86696_01-52.qxd 8/31/05 9:23 AM Page 1 34124 OccasionalPaper NO. 11 AUGUST 2005 MANAGING RISKS AND DESIGNING PRODUCTS FOR AGRICULTURAL MICROFINANCE: FEATURES OF AN EMERGING MODEL Introduction Globally, 1.2 billion people are extremely poor--surviving on less than $1 a day-- and three-quarters live in rural areas.1 Poverty is predominantly a rural phenomenon. The authors of Occasional Extremely poor people spend more than half of their income to obtain (or produce) Paper 11 are Robert Peck staple foods, which account for more than two-thirds of their caloric intake. Most of Christen and Douglas Pearce, these people suffer from nutritional deficiencies, and many go hungry at certain UK Department for International Development times of the year. In recent years, development agencies and national governments have renewed Joao Pedro Azevedo, Amitabh their commitment to reducing poverty, hunger, and other human deprivations, as Brar, and Myka Reinsch helped conduct CGAP's evidenced by the Millennium Development Goals (MDGs). Among other objec- research on agricultural micro- tives, the MDGs aim to halve the proportion of people living on less than $1 a day finance, which was financially by 2015 (from the starting level of 1990). That means cutting the share of extremely supported by the International Fund for Agricultural poor people in low- and middle-income countries from 28 percent to 14 percent. Development (IFAD). George The MDGs also call for halving the proportion of people suffering from hunger Ayee, Frank Rubio, and Fion by 2015. de Vletter, consultants for CGAP, conducted field visits to Rural poverty and hunger fell sharply between 1975 and 1990, but the rate of several of the institutions used poverty reduction has since slowed. Net aid (that is, official development assistance) as examples in this paper. In to developing countries fell from 0.35 percent of OECD countries' gross national addition, CGAP has produced five case studies on agricul- income in 1982­83, to 0.24 percent in 2002­03.2 The real value of net aid disbursed tural microfinance that comple- to agriculture in the late 1990s was only 35 percent of its level n the late 1980s, ment this paper. The authors according to IFAD.3 And although the proportion of the economically active popu- are grateful for valuable com- ments from Richard Meyer, lation engaged in agriculture has been falling in developing regions, it still exceeds J.D. von Pischke, and Mark 50 percent in Africa and Asia (table 1). Wenner, and to Richard Agricultural finance has been one of the most prominent elements of the rural Rosenberg and Brigit Helms of CGAP for their reviews and development strategies used by development agencies and national governments. suggestions. Over the past 40 years, billions of dollars have been provided to support agricultural production and the green revolution.4 But this financing has long been characterized CGAP, the Consultative Group by poor loan repayment rates and unsustainable subsidies.5 Accordingly, agricultural to Assist the Poor, is a credit from some donors and multilateral development banks has dropped dramati- consortium of 31 development cally in recent decades and is now often considered too risky. agencies that support microfinance. More information For example, agriculture accounted for 31 percent of World Bank lending in is available on the CGAP web 1979­81, but by 2000­2001 had fallen to less than 10 percent.6 This drop was site: www.cgap.org. Building financial services for the poor 86696_01-52.qxd 8/31/05 9:26 AM Page 2 Microfinance institutions have generally man- Table 1 Agriculture's Large Share of Economic aged default risk very well, while traditional agri- Activity in Some Developing Regions (percentage of economically active population) culture lenders have developed products that respond well to the cash-flow cycles and market- Region 1961 1980 2001 ing relationships of farming communities. But it is Africa 79 69 57 important to remember that for many small farm- Asia 76 67 56 ers the main source of credit is not a bank or even Eastern Europe 50 28 15 a microfinance institution, but agribusiness actors Latin America and 48 34 19 such as input suppliers (for example, sellers Caribbean of seed or fertilizer), traders, and processors. Source: Buchenau, "Innovative Products and Adaptations for Rural Finance," 2003, Moreover, self-finance continues to play a vital role in agricultural production. partly due to disappointment with large agricul- tural finance projects, and partly to the fact that Risk in Agricultural Lending World Bank rural finance increasingly occurred Agriculture is widely considered more risky than in other areas: through microfinance projects or industry or trade. Thus, it is not surprising as part of community development, infrastruc- that agricultural lending projects have had poor ture, or rural development projects. Lending by repayment performance. Weather, pests, diseases, other multilateral development banks and bilat- and other calamities affect the yield of crops-- eral aid agencies has mirrored this trend. At the substantially in extreme cases. For example, in Inter-American Development Bank (IDB), total 2003 the United Nations Food and Agriculture lending to agricultural credit projects under the Organization (FAO) reported that the third suc- category "global agricultural credit" fell from US cessive year of widespread crop failures in Malawi $1.6 billion between 1986­1990 to no lending at (due to excessive rains, floods, hailstorms, and in all in the period 1991­95. Sector loans to assist some areas, dry spells) had afflicted 176,000 fam- borrowing countries to reform and strengthen ilies in four provinces with food deficits and financial markets became more significant for chronic hunger severe enough to warrant human- IDB,7 and this type of targeted investment rose itarian assistance to prevent starvation.9 Such risks from $410 million in 1986­90 to $2.9 billion are higher for farmers engaged in monoculture of in 1991­95.8 crops that are particularly sensitive to the correct The renewed international emphasis on poverty use of high-quality inputs or the timing of har- reduction has put rural populations, particularly vesting. Risk in agriculture can also be traced to agricultural households, back in the spotlight of farmers seeking to increase their incomes through development efforts. Agricultural development higher-risk, higher-return cropping strategies. programs often include credits for agricultural Markets and prices are additional risks associ- production, which has renewed the debate about ated with agriculture. Many agricultural markets how to provide finance in rural areas. Traditional are imperfect, lacking information and communi- providers of agricultural finance insist that it is cations infrastructure. The prices that crops will time to recognize their role in specialized lending sell for are unknown at the time of planting, and to meet the crop-based, cash-flow cycles of small vary with levels of production (locally and glob- farmers--now that microfinance institutions have ally) and demand at the time of sale. Prices are successfully expanded into rural areas with their also affected by access to markets. As state-owned one-size-fits-all techniques. 2 86696_01-52.qxd 8/31/05 9:26 AM Page 3 marketing organizations are phased out, small that alter returns on specific activities. farmers face much higher price risks in many With the entry of new players, growing compe- countries. And inelastic demand for many agricul- tition in international markets can fundamentally tural products causes small increases in production change the competitiveness of a local industry, as to result in large price swings. with Vietnam's recent entry into the coffee indus- Complicating the scenario is that decision mak- try at the expense of higher-cost producers in ing in agriculture is not an exact science; it Latin America. The result has been millions of depends on many variables that change from year dollars of bad debt in commercial banks that spe- to year and are beyond the farmers' control. cialize in lending to small coffee producers Farmers have no real way of knowing how many throughout Central America.12 others are planting a specific crop or how average The precision of crop schedules generates yields will fare in any given year. Often, a good specific risk for agricultural finance. Loan dis- price one year motivates a lot of farmers to move bursements need to be tailored to irregular cash into the same crop the next year. This shift flows, yet the timing of final crop income may increases production in the face of constant vary, based on when farmers choose to sell. (They demand, driving down the price and making the may delay selling until market conditions are crop much less attractive the following year. favorable.) These characteristics of agricultural This happened in Uganda recently, when a production require lenders to be quite efficient bumper maize harvest in late 2001 and early 2002 and physically close to their farmer clients. Thus, caused maize prices (and farmer incomes) to fall, for banks and other financial institutions, agricul- significantly affecting loan repayment in four tural lending involves a risk of causing default branches of the Centenary Rural Development due to their own inefficiency. The production of Bank.10 Bumper crops can sometimes cause prob- most improved cash crops is relatively complex, lems even for well-run microfinance institutions. involving careful timing of numerous steps-- At Kafo Jiginew (a federation of credit unions in from preparing land through planting, fertilizing, Mali), the portfolio at risk (over 90 days) jumped and harvesting. Mistakes or delays at any step from 3 percent in 1998 to 12 percent in 1999 can substantially reduce returns--or eliminate due to a slump in cotton prices. (Cotton loans them altogether. accounted for a large share of its portfolio.) Market and price risks can also be exacerbated Agricultural Microfinance by international market conditions and public pol- Drawing on a few significant, successful experi- icy decisions, which can lead to political risk. For ences in various developing countries, this paper example, the creation or removal of tariff barriers offers a model, termed agricultural microfinance, in countries where goods are ultimately sold can for providing financial services to poor, rural dramatically change local prices. In the 1990s, the farming households. This model combines the Ghanaian government introduced a limited most relevant and promising features of tradi- exemption from import duties on white maize in tional microfinance, traditional agricultural response to a crop forecast--which later proved finance, and other approaches--including leasing, incorrect--that predicted a major food shortage. area-based insurance, use of technology and exist- As a result, market prices for maize were depressed ing infrastructure, and contracts with processors, in Ghana for two years.11 Similarly, national gov- traders, and agribusinesses--into a hybrid defined ernments can change farming subsidies in ways by 10 main features. 3 86696_01-52.qxd 8/31/05 9:23 AM Page 4 The analysis here has found that successful crop production techniques and markets for agricultural microfinance lenders rely on various farm goods. combinations of these features to mitigate the Feature 3: Savings mechanisms are provided. risks associated with lending to farming house- When rural financial institutions have offered holds, although in no experience were all 10 fea- deposit accounts to farming households, tures present. In fact, this paper does not suggest which helps them to save funds for lean times that to be successful in agricultural microfinance, before harvests, the number of such accounts all 10 should to be present, just that a substantial has quickly exceeded the number of loans. number of them seem to contribute to a well- Feature 4: Portfolio risk is highly diversified. performing portfolio, in diverse combinations, in Microfinance institutions that have success- a variety of circumstances. In general, the first few fully expanded into agricultural lending have features are found in most successful experiences, tended to lend to a wide variety of farming while those that come later on the list have proven households, including clients engaged in important in addressing particular risks or situa- more than one crop or livestock activity. In tions found in lending to specific types of agricul- doing so, they have ensured that their loan tural activities. Most of the features address issues portfolios and the portfolios of their clients specific to financing agriculture, some respond to are better protected against agricultural and the general challenges of operating in rural areas, natural risks beyond their control. and some reflect good practice in delivering small Feature 5: Loan terms and conditions are unsecured loans. adjusted to accommodate cyclical cash flows Feature 1: Repayments are not linked to loan and bulky investments. Cash flows are highly use. Lenders assess borrower repayment cyclical in farming communities. Successful capacity by looking at all of a household's agricultural microlenders have modified loan income sources, not just the income (e.g., terms and conditions to track these cash-flow crop sales) produced by the investment of the cycles more closely without abandoning the loan proceeds. Borrowers understand that essential principle that repayment is expected, they are obliged to repay whether or not their regardless of the success or failure an any indi- particular use of the loan is successful. By 13 vidual productive activity--even that for treating farming households as complex which the loan was used. financial units, with a number of income- Feature 6: Contractual arrangements reduce generating activities and financial strategies price risk, enhance production quality, and for coping with their numerous obligations, help guarantee repayment. When the final agricultural microfinance programs have been quality or quantity of a particular crop is a able to dramatically increase repayment rates. core concern--for example, for agricultural Feature 2: Character-based lending tech- traders and processors--contractual arrange- niques are combined with technical criteria ments that combine technical assistance and in selecting borrowers, setting loan terms, provision of specified inputs on credit have and enforcing repayment. To decrease credit worked to the advantage of both the farmer risk, successful agricultural microlenders have and the market intermediary. developed lending models that combine Feature 7: Financial service delivery piggy- reliance on character-based mechanisms-- backs on existing institutional infrastruc- such as group guarantees or close follow- ture or is extended using technology. up on late payments--with knowledge of 4 86696_01-52.qxd 8/31/05 9:23 AM Page 5 Attaching delivery of financial services to lines the key elements, provides examples, and infrastructure already in place in rural areas, describes the many challenges that remain to be often for nonfinancial purposes, reduces addressed. Concrete examples are provided based transaction costs for lenders and borrowers on the experiences and achievements of leading alike, and creates potential for sustainable organizations pushing the frontiers of finance in rural finance even in remote communities. agricultural communities. Still, success--measured Various technologies show enormous promise in terms of long-term financial sustainability and for lowering the costs of financial services high repayment rates--remains somewhat rare. in rural areas, including automated teller Clearly, success in agricultural microfinance is machines (ATMs), point-of-sale (POS) harder than in general microfinance. devices linked to "smart cards", and loan offi- General performance standards of the microfi- cers using personal digital assistants. nance field were applied, in terms of loan recovery Feature 8: Membership-based organizations levels and financial sustainability, rather than the can facilitate rural access to financial serv- somewhat lower standards of traditional agricul- ices and be viable in remote areas. Lenders tural finance. It should be noted that many of the generally face much lower transaction costs experiences in this paper that met these standards when dealing with an association of farmers as and were judged successful are nevertheless rela- opposed to numerous individual, dispersed tively experimental or less well-tested than those farmers--if the association can administer in the general field of microfinance and other loans effectively. Membership-based organiza- areas of development finance. Strong microfi- tions can also be viable financial service nance institutions have only recently expanded providers themselves. into more difficult rural markets and begun to experiment with providing services to farming Feature 9: Area-based index insurance can protect against the risks of agricultural lend- households. ing. Although government-sponsored agri- A model with all 10 features may not exist in cultural insurance schemes have a poor any financial institution currently serving poor record, area-based index insurance--which farmers, although a few come close. Moreover, provides payouts linked to regional levels of the paper does not suggest that there is broad rainfall, commodity prices, and the like--holds consensus on a potential model for agricultural more promise for protecting lenders against microfinance. Rather, it identifies features that the risks involved in agricultural lending. have worked well in various combinations on the frontier of rural finance in agricultural regions Feature 10: To succeed, agricultural microfi- with many poor families. With luck, this paper will nance must be insulated from political inter- trigger a more comprehensive discussion of what ference. Agricultural microfinance cannot features such a model should include. survive in the long term unless it is protected The purpose of the paper is to provide practi- from political interference. Even the best- tioners, policymakers, and donors with a thor- designed and best-executed programs wither ough overview of agricultural microfinance. in the face of government moratoriums on It is hoped that they can use this information to loan repayment or other such meddling in expand the access of farming-dependent house- well-functioning systems of rural finance. holds to sustainable financial services on a This paper discusses each feature of the pro- massive scale. posed agricultural microfinance model. It out- 5 86696_01-52.qxd 8/31/05 9:23 AM Page 6 Research Methodology path toward financial sustainability, and had the In 2002­03, the Consultative Group to Assist the potential to operate on a large scale or be repli- Poor (CGAP), with funding from the Inter- cated. Some of the institutions that are not dis- national Fund for Agricultural Development cussed in the paper might have had similarly (IFAD), assessed nearly 80 providers of agricul- strong results, but had recently experienced a tural microfinance to identify sustainable particularly bad year (for example, due to price approaches to providing such services. These insti- fluctuations, unfavorable climate conditions, or tutions had been identified as well-functioning political interference) and did not have an ade- agricultural lenders by rural development special- quate risk management strategy or a sufficiently ists and the microfinance literature. This paper is robust model for dealing with the intrinsic risks of informed by this research, as well as innovative agricultural lending. At the same time, some insti- work by other organizations and individuals, tutions that were included may have experienced including the United Nations Food and similar problems since then and may no longer be Agriculture Organization (FAO), Germany's good examples. Gesellschaft für Technische Zusamm-enarbeit The difficulty in finding a large number of (Agency for Technical Cooperation, or GTZ), the examples of successful providers of agricultural US Agency for International Development microfinance shows how susceptible the field is to (USAID), the World Bank, individual microfi- factors beyond its control--and how necessary it nance experts, technical service providers, and is for agricultural lenders to adopt the most financial institutions. important lessons of the burgeoning microfinance The analysis in this paper was conducted with- industry to minimize controllable lending risks. It out bias toward any specific institutional type or also serves as a cautionary tale for microlenders approach, because there is enormous potential for moving into rural areas and lending to households cross-learning between traditional agricultural that depend on agriculture for their livelihoods. finance, agribusiness credit, and microfinance. CGAP has published case studies of representa- Although the paper focuses on lending, it also rec- tive examples online from the list of successful ognizes and explores the importance of deposit, agricultural microfinance providers. This paper insurance, and money transfer services--for both makes extensive use of the research conducted for farming households and financial institutions. these studies. In addition to identifying innova- This paper was produced as a desk review, tions and practices, the research emphasizes the supplemented by correspondence with the institu- importance of developing a supportive enabling tions in the case studies, visits to selected institu- environment for rural finance. tions, and discussions with knowledgeable third parties.14 The data on rural finance prog- Feature 1 rams reported here, particularly repayment Repayments Are Not Linked to Loan rates and financial sustainability levels, come from Use a variety of sources, including reports by agricultural lenders. A fundamental feature of the emerging agricul- Of the nearly 80 agricultural microfinance tural microfinance model is that it delinks loan providers assessed by CGAP, this paper focused on uses from repayment sources and instead treats 30. These institutions were chosen because they the entire farming household as a single economic reportedly had achieved high repayment rates unit, with multiple income sources and multiple over a long period, had reached or were on a financing needs. Even if a loan is supposed to be 6 86696_01-52.qxd 8/31/05 9:23 AM Page 7 used to produce a specific crop, the borrower's Table 2 Non-farm Income and Employment in entire household income is considered when judg- Rural Households (percentage of total) ing repayment capacity. Correspondingly, what- ever the source, agricultural activities must be Region Non-farm share Non-farm share of rural income, of rural full-time financed, and some microcredit most certainly 1998 employment, 2002 ends up supporting traditional cropping and live- Africa 42 11 stock production, directly or indirectly, by freeing East and 45 Southern Africa up funds that would otherwise have to be saved West Africa 36 for that purpose. By delinking loan uses and Asia 32 25 repayments, successful microlenders have far more East Asia 35 forcefully stressed that repayments must be made South Asia 29 regardless of the success or failure of a particular Latin America 40 36 productive activity. This approach has dramatically increased repayment rates, even for loans to farm- Note: Includes landless households. Data are for selected countries ing households. This feature is especially impor- in each region. Income and employment figures were not available for the same year and reflect the latest available data. tant when considering the financing of staple Source: FAO, "The State of Food and Agriculture," 1998; Haggblade et al, "Strategies for Stimulating Poverty-Alleviating Growth in the crops or livestock produced year in and year out, Rural Non-farm Economy in Developing Countries," 2002. regardless of the availability of credit, and that do not require large (relative to annual return) up- day labor, and livestock husbandry, in addition to front investments. producing staple foods and cash crops. Household The development finance community has members may also travel to other parts of the recently begun to better understand how poor country for seasonal employment on farms or households earn, spend, borrow, and accumulate employment in cities, or even go abroad and send money and other assets. For agricultural microfi- back earnings (remittances). Different family nance, the most important finding is that farming members perform these activities and contribute households are savers. In most agricultural com- all or part of their income to the family's savings.16 munities, the fluctuating incomes that accompany Non-farm income and employment are crop cycles require households to save between extremely important for rural (mainly farming) planting seasons in order to eat and have enough households in developing regions. The average money left to pay the costs of replanting in the share of non-farm household income is highest in next season. Farming households also try to diver- Africa (42 percent) and Latin America (40 sify their income sources to tide them over percent), but also significant in Asia (32 percent). between cycles. This variety of income-generating activities Many farming households diversify their provides relatively steady cash flow for many sources of income by engaging in a variety of farm farming households, which is why so many rural and non-farm