEMI's female employees are actively involved in selling deposit and other services. SUSTAINABLE BANKING WI1H THE POOR 43 growth years. Although room for upward mobility has been fairly limited recently, the establishment of BancoADEMI will create new opportunities. • Salaries. Base salaries are closely linked to pay for comparable work in private companies, so that they are generally higher than salaries available in public sector work (although some credit advisors have been able to obtain higher salaries at competing micro- and small scale lenders). The average salary for credit advisors is RD$6,000 (US$425) per month in Santo Domingo and RD$5,500 (US$390) per month in the interior. The average salary of administrative assistants is RD$4,000 (US$28?) per month in the capital and RD$3,500 (US$250) in the interior. These base salaries (beforebonuses) place all credit advisors and assistants in the top half of the income distribution, and yield pre-bonus incomes for credit advisors that are 2.7 times GDP per capita (US$1,600 in 1996). These "efficiency wages" (see Stiglitz [1986]) reflect CAs' significant skills and are designed to ensure a high rate of staff retention. • Incentives. ADEMI has developed a staff incentive scheme that has become a standard for the microfinance industry worldwide. It balances competing objectives, namely outreach, portfolio quality and sustainability. The amount of the bonus is determined by the quality of the portfolio, the number of active microenterprises, and the size of the portfolio, with the greatest weight given to portfolio quality (see Appendix 3). To qualify for bonuses, CAs must have a ratio of contaminated portfolio (i.e. portfolio affected by arrears) to total • portfolio of no more than 4 percent, and at least 130 active clients. These standards have been increased over time to boost both sustainability and outreach to microenterprises. 42 In addition to these individual performance-related bonuses, all ADEMI staff receive Christmas and year-end bonuses equivalent to at least 3 months of base salary. • Insurance benefits. ADEMI is one of a fairly small number of Dominican enterprises that provides a pension plan for its employees. It is an actuarially sound, defined benefit pension plan that entitles employees to up to 80 percent of their last two years of salary after twenty years of service, as well as access to subsidized employee loans for housing or consumption (see Chapter VII and Appendix 8). In addition, employees are insured against accidents and death, and all family members a,r~ covered by health insurance. Finally, all ADEMI staff receive severance payments whether they quit or are fired, since ADEMI does not want to retain dissatisfied employees. • Job Satisfaction. ADEMI's competitive pay and benefit packages are important in attracting and keeping staff-turnover is fairly low, with departure rates of3-11 percent during 1992- 95, and new staff accession rates of 12-18 percent over the same period. However, most CAs enjoy working at ADEMI maiiily because of the high degree of autonomy that their 42 For example, when it was rolled out in 1987/88, CAs could qualify for bonuses with a ratio of arrears to loan portfolio (!!Q! portfolio affected by arrears to total portfolio) of under 10 percent, and with 90 rather than 130 clients. The switch in 1997 to tracking the contaminated portfolio rather than arrears over loan portfolio proved to be immensely successful (see Chapter VD). Other ways in which the incentive scheme has been changeq since 1987/88 to boost productivity include rewarding CAs for increments in their portfolios or number of clients beyond the minimum targets. SUSTAINABLE BANKING wrrn THE POOR 44 work-style offers and because of the satisfaction they derive from assisting entrepreneurs.· For example, most CAs share the view expressed by one CA that: "When our clients succeed, we succeedi. t , s very g- 1 ymg. ,, tif" FOLLOWING THE DOLLAR: THE LENDING PROCESS The lending process involves at least three distinct phases that are discussed below: loan origination, approval and disbursement. Loan origination. Before any client can apply for an ADEMI loan, she or he needs to know that ADEMI exists. Most new clients learn about ADEMI through the best kind of advertising, namely word of mouth from satisfied clients. Indeed, until 1995, _80 percent of ADEMI's clients learned about it this way. That year ADEMI launched a print and radio advertising campaign to head off competition and boost demand for loans after new fman~ing became available from the EIB. This campaign has been stepped up recently in response to the Government's PROMIPYME initiative. ADEMI also retains a press officer who prepares press statements, and its credit advisors distribute brochures in their respective zones. Finally, ADEMI gains exposure through independent press stories, referrals from Banco Popular and promotions by host organisations outs~de the capital. The lending process begins when a_ potential client contacts ADEMI's credit advisor or the ADEMI office, and the advisor visits the client's bu~iness (see Table 6.1). The advisor begins to evaluate the applicant's character and entrepreneurial skills. The CA usually pays a second, surprise visit, after obtaining and verifying information with other entrepreneurs, suppliers, customers, neighbours and family members on.th~ applicant's business acumen, moral reputation and relationship with the community. The CA also constructs a fmancial statement for the enterprise, evaluates key fmancial ratios (e.g. inventory relative to sales, usage of installed capacity, liquidity, solvency, profitability, payment capacity, etc.), conducts an inventory of machinery and equipment, and prepares an investment plan, based on the entrepreneur's proposals. Drawirig up a fmancial statement is in many cases part of the learning process for ADEMI's lower income entrepreneurs (in other cases, they have already learned about the fiscal benefits of maintaining two sets of accounts). A standard application form is used for all loans. Clients are also required to submit information on guarantors, deeds and collateral (if any) as part of their application. There is little difference in the way that microenterprise loans ofUS$100 and US$10,000 are processed, . although more rigourous credit analysis processes are applied to loans to small-scale enterprises. Indeed, small business applicants are required to provide articles of incorporation, a letter from the enterprise's Board of Directors empowering the applicant to seek a loan from ADEMI, a business plan, cash flow and other projections, audited fmancial statements and collateral that can be perfected and enforced. Credit advisors in the small-scale enterprises department receive special training in credit assessment. SUSTAINABLE BANKING WI1H TIIE POOR 45 Table 6.1: How Microenterprise Loans are Made and Recovered at ADEMI iii t¼t\t1t!liRPli~oUtillijl'll{\JimrnI ~IbMtrt!"'iJ,f.l;lyiij'ijrt(~~Jl!!wti r@Mm1P~rac",11;11ec1~iPffic;, fo1t,:t 1 Applicant requests loan 2 CA visits potential applicant, explains loan terms and procedures 3 Applicant prepares eligibility CA gathers information on requirements applicant's character and creditworthiness 4 Applicant fails screening, (process ends), or passes 5 Applicant completes loan CA completes loan application with CA application with client 6 Applicant signs deeds and obtains guarantor pledges 7 CA appraises loan and Application is processed:~ 15 processes application days for new loans; ~ 3 days for repeat loans 8 credit committee assesses application, including the CA's appraisal 9 CA Informs client Loan is rejected, or more information is requested or loan is approved 10 Approv~d application is sent to HQ with proposed terms, & request for cheques & loan contract 11 Dir. Finance receives approved applications; systems staff input data; cheques are printed. 12 Two signing officers sign cheques; couriers carry cheques and documentation to the relevant branches 13 Client collects cheque, reviews For new loans, CA invites pro9edures and regulations & several new clients to a signs contract ceremony at the branch 14 Client makes payments at CA visits client, verifies loan HQ provides monthly -reports on specified commercial bank to use, provides TA dues/dates due, & verifies ADEMI account, brings payments with branch records receipts to ADEMI branch and commercial bank 15 Client repays loan & may apply CA may process new loan, HQ may process new loan, or for new one, or client goes into or may enforce or restructure legal dept. may assist with arrears loan in arrears enforcement of existing loan Source: Ledgerwood and Burnett [1995] Loan approvals. The credit advisor has major incentives to maintain a healthy portfolio. He often turns down applications without submitting them to his managers for approval, e.g. because of what he learns from the community about the prospective client's character or SUSTAINABLE BANKING WITII THE POOR 46 bu~iness prospects. Of course, in most cases the credit advisor will blame any rejection on an internal committee somewhere in ADEMI that did not understand the applicant's needs, rather than on his own judgement. When the credit advisor is persuaded that the loan is worth making, he takes it to his manager for approval. In the case of small repeater loans (under RD$25,000), the regional credit manager makes the final approval. In the case of larger loans, an internal credit committee approves the loan (see Table 6.2). Authority for loan approvals depends on loan sizes and on whether loans are to new or existing clients. The.approval process has been decentralised significantly over time. 43 The credit committee evaluates applications on the basis of objective criteria such as the maturity of the business, its credit history, asset accumulation, cash flow and market potential, the purpose of the loan and the client's repayment capacity (see Lewin [1991]). 44 The legal department also verifies all legal documentation and registration and control of guarantees. However, the approval process relies fundamentally on the credit advisor's "instinct" and incentive to perform in order to build a healthy portfolio. Indeed, while the CA is sometimes asked to re-work an application or to inform the client that the loan has been rejected, over 80 percent of the loans recommended by credit advisors are typically approved. Credit Advisor and Regional Credit ~RD$30,000 Manager CA, RCM and Director of Microenterprises ~RD$30,000 (no formal meeting) CA, RCM and 5 ADEMI Directors >RD$30,000 RD$30,001- (no formal meeting ) RD$500,000 RCM, 5 ADEMI Directors, 1 Board >RD$500,000 . Member (formal credit committee meeting) Source: Jimenez {1997} Loan Disbursements. Once a loan is approved, the application is sent to the finance department, which records the amount of the loan, verifies the availability of funds with the treasury department, and foiwards the application to the systems office for data entry. This office produces a cheque which is signed by two members of senior management. In most cases, cheques on new loans are disbursed within 10 days of application and on repeat loans within three days. Cheques are made out to the entrepreneur or enterprise, are drawable on ADEMI 43 For example, in the 1980s, all loans llllder RD$15,000 were approved by a committee comprising the credit advisor, the branch manager, the Director for Microenteiprises and the Executive Director. Now the cut-off has risen with inflation to RD$30,000, but the Director for Microenteiprises is no longer involved in approving repeater loans under RD$30,000 and the Executive Director is no longer involved in approving smaller loans at all. 44 While credit committees do not necessarily meet fonnally, three of the five Directors, including one from the finance department, must approve each loan. SUSTAINABLE BANKING WITH THE POOR· 47 accounts at local commercial banks (usually Banco Popular) and must be cashed in person by the client. They are handed over to the client at an ADEMI branch upon signature of the loan contract, and are accompanied by legally binding IOUs for each payment due and by deposit slips specifying the amount and date on which repayments into a specified ADEMI account are due at the local commercial bank. In the case of loans to new clients, borrowers are invited to sign their contracts and receive their cheques at a formal ceremony at the nearest ADEMI branch. During this ceremony, the director or manager officiating welcomes the client to ADEMI and reviews the terms and conditions of the loans, ensuring that ADEMI's rules are cl~ar to clients and that clients' expectations are also understood .and corrected if needs be. Since 15 to 30 clients usually attend at a time, this is also an opportunity for clients to meet other new clients in the same area. Loan Renewals. Repeat loans are processed more quickly than initial loans. Established clients are also monitored less intensively than first time borrowers. Approvals of repeat loans· are based not only on the client's repayment performance but also on the growth of the business. If the enterprise stagnates, the borrower's ability to access new loans may be interrupted while the client and advisor figure out ways of galvanising the business. In some cases, the client may be invited to the branch to discuss their situation with the RCM before a subsequent loan is issued. If clients repay their loans in accordance with the terms and conditions of their contracts, they are guaranteed almost instant access to larger loan amounts, with increments tailored on a case-by-case basis. By linking access to larger future loans to current performance, ADEMI implicitly enters into multi-period contracts with clients that enable it to reduce moral hazard problems relative to one-off loans. Moreover, while loans are initially sub-optimally small, they are much larger in subsequent periods than would be possible in the face of moral hazard problems under one-time loans (see Benjamin [1994b]). FOLLOWING THE DOLLAR: THE LOAN RECOVERY PROCESS Loan Recovery. Clients are required to amortise their loans via monthly payments (there are as yet no balloon loans with full payment at maturity, although ADEMI has considered offering this product). First, a client deposits his or her payment mto an ADEMI account at a local commercial bank and obtains a receipt from the bank. The client must make his or her payment to the bank in cash, in order to avoid the possibility of bounced cheques. Next the client takes that receipt to ADEMI, an administrative assistant records the payment and the client is given a receipt corresponding to the IOU that has been paid off. 45 Finally, the payment information is transferred to ADEMI's headquarters in Santo Domingo, cross-checked against 45 Credit advisors are also not allowed to accept cash from the client or to take the bank receipt to ADEMI on behalf of the client, although both of these occur on an exceptional basis (notably when loans are in arrears). In e~ceptional cases, cheques are also accepted from clients, but, due to the need to clear cheques, the corresponding payments are recorded as being received one month late. SUSTAINABLE BANKING WITH THE POOR 48 information from the bank, and included in daily and monthly monitoring reports that are provided to credit advisors and managers, respectively. In most cases the transfer of information between branches and headquarters is via courier and fax, since only two branches -- both in . Santo Domingo --have modem links to the main office. This innovative loan recovery process is one feature of ADEMI's operations that can readily be replicated by other microfinance providers. Its advantages are that: • Clients are brought into closer contact with the formal financial system and are exposed to a range of financial products (e.g. savings, insurance or payment services);. • The banks -- mainly Banco Popular and Bank of Nova Scotia - acquire essentially free information on potential clients as well as service fees on ADEMI' s accounts; • ADEMl's administrative costs of disbursing and recovering loans are reduced by riding on the banking infrastructure; • Loan supervision is easier and the potential for fraud or theft is reduced. ADEMI' s accounting and audit departments have detected only three instances of fraud involving its staff, in almosfl5 years. In each case, ADEMI has prosecuted the parties involved. ADEMI recognizes a greater potential for automation in the disbursement and payment systems, such as direct deposit of loans into client accounts, use of ATMs for payments, and · down-loading account activity information from the banks directly into ADEMI's information system. To date, the condition of local banking technology has restricted the Association's ability to innovate in this manner. Clearly, once BancoADEMI takes over the branch network and clients are transferred to BancoADEMI, the need for outside banks to accept payments on behalf of ADEMI will end, which will simplify the repay_ment process for clients. While cash handling issues will arise for BancoADEMI, it will also be able to sell a broader range of banking services to clients. Loan Monitoring and Supervision. ADEMl's management information system (MIS) is one of the keys to ADEMI's success. It is designed to provide staff with regularly up-dated information to ensure rapid response to loan delinquencies, and to assist with managerial decision-making. The system is sophisticated without being overly bureaucratic. It generates two key monthly reports that guide the lending and supervision process. The first report lists payment due dates for every credit advisor's loan portfolio and is updated daily using information on payment receipts provided by administrative assistants. CAs use the daily updates to arrange their schedule and visit clients who have fallen behind in their payments or who have a payment falling due shortly. The second report is an "Aging of Receivables Report" which is generated by the Head Office for each advisor at the close of every.month and guides procedures for loan collections. Managers discuss both reports with their CAs to diagnose problems, discuss strategies for loan collection and evaluate the performance of CAs. The MIS also tracks information on clients' characteristics, including certain variables of interest to donors, e.g. the gender of borrowers (see Chapter V); on ADEMI's impact on clients (see Chapter VIl); and on the Association's accounting and fmances. A summary of these data are presented in Quarterly Reports (see ADEMI [1997]) that are made available to the public. SUSTAINABLE BANKING WI1H THE POOR 49 The MIS runs under a customised computer program that was written for ADEMI, and ADEMI • retains the seivices of a programmer to update and fine-tune the system. One deficiency of the MIS is that all data are in nominal terms, which is distortionary in years of high inflation. In addition to central monitoring of the portfolio, ADEMI engages in fairly.active field level supeivision of clients through its credit advisors. Unlike at many other development institutions, the credit advisor's work does not end with loan approvals. On the contrary, his income is tied to the success of his clients and he is held directly responsible for the quality of the loans that he has recommended. Consequently he follows up with visits and phone calls. Advisors are required to visit their clients at least once each month. The first visit usually comes two to three days after a client has received a loan, in order to ensure that it has been used for the designated purpose. Thereafter visits are conducted every two weeks or so, often around the time that payment installments are due. Visits may last from five minutes to a couple of hours. The advisors usually visit four to ten clients a day. During these visits, advisors build important individual relationships with their clients which go far beyond that .of a collection agent. Clients inteiviewed indicated that they felt they could speak openly with advisors who, they felt, contributed to the success of their businesses. The situation is different for small business borrowers, who do not receive regular visits from loan officers unless specific problems arise. Loan en/orcement. Clients are enticed to repay by the promise of immediate access to larger loans, with increments that are contingent on the client's overall performance. ADEMI ensures that it has sufficient cash flow to honour this commitment -- otherwise timely repayment would be discouraged. Late payment is penalised by a 5 percent fee after five days. This fee can be waived if there is a good reason for the delay. If an installment is more than 30 days late, a process for loan collection is launched that involves reminder letters from the branch manager after 30 and 60 days, and the automatic involvement of the legal department after 60 days (see Figure 6.2). The "stick" of legal remedies is applied only when other means of encouraging repayment have been exhausted. 