SECTORAL & IT RESOI 1or.r:= r.FNTER FEB 2 3 2000 , WORLD BANK SUSTAINABLE BANKING with the POOR Case Studies in Microfinance GUATEMALA CARE Village Banks . Project --------- . / October 1997 • The World Bank Sustainable Banking with the Poor (SBP) is a collaborative effort of ASTHR Gender and Poverty Team and AGRPW 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 (ASTHR) and Jacob Varon (AGRPW); the technical manager is Carlos Cuevas (AGRPW), and the associate manager is Cecile Fruman (ASTHR). The World Bank ASTHR/AGRPW, 1818 H Street, N.W. Washington, D.C. 20433 Phone (202) 458-0277 Fax (202) 522-1662 Internet: CCuevas@WORLDBANK.ORG or LGomez@WORLDBANK.ORG WEB PAGE: http://www-esd.worldbank.org/html/esd/agr/sbp *Note: In the back cover box, Swiss Development Corporation, should read Swiss Agency for Development and Cooperation (SDC). This report was prepared by Julia Paxton (SBP Consultant). Valuable comments were contributed by Carlos Cuevas, Cecile Fruman (SBP), Mike Goldberg (CGAP) and Tom Dichter (SBP Consultant). A background report is available upon request. TABLE OF CONTENTS COUNTR.Y CONTEXT ........................... .................................................. ..................................................... 1 METHODOLOGY .......................................................................................................................................... 2 PERFORMANCE ............................:............................................................................................................... 2 EMERGING STRATEGIC ISSUES ............................. ..................................................................................5 WHAT IS THE OBJECTNE OF THE VILLAGE BANK PROGRAM? ................... ....................... ........................ .......... 5 SHOULD VILLAGE BANKS "GRADUATE"? ..................... ... ......................... ..................... :................................... 7 WHAT WILL BE THE ROLE OF SAVINGS? .. .. ...................................... ..... ...................... ... .................................... 8 CONCLUSIONS ........ ................................................................................................................................... 12 ANNEX'! - INSTITUTIONAL DATA ........................................................................................................ 14 SUSTAINABLE BANKING WITH THE POOR LIST OF ACRONYMS ABC Asesores de Bancos Comunales {loan officers) FINCA Fundaci6n Integral Campesina MIS Management Information System NGO Non-governmental Organization SDI Subsidy Dependence Index VB Village Banking EXCHANGE RATES US dollar Guatemalan Currency 1991 1 5.0434 1992 1 5.2743 1993 1 5.8152 1994 1 5.6485 1995 1 6.0418 SUSTAINABLE BANKING WITH THE POOR 11 92" 91° 119" ' GUATEMALA s• PAVl:D ROADS UNPAVl:D ROADS f 18" J TRACKS r ! -+-+- RAILROADS ,+. AIRPORTS .t, PORTS DEPARTMENT BOUNDARIES INTERNATIONAL BOUNDARIES : f t::"'-,1 ~l ~ILOMETEltS BELIZE t~ Ir 0 20 40 60 BO I I 10 20 JO 50 ..Zr· MILES J T i N -../ / MEXICO I ( ,. I I • /s._,1: of H,:,,-.d:,r:,;1,• ~ / • .... ..... ,...... I . 1a,;ii.. MUE°MUE'raNA~ ' 50 • QUI ff0 ....alJCM ../ ,I" l 1,- .. ,.:iA.i'l ./· H O N D U R A S ._ , _ ,.._ .... "------- ____ .... ,,__.,. -------- ___. ...... --------- _..,_ ..... ...... ........... ,.,.___ .,, ....... UIMU~) ·-J _..,....,._ ......,.._ ......... ,___,___:_......-.., _,,,,.,,.., 89" ------· ..(~ . CUU, _,_,...,.,, ... . .. .. . .. L SALVADOR 1' PACIFIC OCEAN 92" 91• ,.......... Departments where Women's Village Country Profile Banking program is active Current Planned Economic and Social Context Inflation 1. Chimaltenango 6. Alta Verapaz 2. Sacatepequez 7. Baja Verapaz GDP Per Capita (1995) $1339 1993 11.9% 3. Guatemala Population (1995) 10.62 million 1994 10.9% 4. Jalapa Population Density 97.4 inhab/kni2 1995 8.5% 5. Chiquimula SUSTAINABLE BANKING WITH THE POOR iii CASE STUDIES IN MICROFINANCE GUATEMALA CARE VILLAGE BANKS PROJECT The Women's Village Banking (VB) Program of CARE Guatemala provides particularly interesting insights into some of the issues and challenges facing village banking programs around the world. Like many programs started in the l 980's, the CARE VB program was created as an experiment in the increasingly popular field of microenterprise finance. It modeled the FINCA VB approach and gradually adapted itself to the Guatemalan context. After seven years of operation, the program has reached a crossroad in its evolution as several major decisions related to its future mission and role must be made. While numerous lessons can be learned from· the experience of the VB program of CARE Guatemala, 1 the following focuses on three unresolved strategic issues that will affect the direction of the program in the future. First, will CARE Guatemala break from its coq,orate culture of subsidized humanitarian aid in favor of a more business-like approach to sustainable banking? Secondly, will village banks be encouraged to "graduate" from the program? Thirdly, how will the recent client interest in savings mobilization be incorporated into the program in a safe and sustainable manner? COUNTRY CONTEXT Guatemala's income level ranks as the third poorest in Central America. While its 1994 GNP per capita of $1278 was three times that of Nicaragua and double that of Honduras, the