activities. Non-farm activities microfinance clients can make weekly loan pay- include all rural economic activity outside of ments over the course of a year when they borrow agricultural production15 and often run counter- to invest in agriculture, an activity with a highly cyclically to agricultural activities, with most labor irregular cash flow. and resources tied up in agriculture during the Traditional agricultural lending tends to crop season and available during the off-season. involve a huge variety of production loans that are Household members engage in trading, rudimen- narrowly designed for particular crops and live- tary agricultural processing (such as rice husking), stock activities. For instance, in 1984 Bank Rakyat 7 86696_01-52.qxd 8/31/05 9:23 AM Page 8 Indonesia (which later became the world's most left over from the previous year that it intends to impressive model for good practice microfinance use for planting, but needs a loan for fertilizer by a commercial bank) had 350 subsidized credit later in the production cycle. The lender may also programs for food crops, cattle and poultry pro- discover that the family would prefer to pay off duction, fisheries, tree crops, and the like--with the fertilizer loan prior to the harvest with the an average repayment rate of 57 percent. For each son's wages as a day laborer, in order to clear the program (or loan product), experts had carefully debt (and interest payments) more quickly and worked out the exact nature of the production have the maximum amount of income from the cycle: required inputs, dates inputs were required, harvest saved (hopefully, with the same financial harvesting dates, processes, yields, and likely institution) to see them through the months when marketing channels and sales prices. Loan terms there is no agricultural activity in the area (and and conditions were strictly designed to fit these consequently, no day-labor wages). In this instance, features for each productive activity.17 This a flexible lender might offer a three-month loan approach continues to prevail in most agricultural for the fertilizer purchase, repayable on a weekly finance programs in most countries. basis. It would not look like a traditional agricul- If expected yields fail to materialize, market tural loan, but would clearly have the intended prices are low, or problems develop with market- effect of supporting agricultural production. ing channels, lenders and borrowers usually revisit A central thesis of the microlending methodol- the original plans and calculate how the problems ogy used by IPC (Internationale Projekt Consult), will affect farmers' ability to repay, without refer- a German consulting firm that specializes in bank- ence to their families' other financial flows ing for poor people, is that the household should and income-generating activities (or savings). be treated as one financial unit, and that analysis This incomplete view of poor households and of cash flow and repayment capacity should look their income is largely responsible for the low at this unit, rather than just the income-generating repayment rates in agricultural finance. activity being financed. IPC has applied this approach to its Latin American partners that have Treating the Household as a Unit expanded into rural and agricultural lending. Successful rural lenders recognize that farming Financiera Calpiá in El Salvador, for example, ini- households have multiple sources of income, and tiated operations in 1988 and expanded into rural therefore multiple sources for loan payments.18 areas once its urban centers were fully established. These institutions treat rural clients like the Its agricultural operations are characterized by sophisticated financial managers they are and work treating the farming household as one financial to build a complete financial relationship with unit, basing loan criteria on repayment capacity, them. Moreover, such lenders make clear to their taking a flexible approach to collateral require- clients that repayment is expected regardless of ments, decentralizing decision making by well- whether a crop turns out as expected. By delinking trained loan officers, monitoring clients regularly crop and livestock loans from strict adherence to a to strengthen borrower-lender relationships, and particular production cycle and, instead, treating using a management information system that farming households as financial units, lenders can reports arrears on a daily basis.19 provide flexible loan products that respond to Reflecting the importance of diversified income cycles without creating incentives for default. sources, many microfinance institutions with sta- For example, an agricultural lender might sit ble agricultural lending portfolios find that they down with a family and discover that it has seed have to minimize risk by excluding households 8 86696_01-52.qxd 8/31/05 9:23 AM Page 9 that rely on just one or perhaps two crops and money is fungible, some poor families obviously have no off-farm income. Examples include Caja use loans provided for trading to support agricul- los Andes and PRODEM of Bolivia, and tural activities. But because agricultural activities Financiera Calpiá of El Salvador.20 can be supported under conventional microfi- nance loan terms, microfinance practitioners do Concerns about Loan Use not consider their services agricultural finance. Agricultural finance has traditionally been advo- Moreover, the microcredit industry does not have cated to support crop production, slow rural- good information on how much of its funding urban migration, and improve poor people's lives ends up in these activities, because it generally by increasing food security, providing basic serv- does not consider information on loan use partic- ices, and promoting adoption of new technolo- ularly valuable or reliable. gies. These are vital social priorities, and it is Many Asian clients have long used microcredit appropriate (to an extent) to expect agricultural for livestock and agricultural processing. One of microfinance to serve them. But concerns about the most common uses of microcredit in rural Asia the purposes for which loans are provided have is for agricultural activities, such as purchasing traditionally led to product designs that overem- livestock for fattening (chicks, goats, pigs, cows) phasize the investment activities to be under or daily production (laying hens, and milk cows taken by borrowers--leading to a proliferation of and goats), or supporting rice cropping (especially products with varying terms and conditions, in South Asia).21 These uses are often talked about as with Bank Rakyat Indonesia in the case in group meetings (many microloans are provided mentioned above. under group-lending arrangements, and the Product proliferation can create unnecessary groups meet regularly to discuss loan status and costs for lending institution (costs often covered needs), and are encouraged by program staff. Less by high interest rates or large subsidies), because discussed, and probably less prevalent, are invest- the fungibility of money makes it difficult to pre- ments in agricultural inputs (seed, fertilizer, wages determine how funds will be used or to supervise for day laborers) made with microloans. investments without excessive spending on client monitoring. This is not to say that clients lack Smoothing Household Income clarity about why they borrow and what they Within agricultural communities, microloans are intend to do with loans. Indeed, they know well undoubtedly often used to free up capital for the intended use of loans and other sources of farming activities that would otherwise be needed funds, and often engage in matching behavior. for daily living expenses, especially during lean That is, the clients match loan terms and condi- times. Farming communities usually experience tions with expected revenue streams (from any boom and bust cycles--both before and after har- source), so that the revenue that supports loan vests (in the case of crops) and between seasons payments may have nothing to do with the (due to price fluctuations). After harvests, times intended use of the loan. are good and funds are plentiful. As the year pro- Most microcredit providers do not try to con- gresses, funds become scarcer, especially when the trol the use of their loans. And although micro- next crop cycle begins and necessary investments credit funds a wide variety of other personal and have been made. If farming households have no productive activities in rural areas, rural house- access to finance during the lean times, they must holds also use such loans to finance agricultural hold back a larger share of their capital to meet and livestock activities. For example, given that consumption needs, forward-sell their future 9 86696_01-52.qxd 8/31/05 9:23 AM Page 10 harvest early at a low price in return for cash, or sections of the paper indicate that they have been secure high-cost, short-term trader loans. successfully modified for agricultural microfi- With access to microfinance (savings and remit- nance. Perhaps the key to understanding this tance transfers as well as loans), households can apparent contradiction is to assume that incorpo- invest more confidently in their primary income- rating all these techniques of successful microcre- generating activities because they have more dit should be a starting point for agricultural options for meeting both expected expenditures microfinance, and that modifications should be and unexpected shocks. Microfinance can also free made carefully, respecting the need for the overall borrowers' own capital by performing an income- approach to retain as many of the basic techniques smoothing function, as well as directly fund as feasible. Many of the techniques used by micro- agricultural investments that generate their own finance organizations differ fundamentally from repayment flows (such as milk cows or laying those of traditional agricultural credit schemes hens). The income-smoothing role of microfi- (box 1). nance is particularly important for farming house- Microfinance institutions that have developed holds subject to extreme income variability during successful agricultural loan portfolios use more the course of any given year. flexible collateral requirements for agricultural loans than for their other lending. They use a Feature 2 combination of personal guarantors and pledges Character-Based Lending Techniques on household and enterprise assets (including Are Combined with Technical Criteria titled land and animals), rather than relying on in Selecting Borrowers, Setting Loan land and property titles. Uganda's Centenary Terms, and Enforcing Repayment Rural Development Bank, for example, accepts livestock, personal guarantors, land without titles, If a lender has reliable knowledge of a potential household items, and business equipment as loan client's character, as is the case with a well- collateral. Caja los Andes in Bolivia takes pledged functioning credit bureau, the lender can make a assets, but measures their value to the borrower loan based on that person's history of repaying rather than the recovery value to the bank. In financial obligations and on its assessment of that rural areas, loans for less than US $7,500 can be person's financial situation and plans. But devel- collateralized with farm or household assets, and oping countries almost never have a credit refer- unregistered land titles can be deposited with the ence system with good coverage of poor families. bank as collateral for up to half of the value of Micro-credit techniques were developed as a sub- a loan.22 stitute for microlenders' lack of knowledge about potential clients' characters and willingness to Bringing Specialized Agricultural Knowledge repay debt. To serve small farmers and farmers in into the Credit Process remote or marginal rural areas, group-based Traditional agricultural lenders have long savings and lending techniques may be essential employed specialized staff with training in crop to mitigate risk, reduce operating costs, and and livestock production. Similarly, the few enforce repayment. microfinance programs that have expanded into agricultural activities have found it desirable to Tools and Techniques hire agronomists and veterinarians to support loan Whenever possible, microlenders should rely on decisions and methodologies. Just as urban a number of basic techniques--even if other microenterprise loan officers can quickly tell how 10 86696_01-52.qxd 8/31/05 9:23 AM Page 11 Box 1 Differences between Traditional Agricultural Lending and Microenterprise Credit Traditional agricultural lending Microenterprise credit Borrower selection, credit decisions, product designs Bases credit decisions on projected income from future Bases credit decisions on current repayment capacity crop or livestock sales Typically uses feasibility studies to determine borrowers' Often uses peer group information and past loan perform- capacity to repay ance to determine the creditworthiness of borrowers Funds all or most of a targeted activity based on its merits Uses short-term, incrementally increasing loans to estab- and the borrower's ability to carry it out lish relationships with clients and lower default risk. Thus microloans tend to be far smaller than agricultural loans to households with the same income level Ties repayment to proceeds of the agricultural activity Schedules frequent payments to take advantage of the multiple income sources of a borrower's household Sometimes provides agricultural finance to small groups, Tends to use group mechanisms to gather client informa- which often administer rotating loan funds tion and enforce loan contracts, but retains loan adminis- tration functions* Often ties credit to the adoption of particular technologies, Does not tie credit to other services. Exceptions include inputs, or marketing channels; often requires farmers to programs that require compensating savings balances or join associations or cooperatives provide minimal training on issues of social concern, such as maternal health or childhood nutrition Often sets interest rates to be affordable within (narrowly Sets interest rates to fully cover costs, enabling microfi- defined) projections of returns on agricultural investments nance institu-tions to engage in more operational activi- ties--which lowers risk Relies on trained technical staff (agronomists, husbandry Relies on staff trained in lending methodology, not on specialists) or detailed analytical models (or both) to make client activities loan decisions and monitor investment/ production programs Following through with borrowers Expects loan officers to spend most of their time develop- Expects loan officers to focus on building relationships with ing and enforcing investment plans and ensuring produc- clients, enforcing repayment, and understanding the per- tion formance of farming households' multiple economic activi- ties Expends enormous effort to ensure that loans are used Understands that money is fungible and makes minimal according to predetermined plans attempts to control loan uses Tends to be far more lax in the timing of payments, often Expends great energy enforcing rigid repayment discipline assuming that farmers time their sales to achieve highest possible prices Relies on extensive guidelines for multiple crops and live- Relies on a couple key indicators (such as loan or pay- stock investment programs, expected cash flows, and ment amount) to monitor repayment performance repayment plans Uses more rudimentary loan tracking systems Develops efficient management information systems to facilitate immediate follow-up on late payments * This practice refers primarily to solidarity group lending, rather than individual lending or village banking (which devolve some admin- istration functions to larger groups) well a small shop is managed, specialized staff in the terms and conditions of an agricultural rural areas can ascertain how well a farming activ- microfinance loan to the investment opportunity ity is pursued without generating a complex, thor- presented and the income flows of the farming ough production model for a specific activity. household to minimize risk to the lender. In addi- Specially trained loan officers can optimally adjust tion, models can be developed that systematize 11 86696_01-52.qxd 8/31/05 9:23 AM Page 12 such information to ensure more consistent analy- including farming. Yet most banks--savings, agri- sis and inform loan officer decisions. cultural, and development--actively discourage For example, Uganda's Centenary Rural small deposit accounts, considering them a costly Development Bank trained loan officers in agri- liability. They discourage such deposits by requir- culture and agribusiness to help them understand ing that potential account holders be referred to farming as a business, and thus more effectively the bank by current clients, providing poor serv- monitor farmer clients.23 Such skilled staff can ice at teller windows (meaning that clients must develop sophisticated tools to support the credit wait long periods to conduct transactions in the decision process. Economic Credit Institution banks), requiring minimum balances to open or (EKI in Bosnian), a microfinance institution in maintain accounts to avoid incurring monthly Bosnia and Herzegovina, which holds about half fees, and instituting documentation requirements of its portfolio in agriculture, uses spreadsheets for almost as onerous as those required for micro- key agricultural products compiled by an agrono- loan applications. mist. In addition to using this tool to conduct Many leading microfinance programs have cash-flow analysis of proposed agricultural activi- learned from experience what academics at Ohio ties, EKI uses its experience in various agricultural State University and elsewhere have gleaned from sectors (cattle breeding, agriculture, apiculture) to numerous studies of informal financial markets.27 evaluate potential loans.24 Virtually all rural households, no matter how Successful organizations also build their capac- poor, engage in a number of financial strategies to ity for agricultural microfinance slowly and care- build assets, prepare for life events (such as wed- fully. Before investing in a branch office, they first dings, funerals, and education costs) and emer- test a potential rural market. This step reduces the gencies, and cover daily transactions.28 They save risks involved in expanding rural outreach. Calpiá using a variety of non-financial means, such as (in El Salvador) reduces the risks of opening rural accumulating livestock, jewelry, building materi- branches by first developing portfolios from als, and staple crops. Some of these mechanisms neighboring branches and conducting market have profound cultural roots, especially in the case studies of new regions. Rural branches are set up of livestock. only if their likely portfolios merit the required In times of need, these assets can be sold for investments in infrastructure and human capital.25 cash, though they have certain limitations. They Banco del Estado de Chile spent two years adapt- are often not liquid and can be turned into cash ing its microenterprise lending techniques before only at a significant discount to their market value expanding into farming activities.26 It also adapted (if sold in a hurry). They are not safe--for exam- agricultural finance techniques, for example, by ple, animals can die, get sick, or be stolen. And integrating crop-based analysis into its wider they are not divisible, in case the saver needs only client analysis and adjusting repayment schedules a small part of the value represented by the asset. to take into account seasonal income cycles. Many rural households engage in informal financial relationships among themselves. They Feature 3 may be members of rotating savings and credit Savings Mechanisms Are Provided associations, setting aside small amounts weekly or daily.29 At the end of each collection period, Household savings continue to be the primary one member receives the entire amount con- funding source for most private, smallholder, and tributed by the group and uses it to buy major microenterprise production and trade activities, items or pay for major, planned expenses, such as 12 86696_01-52.qxd 8/31/05 9:23 AM Page 13 school fees or weddings. They also lend to each agricultural development bank into member- other and to family members, and save cash based organizations (multi-service cooperatives). "under the mattress." In fact, poor families have These cooperatives offer tailored agricultural and most of the same financial requirements as better- non-agricultural loan, savings, and insurance off families. No matter how poor they are, they products. They are member-owned and -con- have the same need to manage liquidity, conduct trolled and have an open membership policy transactions, and accumulate assets. To do so, they toward poor farmers, defined as those with have developed multiple informal mechanisms. 0.5­1.0 hectare of land and less than half of aver- Basic deposit facilities would enable farming age national per capita income. The cooperatives households to cover agricultural and household have 73,000 members, a third of whom are expenditures, make the interest payments needed female. They have received technical assistance to service credit obligations, and respond to funded by the International Fund for Agricultural emergencies in a timely fashion. Seen from this Development, Asian Development Bank, and perspective, few such households would not want German Agency for Technical Cooperation.33 access to safe, liquid, savings accounts in formal One of the most successful Nepalese small banking institutions. farmer cooperatives is in Anandavan: by July A few agricultural lenders have successfully 2002, it had 861 members, 86 percent of them taken on the savings challenge. The most notable female.34 In July 2003, its loan portfolio stood at has been Thailand's Bank for Agriculture and 17.8 million rupees (US $240,500), with no past- Agricultural Cooperatives (BAAC),30 which has due loans and 14.6 million rupees ($197,000) in evolved from a specialized agricultural lending savings. In addition, the cooperative has a 2.9- institution into a more diversified rural bank pro- million rupee ($39,000) capital fund, including viding a range of financial services.31 BAAC was paid-up and institutional capital. The cooperative established in 1966 as a government-owned agri- offers 10 savings products to attract different cultural development bank and is unusual among types of members. Similarly, it addresses local rural finance institutions due to its impressive poverty by providing innovative loan products for scale and coverage. landless members (such as rickshaw loans) and In March 2003, BAAC had more than 600 flexible savings products. offices across Thailand, serving over 5 million In southern Brazil, membership in the clients, with outstanding loans of $5.8 billion and Cooperativas de Credito Rural com Interacao savings deposits of $6.2 billion--and providing Solidaria (Cresol) system of small farmer savings more than 90 percent of Thailand's farming and loan cooperatives35 has grown from fewer than households with credit services.32 Although state- 2,000 members in five cooperatives in 1996, to owned (the government remains BAAC's domi- more than 31,000 in 73 cooperatives today. nant shareholder), BAAC is largely self-sufficient, Members are poor, with half living below the and funds 80 percent of its loans through savings poverty line, and 95 percent earning less than half (deposits). BAAC introduced an aggressive sav- of the average annual per-capita gross domestic ings mobilization campaign in 1987, and now product. Before joining these cooperatives, 85 offers a range of deposit products to meet client percent of members had never taken out a loan, needs, including passbook savings, time deposits, and half had never had a bank account.36 and savings for a hadj (pilgrimage to Mecca). Membership in another Brazilian system of farmer In Nepal, Small Farmer Cooperatives, Ltd., or savings and loan cooperatives, SICREDI, has SFCLs, are the result of a long-term reform of an expanded rapidly in recent years, jumping from 13 86696_01-52.qxd 8/31/05 9:23 AM Page 14 210,000 members in 1999 to 577,500 in 2002, but only 6 percent for Bolivia's Caja los Andes, with a total of 129 cooperatives with 767 bran- with a similar level for Uganda's Centenary ches. By the end of 2002, SICREDI had US $518 Bank (although the share is notably higher for million in savings and $315 million in outstanding El Salvador's Calpiá, which follows a similar loans (with a delinquency rate of 8 percent).37 approach to agricultural microfinance).38 These institutions--along with others, such as A number of microfinance institutions that the unit desa system of Bank Rakyat Indonesia, have developed stable agricultural lending portfo- savings and credit cooperatives worldwide, and lios also minimize risk by excluding households select other microfinance institutions--have that rely on just one or perhaps two crops and shown that rural poor people will save if given the have no off-farm income. Caja los Andes and opportunity to do so in a well-organized, efficient PRODEM of Bolivia, Calpiá of El Salvador, and a operation that has well-designed, attractive finan- number of other microfinance institutions that cial instruments. All rural households, regardless have expanded into agricultural lending require of their income level or sources, can use deposit that their clients have diversified sources of facilities to enhance their ability to manage liquid- income. In addition to non-crop income sources, ity and build capital assets. most of Caja los Andes' agricultural clients have two or more growing seasons and access to estab- Feature 4 lished markets for their crops.39 Portfolio Risk Is Highly Diversified This practice is in line with that of successful rural credit unions, which typically cap their agri- Diversification is one of the primary risk mitiga- cultural lending at 10­25 percent of their portfo- tion strategies used by microfinance institutions, lio. The range of activities supported is diverse, so credit unions, and specialized banks located in that if, for example, a disease kills most of the pigs rural areas. To contain their agriculture-related in a region, the crisis does not have a catastrophic risks and operating costs, microfinance institu- effect on the lender's portfolio. The risk of having tions tend to limit agricultural lending to less than an undiversified portfolio is illustrated by Caja one-third of their portfolios. Agriculture accounts Rural San Martin, a rural finance institution in for about 25 percent of the portfolio for Peru (box 2). Confianza (a rural finance institution in Peru), Box 2 Peru: Caja Rural San Martin--Diversifying Its Loan Portfolio Between 1994 and 2000, more than half of Caja Rural San Martin's portfolio involved agriculture, mostly in the form of loans to small and medium-size rice farmers. But in 1998­99 Peru's rice crop was severely damaged by the El Niño phe- nomenon. Heavy losses in crop yields caused a steep rise in prices that attracted many new producers, resulting in over- production and sending rice prices to an all-time low. Then in 2000­01, a plague destroyed the rice crop for many of the bank's clients. At the same time, Alberto Fujimori's regime introduced populist policies promoting debt forgiveness and restricting banks from imposing further loan recovery measures on delinquent farmers. All these events caused a severe decline in the quality of Caja Rural San Martin's loan portfolio. The events of 1998­2001 forced the bank to become more risk averse and diversify its portfolio. After nearly halving new agricultural loans in 2001, the bank later discontinued lending for rice production altogether. Since 2002 it has provided loans only to farmers who have well-established farm enterprises, own irrigated land, and can provide land and chattel guarantees. The bank now has a diversified loan portfolio, with microenterprise, housing, and consumer loans in addition to agriculture loans. Portfolio quality has improved as a result, and Caja Rural San Martin is now less vulnerable to pro- duction and price risks. By November 2002, its outstanding loan portfolio was $16.3 million, with more than 13,000 borrow- ers and a portfolio at risk (with payments more than 30 days overdue) of 8 percent. Source: Rubio, "Caja Rural San Martin," 2002. 