46 The advisors continue to be responsible for collecting the amounts due until the loan is written off, and together with their managers, they will try to work with clients to secure repayment. If the client does not repay an installment within 90 days, the entire loan is immediately due and payable. Clients who are not willing to restructure or pay their loans are contacted by the sheriff, property is embargoed and within one day it is confiscated unless full payment is received-thereafter, the goods are auctioned following a two month redemption period. Clearly, without the legal and cultural acceptance of sheriffs in the loan enforcement process in the Dominican Republic, ADEMI's ability to collect loans would be greatly constrained. 46 From an accounting perspective, loans are classified as delinquent one day after the due date. They are considered uncollectt"ble only if the client is insolvent or has disappeared. Loans that are handed over to the legal department for collection continue to accrue interest, but 100 percent provisions are made against this interest Loans are not written off until at least one year after they first go into arrears. See Chapter VII for an analysis of arrears. SUSTAINABLE BANKING wrrn THE POOR 50 Finally, the no-nonsense approach to loan enforcement has enabled ADEMI to recover over 99 percent of its loan dues (see Chapter VII). If ADEJ-.fl had adopted a more paternalistic and forgiving approach, it would not have been able to serve nearly as many low income entrepreneurs. SUSTAINABLE BANKING WTIH TIIE POOR 51 Figure 6.2: ADEMI Enforces Collection on all Loans within 90 Days Client is eligible for larger Credit advisor visits client to follow-on loan ,----- ------ ./ determine if there is a problem Yes No Client avoids penalty fee I Client is assessed late payment fee of 5% CA tries to help resolve problem Yes Client pays·penalty, A reminder letter is issued by branch can receive new loan manager, loan is restructured for "can't pay'' situations Yes No Client pays penalty, A second, stronger letter is issued by branch can receive new loan manager for "won't pay" situations; client invited to the office; lawyer comes into case Client repays, but may ◄ no longer receive loans .Yes fromADEMI No Client is issued notice of legal proceedings; clients who are not willing to restructure or pay are contacted by the Sheriff; guarantor is contacted; foreclosure proceedings are launched; Court of the Justice of the Peace seals loan contract; ADEMI legal department demands payment or embargoes property; after one day, clerk of the court confiscates equipment; goods are auctioned after 2 months Sources: Ledgerwood and Burnett [1995}, Lewin [1991J SUSTAINABLE BANKING WITH THE POOR 52 VII. ADEMl'S PERFORMANCE THE PROFITABLE NON-PROFIT ADEMI has been exceptionally successful in the world of microfinance. Only one other institution (the Bank Rakyat Indonesia-BR!) can point to five or more consecutive years of self-sufficient provision of financial services to microenterprises. The experience of these pathbreaking institutions has been emulated by microfinance providers in numerous countries around the world. Moreover, ADEMI's and BRI's self-sustained outreach to low-income clients proves that "banking for the poor'' can be a viable proposition (see Benjamin [1994]). The first thing that is striking about ADEMI is the rapid rate at which both its assets and its net worth have grown, (see Figure 7.1, Table 7.1 below, and Appendices 4 and 6). In constant peso terms, ADEMI's assets increased ten-fold during the 1980s, and increased another nine times during the 1990s. The increase in equity was less dramatic during the 1980s (only five- fold in real terms), but ADEMI's equity at end-1996 was 14 times higher than at the start of the decade. As noted earlier, ADEMI's net worth and assets place it among the leading financial institutions in the country. • • Figure 7.1: ADEMl's Assets and Net Worth (Figures in US$ Millions) 35 30 25 20 15 10 5 0 I ■ Equity (US$m) ■ Assets (US$m) Source: ADF.MI [1997] Although ADEMI is a non-profit institution, it has earned remarkable surpluses on its net worth. The return on equity (ROE) peaked at 106 percent in 1991, together with market interest rates (see Chapter II), and has ranged from 30-40 percent in the past 3 years, compared to around 25 percent for commercial banks in the D.R. Its return on assets (ROA) during the last 5 SUSTAINABLE BANKING WITH THE POOR 53 years (14-20 percent) has also far exceeded that on most commercial bank assets (1 to 2 percent). ADEMI does receive subsidies that most for-profit institutions in the D.R. have not enjqyed, including access to grants and cheap loans, exemption from reserve requirements on deposits, and of course exemption from income taxes. In 1996, these amounted to a hefty US$1.3 million, excluding the tax exemption. Moreover, given its legal status, ADEMI does not have to distribute dividends to shareholders with alternative investment opportunities-indeed, the nominal or accounting cost of its equity is zero. Nevertheless; in pre-tax terms, ADEMI has earned sufficient returns each year since 1991 such that if all its borrowings had been raised at market rates (where these also reflect the costs imposed by reserve requirements) and if all its grants had been removed, ADEMI would still have earned a return on shareholders' equity in excess of the market rate of interest (see the calculation of the Subsidy Dependence Index, or SDI, in Appendix 5). 47 In this sense, even though ADEMI benefits from subsidies, ADEMI is so profitable that it does not depend on them for its continued operation: if the subsidies were to be terminated, ADEMI would still be a fully viable financial institution (see Figure 7.2). Figure 7.2: ADEMI's Subsidy Dependence Index Measures 150.0% ~ 120.0% ll( ,, GI .5 90.0% II u ,, C C GI 60.0% GI a. GI C 30.0% ,, >, 'iii .Cl :II 0 .0% II) -30.0% Source: Benjamin [1994], ADE.Ml [1997] 47 The SDI measures the amount by which interest income would have to increase to ensure self-sustaining operations if all subsidies provided to a financial institution were to be eliminated (see Y aron [19921). A negative SDI indicates that, if all subsidies are removed, the institution could lower rather than raise on-lending interest rates and still be completely self-sustaining. In ADEMI's case, the SDI actually turned marginally positive in 1996 (indicating subsidy dependence), however with the index at only 0.1 percent, and with the required change in average on-lending interest rates to ensure viability equal to less than five basis points, ADEMI can be said to be "subsidy-free" in 1996. SUSTAINABLE BANKING WITH THE POOR 54 DECOMPOSING ADEMl'S EXCEPTIONAL RETURNS There are over 1,000 providers ofmicrofinance services around the world that have at least three years' experience and at least 1,000 clients (see Paxton and Cuevas [1996]). Yet only a handful of them have come close to achieving the full self-sustainability that ADEMI has achieved in serving low-income clients. What explains ADEMI's exceptional performance? Earlier chapters considered operational aspects of ADEMI's approach to microfinance. This chapter takes a financial perspective, decomposing ADEMI's return on assets into its constituent elements to elucidate the sources of profitability (Figure 7.3). First, however, it · briefly reviews some of ADEMI's accounting conventions in order to assess the quality of the financial data on which the subsequent analysis is based. Figure 7.3: Analysing a Fin.ancial Institution's Return on Assets •Quality of Portfolio Provisions for Loan Losses •Default Risk Total Revenue •Loss Given Default •Composition of Profit Margin: Interest Expense • Interest Expense Liabilities Net Income . Revenue . Total Total Liabilities . •Interest Cost of Total Revenue Each Liability. ••Salaries & Employee Benefits ROA: •Occupancy Expenses Non-Interest Expense Netlncome - X Total Revenue •Provisions for Foreign Total Assets Exchange Risk •Other Expenses . Asset Utilization: •Yield on Each Asset Interest Income Total Revenue •Volume of Earning Assets Total Assets Total Assets •Composition of Assets •Loan and Guarantee Fees Non-Interest Income • •Investment Securities Gains Total Assets •Grant Income •Other Income Source: Koch [1992] Quality ofFinancial Data and Accounting Conventions. ADEMI has a very good accounting and management information system. Its fmancial statements are audited by an internationally renown auditing firm and ADEMI has received unqualified audits since its inception. There are ~nly three accounting policies that warrant mention as they materially affect ADEMI' s reported fmancial position: SUSTAINABLE BANKING WITH TIIE POOR 55 • International Accounting Standard 29 calls for adjustment of financial statements whenever cumulative inflation over any three-year period exceeds 100 percent (equivalent to an annual 26 percent). During 1984-85, and again_during 1988-91, annual inflation rates exceeded 26 percent, but ADEMI, like all other Dominican financial institution did not adopt inflation accounting. With non-monetary assets equal to only 15-29 percent of equity during 1989-91, inflation adjustment of non-monetary assets and equity would have led to net restatement losses. • Delays in the transfer of data to the Head Office from some branches in the interior lead to an overstatement of arrears under 3 0 days, as timely payments in the interior are not always reflected in real time in the reports drawn up at the Head Office. 48 • Whereas Dominican commercial banks are required to report interest income on a cash basis since the 1990 financial crisis; ADEMI reports interest income on an accrual basis, in accordance with generally accepted accounting principles. However, ADEMI used to accrue interest on non-petforming loans (NPLs) until the legal.department declared them to be non- recoverable, at which point they were fully provisioned. This led to an over-statement of incomes on those non-performing loans that had yet to be written off (see the discussions of portfolio quality and credit risk below). Two years ago, upon the recommendation of the external auditors, ADEMI ceased to report as income any interest accrued on NPLs over 90 days. Even before that, ADEMI's auditors ruled out income recognition in audited financial statements for ail.y interest accrued on NPLs·that were under legal recovery proceedings. With the exception of these caveats, the quality of the financial data is extremely good, therefore the financial statements have not been adjusted for the purposes of this study. ADEMI 's Profit Margin and Asset Utilisation. In 1996, ADEMI enjoyed a profit margin (i.e. net income over total revenue) of 44 percent, up from the 35-40 percent range that has prevailed since 1989 (except for an unusually high 55 percent profit margin in 1991 ). These figures are considerably higher than the average for the commercial banking industry in the D.R. (around 18.2 percent in 1995), oi- for US banks (around 10-11 percent). ADEMI has achieved this both by containing costs (see below) and by generating substantial revenues. For example, in 1994, ADEMI's ratio of total revenues to total assets was 37 percent, compared to 18 percent for Banco Popular, which is widely regarded as one of the best-managed banks in the Dominican Republic, and to under 10 percent for US banks. While the profit margin has been rising, the _ asset utilisation ratio (i.e. total revenues over total assets) has been declining fairly rapidly, due to declining market rates and major shifts in the composition of the portfolio. In particular, the ratio has fallen by 20 percentage points since the inception of the EIB small- and medium scale lending program in 1992 (see Table 7.1). Taken together, this explains the decline in ADEMI's return on assets over the past five years. 48 In the absence of improved data transfer mechanisms, ADEMl has opted to close out monthly statements later in the month, i.e. around 15th-20th. This improves the accuracy of reporting but clearly also affects managers' -ability to respond quickly to sudden changes in financial perfonnance. SUSTAINABLE BANKING WI1H THE POOR 56 Table 7.1: ADEMl's Financial Statements Yield Impressive Financial Ratios 1. Balance Sheet (US$m) 1984 1986 1988 1990 1992 1994 199t - - - - - - 21,524 Loan Portfolio 381 419 770 1,691 5,354 12,775 Total Assets 485 721 11010 21246 61956 16!009 28,98f Borrowings 178 438 664 1,419 3,612 8,871 15,083 Total Liabilities 264 467 697 11520 31963 91371 16 19~ Total Capital 221 254 312 726 21993 61638 12,791 Total Liab. & Capital 485 721 11010 21246 61~6 161009 28,98f , 2. Income Statement (US$m) 1984 1986 1988 1990 1992 1994 1996 Interest Income - - 134 -- 184 - - 359 - - 926 - - 3,340 - - 5,526 - 8,182 Grant Income 226 214 26 78 44 - 9( Other Income - - 19 54 292 461 79: Total Revenue 360 • 398 404 11059 31676 5!987 9 06~ Interest Expense 18 6 31 137 717 1,266 l,01~ Loan Loss Provisions 13 21 20 56 220 379 55: Other E~eilSes 226 193 261 464 1!336 21119 3 52~ Total Exf'!.enses 257 220 313 657 21273 31764 5 10: !Net Income 103 178 l ·• 91 401 1,403 2,223 3,96~ Net Income excl. Grants (123) (36) . 65 323 1,359 2,223 3,87~ 3. Overall Return 1984 1986 1988 1990 1992 1994 199t Return on Equity (ROE) - - - - - - - - - - - 46.7% 70.0% 29.2% 55.3% 46.9% 33.5% 31.0o/c Equity / Assets 45.5% 35.3% 30.9% 32.3% 43.0% 41.5% 44.1 o/c Return on Assets (ROA) 21.2% 24.7% 9.0% 17.9% 20.2% 13.9% 13.7o/c ROE excl. Grants -55.8% -14.2% 20.7% 44.5% 45.4% 33.5% 30.3o/c ROA excl. Grants -25.4% -5.0% 6.4% 14.4% 19.5% 13.9% 13.4o/c Profit Margin J Net Income/Total Revenue 28.6% 44.7% 22.6% 37.9% 38.2% 37.1% 43.7o/c Interest ExpeI1Se/Tot. Rev. 5.0% 1.5% 7.8% 12.9% 19.5% 21.2% l 1.2o/c Interest ExpeI1Se/Borrowings 10.1% 1.3% 4.7% 9.6% 19.9% 14.3% 6.7o/c Loan Loss Provis./Tot Rev. 3.6% 5.4% 5.0% 5.3% 6.0% 6.3% 6.2o/c Non-Interest Exp./Tot Rev. 62.8% 48.5% 64.7% 43.9% 36.3% 35.4% 38.9o/c Asset Utilization Total Revenue/Total Assets 74.2% 55.3% 40.0% 47.1% 52.8% 37.4% 31.3o/c Interest Income/Total Assets 27.6% 25.6% 35.5% 41.2% 48.0% 34.5% 28.2¾ Interest Inc./Loan Portfolio 35.2% 44.0% 46.6% 54.7% 62.4% 43.3% 38.0o/c Non-Interest Inc./Tot Assets 46.6% 29,7% 4.5% 5.9% 4.8% 2.9% 3.0o/c are annual averages, rather than year-end values. Source: Appendices 4 and 6. Note: Balance sheet figures_ Utilizing Assets to Generate Interest Income. ADEMI's loan portfolio grew by an average of 46 percent per year during the 1990s, and now exceeds that-of all 16 development • banks, both mortgage banks, and 14 out of 19 savings and loans associations and two of the 15 commercial banks in the D .R rCTecnicas de Planificaci6n [1997]). In recent years, ADEMI' s SUSTAINABLE BANKING WI1H TIIE POOR 57 : •• I ; .. ...) . . ; -~ ·.: -:~ -1·: .: • •• .. Jo :-·· ~-- - JI.J_;_ .', :;.:: · :.,-.