skewed distribution of income suggests that a vast majority of people are living in poverty. it is estimated that 54 percent of the population was living in extreme 2 poverty and 22 percent was living in poverty in 1989 . Guatemala has one of the most skewed income distributions in the world. The 1995 World Development Report states that the poorest 20 percent of the population earned only 2.1 percent of the national income while the richest 10 percent earned 46.6 percent. In addition, country statistics3 indicate some alarming trends: the deterioration of the distribution of income, the rapid acceleration of urban unemployment, the increase in absolute poverty, and the rapid increase in the informal sector. However, real GDP growth has been favorable in the 1990's (4.3% on avg. 1991-1995) and inflation is gradually coming under control, from a high of 33.2 percent in 1991 to a low of 8.5 percent in 1995. 1 For a complete analysis of the program, refer to: The World Bank, "CARE-Guatemala's Women's Village Banking Program: A Case Study" Sustainable Banking with the Poor, background paper, 1997. Additional information is available in Annex 1. 2 SEGEPLAN. 3 Published in Realidad Socioecon6mica de Guatemala. SUSTAINABLE BANKING WITH THE POOR METHODOLOGY As in many microfinance institutions, the method selected to provide financial services involves certain trade-offs. Village banking is often used in rural lending in areas where no other formal financial intermediaries are present. Due to the formation of the village into a single group, institutional transactions costs can be lowered since a loan officer can visit the village on repayment days and serve multiple members at once. Other advantages of village banking are self selection of members, group solidarity, .internal monitoring and enforcement through the use of peer pressure, and the empowerment of members who manage their own financial assets as well as the village bank. However, village banking often is criticized for its rigid, homogeneous fmancial services. In addition, despite the cost benefits derived from group formation, village banking typically has high training costs since many of its rural clientele are illiterate with little experience with financial services. CARE Guatemala has made inroads into providing microfinancial services to some of the most marginalized people in Latin America. The program has provided loans averaging only $170 in 1995 to over 10,000 rural women cumulatively in Guatemala since its inception in 1989. Figure 1 illustrates the rural areas where CARE Guatemala is active. The context in which the VB program operates is particularly challenging given the high transactions costs of working with a relatively isolated rural clientele in a mountainous country with poor roads. In addition, the lack of social services, including education, makes training a priority for the smooth operation of banks. Finally, the existence of numerous ethnic groups with their own languages and cultural norms further complicates the :financial intermediation process. The VB program resembles the village banking approach used by FINCA in which communities of 30 to 70 individuals were organized into banks. However, the CARE program in Guatemala has adapted to its context by altering the typical VB model. Changes have included smaller financial transactions, a lengthening of the term structure from four to six months, and a greater emphasis on training. In addition, the program is currently considering allowing one its regional branches to split from the program to form an independent NGO. PERFORMANCE The performance of the CARE VB program has been positive when looking at trends in time series data. Nearly perfect repayment rates have prevailed since program inception. All indicators of outreach, cost, sustainability and profit have become more 4 favorable during the period 1991 to 1995 . However, the mere improvement of these 4 See Annex I for a complete compilation of institutional data 1991-1995. SUSTAINABLE BANKING WITH THE POOR 2 figures does not necessarily indicate that the program is on a sustainable trajectory. The program is still burdened by high overhead costs and donor dependency that, unless rectified, will prevent it from becoming sustainable. The Subsidy Dependence Index (SDI/ gives a measur~ of how much interest rates would have to increase hypothetically in order to cover program costs without using subsidized lines of credit. The 1995 SDI was 4.77 (a dramatic improvement from the 1991 level of28.17). In other words, interest rates would have to be increased by 477 percent in order to cover program costs if no subsidization existed. The 1995 measure illustrates an emphatic decline in the index since 1991. Figure 1 compares the rapid decline of CARE Guatemala's SDI measure to an older village banking program in Central America, FIN CA Costa Rica. FINCA Costa Rica was founded in 1984 compared to the 1989 initiation of CARE Guatemala. When comparing the 7th year of operation in both programs, FINCA Costa Rica's SDI measure of 3.76 is lower than CARE Guatemala's SDI measure of 4.77. CARE Guatemala's higher dependence on subsidies can be attributed to the fact that it serves a poorer and more uneducated clientele, offers smaller sized loans, has higher client training costs, and depends on grants rather than loans to finance its operations. Figure 