14 86696_01-52.qxd 8/31/05 9:23 AM Page 15 Portfolio diversification has both facilitated and balloon repayments at harvest and the flexibility limited the expansion of microfinance institutions to avoid situations where households are forced to into agricultural lending. When institutions have sell produce when markets are flooded and prices sought to expand such lending by, for example, are low. exceeding a set proportion of agricultural loans or In many parts of the world, crop cycles produce channeling government funds into farming, they widely varying cash flows that make regular, sig- have sometimes faced dire consequences (such nificant loan payments difficult at certain times of as severe repayment and liquidity crises).40 the year. This is particularly the case in poor, rural Portfolio diversification policies affect only the areas that depend on agricultural production for proportion of agricultural lending relative to non- cash income. In all of these cases, farming house- agricultural lending within a portfolio, not neces- holds must cope with widely varying cash flows sarily the absolute volume of such lending. For that do not match the rigid repayment schedules example, although the share of Confianza's agri- required by many microfinance institutions. cultural portfolio has fallen as it has diversified Promoting Flexible Payment Options from its original focus on agricultural lending, A few microfinance institutions have added true the volume of its lending to agriculture has flexibility to the loan products offered to farming almost quadrupled.41 households. These institutions have adapted loans to the cash flows of agricultural activities and Feature 5 incorporated an agro-economic component of Loan Terms and Conditions Are loan analysis to do so, while not neglecting the Adjusted to Accommodate multiple other potential sources of income of bor- Cyclical Cash Flows and Bulky rowing households. This flexibility relates only to Investments how loans are structured, not the seriousness given to their repayment. Agricultural activities can produce cash flows that In the early-mid 1990s, Caja los Andes in are cyclical (determined by crop or poultry pro- Bolivia faced an increasingly competitive urban duction schedules) or that have long lead-times market and saw an opportunity in the decline of before providing a return (e.g., tree crops or beef agricultural credit provided by state banks. It rec- cattle). This can influence the income and expen- ognized that its loan analysis techniques and diture patterns for the wider rural community repayment schedules were designed for urban- where agriculture is a significant economic activ- based or trade and service activities, and so were ity, with other enterprise activities (and household not appropriate for agricultural activities because budgets) also affected. they could produce delinquency problems and reduce farmers' demand for loans.42 Cyclical Cash Flows Caja los Andes decided to fill the gap left by Agricultural production often requires staggered state banks by offering loans tailored to the needs cash disbursements to meet production schedules, of small farmers, and took steps to mitigate the while allowing for large lump-sum payments at, or risks associated with such lending. In 1995 it soon after, harvest or the slaughter or sale of live- opened its first rural branch in Punata, near Coch- stock. This is particularly true for farmers who use abamba. Today, most of its rural and agricultural modern inputs, such as improved seed, fertilizer, lending is administered by branches located in and pesticide, as well as hired labor for harvesting. towns and larger villages, and its agricultural lend- In such cases, financing arrangements require ing is restricted to certain regions to contain costs. 15 86696_01-52.qxd 8/31/05 9:23 AM Page 16 Caja Los Andes offers the following repayment By learning from microfinance institutions else- options to better suit the cash flows of its clients' where, introducing comprehensive changes to its agricultural activities, relative to the loan products agricultural lending, and diversifying away from a it offers in cities: heavy reliance on agriculture, Confianza recov- One-time payment of capital and interest ered from a disastrous situation in 1999, where more than half of its portfolio was in arrears. It Periodic payments of equal amounts now has a sustainable portfolio and one of the Periodic interest payments, with payment of highest rates of return of any Peruvian microfi- capital at the end of the loan term nance institution (with a 19 percent adjusted Plans with differing, irregular payments (for return on equity). Moreover, Confianza's agricul- clients with several crops or with livestock tural portfolio, with a portfolio at risk (defined as that must be fattened for market) the total value of loans with payments more than Caja Los Andes also offers loans in up to three 30 days overdue as a percentage of total portfolio) installments to better fit the flow of farmers' of just 3.5 percent in 2003, has a lower delin- incomes and expenses. For example, it offers loan quency rate than does its overall portfolio.44 plans with two or three disbursements and one To adapt products to fit agricultural cycles, final payment of capital and interest. monitor their uptake and performance, and PRODEM, another rural microfinance institu- improve their design over time, financial institu- tion in Bolivia uses cuotas personalisadas ("differ- tions need an adequate management information ential" or "personalized" installments), which system (MIS) and a client feedback system to pro- allow members of solidarity groups to tailor repay- vide information on products, service levels, and ments to their individual cash flows. PRODEM's client needs and opinions. Product adaptations market research indicated that not only farmers, should be introduced only after careful market but workers in nonfarm occupations such as research, which is backed by data from both MISs commerce, would benefit from such flexibility. and client feedback systems. As noted, PRODEM For example, the cash flows of rural grocers conducted market research in Bolivia, supple- were found to be significantly higher in months mented by branch-level monitoring and loan when dominant local crops (soya, rice, cane) officer feedback, to assess client needs before were harvested. Similarly, the program allows cof- introducing flexible repayment options and new fee growers to pay only interest in February and products (such as money transfers, microleasing, May, then repay capital in four monthly install- and savings products).45 ments once the coffee harvest begins in June. Many financial service providers, however, PRODEM also reduces risk by capping final receive insufficient client feedback and are unable loan payments at 60 percent of loan amounts and to adequately monitor the performance of individ- limiting the loan portfolios of its branches to ual products. For example, despite taking many 30 percent in any given economic sector (other- positive steps to increase its agricultural lending, wise PRODEM would have to increase loan-loss Uganda's Centenary Bank used an MIS that was provisioning accordingly).43 unable to effectively segregate the bank's loan In Peru, Confianza introduced flexible loan portfolio by product. Partly because its systems terms, disbursements, and payment schedules dur- did not provide adequate information for analysis ing 2000­01, with borrowers able to receive loans and decision making, the bank's attempts to in up to three disbursements and have repayments expand into agricultural lending in the late 1990s partly or fully amortized over the term of the loan. were slowed and eventually undermined.46 16 86696_01-52.qxd 8/31/05 9:23 AM Page 17 Addressing the Challenge of Liquidity Management by offering deposit products, financial institutions Financial institutions that tailor their loan prod- can offer clients the choice of financing seasonal ucts to agricultural cycles may experience chal- needs with savings, loans, or a mix of the two, as lenges in liquidity management (and higher credit well as with remittances transferred from house- risk) and periods of low asset productivity during hold members working in another part of the off-seasons (in effect a form of "seasonal" opera- country or abroad. tion that ultimately lowers the institution's effi- Long-term investment loans (those with repay- ciency). An agricultural finance institution in ment terms more than a year long) also increase Georgia, for example, tailors its loan products to liquidity risk, and so may require lenders to main- crop and livestock production cycles, and most of tain sufficient long-term liabilities, equity, or its one-year loans involve balloon payments. As a other sources of funding. Rural financial institu- result, the institution has had highly seasonal loan tions may be able to use equity or donor grants to cycles, which results in periods of excess and then finance increased long-term lending. (Many such tight liquidity over the year (see figure 1). This institutions have high levels of equity relative to institution is not unique--many other traditional assets.) But if other funding sources--such as agricultural finance institutions have similar peaks client savings, domestic bond issues, bank loans and troughs in cash flows and liquidity.47 (or certificates of deposit), or overseas borrow- Rural lenders can mitigate liquidity constraints ing--are used to finance long-term lending, more by negotiating liquidity facilities with banking sophisticated asset-liability management capacity institutions at times of the year when loans are in is required to manage the resulting interest rate, high demand. Cooperatives in many parts of the liquidity, and foreign exchange risks.48 world use this approach. Fluctuations in liquidity Long-term loans also create challenges for and operating efficiency can also be tackled by liquidity management (see next section). Such maintaining diverse loan portfolios not dominated loans account for more than half the value of out- by agricultural lending, as discussed above. And standing loans in Thailand's BAAC, and it uses Figure 1 Loans Disbursed by an Agricultural Finance Institution in Georgia (December 1991­2001) 800 700 600 Number 500 of 400 Loans 300 200 100 ­ Dec-99 Feb-00 Apr-00 Jun-00 Aug-00 Oct-00 Dec-00 Feb-01 Apr-01 Jun-01 Aug-01 Oct-01 Dec-01 Agricultural Seasons Source: ACDI/VOCA, 2002. 17 86696_01-52.qxd 8/31/05 9:23 AM Page 18 long-term deposits and government-negotiated In this, traditional agricultural loans seek to match loans from international financial institutions such the cash-flow cycle associated with the particular as the World Bank and Asian Development Bank activity from which repayment will be forthcom- to match these assets. About two-thirds of the ing. This approach makes sense from a cash-flow bank's long-term loans are financed by domestic perspective, but it has led to an association, in the borrowing and long-term deposits. BAAC has minds of lenders and borrowers alike, between the actively promoted time deposits, including a use of a loan and the potential to repay it. If a crop three-year fixed deposit product, to fund its long- fails, the household may feel entitled to default on term lending operations. Access to long-term the loan associated with that crop, regardless of loans and client time deposits has allowed BAAC how well the farming household has done in its to better match the profiles of its medium- and other economic activities (including other crops). long-term loans, and to improve the term match Successful long-term lenders in agriculture are few between its assets and liabilities. and far between in developing countries; a recent Food and Agriculture Organization (FAO) study Bulky Investments found only a handful.49 Many of the investment opportunities available to Loans that match long-term investment cash farming households present challenges that do not flows (as opposed to working capital loans for arise in normal microcredit activities. For exam- shorter-term uses, such as inputs and marketing) ple, the value of a capital asset or other investment are not a feature of classic microcredit operations. is often much larger than a household's annual Microcredit uses a range of techniques to lower income (and the portion of that income that can risks and promote high repayment rates, including be used to repay a loan). Acquisition of a traction frequent repayments, short terms, high interest animal or an irrigation pump can provide immedi- rates, and loans for existing rather than new activ- ate income for its owner, but a loan to buy such ities. But these techniques may not be directly an asset could take more than a year to pay back. transferable for larger loans used to finance long- Tree and bush crops do not even have the advan- term investments, particularly when income tage of immediate income, as they often require streams are delayed and loan analysis requires large up-front investments but entail a substantial understanding the activity being financed. The wait before coming into full production--during risk management strategy of setting loan payments which time the farmer must forgo income that at less than a household's income (which limits could have been generated by the land set aside the corresponding loan size), commonly used by for them. According to conventional agricu- microfinance institutions, may also be inappropri- lture finance, such investments should be fun- ate, because the point of long-term agricultural ded (more or less entirely) by long-term loans. investments is often to expand income-earning Few urban microenterprises face investment capacity. In addition, the risk of climatic, political, opportunities that are similarly large relative to or price events that can negatively affect agricul- their current income flows. (Housing acquisition tural activities is higher over several seasons than it is an exception.) is during one, making longer-term financing more risky for lenders. Long-Term Lending The few examples of successful longer-term Long-term loans often involve a series of disburse- lending uncovered by CGAP have been state- or ments to fund the different stages of crop produc- member-owned institutions, where the priorities tion or livestock husbandry, with a single or small of client-members have overcome the institutions' number of repayments due at the end of the cycle. 18 86696_01-52.qxd 8/31/05 9:23 AM Page 19 reluctance to provide what are understood as Nepal's small farmer cooperatives may not be more demanding, risky loans. BAAC is a prime driven by the state (although they do receive some example of state-mandated term finance to loans from the government-owned Agricultural agriculture. BAAC classifies its loan terms as short Development Bank), but as membership-based (6­18 months), medium (up to 3 years), or long organizations, they are more responsive to clients. (up to 15 years). In 1999 medium- and long- The cooperatives offer 18 credit and deposit prod- term loans accounted for 29 percent of the num- ucts tailored to client activities. These products ber and more than 50 percent of the value of include long-term financing (box 3), which the BAAC's portfolio.50 cooperatives fund using a mix of internal savings BAAC has engaged in longer-term lending to and long-term credit lines from the Agricultural agriculture partly because of its state-influenced Development Bank (channeled in some cases mandate to finance agricultural activities (mean- through the Sana Kisan Bikas Bank). ing that it does not have the freedom to select Options for Financing Long-Term Investments only easier-to-finance, shorter-term activities). In the absence of credit, farmers typically fund Somewhat contrary to expectations, given the some long-term investments using savings (or increased probability of negative price changes or remittances),52 a practice that allows them to diver- weather events over time, the portfolio quality for sify into new activities or adopt new technologies BAAC's medium- and long-term loans is better without assuming the greater risk of credit. than for its short-term loans: 6 percent of the for- Incremental Agricultural Microfinance mer have payments overdue more than one year, Microfinance providers have learned that poor compared with 11 percent of the latter.51 But these people typically break down large, long-term figures should be treated with caution because investments into more affordable, less risky delinquency can fluctuate markedly over the stages. For example, when constructing a house, a course of a year. The size, diversity, and country- family may build the first floor initially and the wide coverage of BAAC's portfolio, together with second floor a few years later, or add rooms over the bank's access to term deposits and longer- time. When applied to agriculture, this practice term funds from international financial institu- implies that small farmers may finance invest- tions, facilitate its ability to manage the risks of its ments incrementally, through a series of small longer-term investment loans. Box 3 Nepal: Small Farmer Cooperatives--Tailoring Long-Term Loan Products to Agricultural Activities SFCL Prithvinagar is a small farmer cooperative located in a tea-growing area of Nepal near the Indian border. Previously, its loan products were not sufficiently large or long-term to allow members to invest in tea production. So, it introduced an eight-year loan that covers three-quarters of the average cost of starting a small tea farm (0.6 hectare) and has a grace period of three years. Interest payments are made every three months between the third and fifth years of the loan term, while principal installments are made every six months between the sixth and eighth years. The cooperative also offers tea farmers marketing services to help ensure loan repayments and higher prices for harvests. Tea leaves are collected from the farmers and marketed collectively, and the sales proceeds are returned to them after deducting loan payments. An SFCL in Bhumistan offers a similar loan for the purchase of buffalo. The loan has a term of three years, with principal installments paid every three months for the first nine months and the fourth payment required two years later, when the three-month schedule begins again. This gap in the repayment schedule allows the buffalo to have calves, during which time the borrower would not earn any money from the animal. Source: Wehnert and Shakya, "Are SFCLs Viable Microfinance Organizations?" 2001; Staschen, "Financial Technology of Small Farmer Cooperatives, Ltd. (SFCLs)", 2001. 19 86696_01-52.qxd 8/31/05 9:23 AM Page 20 loans--for example, buying a few cattle each year loan recovery can be faster and cheaper in the or gradually extending the area planted with tree event of repayment problems because interven- crops--rather than going to scale at the start. tion by the courts is often avoided. A study in In addition, lenders may wish to provide long- Bolivia and Ecuador found that leased equipment term loans only to borrowers with whom they have was typically recovered one or two months faster already developed financial relationships through than the year required to recover loan collateral, series of smaller, shorter-term loans for working though in other Latin American countries there capital. This practice is common among microfi- was less or little difference.53 nance institutions that offer long-term loans in Lenders can also acquire expertise in the mar- urban areas, albeit for home improvements, vehi- kets for whatever equipment they lease, enabling cle purchases, or capital asset acquisition. them to sell repossessed assets at higher prices and lower transaction costs than otherwise. Moreover, Leasing lenders may not even need to sell repossessed An inability to produce an effective collateral assets--they can often lease them to other clients guarantee can be a significant obstacle for farming (box 4). households seeking to finance equipment pur- Although leasing is widespread for agricultural chases. In many countries, land may not consti- equipment in developed countries and is increas- tute an effective guarantee, either due to lack of ingly being used by microfinance institutions for land titling or judicial or political reluctance to non-agricultural equipment, it is not widely used enforce legal contracts (e.g., claiming land in for small-scale agricultural equipment in develop- compensation of non-payment on a loan) that ing countries.54 Tax and depreciation rules may would drive poor farmers away from their means not favor leasing, court systems may make repos- of livelihood. Leasing equipment to farming sessing leased items costly or slow, and secondary households offers a low-risk way to finance long- markets for repossessed equipment may be thin. term agricultural investments, and can offer both In 2002 the Development Finance Company a solution to lack of usable collateral and tax of Uganda (DFCU) Leasing, a subsidiary of advantages, depending on the country's tax code. DFCU, Ltd., decided to move down market from When lenders maintain ownership of leased assets, Box 4 Madagascar: CECAM--Providing Microleases for Agriculture CECAM, a network of more than 150 local banks and credit unions in rural Madagascar, has managed to overcome chal- lenges common to agricultural microleasing. Its microleases finance capital equipment for agriculture, livestock rearing, rural crafts, and domestic production (such as sewing). In 2001 the network had 1,800 leaseholders with an average lease of US $450. CECAM has avoided problems associated with leasing to small farmers by: · using flexible repayment schedules that fit clients' production cycles; · requiring larger down payments on new equipment than is common in leasing arrangements (40 percent, instead of 20 percent); and · leasing and releasing used equipment, rather than trying to sell it in thin secondary markets. In addition, CECAM uses group mechanisms for client analysis and monitoring. As noted elsewhere in this paper, the membership-driven nature of cooperatives and credit unions, such as CECAM and Nepal's small farmer cooperatives, appears to make them willing to take greater risks (or make greater efforts to mitigate risks) to meet the financing needs of their members. Source: Wampfler and Mercoiret, "Microfinance and Producers' Organizations: Roles and Partnerships in the Context of Liberaliza- tion," 2001; World Bank, Agriculture Investment Sourcebook, 2004; FAO, "Term Financing in Agriculture," 2003. 