-!:-:: ·: • loan portfolio has accounted for 76-80 percent of assets (Appendix 6). Only in 1996 did the portfolio decline to 70 percent of assets, as ADEMI took an unusually liquid position that was apparently unplanned, with cash and bank deposits more than doubling to 17 percent of assets. While the ratio of earning to total assets-has remained relatively steady, the composition of the portfolio has changed dramatically in recent years: in addition to the sharp swing towards the retail sector and towards loans to women entrepreneurs, there has been a major shift in lending to small-scale enterprises, from under five percent in 1991 to over 20 percent of the portfolio by 1994 (see Chapter V). Since these loans tend to be for longer _terms (up to four years), the ratio of new disbursements to loans outstanding has fallen from 2.9 in 1991 to 1.3 in 1996. Moreover, with generally declining interest rates in the Dominican economy and with small -business loans at much lower effective interest rates than microenterprise loans (28 percent vs. 65 percent p.a.), ADEMI's interest income over loan portfolio fell from 62 percent to 38 percent during 1992-96. However, these trends do not alter the fact that ADEMI's interest income over assets remains substantially higher than that of other financial institutions in the D.R. Utilizing Assets to Generate Non-Interest Income . Non-interest income in ADEMI's case consists of premia on life insurance policies, which are mandatory for unsecured micro enterprise loans and are assessed at a rate of 28 basis points per peso lent (this is a profitable line of business, given the low frequency of indemnifications); income from disposition of acquired assets; fee income from Banco Popular related to the ADEMI Mastercard; and income from grants. While ADEMI charges up-front fees on loans (see Table 4.2), these are classified under interest income on loans.49 The composition of ADEMI's non-interest income has changed dramatically since ADEMI's inception. Initially, all non-interest income was grant income. Indeed, until 1986, grants accounted for more than half of all ADEMI' s income, which is why it is essential to take into account the grant element in any financial performance ratio comparisons with ADEMl's early years of operation. However, since 1992, there has only been one year in which grants have accounted for more than 1.2 percent of total revenues. While other non-interest income has grown slightly as a share of total income, rising to around 9 percent by 1996, it has declined relative to total assets, from 4.2 percent in 1992 to 2.7 percent in 1996. It is not likely to become a growth area for ADEMI, unless major new products are introduced by BancoADEMI. Boosting the Pro.fit Margin by Containing Interest Expenses. With the asset utilisation ratio declining in recent years, ADEMI has worked hard to convert a growing share of its revenues into net income by containing costs. One of the key areas in which ADEMI has consistently aimed to maintain low costs is interest expenses. First, it has maintained a low debt: equity ratio. This has never risen above 2. 5: 1, and in recent years has hovered around 49 • Also classified under interest income is ADEMI's return on deposits with banks. This category should be broken out from interest on loans to clients in future to increase transparency, particularly if cash and banks and other short term investments remain as high as they were in 1996. SUSTAINABLE BANKING WITH THE POOR 58 1.3: 1, compared to an average of 10: 1 in the Dominican banking system and over 14: 1 in US banks. By generating vast profits and ploughing them back into its operations (which it must, given its non-distribution constraints as a non-profit), ADEMI has held total borrowings below 70 percent of its net loan portfolio in recent years and thus substantially reduced interest expenses. Second, ADEMI has, to the extent possible, substituted cheaper for more expensive sources of funds. During the 1980s, this was done by securing grants and long-term, low- interest loans, both internationally from the InterAmerican Development Bank (IDB) and locally from the Technical Secretariat of the Presidency (STP). These cheap funds (at I percent and 8.5 percent per annum) largely replaced the Banco Popular line of credit (around 20 percent per annum), which was used mainly as a liquidity back-stop (see Figure 7A and Appendices 4 and 6). During the early 1990s, ADEMI secured funding from the USAID-sponsored apex lender FondoMicro, but replaced these fairly cosdy funds with other sources, particularly once competition intensified between the two agencies. After 1992, ADEMI began to mobilize local savings in the form of "loans from clients;" in accordance with the Commercial C9de of the D.R. These _now account for 24 percent of all borrowings. It also borrowed abroad from the International Investment Corporation (TIC) at Libor plus 6 percent and from the European Investment Bank (BIB) at 2 percent plus a capped foreign exchange risk (discussed below). 50 The EIB and ITC loans currently account for over 70 percent of ADEMI's borrowings (see Figure 7.4) The net effect of these shifts in the composition of liabilities has been a sharp decline in ADEMI's interest expenses on borrowed funds from 22 percent to 7 percent in the last four years. As a share of income, they have declined to just 11 percent, compared to approximately 40 percent for both domestic and US commercial banks. Factoring required foreign exchange reserves associated with the EIB loans into the cost of funds raises the 7 percent figure to 12 percent of total borrowings, but this is still much lower than the 22 percent cost of funds four years earlier. 51 Boosting Profits by Containing Non-Interest Expenses. The benefits offmancial efficiency can easily be drained away by unproductive administrative use of gross fmancial margins. This has not been the case at ADEMI. Salaries and administrative expenses (or "burden") as a share of total income excluding grants have fallen every single year since 50 By contrast, the 90-day CD rate at end-1996 was 13 .8 percent and ADEMI paid 16 percent on loans from clients. 51 A fuller discussion of foreign exchange risk provisions is postponed to the section on risk, although it is one of ADEMI's two key non-interest expenses, (the other being salaries and administration). SUSTAINABLE BANKING WITH THE POOR 59 ADEMl's inception, from 217 percent in 1983, to 47 percent in 1990, to 28 percent in 1995, until 1996, when they rose again to around 31 percent. As a share of total assets, ADEMI's burden continued to decline in 1996, to 9.6 percent from 10.9 percent a year earlier. These figures remain considerably higher than similar ratios for domestic commercial banks (around five percent) and for US banks (two to three percent), which reflects the high costs of providing door-to-door financial and technical assistance services to ADEMI's target clientele. Figure 7.4: ADEMl's Funding Sources (Expressed as Percent of Total Liabilities) 100% 90% 80% 70% 60% 40% 30% 20% 10% 0% Source: Appendices 4 and 6 The declining cost figures mask productivity concerns that ADEMI will have to address in order to maintain its excellent financial position and continue to be able to reach out to low- income clients. Table 7.2 shows that ADEMI has managed to cut administrative costs significantly relative to total revenues, assets and the outstanding loan portfolio. It has also managed to reduce these costs as a share of new lending from 11-1'4 percent during 1988-91 to 8-9 percent during 1992-95. On the other hand, while the cost per loan made fell by 18 percent in real terms during 1988-91, it more than doubled between 1991 and 1996, rising from RD$613 to RD$1,337 per loan in constant 1990 pesos. One reason is the need to maintain competitive salaries. Another major reason is that a 45 percent increase in staff productivity during 1988-91 (measured by new loans per staff member) was reversed between 1991 and 1996 as the number of new loans per staff member fell from 84 to 67. Thus the only way that ADEMI has been able to maintain the downward trend in the burden ratio is by increasing loan size. Indeed, the real average loan size has tripled since 1990 in terms of new disbursements divided by new loans, and quadrupled in terms of average loan outstanding per active client. SUSTAINABLE BANKING WI1H THE POOR 60 Table 7.2: ADEMI Has Cut Unit Costs by Increasing Loan Size Figures in 1990 pesos 1988 1989 1990 1991 1992 1993 - 1994 1995 199f - - - - - - - - - - 31.8% - - - 30.5% 28.2% Burden/Total Revenue 69.2% 47.3% 46.6% 36.3% 36.1% 34.4% Burden/Total Assets 28.7% 25.1% 26.5% 23.5% 18.8% 14.3% 12.0% 10.9% 9.61¾ Burden/Net Portfolio 37.1% 31.5% 35.5% 30.9% 24.4% 18.2% 15.0% 13.8% 13.01¾ Burden/Disbursements 12.6% 11 .1% 14.4% 10.7% 9.5% 8.0% 8.4% 8.2% 9.7°A Burden per New Loan 749 576 658 613 727 876 1,031 1,088 1,33i New Loans/Staff 58 83 68 84 92 83 74 73 6, Disbursements/Staff 346,767 431 ,131 313,052 479,582 709,422 908,779 902,456 968,156 915,68E Avg Disb./New Loan 5,950 5,206 4,570 5,743 7,670 10,942 12,253 13,233 13,71~ Avg Portfolio/Client 4,118 3,746 2,579 3,972 4,867 6,077 7,385 9,323 9,76~ Source: Appendix 6 These findings do not suggest that existing trends need to continue. On the contrary, ADEMI' s record during 1988-90 shows that it was able to cut administrative costs relative to loan portfolio even as real average loan sizes fell, by increasing the number of loans per staff member. ADEMI recognises that it is operating well within the productivity boundary suggested by its past experience, and has begun to motivate credit advisors to increase the number of loans they hold by setting higher number-of~loan targets for CAs to qualify for individual bonuses. Boosting Profits by Ensuring Portfolio Quality. ·This is the key area in which ADEMI and other successful microfinance providers have excelled relative to unsuccessful microlenders. To be sure, ADEMI has also struggled with non-performing assets in the past. In particular, during the mid-1980s, indiscriminate financing of loans led to a significant problem with arrears, and following the managerial crisis of 1986, several credit advisors abandoned their portfolios, contributing to further collection problems. However, the arrears problem was largely resolved after 1987 by introducing performance-based incentives for credit advisors. Since then, arrears have remained below 10 percent of the loan portfolio, and have been reduced to under 6 percent in 1995 and 1996 (Table 7.3). Within this low ratio, it is worth highlighting that arrears are significantly lower on small business loans than on microenterprise loans. For example, arrears over 30 days at the end of 1996 amounted to only 0.9 percent of the small business loan portfolio but 3.8 percent of microenterprise loan portfolio. This is not surprising, given the greater vulnerability of microenterprises to business shocks. Within the microenterprise portfolio, there are also important regional differences. For reasons that are hard to discern, arrears over 30 days were only 2.4 percent in the North at the end of.1996, but 4.9 percent in two Santo Domingo offices. SUSTAINABLE BANKING WITH THE POOR 61 Table 7.3: ADEMI Has a Healthy Portfolio and Ample Loss Reserves 1991 1992 1993 1994 1995 1996 Volume ofArrears Arrears/ Gross Yr. End Portfolio 10.1% 10.1% 8.4% 9.6% 5.8% 5.7o/c ofwbich: 1-30 days in arrears 4.4% 4.0% 2.9% 2.7% 2.3% 1.9o/c 31-90 days in arrears 3.1% 3.0% 2.2% 2.3% 1.5% 1.3o/c >90 days in arrears 2.7% 3.2% 1.0% 1.8% 1.0% l.4o/c Accumulation ofAccrued Interest Interest Accrued / Loan Portfolio 11.0% 11.7% 6.9% 8.3% 10.3% 11.6o/c Interest Accrued / Total Assets 8.5% 9.0% 5.5% 6.6% 8.1% 8.2o/c l!ncrem. Accrued Int/ Int. Income 16.1% 9.9% -0.4% 8.9% 11.1% 8.9o/c Adequacy ofProvisions Provisions /Net Avg. Loan Portf. 4.4% 4.1% 2.9% 3.0% 2.7% 2.6'¼ Write-Offs /Net Avg. Loan Portf 2.0% 1.3% 1.5% 1.1% 1.9% 0.8o/c Write-Offs/ Start ofYear Arrears n.a. 16.4% 19.5% 15.3% 23.3% 15.0o/c Loss Reserve /Net Avg. Loan Port. 3.3% 4.5% 4.1% 4.6% 4.1% 5.0'¼ Loss Reserve / Arrears 21.5% 34.6% 37.8% 39.7% 58.6% 76.3o/c !Loss Reserve / Arrears >90 days 81.9% 110.4% 95.7% 82.6% 175.7% 177.9'¼ Source: Appendices 4 and 6 ) In a recent innovation, ADEMI has replaced arrears over loan portfolio with contaminated over total loan portfolio as the indicator by which CA' s portfolio quality is assessed. While the aim of the change in indicator was to begin to harmonise ADEMl's asset classification with the prudential regulations governing formal banks, the main effect to date has been a dramatic effect on recoveries: during 1997, the percent of the portfolio affected by arrears declined from over 5 percent to just 2.5 percent. Moreover, the savings in terms of portfolio quality easily offset incremental incentive payments to CAs. This provides an interesting lesson in the design of staff incentive schemes. Overall, ADEMI has managed to maintain a historic loss ratio of less than 2 percent via a blend of"carrots" and "sticks". On the incentive side, ADEMI has emphasized providing repeat loans without delays to clients with good track records, and has tried to foster a relationship based on trust through the outreach work of its credit advisors. ADEMI's staff also have strong monetary incentives to screen and supervise clients adequately. When delinquencies do arise, ADEMI addresses problems early and strictly enforces loan collections even on very small loans. As of December 1996, ADEMI was conducting legal proceedings against 57 of the 15,370 active clients, or on 1.6 percent of the portfolio. Although the cost of legal proceedings can exceed the amount collected, ADEMI takes into account the benefits of deterring wilful defaults by other borrowers who witness loans being enforced. The result has been both low default probabilities and low losses given defaults. •SUSTAINABLE BANKING WI1H THE POOR 62 ADEMI estimates that recoveries in the event of foreclosures are around 80 percent. They would be higher, but unlike many lenders, ADEMI gives back to the client any amounts obtained through foreclosures in excess of the amount owed to ADEMI. While both arrears over 90 days and charge-offs have hovered around 1 to 2 percent of the loan portfolio in recent years, ADEMI's loan loss reserve reached 5 percent of the portfolio at the end of 1996, and greatly exceeds the value of arrears over 90 days. The Association can take advantage of its commendably conservative level of loss reserves to address the issue of high accrued interest receivables. 52 It did so in 1993, which explains the reduction in accrued interest that year (see Table 7.3). However, around 10 percent of reported interest income in the profit and loss statements continues to end up as incremental accrued interest on the balance sheet, and accrued interest over loan portfolio has risen again to four or five times the rates prevailing among commercial banks. Overall Financial Performance Assessment. ADEMI's financial performance is summed up in Figure 7.5. Its return on assets, while declining in recent years because of declining revenues on assets, is still an extremely impressive 15 percent. This is because ADEMI has been exceptionally effective at converting revenues into surpluses by containing administrative costs and bad debts, as well as by raising funds at low rates of interest. Moreover, it is worth noting that, even though ADEMI's cost of funds has generally been negative in real terms, its real average on-lending rate has always been positive, even when inflation has soared to almost 60 percent on an annualised basis. • Figure 7.5: Key Financial Indicators that Explain ADEMl's Profitability Panel 1: Profit Margin & Asset Use Panel 2: Key Underlying Ratios & Inflation 90% 80% 80% 70% 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% ~~~~~~00~~~~~$ ~~~~~~90~~~~~00 Note: All data in Panel 2 except inflation are expressed relative to the net average loan portfolio Source: Appendices 4 and 6 52 Interest income may accrue because the due date for payments by clients comes after the statement closing date, or because the interest becomes due and payable but is in fact not paid before the statement closing date. In ADEMI' s case, virtually all of the accrued interest is past due rather than accrued but not yet payable. Without proper provisions for potential losses, high accrued interest could result in overstated income, assets and net worth. However, ADEMI has in fact made adequate provisions that it could use to write off any accrued interest that is not likely to be recovered. SUSTAINABLE BANKING WITH THE POOR 63 Does ADEMI' s profitability depend on lending to small rather than just microenterprises, i.e. is the small enterprise portfolio cross-subsidising the microenterprise portfolio? This question is hard to answer without an SDI calculation based on disaggregated financial statements for the two product lines --which are not available. However, in 1989, before lending to small enterprises began, ADEMI's average loan size was only US$515, yet ADEMI brought the SDI down to just 4 percent (see Appendix 5). • In 1991, the year before the EIB small-scale lending program began, ADEMI became completely self-sustaining, and earned both its highest ever return on equity.and its lowest ever (i.e. most negative) SDI. Thus ADEMI has certainly proven its ability to make the microfinance self-sustaining, although without more detailed disaggregated data, the extent of any cross-subsidy between the product lines in more recent years remains an open question. Finally, while ADEMI/BancoADEMI is highly likely to remain "profitable" in the next few years, its degree of self-sustainability remains to be seen (see Chapter VIII). RISK There are several critical risks that ADEMI must consider in its operations. Some risks merit less attention: e.g. capital risk (the risk that the institution might become insolvent) is fairly low, given ADEMI's extremely high ratio of Tier 1 capital-Le. paid up capital plus unallocated retained earnings-to total assets. Some risks merit considerable attention but lie largely beyond ADEMI's control: e.g. the political risk associated with government policy decisions that affect ADEMI's market or operations (as in the case of PROMIPYME), or systemic risks associated either with macroeconomic instability (as occurred in 1990) or with a collapse of confidence in banking institutions (as occurred to a certain extent prior to the collapse of Bancomercio). This section focuses on five key risks that ADEMI can and should actively address, namely: credit risk, foreign exchange rate risk, interest rate risk, liquidity risk and operational risk. Credit Risk. Given the large volume of unsecured lending and the important liquidity and solvency constraints on clients, microfinance is highly susceptible to rapid changes in delinquency rates. Moreover, since word often spreads quickly when lenders do not enforce loans~ there is considerable potential for wilful default, particularly since enforcement costs are very high relative to loan amounts. Containing credit risks by developing appropriate incentive structures and early warning systems is therefore essential. ADEMI has been very successful in that regard, using the systems and methodologies described in Chapter VI, and should maintain its loan screening, underwriting and monitoring systems. In addition, it will be important to: • ensure that term structures on loans reflect clients' cash flows; • ensure that real time accounting is extended to its branches in the Interior; and • monitor