1 - Village Banking SDI Measures: 1991-1995 30.00 25.00 20.00 15.00 10.00 5.00 0.00 1991 1992 1993 1994 1995 J-+-- CAR~ Guatemala SDI -a- FINCA Costa Rica SDI j One of the main causes for the. decline in the SDI calculation for CARE Guatemala is the gradual increase in real interest rates charged by the program. In 1991, the real interest rate charged to the VBs was -11.4 percent compared to 1'9 .8 percent in 1995. In nominal terms, the interest rates swung from nearly half the commercial lending 5 Yaron, J., "Assessing Development Finance Institutions, A Public Interest Analysis" Policy Research Working Paper No. 174, World Bank, 1992. SDI calculated using the equation: (A(m- c)+[E*m)-p)+k) I (LP*i). SUSTAINABLE BANKING WITH THE POOR 3 rate in 1991 to 33 percent above commercial rates in 1995. This bold interest rate policy has allowed the VB program to become much more sustainable, allowing for project growth while maintaining excellent repayment rates. Despite its favorable decline in recent years, the SDI measurement for CARE Guatemala- is alarmingly high. After six full years of operation, an SDI of 4. 77 indicates that the program has a long road ahead towards reaching sustainability. After six years of operation, the Grameen Bank had an SDI measure of 1.3 in 1989. 6 There are several • reasons for CARE's high dependence on subsidy. First and foremost, the VB program was born out of CARE Guatemala, an institution that has always relied on external grants. Therefore, rather than using commercial loans, the VB program has easy access to pure grants. Thus, the subsidy based on free or concessional borrowed funds is significant. In addition, the program costs have greatly exceeded its revenues, resulting in negative profits. Finally, the average annual outstanding loan balance is relatively small in comparison to overall costs. The ratio of operational costs to the average annual outstanding loan portfolio exceeds unity with a ratio of 2.8 in 1991 and 1.2 in 1995. The VB program is still reliant on pure grants rather than borrowing at interest rates near market rates. Each year, the program receives enough external funding to entirely cover operational expenses. Figure 2 demonstrates how interest income has become an increasingly important proportion of operating expenses. With positive real interest rates and low delinquency, the program has been able to increase its revolving loan fund, thus lessening its dependence on external funds. However, even in 1995, interest income equaled only 30 percent of operating expenses. Figure 2 - CARE Guatemala VB Program: Evolution of Operating Expenses and Interest Income $800,000 $700,000 $600,000 $500,000 $400,000 $300,000 $200,000 $100,000 $0 1991 1992 1993 1994 1995 ~ operational expenses - a - interest income 6 Yaron, J., "Assessing Development Finance Institutions, A Public Interest Analysis" Policy Research Working Paper No. 174, World Bank, 1992. SUSTAINABLE BANKING WITH THE POOR 4 Despite the technical difficulties associated with measuring program impact, anecdotal evidence suggests that the village banks are a useful tool for poverty alleviation. In a sample of 30 clients, all believed that the credit had a positive impact on their lives, both financially and socially. Another interesting phenomenon is the gradual augmentation of the number of clients as the VB ages. A VB in its· first· loan cycle has an· average of30.3 members whereas the oldest banks, now in their twelfth and thirteenth cycles, have an average of 38 members. This can be interpreted as a positive sign since more and more clients are interested in participating in the VB as they see its perseverance and benefits to the members. On average, 6 members are added and 6 members drop out of the village bank for each 6 month loan cycle. Those members who drop out of the group cite various reasons for their withdrawal including no desire for new loans, lack of time, and expulsion from the group due to arrears. Interestingly, even when some women do not want credit, they often stay in the VB as savers. Across the board in the survey, the women reported that they liked having a place to deposit their savings which allowed them to manage their household resources more effectively. EMERGING STRATEGIC ISSUES As the VB program has matured, new strategic questions have surfaced. The way in which each of these dilemmas is resolved will determine the future development of the program. WHAT IS THE OBJECTIVE OF THE VILLAGE BANK PROGRAM? The historical nucleus of CARE Guatemala has been social and financial humanitarian aid to the poor. Projects related to health, education, training, and the environment were formed as services to Guatemala's poor. The VB program represents a rupture in .this type of philosophy. For the first time, a program is bringing in revenue from the very poor it serves. Like many non-profit organizations that enter into microfinance~ dual objectives pull the CARE Guat~mala VB program in two directions. Often the NGO microfinance programs worldwide were created as a means to improve the lives of the poor regardless of cost, repayment, or sustainability. 