20 86696_01-52.qxd 8/31/05 9:23 AM Page 21 its traditional larger scale operations and also lease portfolio of Uzbek companies was 48 percent to small farmers (assisted by a US $1 million larger than in 2001; for banks the portfolio was matching grant from the UK Department for 30 percent larger, and two new banks had entered International Development). But the company the market.57 encountered portfolio quality problems, with delinquency for its agricultural microleasing oper- Feature 6 ations estimated to be up to three times the 15 Contractual Arrangements Reduce percent for its overall portfolio in 2003.55 Since Price Risk, Enhance Production ultimate loan recovery is made more certain by Quality, and Help Guarantee the ability to repossess equipment and lease it to a Repayment new client, delinquency can be tolerated to some extent. Still, the future of the company's micro- Agricultural lenders have consistently sought to leasing remains in doubt. mitigate the risks inherent in agricultural produc- For lenders the attractiveness of leasing relative tion, many of which cannot be controlled by small to lending, apart from stronger loan recovery, farmers, regardless of their skills. Some of these largely depends on a country's legal, tax, and risks are posed by catastrophic natural disasters accounting systems. Bank regulations may permit (droughts, hurricanes, etc). Some are posed by financial leasing to be conducted only through seasonal weather patterns that vary by year and subsidiaries, and in some developing countries tax change the amount and timing of available water, regimes are disadvantageous to financial leasing the prevalence of pests, and the yields of crops. for clients that do not pay profit or value added Some risks are only relatively controllable, such as taxes (that is, most informal sector clients).56 the quality of seed and fertilizer and the timing of Where steps have been taken to improve legal and certain agricultural activities (planting, harvesting, tax frameworks for leasing, as in several Central and so on). Asian countries in recent years (with support from Agricultural Complexities, Credit, and Contracts the International Finance Corporation), the prac- tice has expanded notably--though it is not clear Because of the complexity of production risks, how much of this is agricultural leasing. many lenders feel that small farmers require far For example, in the second half of 2002, the more support than simply receiving loans, espe- Uzbekistan parliament overhauled leasing legisla- cially if they are engaged in the production of a tion and made taxation on leasing comparable to complex crop. Such lenders offer technical taxation on other forms of financing. Changes assistance and other types of support directly to were made to the civil, tax, and customs codes and farmers, either because they seek to improve farm to the law on leasing. As a result, leasing payments practices as part of an integrated development are no longer subject to the value-added tax and program or to guarantee minimum yields and customs fees, and the value-added tax is no longer quality of commodities for processing or resale. applied to equipment imported for leasing-- Traditional Agricultural Lending resolving two of the biggest obstacles to leasing In many developing countries, farming house- viability. In addition, lessors are allowed to deduct holds receive most of their agricultural credit from their taxable income the interest paid on not from banks or microfinance institutions but loans used to purchase assets for leasing. The leas- from agribusinesses--traders, processors, expo- ing market has grown markedly since these meas- rters, and other product-market actors. These are ures were introduced. By mid-2003, the leasing not traditional lenders, as they are mostly not 21 86696_01-52.qxd 8/31/05 9:23 AM Page 22 financial institutions and are not engaged in lend- agribusinesses may have existing physical infra- ing as a primary activity. Rather, they lend out of structure (such as warehouses), agents, processing necessity (no one else will lend to farmers) or to facilities, information technology systems, farmer generate an additional source of income. networks, and market knowledge. Agribusiness credit may be in cash or in kind Contract Farming (mostly in the form of inputs, such as seed and fer- Because many agroindustries participate in com- tilizer). About three-quarters of trader lending in petitive export markets, some with high entry the Sindh region of Pakistan, for example, is in standards, or in increasingly demanding national kind--primarily seed, fertilizer, and pesticide.58 markets, they require more control over the vol- Credit in the product-market system is closely ume and quality of their product than do linked to transactions, as the length of typical purchasers of agricultural produce sold in local credit arrangements makes evident: ranging from markets. For example, processors, wholesalers, just a few days (for stocks provided by suppliers to and other buyers in a range of agricultural markets traders) to the entire growing season (for input provide inputs on credit (in cash or in kind) to credit to producers). help ensure that farmers generate produce of suf- Leading agribusinesses across Southern Africa ficient quality and quantity, and often tie this are estimated to have provided about US $91 mil- credit to purchase agreements.61 lion in credit to more than 530,000 rural house- This "contract" farming is a formal type of holds between 2001 and 2003.59 Four out of five agribusiness credit. Repayment of the input credit rice mills in India surveyed by the Food and is deducted when the farmer sells the produce. Agriculture Organization (FAO) offer advance Contract farming developed as a private sector payments to farmers to cover input costs; such response to quality and quantity concerns. arrangements cover about half of the total value of Tobacco and seed companies, coffee and sugar the crop. Processors may also channel credit mills, dairies and slaughterhouses, cotton boards, through traders, rather than directly to farmers. and even wholesale buyers for supermarkets have Two-thirds of Indian rice traders surveyed by the developed packages that combine elements of FAO traded on a commission basis, with funding technical assistance, input provision, marketing from millers.60 assistance, price guarantees, and finance as a way Traders, processors, other agribusinesses, and of ensuring the supply of a sufficient quantity and individuals reduce the production and operational quality of a particular product (box 5). By build- risks associated with lending to farmers by linking ing formal contractual relationships with farmers, credit to the provision of technical advice (such as contract farming (including outgrower schemes, on input use or on what crop variety to grow to the most formalized version of contract farming) meet market demands), or timely delivery of reduce the risk for that farmers will side-sell a por- appropriate inputs (seed, fertilizer), or by building tion of the contracted amount to other buyers. relationships with farmers over one or more years. Contract farming--as opposed to trader Many also tie credit to subsequent sales of pro- credit--is the most common form of credit for duce, a practice often called interlocking or inter- small farmers provided by private companies in linked contracts because it provides inputs on parts of East and Southern Africa. The Kenya Tea credit based on the borrower's expected harvest. Development Agency (a private company), for Operating costs for providing credit can be low, example, operates a fertilizer credit scheme involv- because credit is built into crop purchase and ing more than 400,000 small farmers, to whom it input supply transactions with farmers, for which disburses US $15.5 million a year..62 There are 22 86696_01-52.qxd 8/31/05 9:23 AM Page 23 Box 5 Costa Rica: Hortifruti and CSU--Contract Farming by Supermarket Chains Corporación de Supermercados Unidos (CSU), a supermarket chain based in Costa Rica, has a technical assistance and training program to help its suppliers adopt higher quality and safety standards. It relies on a specialized wholesaler, Hortifruti, to procure its fresh fruits and vegetables. Until 1990 most of Hortifruti's suppliers were conventional wholesalers. But as it began to require larger volumes of produce of standardized quality, it developed a network of 200 preferred sup- pliers (farmers and packers). Seventy percent of these preferred suppliers are small farmers--although 80 percent of the volume purchased by Hortifruti is produced by medium-size or large packers (who integrate the grower function). Hortifruti works closely with these growers, providing financing, technical assistance in production and post-harvest han- dling, and packing materials. In return, farmers sign contracts committing to sell an agreed volume of produce to Hortifruti. Each contract specifies a production calendar and required volume and quality of produce, and assigns a bar code for each farmer's produce. Hortifruti's field buyers and agronomists visit suppliers to monitor crop calendars and production practices. In addition, its quality assurance unit enforces quality and safety standards. Hortifruti now buys only about 15 percent of its produce from conventional wholesalers. Moreover, the preferred supplier system has cut costs by about 40 percent due to lower product losses and waste as quality has improved. Source: Alvarado and Charmel, "The Rapid Rise of Supermarkets in Costa Rica," 2002; Berdigué et al, "The Rise of Supermarkets in Cental America," 2003. 200,000 outgrowers in Kenya's sugar industry, viable for basic staple crops in markets with a large and contract farming also underpins the country's number of suppliers or marketing agents, or in tobacco sector. Similarly, in Mozambique, situations where quality control is hard to tobacco and cotton companies provided about $2 enforce. But as discussed below, links to agribusi- million in input credit to an estimated 270,000 nesses might provide microfinance institutions smallholders during the 2003/03 season.63 with vital access to a large, untapped market of potential clients. Challenges of Contractual Arrangements Standardization requirements resulting from Although agribusiness credit is widespread and increasingly demanding buyer practices can, how- seems to have filled much of the "gap" left in a ever, cause a loss of credit from buyers and so number of countries caused by reduced donor marginalize small farmers. For example, while financing of agriculture, it has fundamental limita- farmers consider it beneficial to be part of the net- tions. Agribusiness provides a narrow range of work from which Hortifruti buys produce (see financial products that primarily consist of sea- box 5), small farmers have a harder time meeting sonal credit and short-term advances for certain the volume, quality, and timing requirements. To key crops.64 There is often no explicit interest rate, minimize risk and transaction costs, Hortifruti although rates as high as 5 percent a month are seeks low turnover in its pool of growers. quoted for input loans from rice traders to farm- Similarly, farmers are keen to stay in the pool ers in the Philippines.65 Instead of paying interest because membership reduces their production rates, farmers may be expected to accept discounts and market risks--they receive necessary inputs on on the prices paid for their produce. time, good advice on how to use them, and a Many agribusinesses and individual buyers guaranteed market. Farmers that fail to meet would prefer not to have to offer inputs on credit standards (for example, by overusing pesticide) and lack the skills (possessed by real financial insti- are not immediately delisted, but given training tutions) to price and monitor credit. Buyers often and assistance to achieve required standards. Still, provide input credit simply to secure a sufficient high quality and production requirements result supply of decent-quality product, or to cover the in fairly significant turnover among smaller, less seasonal cash shortages of their clients (such as capitalized growers. traders or suppliers). This option may not be 23 86696_01-52.qxd 8/31/05 9:23 AM Page 24 Agribusiness may also be reluctant to provide branch networks or other delivery mechanisms of credit if there is potential for fraud when employ- financial institutions. ees handle funds, if suitable information and client If a financial institution simply replaces an tracking systems are lacking (making credit provi- agribusiness as a provider of agricultural credit, sion difficult to manage), and if farmers engage in and no link is established to the business's advice, side-selling. For example, after experimenting for market access, and knowledge about inputs, then some time, a leading agribusiness company in the potential for lowering costs and operational Mozambique, Export Marketing, became wary of risks will be lost.67 But several emerging providing input on credit. It no longer engages in approaches offer financial institutions the oppor- contract farming, preferring to buy produce in tunity to build on experiences with agribusiness cash from a network of buying posts linked to credit, and agribusinesses the opportunity to take warehouses. The company said that effective man- advantage of financial service models. Several of agement of contract farming was extremely diffi- these approaches are still evolving, and the learn- cult because of weak contract enforcement and ing process, as they evolve, offers valuable insights transport infrastructure, corruption among its to financial institutions wishing to adopt them. produce buyers when handling money, and lack of Financial Institutions, Contract Loan Officers, or law enforcement options to protect itself against Other Intermediaries side-selling. Cheetah, a Dutch company specializ- These intermediaries select and monitor farmer ing in paprika production, found that operational clients and ensure that they can access services and challenges in Malawi and Zambia (poor roads, inputs from agribusiness buyers and suppliers. This scattered producers, farmers' lack of experience model was used by Banco Wiese in Peru. CES with commercial farming) resulted in much higher Solidaridad, a non-governmental organization operating costs than it had predicted, and its (NGO), acted as an agent (or broker) for Banco 2002­03 pilot contract farming scheme failed in Wiese, selecting and providing technical assistance both countries.66 to groups of two or three farmers located near one another.68 Banco Wiese provided the loans, and Connecting Agribusiness and Agricultural CES Solidaridad received a 2.5 percent commis- Microfinance--Emerging Approaches sion on each loan, with an additional 1.5 percent Financial institutions and specialized microfi- paid upon successful repayment. This arrange- nance providers may be able to capitalize on ment had excellent repayment performance, and the huge potential that agribusiness offers small by 1998, Banco Wiese had an outstanding portfo- farmers in terms of low-cost, large-scale financial lio in such loans in excess of US $3 million.69 services--while compensating for the deficiencies Linked Services Between Financial Institutions described above. Financial institutions have and Agribusinesses the expertise, systems, and technology needed Under this approach, a farmer receives a loan from to offer a range of financial products with a bank or microfinance institution, and the clear prices, tailored information, and effective repayment is deducted from the price that an monitoring systems; agribusinesses know individ- agribusiness pays for the farmer's crop. The ual clients, crops, prices, and markets. In addi- agribusiness may also provide inputs and advice to tion, agribusinesses already have networks to the farmer, or these inputs may be provided by a distribute inputs (including credit) and collect third party--as in the three-way model used by produce (and repayments) from farmers--net- Mahindra Shubhlabh (MSSL), which operates a works that may be much more extensive than the 24 86696_01-52.qxd 8/31/05 9:23 AM Page 25 Box 6 India: Mahindra Shubhlabh--Linking Banks, Agribusinesses, and an Input Provider Mahindra Shubhlabh (MSSL) is part of the Mahindra and the Mahindra empire (the world's third largest tractor maker), and runs commercial agriculture support centers all over India. These centers, and smaller franchises at the village level, serve as one-stop shops where (mostly paddy) farmers can receive loans and technical assistance, rent specialized equipment (harvesters, tillers, and the like), and buy seed and other inputs. Loans range from 15,000 rupees (about US $350) to 100,000 rupees (about $3,000) per season, with an average loan of just over $500. MSSL facilitates farmers' access to credit by acting as an agent for banks, including ICICI Bank (India's second largest), and recommending that the banks provide loans to farmers that it is providing with other services. Agribusiness buyers are also involved, with a three-way agreement whereby MSSL recommends a client to ICICI for credit, and the client (farmer) receives inputs on (ICICI) credit from MSSL after pledging its produce to a buyer. The buyer repays the loan at the end of the season out of the sale proceeds from the farmer's output. MSSL receives 1.5 percent of the loan value for its loan pro- cessing and supervision services, dependent on the loan being repaid. In early 2004, this arrangement was used by 45 MSSL outlets, with 5,600 active clients. Source: Hess, "Innovative Financial Services for India," 2002; correspondence between author (Pearce) and Kairas Vakharia, CEO of Mahindra Shubhlabh, 2003; CGAP data. nationwide network of agriculture service centers 1987 Trisan, an agrochemical wholesaler in Costa in India (box 6). Rica, introduced a credit card scheme (box 7) The three-way arrangement between MSSL, through its finance company to offer less costly ICICI Bank, and agribusiness buyers outlined in credit to customers (farmers and retailers). box 6 reduces operational risks for lenders Agribusinesses Adopting--and Adapting-- through: Microfinance Techniques to Improve Repayment the quality of services and inputs that MSSL Just as traditional agricultural finance providers provides to farmers; have improved repayment rates by taking lessons MSSL's client knowledge in deciding whether from microfinance, evidence suggests that to recommend farmers for loans; and agribusinesses can do the same to tackle their the link to buyers. repayment and operational challenges (box 8). To This kind of arrangement also reduces opera- promote client repayment, some private contract tional risks for farmers associated with markets, farming schemes use techniques similar to those and increases small farmers' limited access to good used by microfinance institutions: group liability, market opportunities. close monitoring, and development of strong Mahindra Shubhlabh has found, however, that trust between the lender and borrower. communicating with banks on behalf of individual A variant on this approach would be for firms clients can be costly. As a result, it is introducing with microfinance expertise to perform back- an alternative model where banks make loans to office processing--that is, operating credit man- its service centers (rather than directly to farmers), agement systems--for agribusinesses. Such firms and it provides farmers with loans for inputs. This could also potentially strengthen the performance model should still provide reduced operational of agribusiness credit portfolios and facilitate their risk for both the lender and the farmers. expansion. But there is little experience with such efforts, and this conjecture remains unproven. Agribusinesses and Finance Companies By establishing finance companies, buyers and Small Farmers Form Groups and Associations to suppliers can provide input credit in a more spe- Improve Credit Access cialized way and on a larger scale. Smart cards or Small or remote farmers may have few market credit cards can be linked to point-of-sale termi- options and depend on a few traders offering nals70 to further facilitate transactions on credit. In unfavorable credit terms--or have no access at all 25 86696_01-52.qxd 8/31/05 9:23 AM Page 26 Box 7 Costa Rica: Financiera Trisan--A Supplier Creates a Finance Company In 1987 Trisan, an agrochemical wholesaler in Costa Rica, formed Financiera Trisan, a finance company designed to pro- vide faster, cheaper access to finance for farmers and retailers. To lower transaction costs and increase credit sales, Financiera Trisan developed a credit card program (first launched in 1992) for retailers of agricultural inputs and individual agricultural producers with predictable cash flows. Two types of cards were developed: Agrimax, for input retailers and farmers with regular income (30-day billing cycles), and Maxicuenta, for farmers with good credit and seasonal cash flows (allowing balloon repayments after harvest). The cards could be used at a range of rural merchants, including input stores, gasoline stations, and auto repair shops. The credit card program allowed Trisan to evolve from providing supplier credit to a wider range of financial services. By 1999 Trisan had issued more than 3,600 cards, and the Agrimax card had US $4.7 million in outstanding loans. But two factors have led Trisan to rethink its credit card business: a government debt pardoning scheme introduced in 1999 severely lowered repayments, and the Superintendency for Banks deemed Trisan's administrative costs and delinquency rates too high and ordered them lowered. Repayment levels plummeted after the introduction of the debt pardoning scheme, and delinquency rates rose to as high as 25 percent. (In 2004 they remained at about 15 percent.) Since 1999 more than 2,200 accounts have been written off. The company has been shifting the Agrimax card to a smart-card system. The smart card is more flexible in terms of inter- est rates, loan terms, and repayment schedules, enabling Trisan to provide different models of credit (unlike the standard Visa model followed earlier), and thus better manage its lending risk. The volume of smart card-based credit rose from 9 percent in 2001 to 14 percent by September 2002, and delinquency rates on these accounts are reported to be less than one-third of those on the traditional card. Source: Email exchange between a CGAP researcher and Charles Spalding, director of Trisan, 2002; see also Wenner, "Making Rural Finance Work," 2001; and Wenner and Quiros, "An Agricultural Credit Card Innovation: The Case of Financiera Trisan," 2000. Box 8 Two Examples from Southern Africa: High Repayment in Agribusiness Input Credit Cottco, Zimbabwe Cottco, a large cotton ginner in Zimbabwe, has adopted microfinance group lending techniques to help reduce side-selling and defaults on input loans to smallholders. Cottco extends credit to farmer groups with joint liability and provides services, including extension advice. Since its inception the scheme has consistently reached more than 50,000 smallholders a sea- son, and achieved repayment rates in excess of 98 percent. CRM Farm, Zambia For several years, CRM Farm, a commercial farm in Zambia, has achieved repayment rates of or close to 100 percent on fertilizer credit extended to about 70 small farmers for maize production. (Maize is a widely traded food crop highly suscep- tible to side-selling.) By contrast, a government fertilizer credit scheme that operated during a comparable period achieved a recovery rate of just 6 percent. CRM Farm borrowers repay their loans at harvest by delivering two bags of maize for each bag of fertilizer received. Small farmers supplying CRM want to continue doing so, as it provides a reliable source of inputs. Thus, they comply with their agreements and do not sell the corresponding amount of maize to other buyers. The option of paying back inputs received on credit in the form of maize also helps poor farmers short of cash. The fact that the CRM Farm supervisor knows the farmers, and has worked with local chiefs, has also contributed to the scheme's success. Source: Gordon and Goodland, "Production Credit for African Small-Holders," 2000; Ruotsi, "Agricultural Marketing Companies as Sources of Smallholder Credit," 2003. to product-market credit. Such farmers can make of distributing inputs, collecting crops, and themselves more attractive to agribusiness and keeping records. Extension services and technical financial institution credit providers by forming assistance designed to enhance small-farmer market-oriented groups and associations. For production to meet buyer requirements (and product-market buyers, dealing with organized potentially increase creditworthiness) can be pro- groups of small farmers, rather than individual vided more efficiently to a group than to individ- small farmers, can lower the cost and complexity uals. A group (or association, or cooperative) can 26 86696_01-52.qxd 8/31/05 10:29 AM Page 27 family and friends. Nearly half of these other Figure 2 Percent of Clients Served by Non- sources are nonfinancial institutions such as retail- financial Institutions in El Salvador ers and agribusiness suppliers, processors, and buyers (figure 2).74 Share of clients served Increasing the supply of agricultural finance thus requires creating institutional capacity. One Non-financial instns. 