portfolio composition to a greater extent under BancoADEMI. . Foreign Exchange Rate Risk. With over 70 percent of its borrowings denominated in foreign currency, exchange ratedsk is arguably ADEMI's largest single risk. The aim has been SUSTAINABLE BANKING WflH THE POOR 64 to engage in interest arbitrage by drawing on lower-priced foreign funds. However, the dangers ofun-hedged foreign exchange exposures following cross-border interest arbitrage have been amply illustrated by the recent financial crisis in Thailand. ADEMI's open loan exposure is capped to a great extent insofar as the foreign exchange risk on EIB funds (65 percent of all borrowings) is limited to the difference between the two percent cost of funds and the rate on Central Bank certificates (currently 16 percent, down from 18 percent). Moreover, ADEMI has made substantial provisions for foreign exchange risk as mandated under its agreement with the EIB. However, the ITC funding, which accounts for 6 percent of borrowed funds is uncapped. In addition, the heavy use of EIB funds results in foreign exchange exposures and potential liquidity risks that may not have been fully recognised. What would happen in the event of a speculative attack on the peso? With Dominican reserves equivalent to ·less than one month of imports, the Central Bank is not in a position to defend the peso against speculative attacks, except by raising interest rates. But a rising Central Bank certificate rate and depreciating peso would mean that ADEMI's foreign exchange exposure would increase precisely when ADEMI needs the cap most. Since the exchange risk exposure limits apply on the date funds are taken out (rather than the date the payments fall due), ADEMI can of course simply postpone withdrawals of disbursements from the EIB when the rate on Central Bank's certificate rate and thus the cap on ADEMI's exposure is unacceptably high. But this of course may merely result in foreign exchange risk being converted into interest rate risk or liquidity risk, since alternative domestic funding sources would have to be tapped in such a situation. With the establishment ofBancoADEMI, the Association is in a position to mobilise increased resources domestically and possibly to reduce open loan exposures via a broader range of hedging mechanisms. To sum up, therefore, while the nominal cost of funds has declined, the true future cost remains uncertain. Interest Rate Risk. ADEMI's cash flows are highly sensitive to changes in the level of interest rates. ADEMI's short term loans to mforoenterprises are not repriced before maturity, but new loans are repriceable; it's medium-term loans are revisable (albeit not automatically adjusted) as market rates vary. Overall, virtually all loans, and_thus around 70 percent of ADEMI's assets, are repriceable within one year. By contrast, most of ADEMI's liabilities ( except for loans from clients, Banco Popular and the ITC) are medium- to long-term loans at fixed rates. Therefore repriceable liabilities are equivalent to around 40 percent of assets. The difference of +30 percent compares to a range of about minus 5 to plus 5 percent for US banks. The negative term transformation also contrasts with typically positive mismatches for commercial lenders. A positive repriceable funds gap such as ADEMI's result in increases/decreases in net interest income as market rates increase/decrease. This is clearly visible in Panel 2 of Figure 7. 5 above, which shows the large rise in ADEMI's spread around 1991, when inflation and market rates peaked, and the decline thereafter as inflation and market rates fell. Clearly the positive repricing gap could help to counteract some of the risks mentioned above under foreign exchange risks, provided that ADEMI can maintain portfolio quality if rates rise. However, ADEMI must also ensure that it can operate profitably in the event of further declines in market rates, particularly by containing non-interest expenditures. Moreover, BancoADEMI should SUSTAINABLE BANKING WITH THE POOR 65 undertake active asset-liability management once it begins to mobilise different forms of liabilities as a development bank. Liquidity Risk. As a non-profit agency, ADEMI's liquidity management has been adequate for its requirements. In particular, ADEMI has ensured that it has always been able to meet its payment obligations, particularly its lending commitments to clients, in a timely manner. However, liquidity management is an area that warrants considerable upgrading as BancoADEMI begins its operations. For example, while ADEMI's finance and treasury departments ensure that short-term cash flow requirements are met, they do not engage in medium term projections (which is essential, given the rate at which the portfolio has grown); ADEMI at times has excess liquidity that reduces returns (as occurred at the end of 1996); and the Treasurer has no staff per se, relying on the finance department staff to do treasury work. Although ADEMI has at times experienced liquidity problems related to insufficient funding for its rapid growth, it has not experienced major problems with volatile liabilities except when FondoMicro called a large share of its lending in 1994. A more common problem has been lost earnings resulting from excess liquidity. Nevertheless, BancoADEMI should go beyond ADEMI's correspondent bank-like relationship with Banco Popular to develop broader borrowing relationships in the interbank market, as well as liquidity and asset management skills to deal with potentially higher leverage ratios, more volatile domestic promissory note issues, and possible repurchase agreements or securitisations. Operational Risk. These include a wide range of risks, such as data processing failures, fires or other destruction of records, computer hackers, theft, fraud or negligent expense controls. Appropriate internal controls are needed to mitigate these risks. Core principles for effective banking supervision issued by the Basie Committee [1997] urge supervisors to ensure that internal controls "include clear arrangements for delegating authority and responsibility; separation of the functions that involve committing the bank, paying away its funds, and accounting for its assets and liabilities; reconciliation of these processes; safeguarding its assets; and appropri~te independent internal or external audit and compliance functions to test adherence to' these controls as well as applicable laws and regulations." Shortcomings in internal controls can result in major failures. For example, a collapse in underwriting standards recently triggered the insolvency of an important bank for microfinance in Colombia, while a microfinance provider in El Salvador suffered massive fraud at the hands of 40 of its employees. By contrast, ADEMI's internal controls have generally been very • effective, enabling it to quickly unearth fraudulent practices by staff on three occasions in the last 15 years, and to prosecute the culprits. Daily transfers of data between branches and headquarters ensure the presence of copies of information, and back-ups of centrally based information are maintained off-site. The MIS also provides daily reports to managers to facilitate monitoring of their respective activities. Finally, ADEMI began to revamp internal audit procedures after hiring a new auditor in 1996. SUSTAINABLE BANKING WITH THE POOR 66 With the creation of BancoADEMI, new security procedures and internal controls have been introduced that reflect the greater use of cash transactions directly within the bank rather than through Banco Popular. Formalisation has resulted in BancoADEMI having to meet security requirements set by the Banking Superintendency (including e.g. vaults). To conclude the review of risks, it is clear that ADEMI' s processes have served it well thus far in managing risk and achieving outreach to low-income clients in a self-sustaining way. The establishment ofBancoADEMI will, however, present new risks and opportunities that will call for increased attention to and sophistication in risk management, notably in the critical areas discussed above. • PROTECTING STAFF: ADEMl'S PENSION PLAN Unlike most Dominican companies, ADEMI has addressed its staffs retirement concerns. In January 1992, ADEMI introduced a pension plan for all fixed staff to provide retirement, disability, death and survivorship benefits, as well as loans to employees and payments upon resignation of staff. The plan is a defmed benefit plan, with contributions of 4 percent of remunerations by staff and 5 percent of payroll by ADEMI. ADEMI has made additional contributions in the initial years to ensure the actuarial soundness of the plan. The· plan is administered by a committee comprising members of ADEMI's staff and management, with general oversight by the Board of Directors. It is audited by a leading external auditing firm and reviewed for actuarial soundness by external actuaries (see Appendix 8 for information on plan benefits and recent fmancial statements). • As of December 1996, the plan had over RD$9 million in assets and a net worth of RD$8.9 million, well in excess of the actuarial present value of accumulated benefits for participants in the plan (RD$2.1 million) or of projected benefits (RD$5.9 million). On the other hand, the fund is not yet well-diversified and thus not prudently invested: 37 percent of assets as of December 1996 are invested in low-interest loans to staff yielding an average of 15.3 percent, and another 25 percent of assets are invested in 18 percent bonds issued by ADEMI. Most of the remainder (34 percent) is invested in dollar notes in only one other fmancial institution. As Figure 7.6 shows, ADEMI has begun to increase the share of investments outside of ADEMI, and in September 1996, the administering committee decided to broaden the plan's investments to real estate and equities, however it had not begun to do so as of the end of 1996. ADEMI's fully funded pension plan is a fairly unique benefit for employees in the Dominican Republic. Diversifying its asset base away from ADEMI's fortunes into a broader spectrum of sound investments would further augment the tremendous value of this plan. SUSTAINABLE BANKING WITH THE POOR 67 Figure 7.6: Diversifying the Asset Base of ADEMl's Pension Fund 1996: RD$9.0m. other Assets ,/ Loans to members other Investment Loans to ADEMI Source: Appendix 8 THE TRUE BOTTOM LINE: ADEMl'S IMPACT ON CLIENTS ADEMI has extended over 120,000 loans worth US$140 million to more than 40,000 clients over the past 14 years. These figures can be related to an overall market of about 330,000 microenterprises, of which an estimated 100,000 are interested in formal credit (see Chapter ID and Cabal [1993]). Until 1990, ADEMI's average loan size was just 50 percent of GDP per capita, a figure that has risen to 114 percent with the expansion into small scale lending since 1992. Forty-six percent of ADEMI's clients are women. The vast majority of the Association's clients had no prior access to formal credit. Thus ADEMI has been fairly effective in terms of commonly monitored outreach indicators. The overall cost-benefit for ADEMI is hard to assess. Many of the benefits of access to an ADEMI loan accrue to a client over a number of years (especially for investment capital loans). Many of the benefits are.non-pecuniary. And in cases for which financial gains can be tracked over time, without-loan counterfactuals are hard to calculate, and information is not available on the incomes of comparable non-borrowers. However there can be little doubt that ADEMI is generating a positive total surplus for the country. In the case of subsidy dependent microfinance providers, it is an open question as to whether the gains to their clients from the services provided exceed the losses to the financial institution, for an overall net gain. In ADEMI's case, the Association is fully self-sufficient, its clients are repaying at extremely high rates (the overall recovery rate is 98 percent) and clients' assets and equity are growing on average by over 20 percent per year. ADEMI is liaving an impact in four key areas: employment; clients' financial positions; borrowers' social situation, and the international arena. SUSTAINABLE BANKING wrm THE POOR 68 Impact on FJnployment. ADEMI tracks the number of jobs created and the number of jobs strengthened by its operations, because impact on employment is one of ADEMI's fundamental objectives. Employment data ~e rout~ely collected by credit advisors as part of their credit analysis. The measurement approach is simple: when a loan is provided to a new client with e.g. five employees, and the client comes back for a repeat loan a few months later with eight employees, the original five jobs are considered to have been strengthened, i.e. made more permanent, and the three new jobs are considered to have been created. In total, eight jobs are considered to have been supported. Clearly these simple ex-ante versus ex-post comparisons fall short of a full impact assessment in four important ways: i) to the extent that some of the job increases might have occurred without ADE1\1I's credit, the impact is overestimated; ii) in other cases, particularly during the macroeconomic crisis of 1988-91, jobs would arguably have been lost without ADE1\1I's support, so that the absence of visible increases in employment does not mean that ADEl.\11 had no impact on jobs; iii) since the data on the impact of a loan are only collected when a new loan application is processed, the methodology may significantly underestimate ADE1\1I's impact on employment by excluding the jobs that were created under loans which had no follow-ons, and iv) there are no data on the impact of ADE1\1I's loans on non-borrowers' employment rates: these could be boosted as a result of positive multiplier effects in the community, or reduced as a result of increased competition from ADEMI's borrowers. Nonetheless, using this simple approach, ADEMI estimates that it has directly affected over 200,000 jobs in the micro- and small enterprise sector since 1983, including the creation of almost 60,000 new jobs (see Figure 7.7 and Appendix 9). Most of this impact comes after taking a series of loans from ADEMI. Indeed the impact on employment with less than five loans is fairly small (see Appendix 9), but rises significantly once 10-15 loans are taken out (see Gomez [1989], Tenn [1997]). ADFlvfl's Cost Per Job Created. ADEMI's data suggest that the cost per supported job is rising rapidly. Since 1984, the number of dollars lent per job has risen by an average of 6.6 percent per year to US$1,140 in 1996.. The cost to ADEMI per job supported has risen by 12.9 percent per year to US$350 in 1996, where this cost includes ADEMI's cost of debt and equity funds (priced at market rates), loan losses and administrative costs (see Appendix 9). 53 However, these figures may overstate the increase in costs, to the extent that the lengthening of the terms of ADEMI's loans has resulted in less immediate repeat loans and therefore less frequent opportunities for ADEMI to update data on the impact ori. employment. Moreover, these costs are remarkably low when compared with numerous other employment generation or enterprise finance programs. For example, in a review of the World Bank's lending for small and medium enterprise development over the period 1973-1987, Webster, Riopelle and Chidzero [1996] find that the average cost per job created in 43 completed Bank projects was US$5,800, measured in constant 1980 US dollars. This is equivalent to US$10,900 per job, measured in 1996 US dollars or 10 times more than the cost per job created by ADEMI in 1996. S3 The overall cost of the job obviously also includ~ wages and investment costs inClllTed by the borrower, for which data are not available. SUSTAINABLE BANKING WITH TIIE POOR 69 Figure 7.7: ADEMl's Impact on Employment 225000 200000 175000 .,, 150000 .a 0 ~ 125000 0 ,! 