'As programs have proliferated and donors have become more emphatic on financial performance, a . secondary goal of sustainability has been introduced. CARE Guatemala is caught in this nexus of outreach versus sustainability. The dual objectives of humanitarian aid to the poorest on the one hand and sustainable banking services on the other hand are central internal culture issues. ,The VB program, despite its performance goals, is embodied by the larger institutional philosophy based on subsidized humanitarian aid. In some respects, the inclusion of the VB program in the larger CARE Guatemala portfolio makes it more difficult for the employees of the VB program to incorporate cost coverage and sustainability into their objectives. As a SUSTAINABLE BANKING WITH THE POOR 5 result, administrative costs are quite high as the VB program has the luxury of intensive training for its clients, up-to-date MIS, and comfortable office space and vehicles. While most of its costs are explicitly disaggregated from the rest of CARE Guatemala, the VB program benefits from shared service_s Within CARE Guatemala which are not directly paid for such as centralized management and accounting. Given the long-standi_ n g culture of donor dependence and the availability of external funds, it is difficult for the VB management to unilaterally reject donor assistance as it expands its operations. Therefore, the objective of self-sustained growth is quickly replaced by c!- push for rapid program expansion, targeting the country's poorest clients. While outreach and sustainability are not necessarily mutually exclusive goals, a rapid expansion of clients at very low income levels can result in decreased sustainability and a decline in institutional performance indicat_ ors. After any loan cycle, VB members can decide that it is not in the~r best interest to continue operations. By examining the total number of inactive banks (closed in any year) within a given region, one sees that a higher percentage of banks close in, the 7 poorest departments. A Rural Poverty Index ranks Jalapa as the fourth poorest department out of22 and Chiquimula the eighth poorest. Meanwhile the departments in the central zone (including Chimaltenango, Escuintla, Guatemala, and Sacatepequez) are the wealthiest departments in Guatemala. Interestingly, the percentage of inactive banks is the highest in the poorest departments. 29 percent of the VBs are inactive in J alapa, 23 percent are inactive in Chiquimula, and only 9 percent are inactive in the central zone. There could be several explanations for this phenomenon. Firstly, in Jalapa and Chiquimula, agriculture is the predominant economic activity. •In client surveys in these regions, almost all women felt that the terms and conditions of the loans were inappropriately suited for their activities. They preferred longer terms and larger loan sizes, timed to the agricultural season. Also, a higher velocity of money and transactions occurs in more urban areas. More financial opportunities, both formal and infomial, provide clients with alternative sources of funding if difficulties arise. In addition, greater economic diversification in urban settings leads to several sources of income. If one activity fails, others may be used to cover the repayment. Despite the higher closure rate of banks in the poorer departments, the VB has plans to move to the poorest part of the country in the West. The management is currently writing proposals to obtain new grants in order to accomplish this. In some areas, such as Chiquimula and Jalapa, it is the desire of managers to spin the VBs off to local NGOs. As the program re~ches more isolated clients, modifications in the terms and conditions may be in order if the VBs are to be successful. In addition, it will be more difficult to keep transactions costs low unless regional offices are opened. It is clear that CARE Guatemala's mandate is to reach marginalized people. However, the degree to which they can expand their outreach while achieving sustainability is still unclear. 7 "Memoria de Labores 1995" Fondo de Inversion Social, Guatemala, 1995. SUSTAINABLE BANKING WITH THE POOR 6 the answer to this dilemma will depend in part on the program's ability to achieve sustainability in the more central, commercialized areas where demand for financial services is strong and economic conditions are more favorable. One of the most innovative ideas that the VB program supports is the spinning off of the Jalapa regional office into an independent local organization. The goal is to create a sustainable institution capable of serving the VBs of Jalapa through increased productivity and diversification. Services outlined in the proposal include technical assistance, profitability studies, and a fine-tuning of the terms and conditions associated ' I with the provision of financial services. In addition, the new venture hopes to link to I I VBs supported by FIS and eventually move into the department of Chiquimula. The ' J I spin-off may come in the form of a new NGO or become part of an existing NGO. Due to the low overhead costs of the regional office of Jalapa, sustainability is attainable. In a 8 recent budget analysis of the Jalapa region, total interest income from the VBs equaled Q600,000 while operating expenses only totaled Q360,000. The surplus can be distributed to strengthen the revolving fund and technical assistance. This development will reveal the prospects of sustainability for one region that will not be burdened by the I large overhead expenses of the central office in Guatemala City. In many respects, the