47% way to do so is by building on existing institu- Microfinance/ tional infrastructure and networks (such as post banks 31% offices, agribusiness agents or collection centers, Moneylenders 22% and state banks) and using technology appropriate to rural areas (such as cellular phones and mobile banking units). All rural lenders need to invest in Source: Rodríguez-Meza, cited in Buchenau and Hidalgo, "Servi- cios financieros privados en al area rural de America Latina: techniques and technologies that deliver financial situación y perspectivas," 2002. services sustainably in areas characterized by also help hold its members to quality and produc- poor transportation and communications infra- tion standards. And formal associations and coop- structure, low client density (dispersion), and low eratives allow for more effective contracts and levels of economic activity (which affect staff pro- contract enforcement, which can improve repay- ductivity and efficiency). ment rates and lower the risk of side-selling.71 These challenges are greatest outside heavily Examples of farmer associations linking with populated rural regions. It should come as no sur- agribusinesses are outlined in feature 8 below. prise that the most successful rural finance programs are in South and Southeast Asia, where Feature 7 population densities are almost 1,000 people per Financial Service Delivery Piggy-backs square kilometer. Bank Rakyat Indonesia, with on Existing Institutional Infrastructure almost 3 million active clients, operates in a coun- or Is Extended Using Technology try where rural population density averages 700 people per square kilometer. In Bangladesh, home One of the greatest constraints facing agricultural to the Grameen Bank and other well-known rural microfinance is the dearth of formal financial finance providers, the figure is even higher: almost institutions serving poor farming households in 900 people per square kilometer. In contrast, the rural areas. In a typical case, in 1998 it was esti- average rural population density in much of sub- mated that in Ghana just 8 percent of small Saharan Africa and Latin America is fewer than 10 clients--such as rural and urban poor people-- people per square kilometer.75 had access to formal credit and savings services.72 Greater client dispersion relative to urban areas, Similarly, formal loans are scarce in most rural together with poor transportation and communi- areas of the developing world, especially for poor cations infrastructure, can make conventional farming households. In Latin America, for exam- branch structures unviable. These conditions ple, rural households' access to formal credit serv- also increase the costs of moving cash and con- ices ranges from 2 percent in Peru to 28 percent ducting loan analysis, and make client monitoring in Mexico (with Costa Rica something of an more difficult. Responses to these challenges anomaly at 40 percent).73 Research in El Salvador fall into three categories: partnering with local indicates that only 35 percent of the rural popula- institutions, developing alternative delivery mech- tion accesses credit from sources other than anisms, and exploiting technology. 27 86696_01-52.qxd 8/31/05 9:23 AM Page 28 Box 9 Georgia: Constanta--Partnering with Rural Banks to Expand Microfinance Services Constanta, a microfinance institution in Georgia, uses temporary service centers--usually rented rooms in branches of rural banks--to lower the costs of expanding into rural areas. Constanta sends a loan officer to each service center a cou- ple days a week to meet with client groups and supervise disbursements and repayments. Constanta pays the banks a fee for each client that uses their cashier function, and transfers funds through the banks to avoid having to transport cash. If the partner bank has a management information system (MIS) that can provide daily client data, Constanta also pays the bank a fee for each transaction that uses this teller function. If the partner bank cannot provide this service, a Constanta teller spends two days a week at the service center disbursing loans and collecting repayments, and transactions are entered directly into Constanta's MIS. Lending began in the first service center in October 2001. By May 2002, the four functioning service centers had 1,700 active clients and US $140,000 in outstanding loans--with minimal startup and operating costs for Constanta. The service centers have opened the door to more sustained expansion, with several locations holding potential for upgrading to per- manent offices. Other financial service providers are following Constanta's lead into these new rural markets. Service centers have low operating costs and can be viable in small towns. For example, a partner bank in the town of Khashuri charges Constanta just $60 a month to rent a service center office, and another $70 a month to use its systems for disbursements and repay- ments. If only direct costs are taken into account, service centers can be profitable with fewer than 300 active clients. Source: Pearce, "Pro-Poor Innovation Challenge Case Study--Constanta (Georgia)," 2002. Partnering with Local Institutions to the United States and other parts of Mexico), When financial institutions establish relationships offers a service where it collects remittances in with local institutions that already have infrastruc- bulk from the branches of two banks (Banorte and ture in rural areas, it is often a win-win situation Banamex) in a nearby town (Tlaxiaco).76 This for all parties. Financial institutions can partner service reduces the cost and time required for with other financial entities, such as rural banks, clients to obtain remittances from a US $5, six- or with non-financial entities, such as clinics, hour round trip by bus (which also requires wait- schools, lottery outlets, post offices, pharmacy ing in a queue for several more hours) to a short chains, or agricultural input suppliers. The holder walk to the microbank office and a fee of just of the local infrastructure can gain additional rev- $1.50.77 The remittance product has increased enue as a result of financial services from the Xuu Nuu Ndavi's liquidity, and as a result it has financial service provider/partner being offered been able to lower the interest rates paid on client through its branches and other outlets, while the deposits. Despite the bank's small size (260 mem- financial service provider avoids the investment bers), it is at least nominally profitable.78 It and and operational costs associated with setting up a other microbanks in the network are also explor- dedicated network (box 9). ing the possibilities of sharing offices with remit- For example, if a bank or money transfer com- tance transfer companies and of providing them pany does not have a rural branch network, agree- with loans to allow them to transmit larger vol- ments with other institutions can provide access to umes of remittances. rural remittances--say, through agreements with Similarly, the World Council of Credit Unions retail stores or links to credit union networks. (WOCCU) has introduced IRNet (International Although not a formal linkage, the network of Remittance Network), which is linked to money microbanks in Oaxaca, Mexico, shows the poten- transfer companies, such as Vigo. IRNet enables tial of informal rural financial institutions linking remittances to be sent from credit unions in the to town-based banks. The microbank, Xuu Nuu United States to credit union networks in Central Ndavi (Money of the Poor People), in San Juan America, Mexico and Jamaica. IRNet's fees-- Mixtepec (a village with a high level of emigration US $10 per transaction under $1,000--are low 28 86696_01-52.qxd 8/31/05 9:23 AM Page 29 Box 10 Kenya: Equity Building Society--Delivering Mobile Banking In 2000 Kenya's Equity Building Society introduced mobile banking in about 20 of the country's most isolated towns and vil- lages. The mobile units visit weekly and offer a range of banking services, including loans, cash transfers, and savings accounts. The units operate as follows: · Existing client data are downloaded onto a laptop computer, and a customized, bulletproof, all-terrain vehicle takes the laptop, cash, staff, and armed security personnel (hired from the government) to the mobile banking location. · The vehicle uses solar power to run the laptop and management information system (MIS), as well as voice and data transmission equipment. · The vehicle transmits data to the main branch, which can serve up to five mobile units simultaneously. These data transmissions use Global Services for Mobile (GSM) communications technology, which is reliable and safe--unlike the telephone lines that may be available in some areas. This technology enables client account information to also be updated directly to mobile units if appropriate. In addition, VHF radio communication is used to link the mobile units to branch offices. Source: Ayee, "Equity Building Society (EBS): Agricultural Lending," 2003; Coetzee, Kabbucho, and Mnjama, "Understanding the Re- birth of Equity Building Society in Kenya," 2002; Craig and Goodwin-Groen, "Donors as Silent Partners in MFI Product Development: MicroSave-Africa and the Equity Building Society in Kenya," 2003; EBS, "Mobile Banking Experience," 2003; PlaNet Rating, "Equity Building Society (EBS)," 2001. relative to the estimated average of 13 percent for program that enabled it to offer a range of finan- remittances to Central and South America, which cial services--including agricultural loans--even would be $26 for the average ($200) transac- in remote rural areas, with full cost recovery. By tion.79 WOCCU estimates that 10 percent of early 2004, these mobile units were serving 29 remittances sent through IRNet go into savings.80 locations and about 12,000 clients (box 10).83 Cooperatives in the Dominican Republic have a In the late 1990s, the World Bank provided more extensive rural branch network than do support for mobile banking by a state bank as part banks, and have taken advantage of this to expand of a large rural finance project in Vietnam. The their membership. San Jose de Matas, for exam- bank used specially equipped vehicles to reach ple, received and distributed US $500,000 in remote and mountainous areas, served more than remittances in one year, and many of those receiv- 300,000 rural clients, and reportedly earned a ing money have joined the cooperative.81 FEDE- profit.84 These and other experiences point to sev- CACES, a federation of credit unions in eral requirements for successful mobile banking, El Salvador, has an extensive rural network for including robust management information sys- poor clients, as well as a wide network of partner tems that provide efficient loan analysis, repay- credit unions abroad that facilitate remittances. In ment information, and arrears control, and rapid 2002 FEDECACES transferred $22 million in data transfer systems that provide sufficient remittances to its clients in El Salvador.82 protection against inaccuracies and fraud. Developing Alternative Delivery Mechanisms Exploiting Technology Flexible alternative delivery mechanisms, such as Technological innovations can significantly mobile banking or renting space from other enti- increase the efficiency and lower the costs ties, can lower the costs of providing financial of financial service providers operating in rural services in remote, sparsely populated areas. areas. Thus technology has the potential to play a Widely used mechanisms include introducing major role in expanding access to rural financial mobile banking and renting space from other services. Among the most practical and increas- entities. In 2000, for example, Kenya's Equity ingly affordable of these technologies are auto- Building Society instituted a mobile banking mated teller machines, smart cards, debit cards, 29 86696_01-52.qxd 8/31/05 9:23 AM Page 30 personal digital assistants (PDAs) and handheld stead, based on hour-long visits. The value of fast, computers, and cellular phones. field-level decisions can be enhanced by incorpo- ATMs, smart cards, and debit cards can provide rating credit scoring into PDA systems. flexible payment options and more convenient Cellular telephones also offer potential for access to client accounts. They can also reduce extending financial services in developing coun- branch infrastructure and employee costs, and tries, including rural areas, as cellular networks are facilitate financial services in areas with poor extended. In some developing countries, the communications and electricity supplies. number of cellular phones already exceeds the In Bolivia, PRODEM has extended its branch number of bank accounts. Cellular phones can be network by installing 20 ATMs. These machines used to check loan balances and repayment sched- have some unusual features: they are equipped ules, and they have the potential to be used for with fingerprint readers for client verification, and stored-value transactions, and to facilitate remit- they provide audio instructions in three lan- tance transfers and payments if linked to point-of- guages, to make financial services more accessible sale devices and other payment points. Cellular to illiterate and semiliterate clients and to those networks could also be used for low-cost deposits who do not speak Spanish. Because the ATMs are and withdrawals if they are linked to local mer- linked to smart cards (which contain information chants and agents. But--and this has been a key on client accounts and previous transactions), they lesson among microfinance institutions--for a only have to update data from the central process- technology to add value, a financial institution ing site twice a day--saving about US $800,000 a must first conduct careful market research and year in internet access charges. Smart cards cost-benefit analysis, then ensure that its informa- cost clients $10 to obtain, and $7 a year in oper- tion systems can provide data in the form and at ating fees. PRODEM's ATMs cost less than the time that the new technology requires. $20,000 each, making their installation economi- cal when compared to the costs of setting-up a Feature 8 branch office.85 Membership-Based Organizations Can In Ecuador, a network of ATMs enables poor and Facilitate Rural Access to Financial rural families to access remittances sent by Services and Be Viable in Remote Areas relatives working in Spain. Banco Solidario, an Ecuadorian bank for poor people, offers a debit card Membership-based organizations have a mixed (La Chauchera) that clients can use to withdraw track record in managing financial services in rural money deposited in Spanish savings banks, including areas, but they can be viable even in remote areas La Caixa, Caja Madrid, and Caja Murcia, as well as because they can make use of voluntary or semi- Banca Sella in Italy. Clients can access remittances at voluntary staff, draw on community knowledge more than 800 ATMs nationwide, or at any of about when making loan assessments, use community 100 cooperatives with whom Banco Solidario has a peer pressure to ensure loan repayments, and rely strategic alliance.86 on low-level institutional systems and infrastruc- PDAs can streamline the work of loan officers ture. Thus, such organizations, formal or infor- and speed decision making--as long as the finan- mal, can expand rural access to loans, savings, and cial institution's loan analysis and client monitor- other financial products. In addition, producer ing systems are sufficiently developed. Chile's (farmer) associations can lower transaction costs Banco del Estado has used PDAs with great suc- for credit providers--both financial institutions cess in generating agricultural loans at the farm- and product-market actors, such as processors and 30 86696_01-52.qxd 8/31/05 9:23 AM Page 31 exporters--because it is easier for these providers The village-based nature of CVECAs results in to deal with a single group rather than numerous lower operating costs. Each village determines its individual, scattered farmers. And for agribusiness own interest rates and loan products, which helps buyers that also provide inputs to farmers, dealing ensure that loan products are suited to local agri- with organized groups of small farmers reduces cultural activities. CVECAs also keep costs low by the cost and complexity of distributing inputs, collaborating with village farmer associations on collecting crops, and keeping records. client appraisals, loan guarantees, and repayment schedules. CVECA members, which helps General Savings and Loan Organizations improve monitoring effectiveness.88 Community-managed village savings and credit Another example where members play a valu- organizations (known as CVECAs) are prevalent able role is the livestock insurance product offered in parts of West Africa and can be viable even in by Nepal's small farmer cooperatives (SFCLs). This difficult, remote areas. For example, the network product uses committees of members--working of CVECAs in the Niono region of Mali has more under the guidance of SFCL managers--to evalu- than 9,000 active borrowers and savers and is ate livestock for the amount to be covered financially sustainable, with very good reported by insurance.89 portfolio quality. CVECAs are organized into Less formal group-based models also have the networks that borrow from an apex bank and potential to operate viably in poor, remote areas onlend to the individual CVECAs. Loan funds are (box 11). Techniques that reduce operating costs also generated from member savings. Fee-based include relying on basic information systems, auditing and training from an independent sup- simple financial products, voluntary or semi- port centre ("CAREC,"87 which took over this voluntary staff, and group knowledge about function from the French NGO CIDR) provides potential borrowers, as well as using the group necessary ongoing support functions. mechanism to enforce repayment. In the absence Box 11 Africa: CARE's Village Savings and Loan Model--Self-Managed Groups for Rural Women CARE's village savings and loan model, designed primarily for women in low-density and poor rural areas, has achieved impressive outreach in several African countries since its initial development in Niger as the Matu Masa Dubara (MMD) program from 1993 onwards. In Niger more than 160,000 rural women belong to 5,500 small, self-managed savings and loan groups, each with about 30 members. Zimbabwe's Kupfuma Ishungu program had 5,000 members in 770 mixed groups (although only about 13 percent of members are men), as of June 2002. The CARE project helps women organize themselves, while a village agent (serving an average of 550 members in the Zimbabwe program) provides basic training on and monitoring of procedures, product designs, the role of group manage- ment committees, and other areas. Group members make weekly contributions, but can also access loans from the group savings fund. After a year or so, most groups distribute the accumulated funds equally to the members--usually when there is a particular need for funds, such as the start of a planting season. Members typically double their savings in a year through the interest income on loans (members set the interest rates, typically at 10 percent a month, with loans typi- cally three weeks in length). Most groups rely on verbal rather than written records, but procedures are simple enough (members each contribute the same amount each week, for example) that this is not a problem. The cost of helping to establish and support these groups is estimated at between US $18­$39 per member, while in Niger the average savings per member is $12.50, and average loan sizes $7. (The value of loans would build up over the cycle, and there could be 10 loans of $7.) The promoters of this program (CARE International) point out that this program functions even in remote and impoverished rural areas, and that a very high percentage of the groups become sustain- able. Moreover, they claim that this cost is much lower that the cost per client for more formal microfinance programs. Sources: Hirschland, "Savings Operations for Very Small or Remote Depositors: Some Strategies," 2003; Allen, "CARE International's Village Savings and Loan Programmes in Africa: Microfinance for the Rural Poor that Works," 2002. 31 86696_01-52.qxd 8/31/05 9:23 AM Page 32 of alternative funding sources, such as bank loans, cotton to ginneries, and reducing the paperwork and in response to the need for safe places for that would otherwise be involved in dealing with clients to store money, these organizations are numerous small farmers. In recognition of these often savings-based. benefits, CANAM pays the associations commis- When these informal models are managed by sions of 5­10 percent (though commissions are members rather than professionals, savings prod- not always paid if market conditions are tight). ucts may have to be less liquid--for instance, V&M Grain Company offers interest-free fixed-term deposits. This lack of liquidity is typi- advances to traders small and large, as well as to cally compensated for by access to loans or groups of producer associations. The company emergency withdrawals from the savings.90 reports an overall repayment rate of 98 percent. Advances to groups of associations are based on Producer Associations half of the crop value at an agreed price, with no Producer associations can enable small farmers to other collateral arrangements, and are provided access credit from agribusinesses and financial for up to 20 days. Part of each advance is used to institutions by reducing the transaction costs of transport the group's produce to a warehouse; the lenders that deal with them, and helping raise the rest is distributed to individual producer associa- quality and volume of their products to meet the tions, who further distribute it to their members. standards required by buyers, as noted earlier in Loans average US $5,000­10,000.91 feature 6. Experiences with producer (farmer) associa- Private processors and exporters in a number of tions have been mixed though, with some prob- countries have recognized the potential offered by lems of lack of member motivation and association producer (farmer) associations. Since the late capacity. Smaller and more marginal farmers may 1990s, outgrower associations have been widely need technical assistance and training in order to promoted by cotton and tobacco processors and establish effective associations. The upfront costs buyers in Mozambique. Lomaco, a cotton ginning may be more than private sector actors are willing company, and Mocotex, which has taken it over, to pay, and so may merit additional donor support have initiated hundreds of producer associations through specialized intermediaries that can (with 20­40 members each) in the districts of provide training, systems support, and other assis- Montepuez, Balama, and Namuno. CANAM is a tance to existing associations and to farmers wish- company that buys seed cotton to process and ing to create producer associations. Examples of market from about 30,000 small, medium-size, support provided by one such intermediary to and large farms. In 2003, it provided farmers with market-oriented farmer associations and coopera- US $500,000 in input credit. Many of its suppli- tives, the Cooperative League of the USA ers are members of farmer associations, with (CLUSA), are outlined in box 12. whom it has formal contracts. Smallholder mem- In addition to increased access for small farmers bers have individual cards that record input cred- to credit, less quantifiable results of this type of its, seed cotton sales, and other transactions. support to producer associations can include CANAM has had a positive experience with longer-term structural changes in farmers' access farmer associations, finding that they help in to finance, markets, and negotiating position, as recovering credit (in 2003 nearly all had 100 per- well as enhanced agricultural skills, market knowl- cent repayment), lowering transaction costs by edge, organizational development, literacy, and enabling it to buy in bulk and delivering seed community lobbying power. 32 86696_01-52.qxd 8/31/05 9:23 AM Page 33 Box 12 Zambia and Mozambique: CLUSA--Support to Farmer Associations and Cooperatives Agriflora, Zambia Agriflora was a private Zambian exporter of flowers, fruits, and vegetables to European and Australian markets. It had annual sales of US $30 million and provides input credit to 7,000 farmers.* Although the company ran its own farms, it also bought produce from nearby farmers small and large. CLUSA and Agriflora worked together, setting up cooperatives and providing them with technical assistance to make small farmers more attractive and competitive and to link them to Agriflora's supply and purchasing arrangements. In addition, Agriflora provided cooperatives with input credit, using the group-guarantee methodology to secure loans. After signing advance purchase contracts, farmers received monthly pay- ments from the company for sale of their produce.