100000 E I 75000 50000 25000 0 1983 1985 1987 1989 1991 1993 1995 gCreated ■ Strengthened Source: Appendix 9 impact on Assets and Net Worth. One of th~ earliest and most thorough impact assessments of ADEMI was conducted by Gomez [l 989]. Analysing a sample of 100 clients over the period 1983-88, she found that assets increased by an average of six percent in real terms for clients with one loan from ADEMI, 20 percent for those with five loans, 314 percent with 10 loans and 375 percent with 15 loans, relative to pre-loan values. A more recent sampling of39 clients who had received two or more loans between September 1991 and September 1997, reveals that, with an average of 3.3 loans totalling RD$57,000 (US$4,000) over a 17-month period, clients were able to boost assets by RD$95,000 or 82 percent, while liabilities rose by only RD$3,900 (84 percent) on average. Therefore net worth rose by over RD$91,300 (US$6,500) per client, i.e. by four times the GDP per capita of the Dominican Republic (see Appendix 9). Where enterprises have not experienced substantial increases in assets, turnover and net worth, the reasons have generally included debts with previous lenders, inefficiencies in production or productivity, price fluctuations, consumption of surpluses rather than reinvestment in the enterprise, investment in unneeded fixed capacity, weak liquidity management, or even a change of enterprise (e.g. due to more attractive alternative enterprises or to formal sector employment opportunities). Non-Financial Benefits ofADFlvfl's Lending. The impact of ADEMI's lending and technical assistance services clearly extends beyond borrowers' enterprises to their households. Some of the most frequently cited benefits include increased economic security, greater investment in family members' health, improved educational opportunities for borrowers' children (particularly investment in private schooling and in college education), purchases of furniture and appliances, and new or upgraded housing (Gomez [1989]; Lewin [1991 and SUSTAINABLE BANKING WITH THE POOR 70 1991b); Tenn [1997)). Another important and often intangible benefit is that clients came to view themselves, and to be viewed by others, as successful entrepreneurs and respected members of the community. Indeed, one of the key benefits enjoyed by tile more than 6,000 holders of ADEMI's Mastercard is the status that pertains to holding a credit card. ADEMI's International Impact. ADEMI' success has not only garnered considerable attention in the local press but also in the international microfinance community. ADEMI has shared its experience via both in-house and on-the-road training sessions with numerous other microfmance lenders from Panama, Jamaica, Haiti, Egypt, Gaza and South Africa, among . others. For example, the Alexandria Business Association in Egypt has applied ADEMI's individual lending methodology with considerable success, serving over 17,000 clients with a 99 percent recovery .rate as ofMay 1995. While ADEMI has been heavily criticised for its interest rate policy in the past, its clients understand the need for fairly high interestrates, and ADEMI's success has muted most criticisms from those who believed that subsidized interest rates were essential for the development of these enterprises. The donor community also drastically changed its outlook on this issue during the 1980s, partly in response to the positive experiences emerging from more • successful microfmance providers, such as PRODEM/BancoSol in Bolivia, BRI in Indonesia and ADEMI . . In the coming years, the challenge·for ADEMI will be to expand its outreach, particularly within the vibrant microenterprise sector rather than among small- or medium-scale enterprises, while boosting productivity so as to contain operating costs and maintain the sterling financial performance that ADEMI has displayed in recent years. The next chapter considers (Banco)ADEMI' s prospects for the future, as well as drawing some lessons for other microfmance providers on what has worked well and what has not SUSTAINABLE BANKING WI1H THE POOR 71 VIII. LESSONS FROM ADEMl'S SUCCESS, AND THOUGHTS ON ADEMl'S FUTURE THE ESSENCE OF THE ADEMI MODEL ADEMI is a successful experiment in microfinance. Several insights and lessons can be drawn from its experience. Much like the microenterprises that it supports, ADEMI started with limited resources, trying out new products (e.g. group lending versus individual lending, working capital loans versus investment loans, lending in Santo Domingo versus lending in small towns, and in-house provision of TA versus referrals to third parties). However after four years in business, ADEMI matured into a sound financial operation that was able to withstand the macroeconomic crisis an • of 1989-91. The year 1992saw ADEMI launch important new line of activity, namely small to •medium scale lending for both graduating microfmance clients and new small-scale clients. That new line of activity is now a critical component of ADEMI's own graduation into the formal fmancial sector through BancoADEMI. Several key strategic decisions were taken in the course of ADEMI's evolution. These decisions shaped ADEMI's culture and experience in critical ways. They were to: • Provide microfinance in a business-like and self-sustaining manner (1982) • Intermediate funds through a commercial bank, rather than receiving payments and disbursing loans in cash (1983) • Provide individual loans rather than group lending (1984, and again in 1986) • Become a nation-wide provider of microcredit (1985) . • Strengthen internal controls and provide an incentive system for staff (1987) • Graduate with the clients, rather than letting them graduate from ADEMI (1990/92) • Establish a formal bank, particularly to deal with larger clients (1997) The ADF.MI Methodology: From these strategic decisions and ADEl\fl's historical evolution, one can distill several principles that constitute the essence of the ADEMI methodology. These are to: • Establish a fundamental objective: i.e. employment generation via the" democratisation of credit" • Operate like a business, not like a project; • Get the fundamentals right: clarify responsibilities and lines of command, and establish proper internal controls, audit procedures and a sound MIS; • Pursue self-sustainability, so as to reach more clients; • Orient operations towards the clients,.not towards donors (work with donors as partners, not as the major decision-makers); • Treat clients as partners, not as beneficiaries; offer them prompt and courteous services; • Target a clearly identified market niche: low-income entrepreneurs with demonstrated potential; • Ensure that services are welcomed in the community by seeking partnerships with local organisations; • Engage in responsible lending: screen risks carefully and actively pursue delinquencies, but do not blame the clients for bad loans; SUSTAINABLE BANKING WITH TIIE POOR 72 • Know and visit clients in their work-places, bringing setvices to their door-steps; • Assess each client on his/her own merits, and lend to him/her individually; • Provide focused services (i.e. credit plus limited TA, and voluntary savings as permitted); • Charge all clients positive real rates of interest, and reflect the higher cost of smaller loans in an accordingly scaled interest rate and fee schedule; • Do not impose unnecessary costs on clients, e.g. group activities or forced savings; • Build a relationship with clients by scaling up loan sizes and maximum loan terms over time, according to the client's repayment history; • Allow clients to build up capital via monthly, not weekly, repayment schedules; • Grow with the clients; • Adapt to market conditions, and never cease to evolve with time. A Good Investment: The successful development and application of this methodology has allowed ADEMI to report positive net income (not counting grants) within five years of starting operations, a positive SDI within eight years, and the full recovery of all subsidies provided to it within eleven years. This is illustrated in Figure 8.1, which shows both annual and cumulative subsidies received or surpluses earned, measured in constant 1997 US dollars (see Appendix 5). It is very clear that the surpluses that ADEMI has earned substantially exceed the subsidies that it has received. 54 Moreover, it is worth repeating that ADEM! has had a major impact on its clients in terms of growth in assets, equity and jobs, and improved quality of life (see Appendix 9). This is the true bottom line. WHAT DID NOT WORK AND WHAT NEEDS MORE WORK Clearly, ADEMI's success to date does not mean that there are not, or have not been, any problems, shortcomings, or issues requiring attention. Several of these have been highlighted in this report. They include: • In 1985-86, ADEMI was awash with liquidity when it obtained a large IDB loan. It therefore focused on rapid disbursement and establishment of new branches without ensuring that systems were in place tq manage the expanded activities. Portfolio quality s~ered and arrears rose to 30 percent. 54 Toe internal rate of return on the stream of subsidies and smpluses is 9. 7 percent (see Figure 8.1 ). This 9.7 percent rate can be thought of as society's dired retmn on investing in ADEMI. Note that while the calculation imputes opportunity costs to ADEMI' s resources, it does not represent a true eamomic rate of return insofar as it does not capture the costs and benefits associated with the gains and losses ofADEMI's clients and their competitms, or of ADEMI's own competitors. There is every reason to believe that the true eamomic rate of return on investing in ADEMI, when one takes into account these factors, would be substantially higher. SUSTAINABLE BANKING WITH TIIE POOR 73 Figure 8.1: ADEMI has Recovered All Initial Subsidies to Earn a Cumulative Surplus 1,500 iii C, C, ? 1,000 Ill ::I ti; 500 en ~ Ill ::, e: . ::, Ill 0 (500) >- ~ (1,000) ,g Ill (1,500) 11111Annual (Subsidy)/ Surplus ■ Cumulatiw (Subsidy)/ Surplus Source: SDI Calculations in Appendix 5 \ • The years 1984 and 1986 saw leadership changes associated with very different approaches to managing .ADEMI. The problems of that era are largely associated with the dissonance between the two management cultures (and consequent rival loyalties). This problem could arise again when Jimenez and/or Lluberes retires, as there is considerable staff loyalty to them and support among Board members, due to their exceptional leadership of the Association. 55 • There is a need for further attention to detailed planning and risk management. Gomez [1989] first highlighted the absence of medium term planning, which continued to receive relatively little attention until .ADEMI commissioned a study in 1996 to assess its options for the future ( Tecnicas de Planificaci6n [1997]). Liquidity management, the sectoral distribution of the portfolio, foreign exchange risk and interest rate risk management will continue to require attention, ideally through improved asset-liability management and value-at-risk monitoring under Banco.A.DEMI (see Jorion [1997]). 56 55 Several years ago, Gomez (1989] noted that: ''Paradoxically, the same factor that contn1>uted significantly to the strengthening and development of the program, also constitutes the factor that weakens the capacity of thcprogramme to surviv~ in the long term: its great dq,endence on the leadership of PedroJimene7., the Exerutive Director." While the institution's survival is no longer in doubt, a seamless transition to new management is unlikely. 56 Value-at-risk (VAR) measurement is increasingly being recommended for financial risk management and analysis by banks, investment houses, other financial intermediaries and regulators. It involves the use of statistical techniques to measure "the worst expected loss over a given time interval under normal market conditions at a given confidence levef' (Jorion [1997], bis italics). Thus managers and shareholders can assess not only the accounting returns but also the exposure of those returns to market risk and the probability of adverse movements. This in tum permits improved resource allocation and better performance evaluation. SUSTAINABLE BANKING WITH THE POOR 74 • Customer satisfaction has generally been very high among ADEMI clients. However, the time involved in taking p.:. yments to Banco Popular, difficulties with the te~ of loans, delays in obtaining Mastercards, the preclusion of concurrent loans, and high interest rates, have at times been sources of frustration for clients. One way to enhance service would be to follow up clients who have stopped borrowing from ADEMI, (e.g. through occasional surveys of a sample of clients who have not borrowed for 18 months since paying off their last loan), to find out why and to adapt services where appropriate and feasibl~ to reflect their comments. • Accounting and disclosure practices for ADE.MI warrant adjustment to reflect the standards required of BancoADEMI, particularly since ADEMI will continue to be a going concern for some time even after BancoADEMI is established, and consistent data will be needed to present a consolidated financial position for .the group. 57 • In recent years the ratio of administrative costs to amounts lent has been kept low due to increasing loan size, rather than by cutting the cost per new loan. One reason for this is that there has been little incentive to contain costs in ADEMI's staff incentive schemes: currently, ADEMI's staff bonuses are related to (a) individual performance on amounts lent and recovered, and (b) ADEMI's overall profits. However, the latter indicator does not allow the staff of a particular branch to internalize the costs and benefits of • increasing efficiency at their branch. 58 ADEMI may be able to enhance productivity by monitoring branch level profitability and incorporating it into the incentive mechanism (as do BRI and Grameen Bank, among others). 59 57 For example, ~) the treatment of accrued interest inADEMI's quarterly reports (but not in its audited financial statements) leads to an overstatement of assets; and (ii)ADEMI' s published ammal reports currently present only the balance sheet but not the profit and loss statement. 58 The benefits of a little extra effort at one branch are ~oyed by staff at all branches, and the costs of a little extra laziness at one branch are borne by all branches. 59 This is somewhat difficult to do under current intemal acoounting arrangements: unlike for branches in the Interior, branch level expense data are not available for the Santo Domingo branches. This is because Santo Domingo expenses are calculated as residuals of total ADEMI expenses minus the sum of reported branch level expenses in the interior (see Appendix 7). ADEMI's management has explained that this is due to the fact that, thus far, the branch network in Santo Domingo has grown by moving staff from one branch to another, therdJy rendering comparative cost data meaningless. However, an acoounting treatment of branches as profit£entres may be introduced forBancoADEMI. SUSTAINABLE BANKING WITH THE POOR 75 • Finally, while ADEMI has made significant progress in boosting staff productivity over time, (e.g. by raising the cut-off for credit advisors to qualify for bonuses from 60 to 140 clients over the last 10 years), further progress will be required to ensure that the Association can continue to meet rising costs and serve lower income clients. 60 Figure 8.2 shows the productivity CUIVe in terms of the loan portfolio per credit advisor. While it is flattened somewhat as older credit .advisors with too many clients regularly hand over portfolios to new credit advisors, the productivity dispersion within a given class of advisors is of interest, as is the share of advisors who are qualifying for bonuses based on this indicator. Control variables that can influence these indicators include the scale of bonuses (and related targets) versus the base salary, the size and location of the blocks that CAs serve, the nature of the target clientele(micro versus small business) and the advisors' skills and drive (whence the importance of initial screening and of on- going training). ADE?vll's management is very aware of these considerations, and has regularly adjusted these control variables to boost productivity. In the coming year, it plans to increase productivity by computerising several processes and making lap-top computers available to all its credit advisors to reduce papeiwork and automate information transfers to headquarters. 61 Figure 8.2: Ademi's Maximum, Average and Minimum Loan Portfolios per Advisor in Millions of Dominican Pesos, as of September 1996 4.5 4.0 E3.5 * i 3.0 .2 2.5 ~ 2.0 0 P.. 1.5 C ! 1.0 .J 0.5 0.0 1 2 3 4 5 6 7 8 9 10 11 12 Years of Service Note: RD$1.5 million was the minimum to qualify for incentive payments in September 1996. This figure has been raised to RD$2.5 million. Source: Appendix 3 60 Chapter VII indicated that increasing average loan size was important in keeping costs down, due to the declining number of loans per staff member after 1991. Nonetheless,ADEMI's average outstanding loan ammmt for microfinance loans (US$1, 107 as ofJlllle 1997) is still not high when compared to members of thcMicrofinance Networlc. ofleading microfinance providers, (US$91-US$1,720). This is particularly true when average loan size is compared to the counny's GDP/capita (see TheMicroFinance Networlc [1997) and Appendix 10). InADEMI's case the ratio of average micro-loan to GDP/capita is 76 percent, compared to between 16 percent and 321 percent for members of theMicrofinance Networlc. 61 One area with important potential productivity increases is services provided by administrative assistants, where several regular processes, e.g. is~ce ofreceipts, are still undertaken manually. SUSTAINABLE BANKING WI1H THE POOR 76 .These considerations, far from detracting from ADEMI's exceptional performance, reveal some of the difficulties associated with the effective provision of microfinance. ADEMI has over thr: . ·ars successfully addressed several of the challenges in microfinance. In order to improve its servi..:.\;'.~, ADEMI has taken the next logical step, setting up a formal bank for micro- and small scale fmance. BANKING ON THE FUTURE On 11th September, the Monetary Board of the Dominican Republic approved a banking licence for the Banco de Desarollo ADEMI, S.A. A new central office was purchased for BancoADEMI, which opened its doors to the public on January 2, 1998. Thus ADEMI has joined a select few microfinance providers at the forefront of the movement to place fmancial services for low-income clients firmly in the formal fmancial sector. The Vision for (Banco)ADFA11. 62 The new development bank began operations with an initial focus on the larger micro-, small- and medium-scale enterprise sectors, making loans in the range of US$2,000 to US$200,000, while ADEMI continued to serve the smaller microenterprises that make up the bulk (85%) of the Association's clientele. 63 However, early in 1998 it was decided that the micro- and small-business lending would be integrated fully into BancoADEMI as soon as feasible. In particular, AD:ap: is phasing out loan originations and all new loans as well as all follow-on loans to current ADEMI loans are being extended by BancoADEMI. Once the transfer of clients is complete, ADEMI itself will simply function as a holding company for BancoADEMI, much like Pro:-Credito does for Caja de las Andes in Bolivia (see Box 8.1 below). In addition, as soon as the proposed new monetary code is approved by the National Assembly, BancoADEMI, like all development banks, will be eligible to (and plans to) convert into a savings and credit bank, within a new fmancial sector structure that would comprise multipie service banks, savings and credit banks, finance corporations, fmancial groups and savings and loan associations. As a development bank (and thereafter as a savings and credit bank), BancoADEMI has begun to provide not only enterprise credit but also formal time deposit services, including both savings accounts and promissory notes (but not checking accounts or certain foreign exchange . services, which are currently limited to commercial banks and will in the future be limited to multiple service banks). BancoADEMI will probably also expand its Mastercard programme and its insurance services, and begin to offer new products such as leasing and personal loans. In addition, BancoADEMI is exploring the possibility of installing automatic teller machines and issuing debit cards to savings account holders, both to increase access to·savings and to facilitate repayment of (Banco)ADEMI loans. Regrettably its lending programme for housing development is facing the properties for which they are taking regulatory constraints: all clients are required to have titles to _ housing loans. 62 Together, ADEMI, the NGO, and BancoADEMI, the development bank that ADEMI has established, are referred to in this paper as the ADEMI Group, or alternatively as (Banco)ADEMI. 