break away of the Jalapa office can help overcome one of the internal culture issues faced by CARE Guatemala. As previously mentioned, the VB program of CARE was created out of an institution that is not geared towards collecting revenue from its patrons. An I . ' independent Jalapa office may be able to create a more business-like environment, amenable to the goal of sustainability. Decision.s regarding institutional evolution and program objectives are made by·a relatively small circle of managers. One of the main concerns of the loan officers (asesores de bancos comunales-ABCs) in an anonymous survey of all ABCs was that management made decisions without consulting the field workers. In fact, many of the ABCs believed that the top man~gement of the program was not in tou~h with the reality of the clients and that input from lower staff levels was not encouraged. A greater harmony of teamwork, knowledge sharing, and common goals is needed from an institutional perspective if the VB program will be able to meet the challenges of the future. SHOULD VILLAGE BANKS "GRADUATE"? One of the deviations from the original goals of village banking has been related -. I to the concept of graduation. Initially, it was hoped that actual banks would develop over I three years with the technical support of CARE. During this time, members would establish relationships with local 'formal banks, would become able to self-manage the VBs, and would develop a substantial amount of internal savings. By the end of the trial 8 "Sostenibilidad de! Prograrna Bancos Cornunales CARE Jalapa" 1995. SUSTAINABLE BANKING WITH THE POOR 7 period with technical assistance, it was hoped that the banks would be completely autonomous and could use their internal savings for on-lending or cooperate with the local banks which had been acting as depository institutions for their savings in order to obtain formal group loans. • ' Nevertheless, in its first years of operation, it has become clear that the VBslare not moving towards graduation as anticipated. Even the most sophisticated urban VBs are reluctant to give up their ties with CARE. In addition, local commercial institutions have cooperated with the women's banks only under constant coaxj.ng from-the ABCs. Without CARE's intervention, it is likely that the relationship between the VBs and the local commercial banks would disintegrate in many instances. Without the participation of a commercial bank network, an individual VB suffers the consequences of market fragmentation and does not benefit from the obvious advantage of economies of scale related to financial intermediation. Not only are the bank members and commercial banks reticent about the concept of graduation, but CARE itself has motives for maintaining its VBs. The VBs that would be eligible for graduation are precisely those that have performed exceptionally well for several years. By encouraging graduation, CARE loses some of its best clients with proven repayment records. Not only do these banks represent the best clients, but they also require much less time to serve. ABCs only visit these ba,nks once per month. The interest income for these banks exceeds the costs of serving them and thus represent the most cost-effective, sustainable banks in the program. At the International Village Banking meeting in 1994, the concept of-graduation was discussed by managers and proponents of village banking from all over the world. The failure to have banks actually graduate from their programs was a phenomenon witnessed by many programs, not only CARE Guatemala. At this meeting, it was decided that the word "graduation" in reference to village banking should be abandoned. Instead, there was an emphasis on establishing ties to as many formal financial institutions as possible. As a result, graduation is no longer viewed as a goal of the VB program at CARE Guatemala Since sustainability achieved through graduation is no longer an objective of the program, other ways to become sustainable must be explored. Numerous potential permutations exist. One alternative is that the banks continue their affiliation with the NGO CARE. Another is that CARE would be transformed into a formal financial intermediary. Yet another possibility is that the VBs use internal funds to become autonomous entities. WHAT WILL BE THE ROLE OF SAVINGS? One 9fthe stand-out features of the VBprogram is its emphasis on s·avings although it does not collect savings itself. Members are required to save 5 percent of the value of the loan in order to qualify for their first loan. Deposits are collected at weekly SUSTAINABLE BANKING WITH THE POOR 8 or bi-weekly meetings and deposited into local commercial banks. The philosophy behind the mandatory savings was that potential clients would signal their seriousness about participating in the program and demonstrate an ability to manage their finances. In 1995, the minimum loan size was $83 for a 6 month term, nearly double the minimum loan size of the previous year. Thus, in order to qualify for the first loan, an individual would have to save $4.15. Each VB can decide how to use its group savings. In some cases, the funds are deposited into a group account at a local commercial bank. In other cases, the money is on-lent to group members at interest rates determined by the group. On-lending of savings is more profitable than placing