** Agriflora had supply contracts with 300 smallholders in eight coopera- tives in 2001. Rural Group Enterprise Development Program, Mozambique CLUSA launched its Rural Group Enterprise Development Program in Mozambique in the mid-1990s. CLUSA began by bringing together existing farmer associations, as well as unaffiliated farmers, to create a network of more than 800 associ- ations in the provinces of Nampula, Niassa, and Cabo Delgado. CLUSA focused on making the associations more attrac- tive to agribusiness buyers and financial institutions by strengthening their capacity to maintain records, coordinate produc- tion, collect produce, and provide information on quality standards required by buyers. CLUSA used decentralized staff to provide associations with onsite training, support, and consultation services. Participatory training techniques were used to teach the skills needed to run market-oriented associations, such as managing budgets and contracts. CLUSA also helped broker credit agreements for associations. By mid-2003, CLUSA was working with nearly 26,000 farmers in 860 associa- tions (which it also calls "rural group enterprises"). CLUSA coordinated with six agribusinesses, and in 2003 two of them-- Export Marketing and V&M Grain Company--provided $136,000 in cash advances to producer associations for ground- nuts, sesame seeds, and pulses. CLUSA also brokered access to finance from GAPI (a partly state-owned non-bank financial institution) for 24 groups of associations. * Since this was written, Agriflora has run into financial difficulties that have caused it to cease operating. This does not affect the valid- ity of including this scheme here, which was small relative to the company's operations. **Pearce, "Buyer and Supplier Credit to Farmers: do Donors Have to Pay?" 2003. Sources: de Vletter, "A Review of Three Successful Rural Finance Cases in Mozambique," 2003; CLUSA, "Quarterly Report, April-June 2003"; CGAP research. The Need for Second-Tier Support Systems loan cooperatives, can address some of these chal- Small, rural, membership-based financial enti- lenges. Audits and benchmarking can promote ties--whether savings and loan organizations or transparency and performance standards. In addi- producer associations--can suffer from weak tion, services can be offered that make it easier for internal controls and monitoring, and may be sus- member organizations to negotiate funds from ceptible to deterioration in portfolio quality, cap- banks and donors, lobby for policy and legal ture by well-educated or influential persons, and reforms, monitor performance, and meet short- even fraud. Some are run by or for the benefit of term cash-flow needs (for example, through a a few members who monopolize access to loans or refinancing facility). provide loans to members as a "right," with loan Building an effective, viable structure, however, amounts simply calculated as multiples of member can be problematic. For example, in Mali, it took savings or shares. More formal membership-based more than 10 years for an institution-building organizations, such as savings and loan coopera- project to achieve technical and financial sustain- tives registered with a country's financial institution ability for the CVECA system.93 In East Africa, supervisors, are less prone to these weaknesses. there is a social and cultural divide between sav- Such organizations, however, have higher cost struc- ings and loan cooperatives and their apex struc- tures and are less suited to marginal rural areas.92 tures, where the apexes are perceived as serving Creating a second-tier institutional support their own interests and insufficiently responsive to structure for small, rural financial organizations, the needs of their members. In Tanzania, the apex such as a network or federation of savings and institution, SCULT (Savings and Credit Union 33 86696_01-52.qxd 8/31/05 9:23 AM Page 34 League of Tanzania), did not provide many serv- each month and balloon payments for the princi- ices for its partner cooperatives; as a result the pal at harvest time. cooperatives began losing interest, and in early 2000 even considered canceling their subscrip- Remaining Challenges tions to the apex.94 The human and financial resources available in Although some cooperative federations have small, rural communities limit management and been ineffective and costly, there have been posi- governance capacity for developing membership- tive experiences--such as Brazil's SICREDI based organizations. These organizations often and Cresol (box 13). SICREDI is a system of sav- require more supervision, assistance, and moni- ings and loan cooperatives for small farming toring than is initially assumed, and in the households. It specializes in agricultural lending, medium- to long-term (if not from the start), fees primarily for the production of rice, wheat, beef, will have to be charged for these services. Where fodder, fish, and vegetables, and for agricultural community resources and capacity are severely equipment. Short-term loans are financed by limited--for example, in remote communities or deposits, and longer-term loans by loans from the where economic activity is at a very low level-- National Development Bank. The size of informal models may be more appropriate. SICREDI loans depends on the potential returns Institutions should not be pushed to grow faster from crop sales, as well as household income and than their capacity allows.95 debt payments, and is limited to half of produc- An important challenge for donors, governments, tion costs. Borrowers make interest payments and others seeking to promote membership-based Box 13 Brazil: SICREDI and CRESOL--Providing Second-Tier Support for Groups of Cooperatives SICREDI follows consistent, agriculture-focused lending practices, and pools and manages liquidity risk at the system level. Uniform, system-wide standards are strictly enforced. To use the SICREDI name and logo, credit unions must meet stringent financial, policy, and product quality standards. The financial details of all members are shared among the sys- tem to ensure peer enforcement of these standards. The high risks associated with narrow dependence on agricultural lending is managed by limiting the percentage of assets in such lending, financing long-term loans with borrowings from the National Development Bank, and buying crop insurance (through PROAGRO, the national crop insurance program). Being part of a system is central to the success of SICREDI cooperatives: they can obtain refinancing, offer a wider range of services than if they were stand-alone entities, benefit from system-level management of liquidity risks, and associate with a brand that requires commitment to high standards. The SICREDI council develops policies and products, and pro- vides training services. A cooperative bank (Bansicredi) enables members to issue credit cards, offer internet banking, issue trade credits (including letters of credit), and provide insurance (life, non-life, and rural). Members can also facilitate forward sales, notably by coffee growers, through the Cedula de Producto Rural instrument. In addition, SICREDI's partici- pation in the PROAGRO crop insurance program, which adds a premium of 3.9 percent to loan rates, enables its mem- bers to provide agricultural insurance. Cresol, another network of small farmer cooperatives in Brazil (which generally serves poorer clients than does SICREDI), also provides the benefits of a second-tier support structure. Cresol has its roots in farmer organizations and movements that built community savings and loans mechanisms in agricultural communities and gradually formalized into coopera- tives. These cooperatives then formed a network (Cresol) with a central cooperative (Cresol-Baser) and regional "Cresol Baser" service centers offering support services, such as audits, bookkeeping, software, and legal assistance; the centers also fulfill a monitoring role and mediate with the Central Bank. The centers serve numerous cooperatives, ensuring more cost-effective operations, and are mainly staffed by members of the cooperatives. Cresol cooperatives also have access to a central liquidity fund. The government provides subsidized credit to Cresol through BNDES (the state development bank, Banco Nacional de Desenvolvimento Econômico e Social) and the PRONAF (Programa Nacional para Agricultura Familiar) program. Source: Branch, "Credit Union Rural Finance: Sicredi Brazil," 2003; CGAP research (including communications with SICREDI and DGRV (German Cooperative and Raiffeisen Confederation); DGRV reports and web site, www.dgrv.org; Bittencourt, "O coopera- tivismo de crédito no Brasil," 2003; and Cresol financial reports. 34 86696_01-52.qxd 8/31/05 9:23 AM Page 35 organizations is to strike a tricky balance between (because they assume the government will cover providing the crucial support needed to reduce the losses). Fraud can also be a problem. In corruption, avoid mistakes caused by poor gover- Mexico, for example, before the national agricul- nance and incompetent management, and limit tural insurance agency was closed, inspectors were financial failure to acceptable levels, while not found to be accepting bribes averaging 30 percent infringing on the ability of small informal associa- of the payouts made to farmers.99 Given the failure tions to operate viably. When membership-based of many government-sponsored schemes and their organizations are overly dependent on external vulnerability to being undermined by political funding or support, it can conflict with the inter- motives, the validity and potential effectiveness of ests of their members and put the safety of mem- state agricultural insurance programs must be seri- bers' deposits at risk.96 Community-managed ously questioned. revolving funds that are not savings-based almost A more promising approach is area-based index never succeed. When the initial or main source of insurance, which can be applied to both produc- funds is external, for instance, from a donor tion and price risks. Such insurance is defined at a agency, these funds tend not to achieve repayment regional level and provided against specific events that is good enough to keep the fund revolving that are independent of the behavior of the for very long. insured farmers. Examples include weather-related insurance policies linked to rainfalls or tempera- Feature 9 tures in a defined area, offering indemnity pay- Area-Based Index Insurance Can ments if the relevant index falls below (or rises Protect against the Risks of above) a certain level, and price-related policies Agricultural Lending with payouts based on crop prices. Such policies enable providers to insure against a specific risk, Governments have long sought to help reduce rather than all agriculture-related risks, and being agricultural production and price risks by provid- defined at a regional level makes them more viable ing livestock or, most often, crop insurance. But and attractive to private insurers because they these programs have tended to suffer from high reduce administrative costs and risks of fraud and administrative costs, unrealistically low premiums, moral hazard. moral hazard, and vulnerability to severe losses. Lenders can take out insurance policies to cover Administrative costs are rarely less than 30 percent their agricultural portfolios and pass the costs of of the income received from premiums.97 the premiums onto their farmer clients through Moreover, insurance has usually covered multiple additional fees or interest charges. Index-based (or all) risks, rather than specific, quantifiable hedging instruments bought on international ones. In the late 1990s, a study of seven public markets can allow lenders to manage potential crop insurance programs found that on average, losses from weather or price risks, giving them losses (payouts) were more than twice income.98 greater confidence to start or expand agricultural Moral hazard affects both insurance providers lending. Financial institutions can buy hedging (administering public-funded schemes) and clients instruments that reduce their exposure to losses under such programs. Farmers are less likely to take from default, bad weather, or delayed interest pay- steps to reduce losses and more likely to take more ments resulting from adverse price movements in risks (such as planting crops in marginal areas), commodities that their clients produce, trade, or while providers are less concerned about following process.100 Hedging can be done for an overall careful insurance practices when assessing losses portfolio, or a hedge can be attached to each loan. 35 86696_01-52.qxd 8/31/05 9:23 AM Page 36 Area-based index insurance has only recently lending, as well as help them negotiate insurance been extended to institutions that lend to or buy and hedging arrangements. from small farmers, and successful examples of it Although brokers would ideally come from are still rare. An emerging example involves the the private sector, Tanzania's KNCU and CRDB Kilimanjaro Native Cooperative Union (KNCU), received such assistance from the World Bank's a large Tanzanian coffee cooperative of small Commodity Risk Management Group (CRMG), farmers that trades about 11 percent of national which helped the cooperative develop a risk coffee production. The cooperative has had some management strategy and negotiate the put success in reducing its exposure to negative coffee options. The group also trained CRDB staff in price movements by buying "put" options that assessing price risks and providing advice on allow it to maintain an agreed floor purchase price hedging trends. The CRMG envisages this facili- with farmers during the trading season. It borrows tating role being performed in the future by a pri- from a domestic bank, the Cooperative and Rural vate broker, requiring only temporary donor or Development Bank (CRDB), to pay for the hedg- government support. ing contract premiums (put options). Thus the Microfinance institutions that insure small cooperative has reduced its exposure to price fluc- farmers and assume the related risks must be very tuations and falls in the value of coffee stocks held careful. When the insured event is relatively rare, during processing or while awaiting sale. Because all is well, and premiums can be an attractive the cooperative has used this approach for only income source. But a catastrophic event--even on one season, it is too early to draw any definitive a local level--may put a microfinance institution conclusions about its effectiveness.101 at risk of bankruptcy and non-compliance with its Index-based insurance has the potential to obligations to its insured clients. Insurance is by reduce both the risks of losses for individual farm- nature a product best built on the back of risk ers and the operational risks of lenders. A basic diversification over the largest possible group of difficulty for insurers in extending such coverage insured clients and the broadest range of circum- to small farmers is the same as that faced by micro- stances that can affect claims. finance institutions: how to profitably service small contracts and transactions. Governments Feature 10 and donors can adopt or support measures that To Succeed, Agricultural Microfinance enhance the potential for index-based insurance Must Be Insulated from Political from the private sector to include small-farmer Interference clients. They can, for example, ensure the existence and availability of accurate, timely, and Government and donor intervention in agricul- comprehensive databases--for example, on tural markets and lending, whether persistent or national or regional rainfall levels and commodity unpredictable, is perhaps the greatest source of prices--that private insurers can use to value risk for agricultural lenders. The provision and instruments for weather and price risks. In addi- design of agricultural finance have largely been tion, donors can encourage brokers to enter the driven by pressures to finance farm production market (for example, by disseminating data on and raise rural living standards, rather than build emerging approaches or even by providing train- sustainable infrastructure for rural finance. When ing). Brokers can help financial service providers government officials face a perceived choice assess and price the risks in their agricultural port- between promoting maximum outreach of folios and the risks of expanding agricultural rural financial services by building sustainable 36 86696_01-52.qxd 8/31/05 9:23 AM Page 37 institutions; and using institutions to channel agricultural subsidies and import duties in the finance in direct support of technology adoption, European Union and the United States, and in subsistence food production, and rural infrastruc- directed credit programs throughout the develop- ture development--regardless of long-term sus- ing world--is that farms, particularly family farms, tainability--they usually opt for the latter, at the should be subsidized to achieve broader social expense of sustainability. goals. Domestic agricultural subsidies in devel- Since the early 1970s, Ohio State University's oped countries are significant relative to their total Rural Finance Program and a host of other aid to developing countries, let alone their aid to academics, evaluators, and program administrators agriculture. US agricultural subsidies in 2003 (US have produced a vast literature detailing the $16.4 billion) were actually larger than total US shortcomings of this traditional approach to aid to developing countries ($16.3 billion).104 agricultural lending.102 At the heart of their critiques lie the propositions that governments Poor Repayment Rates and development agencies can serve a vital When governments in developing countries development purpose by fostering sustainable channel income support programs through loans, financial institutions to serve rural populations, discipline in credit markets suffers and agricultural and that these institutions do not require finance becomes more difficult. In the 1960s and permanent subsidies. 1970s, when evaluators analyzed the weak repay- At the same time, many sub-Saharan African ment performance of credit programs targeted at governments have traditionally maintained low small farmers, they blamed circumstances beyond food prices that favor urban populations--an the control of both borrowers and lenders. To approach that has reduced returns on agriculture explain the seeming inability of poor borrowers to and lowered demand for rural financial services. repay loans, they blamed natural disasters, poor In addition, governments of developed countries market infrastructure, inadequate land tenure, and undermine developing country agriculture by other factors that increase agricultural risks. In "dumping" surplus agricultural products on fact, they suggested that the same factors that pro- developing country markets, often in the name of duced poverty were responsible for loan defaults. aid. These surplus products are cheaper than During the early 1980s, evaluators blamed the locally produced produce and risk undermining intervention strategies for farmers' low repayment local agricultural production and lowering farmer rates. Because farm loans were usually provided as incomes.103 State controls on export crop prices a basic input under an integrated approach that and state intervention in processing and market- included better seeds, farming techniques, and ing have also distorted agricultural markets in marketing structures (such as cooperatives), as many developing countries. In many cases, export well as land reform, the failure of any of these ele- crops have been excessively taxed. ments was deemed sufficient to provoke default. Similarly, delayed disbursements, inappropriate Extensive Agricultural Subsidies loan amounts or terms, and other problems asso- By and large, governments around the world, ciated with providing loans under an integrated including those of the leading economic powers, approach were thought to contribute to default. have not yet accepted these propositions, and What these early studies did not reveal was why seem to take the view that farming families and some farmers in an area repaid their loans while rural communities should be supported through others did not, despite similar crops, yields, income transfers. Their position--reflected in incomes, and risks. These analysts assumed that 37 86696_01-52.qxd 8/31/05 9:23 AM Page 38 because small farmers had never before obtained defer loan repayments or in a situation where both formal sector credit, they did not know how to borrowers and lenders recognize that repayments use it. They believed that such borrowers "misal- will be difficult. This situation is exacerbated if a located" the credit to consumption activities, such lender has a tradition of pardoning farmer debts as weddings, funerals, education of their children, every few years, if such debts threaten the land or even food. The literature exuded a strong tenancy of smallholders. In such scenarios, small moral censorship when discussing these factors farmers have a strong incentive to delay payments even though today it is recognized that many of or roll over debts in the expectation that lenders these activities serve a strong economic purpose will ultimately write them off.107 for the poor by providing a financial safety net for Governments have all too often granted debt parents in old age.105 pardons to farmers in order to secure rural sup- After many years, evaluators began to realize port prior to elections. In Costa Rica, the govern- that the blame for poor repayment rates had to be ment debt pardon in 1999 substantially lowered laid at the doorstep of lenders and the incentives repayments on credit cards offered by Financiera they created for borrowers to comply with loan Trisan, a finance company set up by a wholesaler obligations.106 Although some of the aforemen- of agrochemicals (see box 7). As a result, delin- tioned reasons undoubtedly raised risk levels in quency rates on the cards rose as high as 25 per- agricultural lending, the primary cause of bor- cent, and more than 2,200 accounts were written rower default was the political economy of credit off over the following two years.108 Similarly, in (box 14). Governments provided cheap credit to 2001 Thailand's state-owned Bank for Agriculture small farmers as a way of getting votes, as well as and Agricultural Cooperatives was forced to par- to compensate for low farmgate prices and lack of ticipate in a government debt suspension program investment in rural infrastructure. As a result gov- for farmers who faced difficulties repaying their ernments were reluctant to enforce strict loan loans. More than 2 million farmers owing over recovery, especially in the face of general difficul- $1.7 billion--a third of BAAC's portfolio-- ties faced by large groups of farmers. enrolled in the program. As a result, BAAC's loan write-off rate jumped from 3 percent in 2001 to Debt Pardoning Schemes 12 percent in 2002, and its reserves for bad debt Loan defaults are bound to rise if farming house- rose to 21 percent of its loan portfolio.109 This holds are automatically offered the opportunity to indicated a growing repayment problem, with Box 14 Zambia: Omnia Small Scale--The Unintended Effects of Government Support for Agricultural Production Omnia Small Scale was created by Omnia, a South African fertilizer manufacturer and marketer, to extend fertilizer on credit to small farmers in Zambia. Farmers lacking cash could obtain fertilizer for an agreed amount of produce (such as three bags of maize for one bag of fertilizer). This scheme worked well until the government decided to use Omnia dealers as agents for its subsidized fertilizer credit program. The relaxed attitude toward default in the government's program led to reduced demand for Omnia's program. In response, Omnia shifted to a more sophisticated credit package that included both seed and fertilizer. It focused on farmers with good repayment records under the government scheme and disbursed US $300,000 in credit. But Omnia suf- fered losses with this package due to loan recovery rates of just 80 percent. In addition to crop failures in some areas, the company blamed its losses on the indebtedness caused by the government scheme (because some of its clients also took advantage of the government credit) and an expectation that non-repayment would be tolerated. Omnia no longer offers credit to smallholders, largely because of the government's distorting presence in the fertilizer sector. Source: Ruotsi, "Agricultural Marketing Companies as Sources of Smallholder Credit," 2003. 