63 In order to retain its focus on low~income clients, BancoADEMI plans to place a cap on its loan size of around RD$3 million (US$215,000), which is lower than some of the loans it has provided to date with EIB funds. SUSTAINABLE BANKING WI1H THE POOR 77 In spite of the increased modernisation of its operations and sophistication of its seivices, the ADEMI Group plans to continue building its banking relationships with its clients in their own workplaces and homes, through personalised, door-to-door seivice provided by its credit advisors. Thus the vision for (Banco)ADEMI is to create a modem financial institution while retaining its original mission and methodology. Why Did ADEMI Set Up a New Bank? The establishment of BancoADEMI was a response to changing circumstances, including a clientele that has grown considerably with ADEMI; increasing competition in small enterprise finance; broader e~nomic reforms; and in particular, reforms in the financial sector. There has been a steady tightening of prudential regulations in the Dominican Republic since the financial crisis of the early 1990s and the proposed monetary code would grant the Banking Superintendency the right to regulate any enterprise whose primary activity is financial intermediation, even if it does not mobilise deposits. Since ADEMI had a net worth at end-1997 that placed it among the 10 largest financial intermediaries in the country, it was a very likely candidate for regulation even if it continued to operate just as an NGO. While ADEMI had considered purchasing an existing development bank, the most solid of these were already contemplating mergers with commercial banks in order to become multiple seivice banks under the new financial structure. The alternative of merging with an existjng commercial bank was rejected because, as a junior partner, ADEMI might not be able to maintain its methodology or target group. Therefore, establishing a new development bank was viewed as the best way to formalise ADEMI's fmancial seivices. Who Owns BancoADEMJ? BancoADEMihas an authorised capital ofRD$150,000,000, or about US$10.7 million (see ECOCARIBE [1997]) and an initial paid-up capital ofRD$100,000,000, more than half of which has taken the form of new cash injections, rather than ADEMI transfers. Currently BancoADEMI has four principal shareholder groups: • ADEMI owns 24 percent of the shares in BancoADEMI. ADEMI is financing the initial cash and fixed asset injections and is spinning off its entire portfolio to BancoADEMI, together with certain liabilities. This gives it the largest single shareholding in BancoADEMI. • Another 20 percent of the shares in BancoADEMI are owned by a new company-Inversiones Mutualistas, SA-whose owners are exclusively ADEMI employees, and whose sole purpose is to hold employees' shares in BancoADEMI. This is similar to the approach adopted by K-REP (see Box 8.1 ). Half of the capital injection came from shares that each employee of ADEMI received in the employee ~mpany in proportion to his or her rights to severance payments from ADEMI. Instead of actually making severance payments to employees, the equivalent cash has been made available by ADEMI on their behalf, and through their company, to BancoADEMI. The ,emaining 10 percent was in the form of cash through share purchases by employees. Shares in the employee company are freely tradable among ADEMI employees. • Seventeen percent of BancoADEMI' s shares have been sold to the European Investment Bank for cash (rather than a debt-equity conversion). • The remaining 39 percent of shares are held by several individuals who have been active in the development of ADEMI. Around 14 percent was transferred to members of ADEMI's Board of SUSTAINABLE BANK.ING WITH THE POOR 78 Directors as compensation for almost fifteen years of work without remuneration. The remaining 25 percent was sold to Board members for cash. The Physical Conversion from Semiformal to Formal Micro.finance. ADEMI is not the first microfinance NGO to formalize its operations (see Box 8.1). However, the approach that it has adopted for transferring assets to BancoADEMI is different from previous experiences. Initially it transferred only its headquarters staff to the new bank, (together with larger loans and with a share of the liabilities). However, over a little more than one year, almost all physical assets, all staff time, and the ADEMI pension plan are being transferred to the new bank, and ADEMI will simply become a holding company. ~ Box 8.1: Converting from Semi-Formal to Formal Microfinance - Five Recent Experiences In Colombia, Corposol purchased and became the primary shareholder of an extant finance company, which it renamed Finansol. Corposol vetted loans for financing by Finansol, which mobilised deposits to finance the loans. In the end, this arrangement proved unfortunate for several reasons (see Berenbach and Churchill [1997]). Finansol became insolvent and was restructured, while Corposol was forced into bankruptcy by its creditors. In Bolivia, PRODEM sold entire branches iµ its network to a new formal bank for microenterprises, BancoSol, and became BancoSol's largest shareholder. BancoSol is regulated by the Superintendency under existing banking laws. PRODEM has continued to provide loans to microenterprises in rural areas, while BancoSol has focused on urban microenterprises. Also in Bolivia, Pro-Credito established a private financial fund called Caja de los Andes, an intermediary set up under special regulatory arrangements for microfinance. Pro-Credito transferred its entire portfolio to Caja de los Andes, and is now simply a holding company. In Kenya, K-Rep Holdings was established both to control K-Rep's NGO and to own a new K-Rep commercial bank. The NGO will pass its entire portfolio to the commercial bank, and will focus exclusively on providing technical services related to microfmance. Employees will also hold a I 0 percent share in the bank through a holding company owned exclusively by K-REP employees. • In the Dominican Republic, FondoMicro established and became the largest shareholder of Banco para la Pequefia Empresa, S.A, a formally regulated bank for small enterprises. FondoMicro has continued to offer apex credit to microfmance NGOs (e.g. ADOPEM), while its bank affiliate provides credit directly to clients. Challenges Related to the Transition. From an operational perspective, (Banco)ADEMI' s staff are administering both ADEMI loans and BancoADEMI loans during the transition, which entails much more complex accounting arrangements than occurred for the other NGOs that established formal counterparts. 64 ADEMI will therefore a fortiori have to adopt the accounting 64 Similar staffing and participation issues will of comse arise for ADEMI' s pension plan. SUSTAINABLE BANKING WI1H THE POOR 79 ..... . -~. . '·. • conventions mandated by the Superintendency in order to present a consolidated financial position, as this is likely to be more revealing during the transition than even the best attempts to present financial statements for the bank and the NGO based on arm's-length transactions between the two. ADEMI is already adjusting its accounting conventions accordingly, particularly in the area of asset classification. It is hoped that the ADEMI Group will present not only individual statements but also consolidated financial statements as of end-1998 for ADEMI and BancoADEMI. Asset classification is an especially important issue, on which much ofBancoADEMI's profitability, outreach and market position will rest. FondoMicro's small enterprise bank (see Box 8.1) took the lead in this area: it classified its loans to small businesses as commercial rather than consumer loans, (even though more relaxed asset classification and provisioning standards would have applied to the latter). The idea was to persuade the Superintendency that the special features of microfmance (i.e. small loans with simple documentation to clients with limited accounting and . collateral) have to be addressed within the same framework as loans to larger enterprises, rather than by mis-classifying small business loans as consumer loans to circumvent regulations that are hot appropriate for this sector, The ADEMI Group has adopted the same asset classification approach. Thus far the new asset classification system appears to have presented BancoADEMI with a serious challenge regarding seivice to its smallest clients. In particular, their payment histories under ADEMI are not being recognised by the Banking Superintendency for the purposes of classifying · them into low-risk categories, while these clients tend to lack legal documentation for the collateral they offer. This means that loan loss provisions of 10 percent or more must be made if these clients are to be seiviced. Several microenterprise clients have already switched to other microfinance NGOs, since the Superintendency appears to be reluctant to see an ADEMI holding company nating loans with BancoADEMI staff, Le. they cannot remain with ADEMI beyond a transition origi_ period. Options for addressing this dilemma would appear to be to continue dialogue with the Superintendency regarding recognition of formally documented payment histories that small clients build up outside the formal banking system; or possibly passing on the cost of initial loan loss provisions-to clients-as these provisions would hopefully be needed only in the short term while clients build up payment histories in the formal system. It is, however, a genuine concern that ADEMI might have to sharply reduce its lending assistance to its smallest clients for regulatory reasons and·shift its focus towards larger microenterprises and small to medium scale enterprises. Finally, an issue that the transition cannot resolve (and has not resolved for any NGO that has .. established a formal banking affiliate) is the question of ownership. As an NGO, ADEMI does not • have shareholders, therefore one-fourth ofBancoADEMI has no ultimate owner. This will likely be addressed by phasing out ADEMI' s shareholding over time, particularly since ADEMI will have limited ability to inject further capital into the bank once all its clients are transferred to BancoADEMI. It is a considerable advantage that the shareholdings have been structured to ensure majority ownership for shareholders with a stake in the Bank and with the resources to provide further injections as needed. • SUSTAINABLE BANKING WI1H THE POOR 80 All three of these challenges (transfer modalities; asset classification, and the structure of ownership) are cutting-edge issues in microfinance. Much can therefore be learned from (Banco)ADEMI as it confronts these issues in 1998 and beyond. A Very Promising Start. BancoADEMI has originated more than 3,500 loans through June 1998, for a total of almost RD$ l 80 million. Eighty percent of these loans are for amounts of RD$50,000 (US$3,500) or less, accounting for over RD$58 million in lending. The recovery rates have remained extremely high, with the portfolio affected by arrears at under 4 percent of the total loan portfolio. The bank has taken on additional s~with commercial banking backgrounds to deal with financial aspects, including risk management, and with regulatory accounting and reporting requirements. Promising new areas ofunderseived small business lending have been identified, while savings services are also gradually being expanded. To date, BancoADEMI's activities have been so profitable that it expects to have more than sufficient returns for a dividend payout in its very first year of operations. A Benchmarkfor the Future. ADEMI has grown and evolved in remarkable ways over the last 15 years. It now stands on the threshold of a new era in its development, an era that promises countless challenges and opportunities. The beacon that it should follow, and the mark of its success, should always be to remain true to its fundamental mission of democratising credit and incorporating marginalised groups into the economic life of the nation. J SUSTAINABLE BANKING W11H THE POOR • 81 ··, · ·.: - ·.. - - - - - - - - - - - - ' - - - - BIBLIOGRAPHY ADEMI. 1982. By-Laws of the Association for the Development ofA1icroenterprises, Inc. Santo Domingo: Association for the Development ofMicroenterprises. ADEMI. 1997. lnforme Trimestral (Abril-Junio 1997). Santo Domingo: ADEMI. Banco Central de la Republica Dominicana. 1997. Boletin Trimestral. Enero-Marzo de 1997; Vol. LII, Nos. 1, 2 y 3. Santo Domingo: Banco Central de laRepublicaDominicana. Basle Committee on Banking Supervision. 1997. Core Principles for Effective Banking Supervision. Basle: Basle Committee. Benjamin, McDonald P. Jr. 1994. Credit Schemes for Microenterprises: Motivation, Design and Viability. Ph.D. Dissertation. Washington, D.C.: Georgetown University. Benjamin, McDonald P. Jr. 1994b. Designing Effective A1icrolending Institutions: Theory and Evidence. Unprocessed. Washington, D.C.: Georgetown University. Berenbach, Shari, and Craig Churchill. 1997. Regulation and Supervision ofMicro.finance Institutions: Experience from Latin America, Asia and Africa. Occasional Paper No. 1. Washington, D.C.: The MicroFinance Network. Cabal, Miguel. 1993. Evoluci6n de las Mcroempresas y Pequeiias F-mpresas en la Republica Dominicana, 1992-93. Santo Domingo: Fondo para el Financiamiento de la Microempresa, Inc. (FondoMicro). Calvo, Felix. 1996. Los Hijos de Mercurio en la Quiebra de BANCOlvfERC/0. Santo Domingo: Editorial Tiempo, S.A. Central Intelligence Agency. 1997. Map of the Dominican Republic. The World Factbook. Website: http://www.odci.gov/cia/publications/nsolo/factbook/map-gif/dr-150 .gif Dauhajre, Andres. 1992. Jmpacto Macroecon6mico de las A1icroempresas. Sabado Econ6mico, 5th September. Santo Domingo: Listin Diario. Dauhajre, Andres, Jose Achecar Chupani and Anne Swindale. 1994. Estabilizaci6n, Aperturay Dinamica de la Desigualdady la Pobreza en la Republica Dominicana: 1986-92. Santo Domingo: Fundaci6n Economia y Desarrollo, Inc. (FED). ECOCARIBE. 1997. Estudio de Factibilidad Econ6mica y Financiera para la Creaci6n def Banco de Desarrollo ADEJvfl S.A. Santo Domingo: ECOCARIBE. Economist Intelligence Unit. 1997. Dominican Republic. Country Report, 3rd Quarter 1997. London: Economist Intelligence Unit. Funk and Wagnall's Corporation. 1994. Dominican Republic: History. Website: http ://www.sjbc.org/missions/drinfo3 .htm. Goldmark, Lara. 1996. Business Development Services in Latin America. Washington, D.C.: The Inter-American Development Bank. SUSTAINABLE BANKING WITII TIIE POOR 82 Gomez, Arelis. 1989. ADFMI: Programa de Financiamiento aMicroempresas; lnforme de Evaluacion de lmpacto. Santo Domingo: Asociaci6n para el Desarrollo de Microempresas, Inc. (ADEMI). Inter-American Development Bank. 1994. Republica Dominicana: lnforme Socioeconomico. Washington, D.C. : Departamento de Desarrollo Econ6mico y Social, IDB. International Monetary Fund. 1997. International Financial Statistics Yearbook. Washington, D.C.: International Monetary Fund. Jimenez, Pedro. 1997. Executive Summary of the Association for the Development of Microenterprises, Inc. (ADFMI). Santo Domingo: ADEMI. Jimenez, Pedro. 1997b. Nuevo Plan de lncentivos para Asesores de Credito. Internal Memorandum to all Operational Staff, dated November 30, 1997. Santo Domingo: ADEMI. Jorion, Philippe. 1997. Value at Risk: The New Benchmarkfor Controlling Market Risk. Chicago: Irwin Professional • Publishing. ' Koch, Timothy. 1992. Bank Management. Second Edition. Chicago: Dryden Press. Ledgeiwood, Joanna, and Jill Burnett. 1995. A Case Study Review: ADFMI, The Dominican Republic. Individual Microlending Research Project. Toronto: Calmeadow. • Lewin, A. Christopher. 1991. The ADFMI Approach to Microenterprise Credit. Growth and Equity through Microenterprise Investments and Institutions (GEMINI) Project. Washington, D.C.: USAID. Lewin, A. Christopher. 1991b. Al Borde de la Cumbre: El Programa de Microempresas de ADFMI. Santo Domingo/Wiesbaden: ADEMI/Settlements and Urban Management (SUM) Consult. Liedholm, Carl, and Donald Mead. 1993. The Structure and Growth ofMicroenterprises in Southern and Eastern Africa: Evidence from Recent Surveys. GEMINI Working Paper No. 36. Washington, D.C.: Development Alternatives, Inc. McPherson, M.A. 1992. Growth and Survival ofSmall Southern African Firms. Ph.D. Dissertation. East Lansing: Michigan State University. Microfinance Network. 1997. MicroFinance Network: Basic Statistics (J 2/96). Website: http://www.bellanet.org/partners/mfn/whoswho3 .html . Paxton, Julia, and Carlos Cuevas. 1996. :A Worldwide Inventory ofMicro.finance Institutions. Sustainable Banking with the Poor (SBP) Project. Washington, D.C.: The World Bank. • PRODEM. ·1996. lndicadores de/ Servicio de Credito paraMicroempresarios, Correspondiente a Noviembre de 1996. La Paz: PRODEM. Stiglitz, Joseph. 1986. The Wage-Productivity Hypothesis: It's Economic Consequences and . Policy Implications for L.D.C.s. Working Paper Series (U.S.) No. 1976: 1-34. Washington, D.C.: National Bureau Of Economic Research. SUSTAINABLE BANKING WITH THE POOR 83 Tecnicas de Planificaci6n. 1997. Lineamientos Genera/es Para una &trategia de Desarrollo lnstitucional de ADFl\.11. Santo Domingo: Tecnicas de Planificaci6n. Tenn, Sharyn. 1997. A Preliminary Evaluation ofImpact: ADFMI Client Case Studies. Unprocessed. New York: Columbia University. • Webster, Leila M., Randall Riopelle and Anne-Marie Chidzero. 1996. World Bank Lending for Small Enterprises, 1989-1993. Technical Paper No. 311. Washington, D.C.: The World Bank. World Bank. 1992. Dominican Republic: Updating &anomic Memorandum: !- The Challenge of Sustainability. Report No. 10614-DO. Washington, D.C.: The World Bank. World Bank. 1995. The Dominican Republic, G.rowth with &[uity: An Agenda for Reform. Report No. 13619-DO. Washington, D.C.: The World Bank. World Bank. 1997. Wor.JdDevelopment Indicators CD-ROM Washington, D.C.: The World Bank. Yaron, Jacob. 1992. Successful Rural Finance Institutions. Working Paper 150. Washington, D.C.: The World Bank. Yaron, Jacob, McDonald P. Benjamin Jr. and Gerda Piprek. 1997. Rural Finance: Issues, Design and Best Practices. Environmentally and Socially Sustainable Development Studies and Monographs Series, No. 14. Washington, D.C.: The World Bank. SUSTAINABLE BANKING WI1H THE POOR 84 Appendix 1: Key Macroeconomic and Financial Sector Indicators 1982 1983 1984 198S 1986 1987 1988 1989 1990 1991 1992 1993 . 1994 199S ~ GDP (RDS Millions) 7,964 9,220 11,594 15,702 17,780 . 