the funds in a commercial bank and therefore has become quite popular. Thus, in addition to an "external loan account" with CARE, savings and fines comprise an "internal account9" for additional lending administered and collected by the group itself. The mandatory savings was intended to be used as a screening and enforcement mechanism as well as a source of long term capitalization of the microenterprise, but as the project has grown, it has become an unexpectedly important part of the financial intermediation process. A strong demand for a reliable place to save has led group savings to grow at an unanticipated pace. Not only are savings being used as a compulsory mechanism to qualify for new and larger loans, but they are being used voluntarily. In some cases, the clients are saving when they do not want loans and in other cases, they are saving more than is required for the amount of credit that they desire. The use of savings for on-lending is a recent phenomenon, appearing in a significant amount after the 1992 mid-term evaluation. In fact, no savings were used for loans in 1993. Since then, nearly all of the savings mobilized in the groups are used for on-lending. Figure 3 shows the dramatic increase in group savings in general and also group savings used for on-lending from 1993 to 1995. During those three years, the volume of savings increased 10 fold from $35,000 to $350,000, compared to a 3.5 times increase in the outstanding loan portfolio. The amounts shown in Figure 4 represent an annual volume of savings which is an annual stock figure including the savings of two 6 month cycles 1°. Therefore, the average amount of deposits at any time is approximately half of the amount reported below. 9 Holt, S. "The Village Bank Methodology: Perfonnance and Prospects" The New World of Microenter:prise Finance: Building Healthv Financial Institutions for the Poor, ed. Otero, M., and Rhyne, E., Kumarian Press, 1992. 10 The accounting software used by CARE Guatemala totals savings over the entire year for its annual stock figure. However, this figure exaggerates the volume of savings since savings may be carried over from one six month loan cycle to the next if the client has repetitive loans, and hence is double counted in the annual volume of savings. SUSTAINABLE BANKING WITH THE POOR 9 Figure 3 - Growth in Group Savings $400,000 $350,000 $300,000 $250,000 $200,000 $150,000 $100,000 $50,000 $0 1993 1994 1995 j- ♦ - Volume of savings --savings on-lent J The tremendous growth of savings is not due entirely to expansion of the program. In fact, the number of depositors grew from 3211 at the end of 1993 to 40?0 at the end of 1995. However, the depositors are saving more on average. The average deposit size in 1993 was $15 and grew to $53 in 1995. There are several reasons linked to the increase in savings. First, a greater volume of deposits leads to greater loan amounts. CARE Guatemala has steadily increased loan amounts as the clients request larger loans and successfully repay smaller ones. Secondly, after several years of : experience with the VB program, the women have witnessed several loan cycles an~ feel more comfortable about the prospect of program longevity. A history of failed credit programs in the past can lead to skepticism, especially related to deposit mobilization. Also, a trust has developed within the groups themselves. After establishing the gro:up norms and regulations, the women are more confident to on-lend their own funds to :the group and reap the rewards of favorable interest rates on their internal account. Finally, as on-lending has become more prevalent, savings provides a relatively high return , endeavor with nominal interest rates around 30 percent compared to only 5 percent in local financial institutions. From 1993 to 1995, the outstanding loan balance grew at a staggering annual rate of 78 percent, primarily due to an enormous increase in lending in 1994. However, an even more startling finding is that during this same period, savings mobilization skyrocketed with an average annual increase of215 percent. Another impressive trend is the increase in on-lending of internal savings. Rather than depositing their funds in 1ocal banks, more experienced VBs are on-lending their funds to group members. In 1993, none of the deposits were used for on-lending. By 1995, 86 percent of the deposits were used for on-lending. SUSTAINABLE BANKING WITH THE POOR This phenomenon is important for several reasons. First, it shatters the myth that poor people do not have the resources or the interest in savings. A mid-term evaluation of the VB program in 1992 states that: "it is not realistic to expect that village bank members will be willing to continue lending through the use of their own savings after the project's intervention ... (They) are thus unable and unwilling to risk losing their savings, regardless of the possible greater returns offered through on-lending. " 11 Clearly the recent growth in savings indicates that the clients of the VB program are willing to mobilize savings and on-lend them. This finding is in Jine with the observations made by USAID about rural small farmers: "Not only do small farmers save, in many cases they do not need or want credit ... What small farmers most often need and want is a safe repository for their savings that is convenient, allows quick access to