38 86696_01-52.qxd 8/31/05 9:23 AM Page 39 additional long-term implications on BAAC's client Directed Lending image and capacity to enforce future repayments. The requirement that lending be directed to cer- tain crops or uses has often complemented subsi- Subsidized Interest Rates dized interest rates. Given the fungible nature of The traditional agricultural lending paradigm money and the integrated nature of income remains widespread and has a number of funda- streams and expenditures in a poor household's mental flaws--none more damaging than highly budget, any insistence on directed lending subsidized interest rates. Cheap credit triggers a requires high monitoring costs to be effective. vicious cycle of credit rationing that favors better off rural inhabitants able to use status and connec- Failure of Credit to Reach Poor People tions to siphon off the available cheap credit, Most agricultural lending in developing countries deprives lenders of the budgets needed to is still subject to political interference, subsidized adequately follow through on and recover loans, interest rates, directed lending, and poor services. politicizes credit allocation and collection, and This paradigm does not work. Local elites often ultimately requires continued operating subsidies capture the loans that are supposed to go to poor for lenders.110 This unending cycle, combined with farmers, and default rates are unsustainably high periodic government interventions that under- (often well above 40 percent). These results have mine loan repayment, has convinced many continued despite 30 years of experience in thou- analysts that agricultural lending is too risky and sands of subsidized rural credit programs. cannot be undertaken commercially. In 1974, for example, the largest 10 percent of rural borrowers captured 80 percent of the highly High Transaction Costs subsidized agricultural credit offered by the Lenders often impose high transaction costs on National Bank of Costa Rica's agricultural credit low-interest loans as a rationing mechanism in the department. The smallest 50 percent obtained face of excess demand because they are unable to only 5 percent of this credit (in loans averaging adjust for this demand by raising interest rates. US $585), even though the program was designed For example, lenders require borrowers to make to promote small farmers' access to cheap credit. several trips to the bank or other agencies to put These results were in line with those in all the together necessary paperwork, check on the status other Costa Rican banks studied at the time.112 of the application, and meet other pre-conditions A decade later, the findings were similar. An for approval. These high transaction costs reduce Inter-American Development Bank review of the value of loans for clients and make them less directed and subsidized agricultural credit projects likely to repay, as they do not fear being disquali- in Latin America pointed to the continuing nega- fied for another loan the following year. When tive effects that cheap credit had on resource loans are relatively complex, borrowers may allocation, income distribution, macroeconomic rightly fear that approval in the coming year might management, and financial market development. be held up for arbitrary reasons. In part, high Evaluations of agricultural credit programs in transaction costs result from the fact that lenders Ecuador, Panama, and Peru in 1983 shared the do not have much revenue from which to develop following findings: a robust operational budget. Thus, low interest Outreach was limited. Peru's Banco Agrario rates fail to provide lenders with the income they accounted for more than 80 percent of credit need to build the infrastructure required to offer disbursed by the financial sector, but reached quality financial services.111 only 7 percent of farmers. 39 86696_01-52.qxd 8/31/05 9:34 AM Page 40 Most loans went to upper-income clients. households and agricultural investments. This Interest rates were generally negative in real final section suggests an emerging model for terms. agricultural microfinance. Borrower transaction costs were high. In General Features of Successful Agricultural Ecuador, the total transaction costs of a bor- Microfinance rower qualifying for and repaying a small loan were estimated to be equivalent to a 5 percent Over the past 20 years, one of the most important monthly interest rate. conclusions reached by development finance specialists is that poor families--and especially poor Portfolio quality was poor, with arrears rang- farming families--engage in a number of income- ing from 14 percent to 26 percent.113 generating activities, have a number of financial Such results are not limited to Latin America. coping strategies, and use a variety of formal An analysis of the state-controlled Agricultural and informal financial instruments to manage their Development Bank of Pakistan (ADBP) in the affairs. Although agricultural production may mid-1990s found that the bank's social costs be the main source of revenue, it is seldom the exceeded its social benefits by as much as 35 per- only source. cent. Moreover they found that 69 percent of This revelation has allowed development rural households received just 23 percent of for- finance initiatives, especially in the microcredit mal sector loans of the kind provided by the movement, to move away from the concept of ADBP (which reportedly provided the majority of loan repayments being tied to specific investment formal sector lending to agriculture)--while 4 activities. Accordingly, the entire borrowing percent of households (the wealthiest) received 42 household is seen as an economic unit where percent. Loans to politically connected borrowers income from different activities is mixed together also had a far higher default rate. In 1996 the to meet a wide variety of daily needs and repay- ADBP's loan recovery rate was 45 percent, down ment obligations. This shift has dramatically from 59 percent in 1991. The authors concluded improved loan recovery rates and is probably a that political factors had played a large role in the precondition for financial sustainability in most bank's declining worsening loan recovery rate.114 credit for poor people. An Emerging Model for Agricultural Successful agricultural microfinance providers Microfinance have married the core principles of the microcre- dit movement--peer-based borrower selection This paper has discussed a number of features that and repayment enforcement, incremental lending, are found in relatively successful attempts to pro- close follow-up on repayment, and so on--with vide financial services to poor farmers. No the technical expertise required to assess the agri- program incorporates all of them, nor is there any cultural competence of potential borrowers. In suggestion that every program should. But each this, these providers do not differ much from feature played a prominent role in a number of many of their urban counterparts operating indi- programs that had attained high repayment levels vidual lending schemes, which require that loan over a significant time span, and had reached officers have a certain degree of familiarity with a profitability or were well on their way to it. Some substantial number of business activities. of the features are practices that are relevant to Agricultural microfinance providers may mod- any kind of microfinance, while others respond ify loan terms and conditions to better accommo- to the particular challenges of serving farming date the more cyclical cash flows in farming 40 86696_01-52.qxd 8/31/05 9:36 AM Page 41 households and the more demanding timing of permanently subsidize sector-targeted credit pro- credit needs for crop or livestock activities. They grams. By adopting the risk management strate- have been able to do so, however, without sug- gies of microfinance institutions and loan analysis gesting that repayments are linked solely to the techniques that take into account the range of outcomes of specific investment activities. household economic activities and income One of the hallmarks of successful microcredit sources, traditional lenders can increase their sus- is that lenders diversify their loan portfolios into a tainability and the value of their services to poor large number of discrete, unrelated economic families. For example, by raising their interest activities. Similarly, organizations seeking to rates, such lenders can increase their budgets, engage in or expand lending in farming areas can offer higher-quality services, and improve their limit their exposure by lending to a large number repayment performance. of households engaged in many distinct agricul- But nowhere can agricultural microfinance tural and other economic activities. prosper in the face of political interference. Even Deposit facilities should be considered an the best-designed and -executed programs find it essential component when supplying microfinance almost impossible to maintain high repayment to farming households. Evidence suggests that rates in the face of widespread debt pardoning most rural poor people would prefer to use sav- (loan forgiveness) programs, massive provision of ings accounts instead of loans for bulky invest- highly subsidized credit, and the repressive ments, and in fact, for most financial service interest rates that characterize most government- needs, given that poverty tends to enforce a con- sponsored approaches. servative (in the sense of prudent) approach to financial management. Special Features for Specific Challenges In sum, with some significant--but not This paper has discussed a number of particularly daunting--adjustments, a lot of agri- approaches--linking loans to contractual farming cultural production can be financed using stan- arrangements, buying index-based insurance, dard microcredit principles. Supporting the steady making use of technology and existing institu- expansion of successful microfinance institutions tional infrastructure in rural areas--that successful into rural areas will almost inevitably increase the agricultural microfinance organizations (and oth- funds available for agriculture as these institutions ers) have used to address specific challenges. strive to serve the financial needs of farming These challenges are not common to all agricul- households. The research conducted for this paper tural finance situations, so not all programs need identified a significant number of successful to make use of all these features. Moreover, these efforts to adjust traditional microcredit products features are less mature, and the experiences back- to the needs of agricultural borrowers, although ing up their success are more tenuous, than those the total loan volumes remain small and the track generally applicable to microfinance. Some are record brief. still experimental, but have been included here Moreover, traditional agricultural finance because they try to address core issues in the fund- institutions can provide financial services more ing of agricultural production, and because they sustainably by adopting good practice microfi- provoke considerable interest in the agricultural nance techniques that reduce risks and enable finance community. financial sustainability. Such techniques will promote Contractual farming arrangements have proven institutional survival in a political climate where to be a powerful tool for managing risk and many donors and governments are unwilling to financing the production of complex crops, crops 41 86696_01-52.qxd 8/31/05 9:34 AM Page 42 that require a high level of standardization, or schemes, and knowledge has been gained about where a minimum volume of production is what types of insurance and guarantee funds do needed. The client information held by traders not work, considerable research is needed before and processors that provide credit through such reaching a general recommendation on crop arrangements is also of immense potential value to insurance for small farmers. rural lenders. There are a number of emerging Nevertheless, all these specific features have approaches for linking with or adopting contrac- shown definite potential in pilot schemes around tual arrangements, while obtaining the corollary the world in recent years. More important, they support required to meet production require- address the thornier challenges of agricultural ments. But agribusiness finance does not address finance: large and long-term investments, the the constraints of long-term finance because con- combination of price and yield risks, the relatively tractual arrangements are generally made on a sea- high costs of operating in rural areas with low sonal basis, and not for long-term investments in population density, and lending to clients with no agricultural infrastructure. credit records. These "extreme" challenges are Long-term financing of investment activities is unlikely to be addressed effectively by agricultural one of the least common types of agricultural microfinance institutions that restrict themselves finance. Very few successful programs exist. to the more general features closer to the microfi- Leasing has been tried on a limited basis, but its nance paradigm. And if agricultural finance results should still be considered experimental. providers do not make significant advances in The demand for long-term finance, however, may solving them, such providers will continue to be not be as high as the literature on agricultural viewed as an ineffective component of the broader finance assumes. Most long-term agricultural financial sector. investments are not financed primarily by loans One of the main reasons that most financial from financial institutions, but by household sav- institutions avoid agricultural lending--in addi- ings and by funds borrowed from friends and fam- tion to its high perceived risk--is that a history of ily members. By using multiple sources to finance cheap, subsidized credit and associated client long-term bulky investments, poor households expectations makes it extremely difficult to price reduce the weighted cost of overall borrowing and loans at viable (that is, profitable) levels. At the diversify the risk of repayments falling due when same time, poor farming households may be unexpected events affect anticipated income. unwilling or unable to pay the high interest rates Efforts are under way to piggy-back the provi- needed to cover the inefficiencies of many small sion of financial services for rural and farming rural lenders or the high operating costs of lenders households on existing commercial infrastructure. located far from urban centers. But by following In addition, technology is being used to improve the features of the emerging model presented in service access and quality through ATMs, cellular this paper, the costs for both lenders and borrow- phones, PDAs and handheld computers, and ers should be reduced, leading to sustainable agri- smart cards. Again, most of these attempts to dra- cultural microfinance. matically reduce the cost of service provision in The question remains of whether there are rural areas should be considered experimental-- certain groups of farmers, such as those with very though promising. small landholdings or those dependent on The same can be said of crop insurance against marginal low-return rainfed agriculture, for whom the general risks of agricultural lending. Although subsidies through agricultural finance might efforts are being made to develop insurance be justified. There is a case for initial subsidies 42 86696_01-52.qxd 8/31/05 9:34 AM Page 43 intended to lower the costs of the financial insti- risk management strategies, which may value a tution serving them--for example, through more balanced portfolio of activities) could be gained efficient operations and better systems or proce- from an alternative activity such as a microenter- dures (rather than subsidizing interest rates). But prise or working on someone else's farm. any use of subsidies needs to be balanced against Many millions of rural people would be much the limited budgets available to developing coun- better served if more financial institutions applied try governments, and the value of subsidizing a the features of the emerging agricultural microfi- farming activity that is not producing a viable nance model demonstrated by the minority of return relative to spending on hospitals, schools, relatively successful programs and set out in this roads, and other pressing needs. Put simply, a paper. The authors hope that donors and govern- person or household should not be encouraged to ments will abandon the old paradigm as a failure, go into debt for a particular crop or livestock and will continue to invest in the development of activity that is not likely to produce a profit, or if the elements of this new model. a better return (taking into account household 43 86696_01-52.qxd 8/31/05 9:23 AM Page 44 Endnotes Gonzalez-Vega, "Innovative Approaches to Rural Lending: Fi- nanciera Calpiá in El Salvador, " 2000. 1 IFAD, Rural Poverty Report 2001: The Challenge of Ending 20 Financiera Calpiá has recently converted to a bank, Banco Rural Poverty; and World Bank, Rural Finance Services: Im- ProCredit. In January 2005, Caja los Andes (legally, a "private plementing the Bank's Strategy to Reach the Rural Poor, financial fund") also converted to a bank--Banco los Andes 2003. The most recent World Bank data ("Global Monitoring ProCredit. In this paper, the original names will be used be- Report 2005") indicate that this number has fallen to 1.1 billion. cause the research was conducted while both were under their previous legal forms and names. 2 Here aid is measured as a share of average gross national in- come (GNI) for members of the Development Assistance 21 Wright, Microfinance Systems: Designing Quality Financial Committee (DAC) of the Organisation for Economic Co-op- Services for the Poor, 2000. eration and Development (OECD). See OECD, "2003 OECD 22 Pearce and Reinsch, "EDPYME Confianza, Peru," 2005. Development Co-operation Report: Statistical Annex." 23 Ayee, "Centenary Rural Development Bank: Agricultural 3 IFAD, Rural Poverty Report 2001. Lending," 2002. 4 DFID, "The Importance of Financial Sector Development for 24 Shrader, "Sustainable Expansion Strategies," 2003; CGAP re- Growth and Poverty Reduction," 2004. searcher correspondence with Richard Reynolds (World Vi- 5 Von Pischke, Finance at the Frontier: Debt Capacity and the sion) and Sabina Bina (EKI finance director) during 2003; EKI Role of Credit in the Private Economy, 1991. "Financial Statements, 31-12-2002." 6 World Bank, Implementing the Bank's Strategy to Reach the 25 Buchenau, 2003, "Innovative Products and Adaptations for Rural Poor, 2003. Rural Finance," 2003. 7 "Sector loans" were used to assist borrowing countries to re- 26 Interviews with Banco del Estado management team by au- form financial markets (introducing market-based interest thor (Christen), 1997­98. rates and more banking competition, and strengthening reg- 27 See, for example, Adams and Fitchett, Informal Finance in ulatory frameworks, for example), and were not directed to- Low-Income Countries, 1992. wards agricultural lending. 28 Sebstad and Cohen, Microfinance Risk Management and 8 Wenner, "Lessons Learned in Rural Finance: The Experience Poverty, 2001. of the Inter-American Development Bank," 2002. 29 See Rutherford, The Poor and Their Money, 2000. 9 FAO, "Crop and Food Supply Assessment Mission to Malawi," 2003. 30 BAAC is included here due to its massive scale, although its state-owned character clouds its sustainability. 10 Ayee,"Centenary Rural Development Bank: Agricultural Lend- ing," 2002. 31 The diversity referred to here relates particularly to the intro- duction of savings services. BAAC's lending operations are 11 Onumah, "Improving Access to Rural Finance through Reg- still dominated by agriculture, if less so than previously. ulated Warehouse Receipt Systems in Africa," 2003. 32 These data come from BAAC's Annual Report 2002, citing 12 Varangis et al, "Dealing with the Coffee Crisis in Central data from Thailand's National Statistics Office. Clients re- America," 2003. ceived credit from BAAC either directly or through coopera- 13 The main exceptions are certain types of credit arrangements tives and associations. Note that BAAC's fiscal year runs from between farmers and agricultural processors or traders, in April to March. The data here are for fiscal 2002 and as of 31 which loan repayments are deducted from the prices paid for March 2003, unless otherwise stated. the resulting production. 33 Wehnert and Shakya, "Are SFCLs Viable Microfinance Orga- 14 The visits, conducted by consultants contracted by CGAP, nizations? A Comprehensive Financial Analysis of 33 SFCLs," involved institutions and programs in Bolivia, India, Kenya, 2001; Siebel, "Agricultural Development Bank Reform," 2001. Mozambique, Peru, and Uganda. 34 The performance of many SFCLs has recently suffered due 15 Haggblade et al, "Strategies for Stimulating Poverty-Alleviat- to Maoist activities. ing Growth in the Rural Non-farm Economy in Developing 35 The term "savings and loan cooperatives" here is used inter- Countries," 2002. changeably with "credit unions." 16 Hulme and Mosley, "Finance for the Poor or the Poorest?" 1997. 36 Bittencourt, "O cooperativismo de crédito no Brasil," 2003; 17 Robinson, The Microfinance Revolution, Volume 2: Lessons Cresol reports. from Indonesia, 180­82. 37 CGAP research (including communications with SICREDI and 18 One exception is the credit arrangement between farmers DGRV); DGRV web site, www.dgrv.org. and agricultural processors or traders in which loan repay- 38 Pearce, Goodland, and Mulder, "Investments in Rural Finance ments are deducted from the prices paid for the resulting for Agriculture: Overview," 2004; Ayee, "Centenary Rural De- production. velopment Bank (CERUDEB): Agricultural Lending," 2002. 19 Navajas, "The Process of Adapting Lending Technology: 39 Pearce and Reinsch, "Caja los Andes, Bolivia," 2005. Financiera Calpiá in Rural El Salvador,"1999; Navajas and 44 86696_01-52.qxd 8/31/05 9:23 AM Page 45 Endnotes continued 40 Christen and Vogel, "The Importance of Domestic Resource 64 It is probable that less specialized traders may offer loans for Mobilization in Averting Financial Crisis: The Case of Credit other purposes. Unions in Honduras," 1984. 65 Shepherd, "Financing of Agricultural Marketing: The Asian 41 Rubio, "EDPYME Confianza," 2002; Pearce and Reinsch, Experience," 2004. "EDPYME Confianza, Peru," 2005. 66 Ruotsi, "Agricultural Marketing Companies as Sources of 42 Buchenau, "Financing Small Farmers in Latin America," Smallholder Credit: Experiences, Insights, and Potential 1997. Donor Role," 2003; CLUSA , "Quarterly Report, April-June 2003"; de Vletter, "A Review of Three Successful Rural Fi- 43 Lee, "Client Based Market Research: The Case of PRO- nance Cases in Mozambique," 2003. DEM," 2000; Rubio, "EDPYME Confianza," 2002. 67 Pearce, "Buyer and Supplier Credit to Farmers: Do Donors 44 Rubio, "EDPYME Confianza," 2002; Pearce and Reinsch, Have a Role to Play?" 2003. "EDPYME Confianza, Peru," 2005. 68 After the bank's merger with a European bank, its involvement 45 Lee, "Client Based Market Research: The Case of PRODEM," 2000. was discontinued and CES Solidaridad set up a microfinance 46 Ayee, "Centenary Rural Development Bank (CERUDEB): institution to provide the financing instead. Agricultural Lending," 2002. 69 Alvarado, Galarza, and Cajavica, "El Linking ONG--Banco del 47 This institution's technical partner, ACDI/VOCA, developed CES Solidaridad de Chiclayo," 2000; Wenner, Alvarado, and a portfolio diversification strategy to address this issue, and Galarza, "Promising Practices in Rural Finance: Experiences provided technical assistance to implement the strategy. from Latin America and the Caribbean," 2003. 48 See Westley, "Equipment Leasing and Lending: A Guide for 70 Point-of-sale devices are used for payments and disburse- Microfinance," 2003, for a fuller discussion of the asset-lia- ments and located at retail outlets. bility management implications of term lending. 71 Pearce, "Buyer and Supplier Credit to Farmers: Do Donors 49 FAO, "Term Financing in Agriculture: A Review of Relevant Have a Role to Play?" 2003. Experiences," 2003. 72 IFAD (2003). For more information on savings rates, see Ary- 50 FAO, "Term Financing in Agriculture." Changes in institutional eetey and Udry, 2000. loan classification after 1999 obscure the more recent pro- 73 Wenner and Proenza, 2000. portion of term loans. 74 Rodriguez-Meza, as cited in Buchenau and Hidalgo (2002). 51 BAAC annual reports; Haberberger et al, "The Challenge of Sustainable Outreach: The case of Bank for Agriculture and 75 www.studentsoftheworld.info/infopays/rank/densite2.html; Agricultural Cooperatives (BAAC)," 2003. World Bank, World Development Indicators 2001. 52 FAO, "Term Financing in Agriculture," 2003. 76 CGAP provided a Pro-Poor Innovation Challenge grant of US $50,000 to support two microbanks that are piloting remit- 53 Westley, "Equipment Leasing and Lending: A Guide for Micro- tance services, as well as working with migrants in California finance," 2003 on the remittance-sending side. 54 See FAO, "Term Financing in Agriculture," 2003. 77 This is a significant percentage of savings, given that remit- 55 Ayee, "DFCU Leasing," 2002. tances usually average between $100 and $250. 56 Westley, "Equipment Leasing and Lending: A Guide for Mi- 78 AMUCSS (Asociación Mexicana de Uniones de Crédito del crofinance," 2003 Sector Social) reports; CGAP research and field visit by au- thor (Pearce), October 2003. 57 IFC, "Leasing in Central Asia," 2003); USAID, . "Leasing: An Overview from the IFC," 2003. 79 World Bank, "Workers' Remittances: An Importance and Sta- ble Source of External Development Finance," 2003. 58 Smith, Stockbridge, and Lohano, "Facilitating the Provision of Farm Credit: The Role of Interlocking Transactions be- 80 Sanders, "Migrant Remittances to Developing Countries. A tween Traders and Zamindars in Crop Marketing Systems in Scoping Study: Overview and Introduction to Issues for Pro- Sindh," 1999. Poor Financial Services," 2003. 59 Gore, "Rural Trade Finance in Southern Africa," 2003. 81 Orozco, "Remittances, the Rural Sector, and Policy Options in Latin America," 2003. 60 Shepherd, "Financing of Agricultural Marketing: The Asian Experience," 2004. 82 Ibid. 61 This type of arrangement is also referred to as "interlocking" 83 Pearce and Reinsch, "Equity Building Society, Kenya," 2005. because it provides inputs on credit on the basis of the bor- 84 World Bank, Agriculture Investment Sourcebook, 2004. rower's expected harvest. 85 Rubio, "Fondo Financiero Privado PRODEM S.A.," 2003; 62 Ruotsi, "Agricultural Marketing Companies as Sources of Whelan and Echange LLC , "Automated Teller Machine Tech- Smallholder Credit: Experiences, Insights, and Potential nology," 2003. Donor Role," 2003. 