22,404 32,850 42,393 60,305 96,333 112,698 121,808 136,206 160,587 181,466 GDP Deflator(1990=100) 16.l 17.4 21.6 29.9 32.7 37.5 53.8 66.5 100.0 158.2 171.3 179.8 192.7 216.9 228.3 Real Growth in GDP (%) . 7.1% 1.3% -2.2% 3.5% 9.9% 2.2% 4.4% -S.4% 1.0% 8.0% 3.0% 4.3% 4.7% 7.4% Population (Millions) 5.98 6.12 6.27 6.42 6.56 6.71 6.86 7.01 7.17 7.32 7.47 7.62 7.77 7.91 8.05 Exchange Rate (RD$/US$) 1.00 1.00 1.00 2.94 3.08 4.96 6.34 6.34 11.35 12.66 12.58 12.77 13.06 13.47 14.06 GDP per Capita (US$) 1,332 1,507 1,849 832 880 673 755 954 741 1,040 1,199 1,252 1,342 1,507 1,603 Real Effective Exch. Rate (1990=100) 158.30 155.50 110.10 123.00 116.80 100.00 83.50 101.00 100.00 101.00 101.60 105.90 110.00 113.00 117.70 Exports f.o.b. (USS Millions) 768 785 868 739 722 711 890 924 735 658 563 2,738 3,296 3,052 3,167 Imports f.o.b. (USS Millions) 1,257 1,279 1,257 1,286 1,352 1,592 1,608 1,964 1,793 1,729 2,174 4,590 5,015 4,844 5,350 Current Account (US$ Millions) (443) (418) (163) (108) (183) (364) (19) (327) (280) (157) (708) (447) (68) (125) (517) Reserves excl. Gold (USS Millions) 129.0 171.3 253.5 340.l 376.3 182.2 254.0 164.0 61.6 441.9 499.8 651.2 252.1 365.6 350.3 Reserves: Months of Imports (mos.) 1.2 1.6 2.4 3.2 3.3 1.4 1.9 1.0 0.4 3.1 2.8 1.7 0.6 0.9 0.8 Total External Debt (USS Billions) . . . . . . . . . 4.5 4.6 4.9 4.2 4.3 4.3 Debt-Service Ratio (%) . . . . . . . . . 13.1% 15.3% 6.5% 8.7% 7.3% 7.9% Govt. Surplus or (Deficit) (RDSm) (232.1) (214.2) (71.l) (205. l) 125.8 (69.5) (73.5) 143.7 367.3 1,013.7 . . (690.6) 1,720.3 540.6 Ratio of Surplus (Deficit) to GDP (%) -2.9% -2.3% -0.6% -1.3% 0.7% -0.3% -0.2% 0.3% 0.6% 1.1% . . -0.5% 1.1% 0.3% Government Expenditures (RDSm) 988.8 1,136.1 1,236.9 1,829.3 2,222.7 3,009.3 4,626.7 5,701.4 6,548.2 9,165.2 . . 22,190.5 23,170.5 26,593.0 Govt. Expenditures/ GDP (%) 12.4% • 12.3% 10.7% 11.7% 12.5% 13.4% 14.1% 13.4% 10.9% 9.5% . . 16.3% 14.4% 14.7% Money (RDSm) 784 842 1,247 1,456 2,080 2,773 4,324 5,374 7,587 9,448 11,047 13,549 13,742 16,092 20,884 Index of Ml (1990=100) 10.33 11.10 16.44 19.19 27.42 36.55 56.99 70.83 100.00 124..53 145.60 178.58 l8l.l3 212.10 275.26 Money + Quasi-Money (RDSm) 1,434 1,654 2,144 2,492 4,335 5,061 7,625 9,954 13,849 19,058 24,368 30,775 33,717 39,629 48,043 Index ofM2 (1990=100) 10.35 11.94 15.48 17.99 31.30 36.54 55.06 71.88 100.00 137.61 175.95 222.22 243.46 286.15 346.91 M2/GDP(%) 18% 18% 18% 16% 24% 23% 23% 23% 23% 20% 22% 25% 25% 25% 26% Consumer Price Index (1990=100) 12.8 13.4 17.l 23.5 25.8 29.9 43.1 62.7 100.0 153.9 160.9 169.4 183.4 206.4 214.6 Inflation (%) 7.6% 4.7% 27.6% 37.4% 9.8% 15.9% 44.1% 45.5% 59.5% 53.9% 4.5% 5.3% 8.3% 12.5% 4.0% Market Reference Interest Rate(%) . 20.0% 22.0% 20.1% 18.4% 19.3% 26.7% 27.6% 36.4% 30.0% 25.0% 22.5% 20.8% 23.5% 21.0% Real Market Reference Rate (%) . 14.6% -4.4% -12.6% 7.8% 2.9% -12.1% -12.3% -14.5% -15.5% 19.6% 16.4% 11.5% 9.7% 16.4% (Highest) Cap on Lending Rates (%)* 7.5% 7.5% 7.5% 12.0% 12.0% 12.0% 12.0% 12.0% 12.0% 43.0% 30.0% . . . . Avg. Commercial Lending Rate(%) . . . . . . . . . 35.3% 28.3% 29.9% 28.7% 30.8% 25.3% Ceiling on CD Rates(%)* 14.0% 14.0% 14.0% 18.0% 16.0% 16.0% 18.0% 24.0% 36.0% . . . . . 90-day CD Rate(%) . . . . . . - . - 20.7% 16.3% 15.0% 13.6% 15.8% 13.9% Deposit Rate (%) 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 5.0% 10.5% 6.2% 5.6% 5.0% 4.9% 4.7% 5.0% • Note: During 1985-91, commissions on loans were also capped, initially at 6% on new lending, rising to 24% on new lending by 1990. Lending rates were deregulated in 1992. Interest rates on liabilities became freely negotiable in January 1991. Source a: Banco Central de la Repriblica Dominicana [1997], Benjamin [1994], Economist Intelligence Unit [1997]. International Monetary Fund { 1997] Appendix 2: Poverty Indicators Table A2.1: Poverty in the Dominican Republic Percent Year Total Urban Rural Total Population Poverty 1986 18.3 11.7 24.5 1989 24.5 23.3 27.4 1992 20.6 10.9 29.8 Extreme poverty 1986 10.5 7.1 14.1 1989 13.7 12.7 16.3 1992 9.0 5.0 12.3 Children in Poverty Poverty 1992 26.8 14.0 37.3 Extreme poverty 1992 11.8 6.5 15.6 Chronic malnutritior 1991 19.4 14.6 26.3 Severe malnutrition 1991 5.9 3.1 9.8 Gender and Poverty Female po01 1986 22.6 15.4 32.7 1992 15.8 10.8 24.4 Male POOi 1986 14.8 9.8 18.8 1992 16.7 8.5 23.6 Female extreme poor 1986 12.5 8.2 18.5 1992 8.5 7.1 10.0 Male extreme poor 1986 8.6 6.5 10.2 1992 7.4 2.7 10.9 Source: World Bank [1995), lnter-Amencan Development Bank [1994), Dauhajre et al. [1994) Note: Figures for the Poor include the Extreme Poor Table A2.2: Income Distribution in the D.R., 1986-92 Percent of Total Income 1986 1989 199~ Poorest Decile 1.1 0.8 1.6 Second Decile 2.7 1.9 2.8 Middle Six Deciles 43.6 36.6 39.1 Ninth Decile 16.6 16.4 14.5 Richest Decile 36.0 44.3 42.0 Gini Coefficient 0.47 0.55 0.49 Source: World Bank [1995), Dauhajre et al. [1994) SUSTAINABLE BANKING WITH THE POOR 86 Appendix 2: Poverty Indicators (continued) Table A2.3: Income Sources of the Rich and Poor, 1992 Poorest Richest 20% 20% Wages 59.4 30.9 Family business 4.3 10.9 Self employment 12.2 38.4 Domestic transfers 4.3 0.3 Foreign remittances 1.5 5.6 Sale of agricultural products 11.1 6.7 Other 7.2 7.2 Total 100.0 100.0 Sources: World Bank [1995], Dauhajre et al. [1994] Table A2.4: Expenditure Patterns of the Rich and Poor, 1992 Poorest Richest 20% 20% Food 60.6 41.2 Transport 4.7 7.7 Clothing and Shoes 4.8 6.0 Cigarettes and Beverages 2.7 2.5 Health 2.6 2.8 Schooling 1.9 2.7 Electricity and Water 1.1 5.0 Other 21.6 32.1 Total 100.0 100.0 Source: World Bank [1995], DauhaJre et al. [1994] SUSTAINABLE BANKING WITH THE POOR 87 Appendix 3: ADEMl's Incentive Scheme for Credit Advisors 1. Contaminated portfolio 4% 1,400 over total loan portfolio 3% 1,800 2% 2,400 1% 3,000 2. Number of clients with 130-140 1,000 active loans 141-150 1,200 151+ 1,500 3. Loan portfolio (RD$m.) 3.0 700 4.0 1,000 5.0+ 1,500 Source: Jimenez [1997b] Notes: 1) The above incentive scheme went into effect on January 1, 1998 and applies to all credit advisors. 2) Since 1997, ADEMI has replaced arrears over total loan portfolio with contaminated portfolio over total loan portfolio as the measure of portfolio quality by which CAs' performance is assessed-the contaminated portfolio is defined as the total volume of loans outstanding affected by arrears. 3) In order to qualify for. any incentive payments, CAs must meet minimum requirements for both the contaminated portfolio (under four percent) and the number ofactive borrowers (at least 130), but not necessarily for size of loan portfolio (over RD$3 million). 4) These incentive payments are in addition to the Christmas bonus (13th month), an automatic second month's salary in December (14th month), leave conipensation (up to one month's salary), and a bonus related to ADEMI's overall performance (that can be equivalent to three months' minimum salary). SUSTAINABLE BANKING Wl1H TIIE POOR 88 Appendix 4: ADEMl's Financial Statements (Dominican Pesos '000s) 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993· 1994 1995 1996 Assets - - - - - - - - - - - - - - • Cash & Banks: 42 37 544 832 837 556 1,901 2,466 4,137 5,839 14,118 17,374 26,140 77,99'i Accounts Receivable: n.a. n.a. n.a. n.a. n.a. 94 210 349 638 2,242 2,446 2,479 2,464 3,344 Accrued Interest Receivable: n.a. n.a. n.a. n.a. n.a. 74 512 1,339 5,643 9,803 9,563 15,981 27,564 37,78'i Gross Loan Portfolio: 6,006 11,898 17,933 52,225 86,960 142,398 200,341 277,645 340,225 Less: Provisions For Loan Losses: (109) (283) (331) (1,133) (3,028) (4,531) (7,649) (9,462) (14,893 Net Loan Portfolio: 247 514 944 1,595 3,022 5,897 11,615 17,602 51,092 83,932 137,867 192,692 268,183 325,332 Land, Buildings And Equipment: 93 213 443 564 2,371 4,109 6~739 8,262 11,268 13,832 13,821 Other Assets: 28 102 168 186 405 14 95 182 493 760 1,133 1,034 1,860 1,930 Total Assets: 317 653 1,656 2,706 4,477 7,078 14,897 24,309 66,112 109,315 173,389 240,828 340,043 460,211 Liabilities And Capital Liabilities BPD Overdraft: 65 1,089 Banco Popular Dominicano: 75 50 30 500 1,300 3,300 4,500 9,500 1,000 2,000 1,000 23,500 3,500 Loans From Microenterprises: 13 13,835 23,569 23,087 37,399 52,776 Banco Mercantil / Jnteram. Inv. Colp, 13,450 14,000 Inter-American Development Bank: 231 1,030 1,065 1,065 1,065 4,086 4,252 4,252 4;252 4,252 4,216 4,181 4,145 President's Technical Secretariat: 525 1,250 2,450 2,450 4,988 7,700 7,621 1,050 1,050 1,050 1,050 Fondomicro: 10,000 28,800 47,500 54,600 European Investment Bank: 17,792 43,895 89,878 143,763 Add: EIB Exchange Risk Provision: 2,451 7,698 16,956 Other Loans, Incl. Revolving Fund: 737 1,515 1,515 1,515 1,515 1,515 1,453 Other Liabilities: 73 99 105 64 178 201 721 1,002 4,216 4,633 4,341 8,617 12,934 17,821 Total Liabilities: 148 380 1,165 1,654 2,993 5,016 10,622 15,479 37,196 62,745 102,019 140,431 191,605 255,464 Capital Equity: 169 272 491 1,052 1,485 2,062 4,274 8,830 28,916 46,569 71,370 100,398 148,438 204,747 Total Capital: 169 272 491 1,052 1,485 2,062 4,274 8,830 28,916 46,569 71,370 100,398 148,438 204,747 Total Liabilities And Capital: 317 652 1,656 2,706 4,478 7,078 14,896 24,309 66,112 109,314 173,389 240,829 340,043 460,211 Source: ADEMI's Audited Financial Statements, ADE!v.i {1997], Benjamin {1994] Appendix 4: ADEMl's Financial Statements (continued) (Dominican Pesos '000s) 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 Income - Income From Loans: 53 134 336 568 1,109 2,274 5,616 10,506 26,703 42,018 , 53,657 72,166 104,206 115,042 Guarantee Fees: 221 1,024 1,816 • 2,502 2,125 2,464 3,331 Income From Grants: 238 226 415 659 450 168 350 890 7,336 553 .6,660 50 1,271 Other Income: 118 210 397 1,561 1,855 2,470 3,898 6,148 7,794 Total Income: 291 360 751 1,227 1,559 . 2,560 6,176 12,014 36,624 46,242 65,289 78,189 112,868 127,444 Expenses Total Interest Expense: 5 18 11 18 75 199 799 1,553 4,234 9,022 16,713 16,540 21,166 ~ Comprising - Banco Popular Dominicano: 5 18 5 6 59 176 673 1,099 965 411 1,343 336 2,235 1,718 f: Loans From Microenteq,rises: 2 3,877 5,980 4,912 5,898 . 8,479 '"•" Banco Mercantil / Interam. Inv. Corp'. 1,446 1,862 }":- Inter-American Development Bank: 6 12 6 11 32 11 62 45 43 43 58 42 President's Technical Secretariat: 10 12 94. 441 965 1,025 481 · 95 94 95 .Fondomicro: 2,238 3,654 8,749 10,439 8,680 European Jnvestment Bank: 108 687 2,664 2,162 Other Loans, Incl. Revolving Fund: 2 2 10 9 28 28 18 Salaries And Administrative Costs: 115 226 446 595 990 1,656 2,754 5,185 10,625 16;503 20,152 24,849 31,776 38,518 ,-:- · Guarantees Paid: 9 EIB Exchange Risk Provision: 2,451 5,247 9,258 Loan Loss Provisions: 3 13 76 66 61 127 334 635 1,503 2,762 3,207 4,949 6,305 7,846 Depreciation: 77 75 175 303 415 374 857 1,809 Total Expenses: 123 257 533 679 1,126 1,982 3,964 7,457 16,537 28,590 40,487 49,163 65,351 71,728 Net Income: 168 103 218 548 433 578 2,212 4,557 20,087 17,652 24,802 29,026 47,517 55,716 Other Adjustments To Net Worth: 523 593 Change In Net Worth (Capital): 168 103 218 548 433 578 2,212 4,557 20,087 17,652 24,802 29,026 48,040 56,309 Source: ADF,M/'s Audited Financial Statements, ADF,M/ {1997], Benjamin [1994} '----- SUSTAINABLE BANKING WITII THE POOR 90 Appendix 5: Calculation of the Subsidy Dependence Index for ADEMI (Dominican Pesos '000s) 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 - - - - - - - - - - - - - - - - - - - - - - - - - - Derivation Of The Market Reference Rate Marginal Financial Cost Of Funds: 9.5% 9.5% 10.5% 11.5% 12.5% 14.0% 22.0% 18.0% 17.0% 15.0% 13.6% 15.8% 13.8% Marginal Administrative Cost Of Funds: 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% Reference Rate Without Required Reserves: 12.5% 12.5% 13.5% 14.5% 15.5% 17.0% 25.0% 21.0% 20.0% 18.0% 16.6% 18.8% 16.8% IAverage Reserve Requirement: 43.2% 37.9% 26.6% 24.7% 41.9% 38.4% 31.4% 30.1% 20.0% 20.0% 20.0% 20.0% 20.0% Rate Of Interest On Required Reserves: 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Market Reference Rate: 22.0% 20.1% 18.4% 19.3% 26.7% 27.6% 36.4% 30.0% 25.0% 22.5% 20.8% 23.5% 21.0% Identification Of The Sources Of Subsidy Cost Of Concessional Liabilities: 0.0% 1.0% 0.9% 0.8% 0.8% 2.5% 5.5% 19.5% 14.4% 16.4% 12.7% 11.4% 1.9% IAverage Concessional Liabilities: 116 631 1,310 1,953 2,915 5,026 8,257 16,722 32,828 57,149 · 88,693 100,950 123,518 Subsidy On Concessional Liabilities: • 25 121 229 360 755 1,261 2,555 1,757 3,473 3,468 7,112 12,199 23,622 Reference Rate For Nonconcessional Funds Exempted From The Reserve Requirement: 28.8% 12.5% 40.0% 23.6% 19.6% 28.9% 27.9% 13.8% 33.7% 35.3% 21.1% 19.5% 16.7% lAverage Nonconcessional Liabilities: 63 40 15 250 900 2,333 3,933 7,007 12,719 20,747 24,828 49,218 72,313 Subsidy On Nonconcessional Liabilities: (4) 3 (3) (11) 64 (29) 334 1,138 (1,108) (2,655) (96) 1,987 3,12 Opportunity Cost Of Equity: 22.0% 20.1% 18.4% 19.3% 26.7% 27.6% 36.4% 30.0% 25.0% 22.5% 20.8% 23.5% 21.0% Accounting Cost Of Capital: 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% lAnnual Average Equity: 221 382 772 1,269 1,774 3,168 6,552 18,873 37,743 58,970 85,884 124,418 176,593 Subsidy On Equity: 49 77 142 244 473 874 2,388 5,670 9,436 13,268 17,821 29,238 37,084 Miscellaneous Grants And Benefits: 226 415 659 450 168 350 890 7,336 553 6,660 0 50 1,271 Calculation Of The SDI Total Subsidy Received By The DFI: 296 616 1,027 1,043 1,460 2,456 6,167 15,901 12,353 20,742 24,836 43,475 65,104 Profit (Loss), Adjusted For Forex Provisions: 103 218 548 433 578 2,212 4,557 20,087 17,652 24,802 31,477 52,764 64,974 Net Subsidy: 193 398 479 610 882 244 1,610 (4,186) (5,299) (4,060) (6,641) (9,289) 130 Interest Income: 134 336 568 1,109 2,274 5,616 10,506 26,703 42,018 53,657 72,166 104,206 115,042 Subsidy Dependence Index (SDI): 143.8% 118.4% 84.3% 55.0% 38.8% 4.3% 15.3% -15.7% -12.6% -7.6% -9.2% -8.9% 0.1% Requisite Interest Rate Adjustment Current Average On-Lending Rate: 35.2% 46.1% 44.7% 48.0% 51.0% 64.1% 71.9% 77.7% 62.2% 48.4% 43.7% 45.2% 38.8% Increase In The On-Lending Rate Required To Eliminate Subsidy Dependence: 50.6% 54.6% 37.7% 26.4% 19.8% 2.8% 11.0% -12.2% -7.8% -3.7% -4.0% -4.0% 0.0°/4 Lowest Subsidy-Free On-Lending Rate: 85.9% 100.7% 82.4% 74.5% 70.8% 66.9% 82.9% 65.6% 54.4% 44.7% 39.6% 41.2% 38.8% Source: ADEA,1/'s Audited Fmanc,al Statements, ADEA,11 {1997], Banco Central de laRepublicaDominicana [1997], Benjamin [1994} Appendix 6: ADEMl's Financial Performance Ratios 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 Market Indicators - Rate Of Consumer Price Inflation: 4.7% 27.0% 37.5% 9.7% 16.0% 44.4% 45.4% 59.5% 53.9% 4.3% 5.3% 8.3% 12.5% 5.4o/c Market Reference Rate: 22.0% 20.1% 18.4% 19.3% 26.7% 27.6% 36.4% 30.0% 25.0% 22.5% 20.8% 23.5% 21.0o/c Exchange Rate (Peso / US Dollar): 1.00 1.00 2.94 3.08 4.96 6.34 6.34 11.35 12.66 12.58 12.77 13.06 13.47 ''6 Key Financial Performance Ratios Subsidy Dependence Index 143.8% 118.4% 84.3% 55.0% 38.8% 4.3% 15.3% -15.7% -12.6% -7.6% -9.2% -8.9% 0.1 o/c - Return On Equity 46.7% 57.1% 71.0% 34.1% 32.6% 69.8% 69.6% 106.4% 46.8% 42.1% 33.8% 38.2% 31.6o/c Growth In Equity 60.9% 80.5% 114.3% 41.2% ·38.9% 107.3% 106.6% 227.5% 61.0% 53.3% 40.7% 47.8% 37.9o/c ROE w/o. Subsidies=(Profit-Subsidies}'Equity -87.4% -104.3% -62.0% -48.1% -49.7% -7.7% -24.6% 22.2% 14.0% 6.9% 7.7% 7.5% -0.1 o/c Return On Assets 21.2% 18.9% 25.1% 12.1% 10.0% 20.1% 23.2% 44.4% 20.1% 17.5% 14.0% 16.4% 13.9o/c Growth In Assets 106.0% 153.6% 63.4% 65.4% 58.1% 110.5% 63.2% 172.0% 65.3% 58.6% 38.9% 41.2% 35.3o/c t/· ROA w/o. Subsidies=(Profit-Subsidies)iAssets -39.7% -34.5% -21.9% -17.0% -15.3% -2.2% -8.2% 9.3% 6.0% 2.9% 3.2% 3.2% 0.0o/c it Total Income / Total Expenses 236.6% 140.1% 140.9% 180.7% 138.5% 129.2% 155.8% 161.1% 221.5% 161.7% 161.3% 159.0% 172.7% 177.7o/c -:~ .· Total Income w/o. Grants / Total Expenses 43.1% 52.1% 63.0% 83.7% 98.5% 120.7% 147.0% 149.2% 177.1% 159.8% 144.8% 159.0% 172.6% 175.9o/c ,- Balance Sheet Ratios C!ISh And Banlcs / Total Assets 13.2% 5.7% 32.9% 30.7% 18.7% 7.9% 12.8% 10.1% 6.3% 5.3% 8.1% 7.2% 7.7% 16.9o/c · - Net Loan Portfolio / Total Assets 77.9% 78.7% 57.0% 58.9% 67.5% 83.3% 78.0% 72.4% 77.3% 76.8% 79.5% 80.0% 78.9% 70.7o/c --- Fixed Assets/ Total Assets 0.0% 0.0% 0.0% 3.4% 4.8% 6.3% 3.8% 9.8% 6.2% 6.2% 4.8% 4.7% 4.1% 3.0o/c ~t'- : Cash And Banlcs / Loans from Microents. 31,823% 42.2% 59.9% 75.3% 69.9% 147.8o/c ,,(,.. Loans from Microents. / Net Loan 0.0% 16.5% 17.1% 12.0% 13.9% 16.2o/c c~ - Portfolio Total Borrowings/ Net Loan Portfolio 30.4% 54.7% 112.3% 99.7% 93.2% 81.7% 85.2% 82.2% 64.6% 69.2% 70.8% 67.1% 63.8% 67.8o/c New Annual Lending /Net Avg Loan Portf. 