funds, and protects 12 the value of their deposits from being eroded by inflation. " Another reason why this increase in lending from the internal account is important is that new issues have to be considered by progrmn administrators. One such issue is risk. As some VBs empty their commercial bank savings accounts and opt for on- lending, the threat of covariant income shocks arises. Especially in areas that are heavily dependent on agricultural activities, the possibility of widespread default linked to poor weather or natural disaster is real and could easily lead to the downfall of an individual bank. One of the benefits of financial intermediation on a larger scale is to spread risk throughout geographic areas so that a bank is not irreparably damaged by widespread default in one locale. However, this advantage is no longer present when each VB operates using a high percentage of internal funds. Not only are the individual VBs operating in a fragmented market, but CARE itself is not recognized as a financial institution, but rather has the status of an NGO. Therefore, the clients of the VB program are not protected by the type of insurance linked to the formal financial sector. Risk, controls, legal status, and insurance must be more carefully considered as the volume of savings continues to increase. Savings have been shown to be quite delicate in other settings. Once a system collapses and people lose money, it is difficult to build the trust needed to establish deposit mobilization once again. Furthermore, once one group defaults, a domino effect can occur leading to widespread default across banks as confidence decreases. Therefore, a tight management and monitoring ofVBs, including audits, must be in place to prevent the collapse of groups. 11 T. Tiffany, "Mid-Term Evaluation: Village Banks Project CARE Guatemala" CARE Regional Technical Advisor, September 14, 1992, pp.6. 12 USAID, "Mobilizing Savings and Rural Finance: The AID Experience" AID Science and Technology in Development Series, 1991. (pp. 4) SUSTAINABLE BANKING WITH THE POOR 11 Another consideration of the VB management is how to better tailor their financial services to meet the demands of their clients. In each of the thirty client ' interviews, the respondents supported the use of deposit mobilization. In fact, it wa~ often their favorite part of the VB program. Obviously, the creation of mandatory ' savings in order to receive loans revealed an unexpected, untapped resource. The dients' interest in obtaining access to savings tools that pay positive returns indicates a sho~age 1 of alternatives for savings. As the VB program continues to tap into this demand, it is important that they better understand the types of savings services that are most appropriate for their clientele. The safekeeping of funds and the lowering of transactions costs for savers are two topics that merit attention. ' Several advantages exist to creating rural financial systems that rely equally on credit and savings mobilization. Savings mobilization can foster healthy financial systems since the use of internal funds creates a certain cautious atmosphere regarding lending. Institutions that collect savings are more compelled to increase profitability through sound banking practices, reduction of operating costs, and appropriate loan . screening 13 . In addition, clients are less likely to default on loans when they know the funds come from their neighbors' savings rather than an international donor. The ~iting of savings and credit to form a strong, sustainable financial system, whether it be through the VBs or through a linkage with established financial institutions, is one of the challenges facing the VB program. T~e one unequivocal message from the clients ~fthe VB program is that given an opportunity, even riddled with risk, the clients demonstrate a clear demand for savings facilities. CONCLUSIONS As the CARE Women's VB program enters into its next phase of development, it is clear that certain philosophical issues related to program objectives and strategies must be elucidated. While the VB program has had success in reaching a marginalized : population well outside the formal financial frontier, high overhead costs and donor dependency create an impasse for attaining sustainability. One of the obstacles to self- sufficiency has been the presence of a corporate culture of subsidized aid, exemplified by the various other non-financial branches of CARE Guatemala. A clearer mission statement regarding outreach versus sustainability is crucial. In addition, since the concept of graduation is no longer a program goal, a vision of how a village bank should mature is in order. Finally, a great interest in savings mobilization has been demonstrated in recent years. While this phenomenon has taken program managers off-guard, it ; illustrates an impressive demand for savings instruments. However, the current system I of savings mobilization is a fragile one since the deposits typically are collected and on-lent in the same small village. Any covariant shock could lead to the collapse of the VB. The 13 USAID, "Mobilizing Savings and Rural Finance: The AID Experience" AID Science and Technology in Development Series, 1991 . SUSTAINABLE BANKlNG WITH THE POOR 12 CARE VB program must decide how to provide deposit services, either themselves or through other financial