86 Informal interview by author (Pearce) with Juan Rosás, Direc- 63 Ibid. tor of International Institutions for La Caixa, December 2003; 45 86696_01-52.qxd 8/31/05 9:23 AM Page 46 Endnotes continued see also Banco Solidario (2003) [need citation--missing from 100 International markets are able to pool risk globally, enabling biblio]; Sanders, "Migrant Remittances to Developing Coun- hedging contracts to be more accessibly priced. tries: A Scoping Study: Overview and Introduction to Issues 101 Bryla et al, "The Use of Price and Weather Risk Manage- for Pro-Poor Financial Services," 2003; Quesada, "Profitable ment Instruments," 2003; Mwakalukwa, "Price Risk Man- Solidarity: An Ecuadorian Bank with Social Aims and Healthy agement and Access to Credit," 2003. Profits," 2003. 102 See, for example, Donald, Credit for Small Farmers in De- 87 CAREC is the acronym in French for "Support Center for Net- veloping Countries, 1976; Von Pischke, Adams, and Don- work Savings and Loan Banks." ald, Rural Financial Markets in Developing Countries, 1983; 88 Wampfler and Mercoiret, "Microfinance and Producers' Or- and Von Pischke, Finance at the Frontier: Debt Capacity and ganizations: Roles and Partnerships in the Context of Liber- the Role of Credit in the Private Economy, 1991. alization," 2001; Wampfler, "Coordination et Pérennisation 103 Although evidence is mixed, and suggests that the largest des Services autour du Financement de l'Agriculture Famil- effect may be import displacement (rather than affecting do- iale dans la zone Office du Niger (Mali)," 2003; CGAP re- mestic agricultural production). See Lowder, Southgate, and searcher correspondence with Renée Chao-Béroff of CIDR, Rodriguez-Meza (2004 (forthcoming) [need full citation, 2004. missing from biblio] and DiGiacomo, "U.S. Foreign Agricul- 89 CGAP correspondence with GTZ Nepal, 2003; Staschen, "Fi- tural Policy," 1996. nancial Technology of Small Farmer Cooperatives, Ltd. (SF- 104 EWG farm subsidy database, www.ewg.org/farm/region- CLs): Products and Innovations," 2001. summary.php?fips=00000); OECD, "2003 OECD Develop- 90 Hirschland, "Serving Small Rural Depositors: Proximity, In- ment Co-operation Report: Statistical Annex," 2003. novations, and Trade-Offs," 2003. 105 Christen, "Take the Money and Run," 1984. 91 de Vletter, "A Review of Three Successful Rural Finance 106 See, for example, Vogel, "Rural Financial Markets: Implica- Cases in Mozambique," 2003; Pearce, "Buyer and Supplier tions of Low Delinquency Rates," 1981. Credit to Farmers: Do Donors Have a Role to Play?" 2003; Ruotsi, "Agricultural Marketing Companies as Sources of 107 Vogel, "Rural Financial Markets: Implications of Low Delin- Smallholder Credit: Experiences, Insights, and Potential quency Rates," 1981. Donor Role," 2003. 108 Email exchange between CGAP researcher and Charles 92 Pearce, Goodland, and Mulder, "Investments in Rural Fi- Spalding, Director of Trisan, December 2002; Wenner and nance for Agriculture: Overview," 2004. Quiros, "An Agricultural Credit Card Innovation: The Case of Financiera Trisan," 2000. 93 Pearce and Helms, "Financial Services Associations: The Story So Far," 2001. 109 BAAC annual reports; Haberberger et al, "The Challenge of Sustainable Outreach: The case of Bank for Agriculture and 94 Chao-Beroff et al, "A Comparative Analysis of Member- Agricultural Cooperatives (BAAC)," 2003. Based Microfinance Institutions in East and West Africa," 2000. 110 See, for example, Gonzales-Vega, "Arguments for Interest 95 Adapted from Pearce, Goodland, and Mulder, "Investments Rate Reform," 1983. in Rural Finance for Agriculture: Overview," 2004. 111 Adams, Gonzalez-Vega, and Von Pischke, Crédito agricola 96 Not surprisingly, members manage their own savings more y desarrollo rural: la nueva vision, 1987. carefully than external-sourced money. If the amount of ex- ternal funding is significant in relation to the amount of mem- 112 Vogel, "The Impact of Subsidized Credit on Income Distri- ber savings, loan analysis, monitoring, and collection tend to bution in Costa Rica ," 1984. suffer. 113 Wenner, "Lessons Learned in Rural Finance: The Experience 97 Yaron, "Rural Finance: Issues, Design, and Best Practices," 1998. of the Inter-American Development Bank ," 2002. 98 Skees, Hazell, and Miranda, "New Approaches to Crop Yield 114 Khandker and Faruqee, "The Impact of Farm Credit in Pak- Insurance in Developing Countries," 1999. istan," 1998. 99 Ibid. 46 86696_01-52.qxd 8/31/05 9:23 AM Page 47 Bibliography ACDI/VOCA. "Quarterly Institutional Summary: SEED-Georgia." Rural Finance and Enterprise Development Unit, Washington, DC, 2002. ------. "Bulgaria--Agribusiness Support Project." Washington, DC, 2003. www.acdivoca.org/acdivoca/acdiweb2.nsf/whatwedo/bul- gariaasp?opendocument. Adams, D. W., and D. Fitchett. Informal Finance in Low-Income Countries. Boulder, Colo.: Westview Press, 1992. Adams, D. W., C. Gonzalez-Vega, and J. D. Von Pischke. Crédito agricola y desarrollo rural: la nueva vision. Columbus, Ohio: The Ohio State University, 1987. Asian Development Bank. "Project Performance Audit Report on the Agricultural Development Project." PPA: NEP161354. Manila, the Philippines: ADB, 1997. Allen, H. "CARE International's Village Savings and Loan Programmes in Africa: Microfinance for the Rural Poor that Works." Tanzania: CARE International, 2002. Alvarado, I., and K. Charmel. "The Rapid Rise of Supermarkets in Costa Rica: Impact on Horticultural Markets." Development Policy Review 20, no. 4 (2002): 473-85. Alvarado, J., F. Galarza, and J. Cajavica. "El Linking ONG--Banco del CES Solidaridad de Chiclayo." Paper presented at the Inter-Amer- ican Development Bank seminar, "Promising Practices in Rural Finance," Washington, DC, 22 May 2000. Aryeetey, E., and C. Udry. "Savings in Sub-Saharan Africa." CID Working Paper 38. Cambridge, Mass.: Harvard University, Center for International Development, 2000. Ayee, G. "Centenary Rural Development Bank (CERUDEB): Agricultural Lending." Consultant report prepared for CGAP, Washington, DC, 2002. ------. "DFCU Leasing." Consultant report prepared for CGAP, Washington, DC, 2002. ------. "Equity Building Society (EBS): Agricultural Lending." Consultant report prepared for CGAP, Washington, DC, 2003. Bank for Agriculture and Agricultural Cooperatives. BAAC Annual Reports 2000/2001, 2001/2002, and 2002/2003. Bangkok: BAAC, 2001, 2002, 2003. Berdegué, J., F. Balsevich, L. Flores, and T. Reardon. "The Rise of Supermarkets in Central America: Implications for Private Standards for Quality and Safety of Fresh Fruits and Vegetables." Draft report prepared for the US Agency for International Development's RAISE/SPS (Worldwide Sanitary and Phytosanitary Support for Trade Capacity Building Activities) project, Washington, DC, 2003. Bittencourt, G. "O cooperativismo de crédito no Brasil." Report prepared for the United Nations Food and Agriculture Organization (FAO), Rome, Italy, 2003. Branch, Brian. "Credit Union Rural Finance: Sicredi Brazil." Paper presented at the World Bank's Rural Finance Week, Washington, DC, 12 March 2003. Bryla, E., J. Dana, U. Hess, and P. Varangis. "The Use of Price and Weather Risk Management Instruments." Paper presented at the US Agency for International Development-World Council of Credit Unions conference, "Paving the Way Forward for Rural Finance: An International Conference on Best Practices," Washington, DC, 2­4 June 2003. Buchenau, J. "Financing Small Farmers in Latin America." Paper presented at the "First Annual Seminar on New Development Finance," Frankfurt, Germany, September 1997. ------. "Innovative Products and Adaptations for Rural Finance." Paper presented at the US Agency for International Development- World Council of Credit Unions conference, "Paving the Way Forward for Rural Finance: An International Conference on Best Prac- tices," Washington, DC, 2­4 June 2003. Buchenau, J., and A. Hidalgo. "Servicios financieros privados en al area rural de America Latina: situación y perspectivas." Paper pre- sented at an Inter-American Development Bank conference, Fortaleza, Brazil, 7 March 2002. Chao-Beroff, R. "Self-Reliant Village Banks, Mali (Case Study)." Eschborn, Germany: GTZ, BMZ, and CGAP, 1999. Chao-Beroff, R., T. Houmard, J. Vandenbroucke, M. Musinga, E. Tiaro, and L. Mutesasira. "A Comparative Analysis of Member-Based Microfinance Institutions in East and West Africa." Nairobi, Kenya: MicroSave, 2000. Christen, Robert P. "Take the Money and Run." Master's thesis. The Ohio State University, Columbus, Ohio, 1984. Christen, Robert P., and Robert C. Vogel. "The Importance of Domestic Resource Mobilization in Averting Financial Crisis: The Case of Credit Unions in Honduras." Conference Paper 13. Presented at the conference, "Financial Crisis, Foreign Assistance, and Domes- tic Resource Mobilization in the Caribbean Basin," at The Ohio State University, Columbus, Ohio, 30 April­1 May 1984. Christen, Robert P., Richard Rosenberg, and Veena Jayadeva. Financial Institutions with a Double-Bottom Line: Implications for the Fu- ture of Microfinance. CGAP Occasional Paper, no. 8. Washington, DC: CGAP, 2004. 47 86696_01-52.qxd 8/31/05 9:23 AM Page 48 Bibliography continued Coetzee, G., K. Kabbucho, and A. Mnjama. "Understanding the Rebirth of Equity Building Society in Kenya." Nairobi, Kenya: MicroSave- Africa, 2002. CLUSA (Cooperative League of the USA). "Quarterly Report, April-June 2003." Rural Group Enterprise Development Program., Mozam- bique: CLUSA, 2003. Coulter, J., and G. E. Onumah. "The Role of Warehouse Receipt Systems in Enhanced Commodity Marketing and Rural Livelihoods in Africa." Food Policy 27: 319­37. Craig, K., and R. Goodwin-Groen. "Donors as Silent Partners in MFI Product Development: MicroSave-Africa and the Equity Building Society in Kenya." Case Studies in Donor Good Practices, no 8. Washington, DC: CGAP, 2003. Cresol. Financial and Performance Reports for the Cresol Cooperative System (consolidated balance sheet and comparative indicators table). Reports provided by Cresol-Baser, Francisco Beltrao, Brazil, 2003. de Vletter, F. "A Review of Three Successful Rural Finance Cases in Mozambique." Report prepared for CGAP, Washington, DC, 2003. DFID (UK Department for International Development). "The Importance of Financial Sector Development for Growth and Poverty Re- duction." Policy Division Working Paper. Financial Sector Team, London, 2004. DiGiacomo, G. "U.S. Foreign Agricultural Policy." Foreign Policy in Focus 1, no. 10 (1996). Donald, G. Credit for Small Farmers in Developing Countries. Boulder, Colo.: Westview Press, 1976. EBS (Equity Building Society). "Mobile Banking Experience." Report prepared for CGAP, Washington, DC, 2003. EKI (Economic Credit Institution). "Financial Statements, 31-12-2002." Sarajevo, Bosnia and Herzegovina: EKI, 2002. FAO (Food and Agriculture Organization of the United Nations). The State of Food and Agriculture. Rome, Italy: FAO, 1998. ------. "Term Financing in Agriculture: A Review of Relevant Experiences." FAO Investment Centre Occasional Paper 14. Rome, Italy: FAO, 2003. FAO (Food and Agriculture Organization of the United Nations) Global Information and Early Warning System and WFP (World Food Pro- gramme). "Crop and Food Supply Assessment Mission to Malawi." Special report, Rome, Italy, 2003. Gonzalez-Vega, C. "Arguments for Interest Rate Reform." In ed. Rural Financial Markets in Developing Countries, J. D. Von Pischke, D. W. Adams, and G. Donald. Baltimore, Md.: The Johns Hopkins University Press, 1983. Goodwin-Groen, R. "Success in Rural Savings: How One Donor Led the Way." Case Studies in Donor Good Practices, no 1. Washing- ton, DC: CGAP, 2003. www.microfinancegateway.org/content/article/detail/3779. Gordon, A., and A. Goodland. "Production Credit for African Small-Holders: Conditions for Private Provision." Savings and Develop- ment 1 (2000): 55­83. Gore, C. "Rural Trade Finance in Southern Africa." Presentation of research findings to the US Agency for International Development, Washington, DC, 2003. Haberberger, M.L., L. Wajananawat, and N. Kuasakul. "The Challenge of Sustainable Outreach: The Case of Bank for Agriculture and Agricultural Cooperatives (BAAC)." Eschborn, Germany: GTZ, 2003. Haggblade, S., P. Hazell, and T. Reardon. "Strategies for Stimulating Poverty-Alleviating Growth in the Rural Non-farm Economy in De- veloping Countries." Environment and Production Technology Division Discussion Paper 92. Washington, DC: International Food Policy Research Institute and World Bank, 2002. Hess, U. "Innovative Financial Services for India: Monsoon-Indexed Lending and Insurance for Smallholders." Agriculture and Rural De- velopment Working Paper 9. Washington, DC: World Bank, 2003. Hirschland, Madeline. "Savings Operations for Very Small or Remote Depositors: Some Strategies." Summary of discussions from the virtual conference, April 29-May 17, 2002. www.microfinance gateway.org/viewpoint_savings.htm. ------. "Serving Small Rural Depositors: Proximity, Innovations and Trade-offs." Paper presented at the U.S. Agency for International Development-World Council of Credit Unions conference, "Paving the Way Forward for Rural Finance: An International Conference on Best Practices," Washington, DC, 2­4 June 200. Hulme, D., and P. Mosley. "Finance for the Poor or the Poorest? Financial Innovation, Poverty, and Vulnerability." In ed. Who Needs Credit? Poverty and Finance in Bangladesh., G. D. Wood and I. Sharif, Dhaka, Bangladesh: University Press Limited, 1997. IFAD (International Fund for Agricultural Development). Rural Poverty Report 2001: The Challenge of Ending Rural Poverty. 2001. ------. "Ghana--Women's Access to Formal Financial Services." Rome, Italy: IFAD, 2003. www.ifad.org/gender/learning/sector/fi- nance/42.htm. IFC (International Finance Corporation). "Leasing in Central Asia." Washington, DC: IFC, 2003. IMF (International Monetary Fund). Balance of Payments Yearbook 2003. Washington, DC: IMF, 2003. Khandker, S. R., and R. R. Faruqee. "The Impact of Farm Credit in Pakistan." Washington, DC: World Bank, 1998. Lee, N. "Client-Based Market Research: The Case of PRODEM." Toronto, Ontario, Canada: Calmeadow, 2000. 48 86696_01-52.qxd 8/31/05 9:23 AM Page 49 Bibliography continued Lowder, S. "A Post-Schultzian View of Food Aid, Trade and Developing Country Cereal Production: A Panel Data Analysis." PhD disser- tation, The Ohio State University, Columbus, Ohio, 2004. Meyer R. "Individual Lending in Rural Finance: The IPC Model." Paper presented at the Microfinance Regulatory Council of South Africa seminar, "Current Issues in Microfinance," Johannesburg, South Africa, 12­14 August 2003. Mwakalukwa, R. "Price Risk Management and Access to Credit." Paper presented at the "CASS School of Business International Con- ference on Commodity Risk Management," London, 4-5 June 2003. Nagarajan, G., and W. Brown. "Bangladeshi Experience in Adapting Financial Services to Cope with Floods: Implications for the Micro- finance Industry." Microenterprise Best Practices Project, US Agency for International Development, Washington, DC, 2000. Navajas, S. "The Process of Adapting Lending Technology: Financiera Calpiá in Rural El Salvador." Columbus, Ohio: The Ohio State Uni- versity, 1999. Navajas, S., and C. Gonzalez-Vega. "Innovative Approaches to Rural Lending: Financiera Calpiá in El Salvador." Economics and Soci- ology Occasional Paper No. 2571. Columbus, Ohio: The Ohio State University, 2000. OECD (Organisation for Economic Co-operation and Development). "2003 OECD Development Co-operation Report: Statistical Annex." Paris: OECD, 2003. www.oecd.org/document/9/0,2340,en_2825_495602_1893129_1_1_1_1,00.html. Onumah, G. "Improving Access to Rural Finance through Regulated Warehouse Receipt Systems in Africa." Paper presented at the US Agency for International Development­World Council of Credit Unions conference, "Paving the Way Forward for Rural Finance: An International Conference on Best Practices," Washington, DC, 2­4 June 2003. Orozco, M. "Remittances, the Rural Sector, and Policy Options in Latin America." Paper presented at the US Agency for International Development­World Council of Credit Unions conference, "Paving the Way Forward for Rural Finance: An International Conference on Best Practices," Washington, DC, 2­4 June 2003. Pearce, D. and B. Helms. "Financial Services Associations: The Story So Far." Washington, DC: CGAP, 2001. Pearce, D. "Pro-Poor Innovation Challenge Case Study--Constanta (Georgia). Service Points: Low Cost and Low Risk Expansion into Secondary Markets." Report prepared for CGAP, Washington, DC, 2002. ------. "Buyer and Supplier Credit to Farmers: Do Donors Have a Role to Play?" Paper presented at the US Agency for International Development­World Council of Credit Unions conference, "Paving the Way Forward for Rural Finance: An International Conference on Best Practices," Washington, DC, 2­4 June 2003. Pearce, D., and M. Reinsch. "Caja los Andes, Bolivia." Agricultural Microfinance Case Study. Washington, DC: CGAP, 2005. ------. "EDPYME Confianza, Peru." Agricultural Microfinance Case Study. Washington, DC: CGAP, 2005. ------. "Equity Building Society, Kenya." Agricultural Microfinance Case Study. Washington, DC: CGAP, 2005. Pearce, D., A. Goodland, and A. Mulder. "Investments in Rural Finance for Agriculture: Overview." In the "Rural Finance for Agriculture" module of the Agriculture Investment Sourcebook. Washington, DC: World Bank, 2004. www-esd.worldbank.org/ ais/index.cfm?Page=mdisp&m=7&p=0. ------. "Microfinance Institutions Moving into Rural Finance for Agriculture." In the "Rural Finance for Agriculture" module of the Agri- culture Investment Sourcebook 2004. Washington, DC: World Bank 2004. www-esd.worldbank.org/ais/index.cfm?Page= mdisp&m=7&p=0. Pearce, D., A. Mulder, and A. Goodland. "Membership-Based Financial Organizations." In the "Rural Finance for Agriculture" module of the Agriculture Investment Sourcebook 2004. Washington, DC: World Bank, 2004. www-esd.worldbank.org/ais/index.cfm?Page= mdisp&m=7&p=0. Pearl, D., and Michael M. Phillips. "Grameen Bank, Which Pioneered Loans for the Poor, Has Hit a Repayment Snag." The New York Times, 27 November 2001: 1. PlaNet Rating. "Equity Building Society (EBS)." Nairobi, Kenya: PlaNet Finance, 2001. Quesada, C. "Profitable Solidarity. An Ecuadorian Bank with Social Aims and Healthy Profits." Washington, DC: Inter-American Devel- opment Bank, 2003. Reille, X., and A. Vega Acosta. "CGAP Appraisal Report: EMT, Cambodia." Washington, DC: CGAP, 2002. Robinson, M. S. Sustainable Finance for the Poor. Vol. 1 of The Microfinance Revolution. Washington, DC: World Bank and Open Soci- ety Institute, 2001. ------. Lessons from Indonesia. Vol. 2 of The Microfinance Revolution. Washington, DC: World Bank and Open Society Institute, 2002. Rubio, F. "Caja Rural San Martin." Consultant report prepared for CGAP, Washington, DC, December 2002. ------. "EDPYME Confianza." Consultant report prepared for CGAP, Washington, DC, 2002. ------. "Fondo Financiero Privado PRODEM S.A." Consultant report prepared for CGAP, Washington, DC, 2003. Rutherford, S. The Poor and Their Money. Oxford: Oxford University Press, 2000. 49 86696_01-52.qxd 8/31/05 9:23 AM Page 50 Bibliography continued Ruotsi, J. "Agricultural Marketing Companies as Sources of Smallholder Credit: Experiences, Insights, and Potential Donor Role." Re- port prepared for the International Fund for Agricultural Development, Rome, Italy, 2003. Sanders, C. "Migrant Remittances to Developing Countries. A Scoping Study: Overview and Introduction to Issues for Pro-Poor Finan- cial Services." Report prepared for UK Department for International Development, London, 2003. Scherr, S. J. "Poverty-Environment Interactions in Agriculture: Key Factors and Policy Implications." Poverty and Environment Issues Series 3. New York: United Nations Development Programme and European Commission, 1999. Sebstad, J., and M. Cohen. Microfinance Risk Management and Poverty. Washington, DC: Pact Publications, 2001. Shepherd, A. "Financing of Agricultural Marketing: The Asian Experience." Agricultural Management, Marketing and Finance Service Occasional Paper 2. Rome, Italy: FAO, 2004. Shrader, Leesa Wilson. "Sustainable Expansion Strategies: Case Studies of 18 Regional Leaders in Microfinance." Research Study. War- saw, Poland: Microfinance Centre, 2003. Siebel, H. D. "Agricultural Development Bank Reform." Working paper, University of Cologne, Germany, 2001. Skees, J., P. Hazell, and M. Miranda. "New Approaches to Crop Yield Insurance in Developing Countries." Environment and Production Technology Division Discussion Paper No. 55. Washington, DC: International Food Policy Research Institute, November 1999. Smith, L., M. Stockbridge, and H. Lohano. "Facilitating the Provision of Farm Credit: The Role of Interlocking Transactions between Traders and Zamindars in Crop Marketing Systems in Sindh." World Development 27, no. 2 (1999): 403­18. Staschen, S. "Financial Technology of Small Farmer Cooperatives Ltd (SFCLs): Products and Innovations." Working Paper 2. Kathmandu, Nepal: Rural Finance Nepal, 2001. USAID (U.S. Agency for International Development). "Leasing: An Overview from the IFC." Enterprise Development Project, Washing- ton, DC, 2003. Varangis, P., P. Siegel, D. Giovannucci, and B. Lewin. "Dealing with the Coffee Crisis in Central America: Impacts and Strategies." Pol- icy Research Working Paper 2993. Washington, DC: World Bank, 2003. Vogel, R. C. "Rural Financial Markets: Implications of Low Delinquency Rates." American Journal of Agricultural Economics 63, no.1 (1981): 58­65. ------. "The Impact of Subsidized Credit on Income Distribution in Costa Rica." In ed. Undermining Rural Development with Cheap Credit, D.W. Adams, D.H. Graham, and J.D. von Pischke, Boulder, Colo.: Westview Press, 1984. ------. "Savings Mobilization: The Forgotten Half of Rural Finance." In ed. Undermining Rural Development with Cheap Credit, D. W. Adams, D.H. Graham, and J.D. von Pischke, Boulder, Colo.: Westview Press, 1984. Von Pischke, J. D. Finance at the Frontier: Debt Capacity and the Role of Credit in the Private Economy. Development Studies. Wash- ington, DC: World Bank Economic Development Institute, 1991. Von Pischke, J. D., D. W. Adams, and G. Donald, eds. Rural Financial Markets in Developing Countries. Baltimore, Md.: The Johns Hop- kins University Press, 1983. Wampfler, B. "Coordination et Pérennisation des Services autour du Financement de l'Agriculture Familiale dans la zone Office du Niger (Mali)." CIRAD and Office du Niger URDOC, July 2003. Wampfler, B., and M. R. Mercoiret. "Microfinance and Producers' Organizations: Roles and Partnerships in the Context of Liberaliza- tion." Paris: CIRAD, 2001. Wehnert, U., and R. Shakya. "Are SFCLs Viable Microfinance Organizations? A Comprehensive Financial Analysis of 33 SFCLs." Work- ing Paper 1. Kathmandu, Nepal: Rural Finance Nepal, 2001. ------. "Microfinance and Armed Conflicts in Nepal: The Adverse Effects of the Insurgency on the Small Farmer Cooperatives, Ltd. (SCFLs)." Working Paper 3. Kathmandu, Nepal: Rural Finance Nepal, 2003. Wenner, M. "Making Rural Finance Work." Microenterprise Development Review 3 (2). ------. "Lessons Learned in Rural Finance: The Experience of the Inter-American Development Bank." Technical Paper. Sustainable Development Department. Washington, DC: Inter-American Development Bank, 2002. Wenner, M., and F. Proenza. "Rural Finance in Latin America and the Caribbean: Challenges and Opportunities." Working Paper. Wash- ington, DC: Inter-American Development Bank, 2000. Wenner, M., and R. Quiros. "An Agricultural Credit Card Innovation: The Case of Financiera Trisan." Sustainable Development Depart- ment. Washington DC: IDB, 2000. Wenner, M., J. Alvarado, and F. Galarza, eds. "Promising Practices in Rural Finance: Experiences from Latin America and the Caribbean." Academia de Centroamerica, Centro Peruano de Estudios Sociales. Lima, Peru: Inter-American Development Bank, 2003. Westley, G. "Equipment Leasing and Lending: A Guide for Microfinance." Best Practice Series. Sustainable Development Department. Washington, DC: Inter-American Development Bank, 2003. 50 86696_01-52.qxd 8/31/05 9:23 AM Page 51 Bibliography continued Whelan, Steve, and Echange LLC.. "Automated Teller Machine Technology." IT Innovations in Microfinance Series. Washington, DC: CGAP, 2003. World Bank. World Development Indicators 2001. Washington, DC: World Bank, 2001. World Bank. "Workers' Remittances: An Importance and Stable Source of External Development Finance." In Global Development Fi- nance: Striving for Stability in Development Finance. Washington, DC: World Bank, 2003. ------."Reaching the Rural Poor: A Renewed Strategy for Rural Development." Washington, DC: World Bank, 2003. ------.Rural Finance Services: Implementing the Bank's Strategy to Reach the Rural Poor. Washington, DC: World Bank, 2003. ------.Agriculture Investment Sourcebook. Washington, DC: World Bank, 2004. ------. "Madagascar: Microleasing for Agricultural Production." Washington, DC: World Bank, 2004. Wright, G. Microfinance Systems: Designing Quality Financial Services for the Poor. Dhaka: Bangladesh: University Press, 2000. Yaron, J. "Rural Finance: Issues, Design, and Best Practices." Environmentally and Socially Sustainable Development (ESSD) Studies and Monographs Series 14. Washington, DC: World Bank, 1998. 51 86696_01-52.qxd 8/31/05 9:23 AM Page 52 Occasional Paper No. 11 Please feel free to share this Occasional Paper with your colleagues or request extra copies of this paper or others in this series. CGAP welcomes your comments on this paper. All CGAP publications are available on the CGAP web site at www.cgap.org. CGAP 1818 H Street, NW MSN Q4-400 Washington, DC 20433 Tel: 1-202-473-9594 Fax: 1-202-522-3744 E-mail: cgap@worldbank.org Web: www.cgap.org