5.52 3.94 3.74 3.70 2.95 2.84 2.46 2.90 2.58 2.27 1.79 1.68 1.33 Debt/ Equity 0.9 1.4 2.4 1.6 2.0 2.4 2.5 1.8 1.3 1.3 1.4 1.4 1.3 1.2 Equity/ Net Loan Portfolio 68.4% 52.9% 52.0% 66.0% 49.1% 35.0% 36.8% 50.2% 56.6% 55.5% 51.8% 52.1% 55.3% 62.9o/c Fixed & Other Assets / Equity 16.6% 37.5% 34.2% 26.5% 41.6% 22.2% 15.4% 28.9% 15.9% 16.1% 13.2% 12.3% 10.6% 7.7o/c Portfolio Quality Ratios Provisions at Start ofYear /Net Av. Loan Port 1.2% 1.9% 1.0% 1.7% 2.7% 2.7% 3.3% 3.2o/c Annual Provisions /Net Avg. Loan Portfolio 3.4% 10.4% 5.2% 2.6% 2.8% 3.8% 4.3% 4.4% 4.1% 2.9% 3.0% 2.7% 2.6o/c Write-Offs / Net Avg. Loan Portfolio 1.6% 1.8% 4.0% 2.0% 1.3% 1.5% 1.1% 1.9% 0.8o/c 2.4% 3.2% 2.3% 3.3% 4.5% 4.1% 4.6% 4.1% - 5.0o/c Year-end Cumul. Provisions /Net Avg. Loan Portf. Arrears/ Gross Y.E. Loan Portfolio 10.1% 10.1% 8.4% 9.6% 5.8% 5.7o/c Yr.End Provisions/ Gross Y.E. Loan Portfolio 1.8% 2.4% 1.9% 2.2% 3.6% 3.3% 4.0% 3.5% 4.6o/c Cumul. Provisions / Arrears 21.5% 34.6% 37.8% 39.7% 58.6% 76.3o/c Cumul. Provisions / Arrears Over 30 Days 38.0% 57.3% 58.1% 55.3% 97.3% 114.8o/c SUSTAINABLE BANKING wrm TIIE POOR 92 Appendix 6: ADEMl's Financial Performance Ratios (continued) 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 Portfolio Quality Ratios (Continued) - - -- - - - - - - - - - - - - - - - - - - -- -- Interest Accrued/ Net Year-end Loan Portfolio 1.3% 4.4% 7.6% 11.0% 11.7% 6.9% 8.3% 10.3% 11.6% Interest Accrued / Total Assets • 1.0% 3.4% 5.5% 8.5% 9.0% 5.5% 6.6% 8.1% 8.2o/i Incremental Accrued Interest/ Int. Income 7.8% 7.9% 16.1% 9.9o/o -0.4% 8.9% 11.1% 8.9o/i Financial Margin Ratios Interest Income / Net Loan Portfolio 35.2% 46.1% 44.7% 48.0% 51.0% 64.1% 71.9% 77.7% 62.2% 48.4% 43.7% 45.2% 38.8¾ Interest Expense / Net Loan Portfolio 4.7% 1.5% 1.4% 3.2% 4.5% 9.1% 10.6% 12.3% 13.4% 15.1% 11.5% 11.5% 7.9o/i Interest Expense / Borrowed Funds 10.1% 1.6% 1.4% 3.4% 5.2% 10.9% 12.7% 17.8% 19.8% 21.5% 14.6% 14.1% 7.3o/i Interest Spread/ Net Loan Portfolio 30.5% 44.6% 43.3% 44.8% 46.5% 55.0% 61.3% 65.4% 48.9% 33.3% 33.7% 36.0% 33.9o/c Int. + Foreign Exchange Provisions / Borrowings 10.1% 1.6% 1.4% 3.4% 5.2% 10.9% 12.7% 17.8% 19.8% 21.5% 16.7% 17.6% 12.0¾ Spread After Forex Provisions / Loan Port. 30.5% 44.6% 43.3% 44.8% 46.5% 55.0% 61.3% 65.4% 48.9% 33.3% 32.2% 33.8% 30.8¾ Lowest Subsidy-Free On-Lending Rate 85.9% 100.7% 82.4% 74.5% 70.8% 66.9% 82.9% 65.6% 54.4% 44.7% 39.6% 41.2% 38.8¾ Real Average On-Lending Rate 6.5% 6.2% 31.9% 27.7% 4.6% 12.9% 7.8% 15.5% 55.6% 41.0% 32.7% 29.0% 31.7¾ Asset Utilisation Ratios Total Income / Total Assets 74.2% 65.0% 56.3% 43.4% 44.3% 56.2% 61.3% 81.0% 52.7% 46.2% 37.8% 38.9% 31.9¾ Income From Loans / Total Income 18.2% 37.2% 44.7% 46.3% 71.1% 88.8% 90.9% 87.4% 72.9% 90.9% 82.2% 92.3% 92.3% 90.3¾ Grant Income/ Total Income 81.8% 62.8% 55.3% 53.7% 28.9% 6.6% 5.7% 7.4% 20.0% 1.2% 10.2% 0.0% 0.0% l.0o/c Profit Margin Ratios Operating Income/Tot Income w/o. Grants 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0¾ Interest Expense/Tot. Income w/o. qt-ants -9.4% -13.4% -3.3% -3.2% -6.8% -8.3% -13.7% -14.0% -14.5% -19.7% -28.5% -21.2% -18.8% -11.3¾ Salaries & Admin/fot. Income w/o. Grants -217.0% -168.7% -132.7% -104.8% -89.3% -69.2% -47.3% -46.6% -36.3% -36.1% -34.4% -31.8% -28.2% -30.So/i -9.7% -22.6% -11.6% -5.5% Provis. & Other Exp./Tot. Inc. w/o. Grants -5.7% - - -5.3% - -7.1% - - -6.5% - - - -6.7% - -5.7% -6.2% - - -9.9% -11.0% -15.0¾ - Net Income/ Total Income w/o. Grants -132.1% -91.8% -58.6% -19.5% -1.5% 17.1% 32.0% 33.0% 43.5% 37.4% 30.9% 37.1% 42.1% 43.2¾ ~et Income / Total Income 57.7% 28.6% 29.0% 44.7% 27.8% 22.6% 35.8% 37.9% 54.8% 38.2% 38.0% 37.1% 42.1% 43.7¾ "Burden" and Productivity Ratios Salaries & Administration/ Total Assets 46.6% 38.6% 27.3% 27.6% 28.7% 25.1% 26.5% 23.5% 18.8% 14.3% 12.0% 10.9% 9.6o/i Salaries & Admin. /Net Loan Portfolio 59.4% 61.2% 46.9% 42.9% 37.1% 31.5% 35.5% 30.9% 24.4% 18.2% 15.0% 13.8% 13 .0¾ Salaries & Admin./Annual Disbursements 11% 11% 16% 13% 12% 13% 11% 14% 11% 9% 8% 8% 8% JO¾ Salaries & Admin I New Loan (RD$'90) 287 479 1,059 754 643 749 576 658 613 727 876 1,031 1,088 1,337 No. Loans Approved/ No. Of Staff 426 138 58 61 59 58 83 68 84 92 83 74 73 67 Annual Dish. ('90RD$000s)'No. Of Staff 1,098 615 395 368 325 347 431 313 480 709 909 902 968 916 Avg Loan (RD$'90): Disb./Loans Made 2,580 4,451 6,822 6,010 5,555 5,950 5,206 4,570 5,743 7,670 10,942 12,253 13,233 13,712 Avg Loan(RD$'90): Portf./No. Of Clients 4,118 3,746 2,579 3,972 4,867 6,077 7,385 9,323 9,763 Source: AD.EA.i's Audited Frnanc1al Statements, ADEA1l {1997], Banco Central de la Republica Dominicana (1997], Be,yamin [1994}, Economist Intelligence Unit (1997}, DvlF [1997] - -- - · - -- - -- - - ---·- Appendix 7: Branch Level Income Statements and Portfolio Data as of Septem:t>er 1996, in Dominican Pesos '000s *1 Total Total Net Income/ Loan Arrears Active Income Expenses Income Expenses Portfolio >30 days Clients 1 Boca Chica *2 •1,285 132 1,153 976% - - - 2 San Francisco 1,131 147 983 767% 2,409 50 14~ 3 Monte Plata 2,296 392 1,905 587% 6,134 401 472 4 San Pedro 4,192 751 3,441 558% 10,846 258 78f 5 San Juan 2,671 533 2,138 501% 6,312 723 518 6 La Vega 1,838 369 . 1,469 498% 5,992 426 392 7Neyba 933 198 735 471% 2,858 46 157 8 Santiago 4,232 906 3,326 467% 10,947 362 677 9 Sabana de la Mar 1,074 233 841 461% •3,372 266 265 10 Higuey 772 168 603 458% 2,067 125 145 11 San Cristobal 2,427 536 1,891 453% 6,305 146 49S 12 El Seybo 840 205 634 409% 2,215 68 23S 13 Bonao 1,240 305 935 407% 3,446 11 24'i 14 Dajabon 1,008 258 750 390% 2,537 91 353 15 Azua 1,108 295 813 375% 2,729 116 169 16 Bani 905 263 642 344% 2,296 190 173 17Mao 486 163 323 298% 1,439 26 99 18 Hato Mayor 205 80 124 255% 384 21 59 19 La Romana 2,308 1,100 1,208 210% 5,555 418 430 20 Santo Domingo *3 57,621 21,130 36,491 273% 214,275 6,272 8,595 ADEMI Subtotal *4 88,570 28,166 60,405 314% 292,118 10,015 14,423 ADEMI Overheads 18,198 6,101 3,939 53 ADEMI Total *5 88,570 46,364 42,206 191% 298,219 13,953 14,47€ Notes: * 1: Income, expense and net income data cover the period January-September 1996. *2: Boca Chica was a very new branch in September 1996, still administered from Santo Domingo. In addition, branches in Barahona and in Santo Domingo's Villa Mella were only established later. • *3: Totals for Santo Domingo are calculated as a residual of totals for ADEMI minus totals for the Interior. Subtotal for Interior: 30,949 7,036 23,913 440% 77,842 3,742 5,828 *4: The subtotal for ADEMI excludes the following overheads for comparability across branches, as these data are not kept at the branch level: interest and insurance fund expenses, depreciation, and forex reserves. It also excludes loans under judicial administration. *5: The total for ADEMI includes these overheads. Source: ADFlv1/'s Management Information System SUSTAINABLE BANKING WITH THE POOR - •94 Appendix 8: ADEMI's Pension Plan Employees are fully vested in ADEMI's pension plan after 15 years. The normal pension benefit equals 2.8 percent per year of employment, multiplied by the average pay received by the employee, including bonuses, during the last two years of employment, with an 80 percent cap and a minimum benefit equal to 50 percent of the legal minimum wage. Disability benefits are quite similar, although staff become eligible sooner, while death and resignation prior to retirement entitle the employee to reimbursement of contributions augmented by IO percent per year over 5 years of employment, plus interest (currently also at IO percent). In addition, staff are eligible for low-interest loans that are repaid via salary deductions. The size and duration of these loans is related to years of service with ADEMI. Finally, .as of 1996, a generous bonus for employees with more than IO years service at ADEMI was still being funded by the pension plan, •although Management has rightly proposed to transfer this bonus out of the pension plan and into ADEMI's own administrative cost structure. The following table presents the financial position of the plan in recent years. ADEMl's Pension Fund: Balance Sheet and Statement of Net Fund Increase (in Dominican Pesos '000s) Balance Sheet 1994 1995 199E Cash and banks 241.2 124.3 141.0 Loans to members 2,065.4 2,436.5 3,342.4 Other receivables 182.4 261.1 Investments 1,907.3 3,933.7 5,276.9 Accrued interest 0.5 0.2 0.4 Total Assets 4,214.4 6,677.1 9,021.8 Liabilities 60.0 28.0 72.0 Net Fund Assets 4,154.4 6,649.1 8,949.8 Statement of 1994 1995 199E Net Fund Increase Income from investments 453.2 877.8 1,067.0 Employer contributions 1,207.0 1,484.1 1,029.7 Employee contributions 522.7 608.0 744.3 Less: Interest on contributions 72.8 130.9 187.0 General expenses 162.3 267.6 223.7 Benefit payments 49.6 76.6 129.6 Net Fund Increase 1,898.2 2,494.8 2,300.7 SUSTAINABLE BANKING WITH THE POOR 95 Appendix 9: ADEMl's Impact on Clients Table A9.1: ADEMl's Outreach and Impact on Employment .Year Number of Volume of Outstanding Number of Percent of Jobs Created Cumulative Jobs Cumulative Total Direct Loans Lending in Portfolio in Clients Female Annually Total Jobs Strengthened Jobs Impact on RDS'000s RDS'000s Clients Created Annually Strengthened Employment 1983 2,980 1,034 247 787 n.a. 596 596 2,202 2,202 2,798 1984 2,762 2,099 - 514 n.a. n.a. 418 1,014 3,549 5,751 3,967 1985 1,793 2,872 944 n.a. n.a. 211 1,225 2,344 8,095 2,555 1986 3,065 4,745 1,595 n.a. n.a. 1,341 2,566 5,224 13,319 6,565 1987 5,152 8,548 3,022 n.a. n.a. 2,362 4,928 4,519 17,838 6,881 1988 5,129 13,162 5,897 3,320 n.a. 1,806 6,734 4,867 22,705 6,673 1989 7,619 24,875 11,615 4,944 n.a. 2,022 8,756 8,449 • 31,154 10,471 1990 . 7,877 36,001 17,602 6,824 n.a. 2,304 11,060 ~,029 37,183 8,333 1991 11,273 99,616 51,092 8,360 n.a. 8,418 19,478 16,918 54,101 25,336 1992 14,152 174,119 83,932 10,750 37% 7,796 27,274 21,795 75,896 29,591 1993 13,621 251,637 137,867 13,436 45% 7,697 34,971 17,720 93,616 25,417 1994 13,184 295,270 192,692 14,274 43% 8,264 43,235 14,418 108,034 22,682 1995 14,193 386,362 268,183 13,984 44% 8,008 51,243 18,880 126,914 26,888 1996 13,289 395,083 325,332 15,370 46% 7,173 58,416 17,559 144,473 24,732 Totals: 116,089 1,695,423 325,332 15,370 46% 58,416 58,416 144,473 144,473 202,889 Growth/yr 1984-93: 14.0% 54.7% 71.2% n.a. n.a. 26.7% 40.2% 14.3% 30.8% 38.8¾ Source: ADEMI [1997] SUSTAINABLE BANKING WITH THE POOR 96 Appendix 9: ADEMl's Impact on Clients (continued) Table A9.2: ADEMl's Cost Per Job and Average Loan Amounts Year ADEMI's Pesos Lent Peso Cost Dollars Dollar Cost Number of Average Average Average Loan / Memo: Memo: Memo: Full Cost Per Job Per Job Lent Per Per Job Jobs Per Loan Size Loan Size (GDP/Capita) GDP/Capit CPI Pesos Per (RD$'000s) Job Loan in RDS in '90 RD$ a in Pesos 1990=100 US Dollar 1983 369.55 0.94 . 347 2,589 23% 1,507 13.4 1.00 1984 331 529.12 83.42 529.12 83.42 1.44 760 4,444 41% 1,849 17.l 1.00 1985 731 1,124.07 285.99 382.34 97.28 1.42 1,602 6,816 65% 2,446 23.5 2.94 1986 1,050 722.77 159.91 234.67 51.92 2.14 1,548 6,000 57% 2,710 25.8 3.08 1987 1,730 1,242.26 251.45 250.46 50.70 1.34 1,659 5,549 50% 3,339 29.9 4.96 1988 3,210 1,972.43 481.01 311.11 75.87 1.30 2,566 5,954 54% 4,789 43.l 6.34 1989 6,099 2,375.61 582.48 374.70 91.87 1.37 3,265 5,207 54% 6,048 62.7 6.34 1990 12,400 4,320.29 1,488.02 380.64 131.10 1.06 4,570 4,570 54% 8,411 100.0 11.35 1991 23,964 3,931.80 945.84 310.57 74.71 2.25 8,837 5,742 67% 13,160 153.9 12.66 1992 41,499 5,884.19 1,402.40 467.74 111.48 2.09 12,303 7,647 82% 15,087 160.9 12.58 1993 57,224 9,900.34 2,251.39 775.28 176.30 1.87 18,474 10,906 116% 15,985 169.4 12.77 1994 71,645 13,017.81 3,158.66 996.77 241.86 1.72 22,396 12,212 128% 17,530 183.4 13.06 1995 101,478 14,369.31 3,774.12 1,066.76 280.19 1.89 27,222 13,189 134% 20,302 206.4 13.47 1996 123,255 15,974.57 4,983.63 1,136.17 354.45 1.86 29,730 13,854 132% 22,542 214.6 14.06 Growth/yr 1984-93: 63.8% 32.8% 40.6% 6.6% 12.9% 2.2% 35.7% 9.9% 10.2%1 23.2% 23.5% 24.6¾ Note: ADEJ,.,f.l's full co~ts include administrative and other operating costs (e.g. depreciation and loan loss provisions), interest expenses on non-concessional funds, and the opportunity cost ofconcessional liabilities and ofequity. Source: ADEMI {1997], Appendices 1, 4, 5 and 6 Appendix 9: ADEMl's Impact on Clients (continued) Table A9.3: Summary Data on the Impact of ADEMI on a Random Sample of 39 Borrowers First Loan Last Loan Elapsed Date Date time (mo.s) Average 09-May-94 18-Oct-95 17 Maximum 15-Jan-97 12-Sep-97 55 Minimum 25-Sep-91 13-0ct-92 3 Standard Dev. 28-Jul-01 03-Jul-01 13 ADEMI Loan Client Employees First Loan Last Loan Change Percent First Loan Last Loan Change Percent Average 11,475 23,525 12,050 105% 2.0 2.1 0 4% Maximum 43,011 61,790 45,257 6.0 9.0 3 Minimum 1,609 3,233 (18,000) - - (2) Standard Dev. 11,345 17,458 12,488 54% 1.9 2.3 1 11~i:, Client Assets Client Liabilities First Loan Last Loan Change Percent First Loan Last Loan Change Percent Average 116,710 211,932 95,222 82% 4,616 8,479 3,863 84% Maximum 482,000 1,164,800 934,800 35,000 70,000 70,000 Minimum 13,700 18,300 (4,300) - - (12,000) Standard Dev. 104,986 232,355 165,533 121% 7,763 16,961 15,285 118% Notes: The original sample consisted of JOO clients, however, since ADFJ.,fJ has only computerised these data since 1992, only 39 of the JOO clients had taken out two or more loans and had a full set of data recorded in the computers. Moreover, given the size ofthe data set, more rigorous statistical methods could not be applied to the data, e.g. to compare impact by gender or by location or type ofactivity. Source: Authors' calculations, based on data provided by ADFJ.,fJ's Statistical Department SUSTAINABLE BANKING WITII 1HE POOR 98 Appendix 10: Selected Indicators for ADEMI and for Members of the MicroFinance Network1 (as of December 1996) No. Country GDP/ Institution Number of Female Average Avg. Loan / Portfolio Outstanding Number of Value of capita Borrowers Borrowers Loan Size GDP per at Risk Portfolio Deposit Deposits (USSm) (%) (USSm) capita (%) (USSm) Accounts (USSm) 1 Argentina $8,030 Emprender 2,700 42% $1,250 16% 9.3% $2.9 n.a n.a 2 Bangladesh $240 BRAC 1,600,000 92% $91 38% 2.0% $78.0 1,800,000 $30.0 3 Bolivia $800 BancoSol 71,800 70% $661 83% 2.6% $47.5 49,500 $11.0 4 Bolivia $800 PRODEM 27,500 65% $362 45% 0.9% $8.3 n.a n.a 5 Chile $4,160 Banco del Desarrollo 16,800 45% $1,250 30% 3.1% $18.9 9,400 $3.0 6 Colombia $1,910 Cooperativa Emprender 32,400 56% $500 26% 5.4% $18.2 n.a. n.a. 7 Colombia $1,910 Finansol 16,000 48% $931 49% 35.8% $11.0 150 $9.1 8 Ecuador $1,390 FED 12,300 59% $501 36% 9.8% $3.4 n.a $0.6 9 Egypt $790 ABA 13,000 12% $806 102% 1.5% $9.9 n.a n.a. 10 Ghana $390 Citi Savings and Loans 500 70% $1,200 308% 32.0% $1.0 10,500 $1.3 11 India $340 SEWA Bank 5,500 100% $250 74% 4.0% $1.5 55,000 $3.0 12 Indonesia $980 BRl Unit Desa 2,560,000 50% $507 52% 3.7% $1,700.0 15,760,000 $2,800.0 13 Kenya $280 K-REP 14,000 57% $900 321% 20.0% • $5.0 14,000 $3.0 14 Peru $2,310 ACP 26,800 59% $527 23% 5.7% $9.7 n.a n.a. 15 Philippines $1,050 TSPI 7,000 76% $492 47% 5.3% $3.5 7,000 so., 16 Senegal $600 ACEP 4,200 30% $1,720 287% 2.0% $6.7 1,300 $0.8 17 South Africa $3,160 Get Ahead 10,000 92% $180 6% 10.0% $1.5 10,000 $0.2 TOTALS MFNNetwork 4,417,800 66% $713 $1,924.1 17,716,850 $2,862.7 Dominican Rep. $1,460 ADEMl - Microfinance 16,010 44% $1,107 76% 4.0% $17.7 n.a n.a. Dominican Rep. $1,460 ADEMl - Total Lending 16,185 44% $1,662 114% 3.0% $26.9 n.a. n.a. Notes: Data on GDP/capita are as ofDecember 1995; ADFMI 's data are as ofJune 1997; ADFMI is not a member of the MicroFinance Network. Sources: ADFMI [1997], The MicroFinance Network [1997], World Bank [1997] The Microfinance Network [1997) describes the Network as " ... a global association of leading microfinance practitioners. The members of the MFN are committed to improving the lives of low-income people through the provision of credit, savings and other financial services. The Network believes that this sector of the population should be served by sustainable and profitable microfinance institutions (MFls)." APPENDIX 11. List of Case Studies SUSTAINABLE BANKING with the POOR List of Case Studies • Benin: FECECAM. Cecile Fruman. June 1997. {Also in French). • Bolivia: Assessing the Performance of Banco So/idaro. Peter Fidler. August 1998. {Also in Spanish). • Burkina Faso: Le Projet de promotion du petit credit rural - PPPCR. Julia Paxton. August 1997. • Colombia: Banco Caja Social. Julia Paxton. March 1999. • Colombia: Cupocredito Credit Union. Gloria Almeyda. March 1998. • Colombia: So/idarios Financial Cooperative (Cali). Gloria Almeyda. April 1999. • Colombia: Women's World Banking. Julia Paxton. December 1998. • Costa Rica: FINCA Village Banking. Julia Paxton, March 1998. • Costa Rica: FINCA Village Banking. Julia Paxton, March 1998. • Egypt: Alexandria Business Association. Tom Dichter. December 1997. • Guatemala CARE Village Banks Project. Julia Paxton. October 1997. {Also in Spanish). • Guatemala: The Case of Union Popular and Union Progresita Amatitlaneca (UPA) Credit Unions. Gloria Almeyda and Brian Branch. April 1999 •• Indonesia: Bank Rakyat Indonesia (BR/) Unit Desa 1970-1996. Stephanie Charitonenko, Richard H. Patten, and Jacob Varon. June 1998. • Kenya: KREP. Stephanie Charitonenko, Cecile Fruman, and Glen Pederson. August 1998. • Mali Self-Managed Village Savings and Loans Banks. Cecile Fruman. May 1998. • Niger: Credit Unions (Caisses Populaires d'Epargne et de Credit). Korotoumou Ouattara, Mayada Baydas and Julia Paxton. April 1998. • Pakistan: Aga Khan Rural Support Program 1982-1994. Tom Dichter, and Stephanie Charitonenko. March 1999. • Philippines - TSP/. Tom Dichter. August 1998. • South Africa: Get Ahead Foundation. Craig Churchill. January 1998. • Thailand: BAAC- The Thai Bank for Agriculture and Agricultural Cooperatives. Tetsutaro Muraki, Leila Webster, and Jacob Varon. April 1998. • Zimbabwe: Zambuko Trust. Peter Fidler and Mohini Malhotra. April 1997. SUSTAINABLE BANKING WITH TIIE POOR 100 APPENDIX 12. List of Discussion/Technical Papers SUSTAINABLE BANKING with the POOR List of Discussion/Technical Papers In English: • Microfinance Practical Guide for World Bank Staff • A Worldwide Inventory of Microfinance Institutions • An Inventory of M icrofinance Institutions in Western and West Central Africa _ • An Inventory of Microfinance Institutions in East, Central and South Africa • An Inventory of Microfinance Institutions in East Asia and the Pacific • An Inventory of Microfinance Institutions in South Asia • An Inventory of Microfinance Institutions in latin America and the Caribbean • Financial Sustainability for Credit Programs: A Travel Survival Guide • Indonesia's Ruraly Financial System: The Role of the State and Private lnsitutions • Outreach and Sustainability of Savings-First vs. Credit-First Financial Institutions: A Comparative Analysis of Eight Microfinance Institutions in Africa • Outreach and Sustainability of-Member-Based Rural Financial Intermediaries in Latin America • Microfinance Handbook: An Institutional and Financial Perspective • Credit Unions in Latin America: Recent Performance and Emerging Challenges In French • lnventaire' mondial des institutions de microfinance • lnventaire mondial des institutions de microfinance en Afrique de l'Ouest et du Centre • Le systeme financier rural indonesien : role de l'Etat et des institutio~s privees • Taux de penetration et viabilite financiere des institutions de micro-finance privilegiant l'epargne ou le credit: une analyse comparee de huit institutions africaines In Spanish • lnventario mundial de institutiones de microfinanzas • lnventario de instituciones microfinancieras en Latinoamerica y el Caribe • La autosuficiencia financiera: una guf a basica para program as de credito en America Latina SUSTAINABLE BANKING WITII THE POOR 101 \{ . ... P) is a collaborative effort of ·, •Sustainable Banking with Jhe Poor (SB: ASTHR Gender and Poverty Tearn,onfAGRPW RuralJinonce of the World Bonk, funded by the World Bank; the Royal Ministry of Foreign Affairs of Norway, the Swiss DevelopmenHorporation, and the Ford Foundation. <, Xhe study aims at improving the ability of donors •governments and pmctitioners to design and impleiue·nt policies-and programs to build sustainable financial institutions that eJfegively reach the poor. The SBP , .task managers ore Lynn Bennett (A,! A,):.ana)acob.Yaron (AG~PW); the technical manage~ is Ca~los ~iJ,Y9~ if~~Jpw), and the administrative assistant ,s Loura G'ornel'(ASTHR). <:,)~\; '~ :g The World Bonk ASTHR/AGRPW, 1818 HStreet, N :w: Wash- ingto~, D.C. 20433 Phone (202) 458-0277 • Fax (202) 522-1662 Internet: CCuevas@W0RLDBANK.0RG or LGomez@W0RLDBANK.0RG