intermediaries, that fit the market demand while ensuring a degree of security and allowing for market integration. Less than one decade after its creation, the CARE Women's Village Banking Program has pulled away from many of its initial goals and procedures. As each of these issues is addressed, the program is undergoing a transformation that will determine its ultimate potential as a sustainable form of poverty alleviation. SUSTAINABLE BANKING WITH THE POOR 13 ANNEX 1 - INSTITUTIONAL DATA CARE GUATEMALA Summary tables of outreach and performance indicators 1991 1992 1993 1994 1995 Deposits outstanding in US$ 51,279 41,550 · 35,192 105,084 350,491 Average size of deposit in US$ (as% of GDP per capita) 12 (1.2%) 8 (0.7%) 7 (0.6%) .15 (1.2%) 53 (4%) Average deposit term na na na na na Nominal interest rate on deposits** variable variable variable variable 5% Number ofloans outstanding at year's end 2,415 2,779 2,825 3,451 3,494 Loan portfolio outstanding in US$ - 115,929 154,556 221,766 526,998 625,198 Number ofloans disbursed 4,102 5,347 4,756 6,703 6,472 Total amount disbursed in the period in US$ 185,403 278,825 329,305 911,067 1,103,990 Average loan portfolio for the period in US$ na 135,242 188,161 374,382 576,098 Growth of outstanding portfolio (in real terms) na 21% 28% 114% 9% Average outstanding loan size in US$ 48 56 78 153 179 Average outstanding loan size as percentage of GDP/capita 4.8% 5.3% 7.1% 12% 13.4% Average loan size disbursed in US$ (as percentage of GDP 45 52 69 136 171 per capita) Average loan size disbursed as percentage of GDP per capita 4.5% 4.9% 6.3% 10.6% 12.7% Loan distribution (maximum/ minimumn in US$) 20/ 397 28/ 379 43/516 44/708 83/993 Average loan term (in months) 6 6 6 6 6 Nominal annual interest rate on loans 18% 24% 24% 30% 30% Real interest rate -11.4% 12.7% 10.8% 17.2% 30% Annualized effective interest rate 33% 41.3% 41.3% 50% 50% Annualized real effective interest rate -0.15% 28.4% 26.2% 35.2% 38.2% Volume of deposits/ volume of loans outstanding 44% 27% 16% 20% 56% * clients are all borrowers and depositors at the same time. ** rates vary depending on which commercial institution is used for deposits. In 1995, clients received approximately 5% nominal interest on commercial deposits, but are increasingly on-lending finds to the village bank for a higher return. SUSTAINABLE BANKING WITH THE POOR 14 Gross financial spread/average loan portfolio na 32% Operating costs ratio (percentage of ALP) na 143% 125% Loan loss provision ratio (percentage of ALP) 0 0 4% Operational self-sufficiency 6% 6% 7% 23% 29% Financial self-sufficiency 6% 6% 7% 23% 29% Subsidy Dependence Index 28.17 16.50 15.26 5.18 4.77 Adjusted return on assets (ROA) -160% -180% -160% -60% -50% Number of depositers per staff Outstanding volume of deposits per staff (in US$) 783 642 2,297 10,704 Number of outstanding loans per staff 99 86 144 194 Number of outstanding loans per loan officer 302 214 157 230 350 Outstanding volume ofloans per staff (in US$) 5,040 5,520 6,720 21,960 34,730 Outstanding volume ofloans per loan officer (in US$) 14,490 11,890 12,320 35,130 62,520 * CARE reschedules loans rather than classifying them as arrears. 2 percent of outstanding loans had been renegotiated in March 1996. Portfolio in arrears is not calculated. DEFINITIONS: • Percentage of arrears: amount of delinquent principal payments as a percent of outstanding principal portfolio • Gross financial spread: financial income (interest and fees on loans and on investments) - financing costs (interest and fees on debt and deposits). • Operating costs: personnel expenses, other administrative expenses, including depreciation of fixed assets. Operating costs exclude financing costs and loan loss provisions. • Operating income: interest and fee income from loans and from other finance-related services (same as financial income above). • Operational seif-sufficiency = Operating income / sum of operating costs and loan loss provisions. • Financial self-sufficiency: operating income / (operating costs + loan loss provisions + financing costs + inflation adjustment+ subsidized cost of funds adjustment+ in-kind donation adjustment) • SDI: See text. • Adjusted operating profit = operating income - ( operating costs + loan loss provision + financing costs + inflation adjustment+ subsidized cost of funds adjustment+ in-kind donation adjustment). • Adjusted ROA = Adjusted operating profit / average total assets for the period. SUSTAINABLE BANKING WITH THE POOR 15 Sustainable Banking with the Poor (S&P}is a colloborati,ve effort of ASTHR Gender and Poverty Team and AGRPW Rural Finance-at the World Bank, funded by the World Bank, the Royal Ministry of Foreign Affairs of Norway, the Swiss Development Corporation, and the Ford foundation. - :' Tbe study aims at improving the ability of doq~rs 'governments and practitioners to design and implement policies and progra!lis to build sustainable financial institutions that effe~ively reach th~ poor: .T!,e SBP task managers are Lynn Bennett (A~THR,land Jacob. Yaron lAGRPW); the technical manager is Carlos Cuevas ,(AGRPW), and the administrative assistant is Lauro Gomei (ASJHR). • ' .~·,; ' :}~:{ ;, • • :' ··.. ~}•;. _..,,.,,, The World Bank •ASTHR/AGRPW, 1818 HStreet, N.W. Washingto~, D.(. 20433 Phone (202) 458-0277 • fax (202) 522-1662 Internet: .... CCuevas@WORLDBANK.ORG.or .. LGomez@WORLDBANK.ORG ,