MICROBANKING BULLETIN - Microfinance Information Exchange

THEMICROBANKINGBULLETINFocus: Reaching the PoorA CALMEADOW PUBLICATIONISSUE NO. 5SEPTEMBER 2000A SEMI-ANNUAL PUBLICATION DEDICATED TO THE FINANCIAL PERFORMANCE OFORGANIZATIONS THAT PROVIDE BANKING SERVICES FOR THE POOR

The MicroBanking Standards ProjectThe MicroBanking Bulletin is one of the principaloutputs of the MicroBanking Standards project,which is funded by the Consultative Group to Assistthe Poorest (CGAP) and managed by CALMEADOW.Project PurposeBy collecting financial and portfolio data providedvoluntarily by leading microfinance institutions(MFIs), organizing the data by peer groups, andreporting this information, this project is buildinginfrastructure that is critical to the development ofthe industry. The primary purpose of this databaseis to help MFI managers and board membersunderstand their performance in comparison withother MFIs. Secondary objectives includeestablishing industry performance standards,enhancing the transparency of financial reporting,and improving the performance of microfinanceinstitutions.In return, we prepare a confidential financialperformance report for each participating institution.These reports, which are the primary output of thisproject, explain the adjustments we made to thedata, and compare the institution’s performance toits peer group as well as to the whole sample ofproject participants. These reports are essentialtools for MFI managers and board members tobenchmark their institution’s performance.The third core service is to work with national andregional associations of microfinance institutions toenhance their ability to collect and manageperformance indicators. This service is provided in avariety of different ways, including teaching thesenetworks to collect, adjust and report data at thelocal level, collecting data on behalf of a network,and providing customized data analysis to comparemember institutions to external peer groups. Thisservice to networks allows us to help a wider rangeof MFIs to improve their financial reporting.Project ServicesTo achieve these objectives, the MicroBankingStandards project provides three services: 1)customized financial performance reports; 2) theMicroBanking Bulletin; and 3) network services.MFIs participate in this project on a quid pro quobasis. They provide us with information about theirfinancial and portfolio performance, as well asdetails regarding accounting practices, subsidies,and the structure of their liabilities. ParticipatingMFIs submit substantiating documentation, such asaudited financial statements, annual reports,program appraisals, and other materials that help usunderstand their operations. With this information,we apply adjustments for inflation, subsidies andloan loss provisioning to create comparable results.We do not independently verify the information.Neither CALMEADOW nor CGAP can acceptresponsibility for the validity of the informationpresented or consequences resulting from its use bythird parties.New ParticipantsOrganizations that wish to participate in theMicroBanking Standards project, either to receivecustomized reports or network services, shouldcontact CALMEADOW's Washington office: emailmicrobanking@calmeadowdc.com, Tel (202) 347-0039, Fax (202) 347-0078. Currently, the onlycriterion for participation is the ability to fulfill fairlyonerous reporting requirements. We reserve theright to establish minimum performance criteria forparticipation in the Bulletin.Bulletin SubmissionsThe Bulletin welcomes submissions of articles andcommentaries, particularly regarding analytical workon the financial performance of microfinanceinstitutions. Submissions may include reviews orsummaries of more extensive work elsewhere.Articles should not exceed 2,500 words. We alsoencourage readers to submit responses to thecontent of this and previous issues of the Bulletin.The MicroBanking Bulletin can be downloaded from CALMEADOW's website: www.calmeadow.com, and it isavailable in hard copy from PACT Publications—Email: books@pactpub.org, Website: www.pactpub.com,Tel: (212) 697-6222, Fax (212) 692-9748.

CONTENTSFrom the Chair..................................................................................................................................1FEATURE ARTICLESReassessing the Financial Viability of Village Banking: Past Performance and Future Prospects.....3Gary WollerA Business Model for Going Down Market: Combining Village Banking and Credit Unions ..............9Kathleen Stack and Didier ThysCOMMENTARYServing the Poorest Sustainably.....................................................................................................13An Interview with David Gibbons, CASHPORSix out of Seven Ain’t Bad (Credit Unions, Continued)....................................................................17Elisabeth RhyneCASE STUDIESIn Their Own Words: FINCA, Uganda.............................................................................................19Michael McCordIn Their Own Words: Pro Mujer Bolivia ...........................................................................................21Nancy NatilsonBulletin Case Study: BURO, Tangail, Bangladesh ..........................................................................23Geetha NagarajanBulletin Case Study: BASIX, India ..................................................................................................27Geetha NagarajanBULLETIN HIGHLIGHTS AND TABLESBulletin Highlights...........................................................................................................................29Craig ChurchillAn Introduction to the Peer Groups and Tables ..............................................................................36Index of Ratios and Tables .............................................................................................................39Peer Group Tables (Tables 1-4) .....................................................................................................40Additional Analysis Tables (Tables A-B) .........................................................................................47APPENDICESAppendix I: Notes to Statistical Section...........................................................................................54Appendix II: Description of Participating MFIs.................................................................................58

From the ChairThe 5 th issue of The MicroBanking Bulletin focuseson microfinance programs that reach poorer clients.We are delighted to present a number ofcontributions about MFIs fulfilling their dualobjectives: to serve relatively poorer clientssustainably. Their task is undoubtedly a difficultone. In many countries they work in rural areas.Many of their clients have a very limited capacity toabsorb increased loan amounts. Often their clientsregard higher rates of interest with great suspicion,complicating the institution’s path to sustainability.The decision to seek out a particularly poor or hardto reach market has fundamental implications forthe financial structure of MFIs. These programs,referred to in the Bulletin as Low-end MFIs, aredifferent than organizations that serve a broadermarket. Comparisons between institutions need totake into account the markets that they are serving.Unfortunately, we have not yet found whollysatisfying indicators that can be easily gathered,unambiguously interpreted, and readily comparedacross regions to measure the relative poverty of anMFI’s clients.To define our peer groups, the Bulletin uses twoproxies for client poverty: 1) Average LoanBalance (ALB) (total outstanding portfolio / numberof active borrowers) and 2) Depth (average loanbalance as a percent of GNP per capita). 1 While farfrom perfect, especially in fast growing programsthat utilize a strongly incremental approach tomicrolending, these indicators are among the bestavailable in expressing something about absoluteand relative poverty. While we would not want toinfer much about the difference between anoutstanding balance of US US$200 and US$300,there is an important difference between anaverage loan balance US$50 and US$500.An important challenge facing the microfinanceindustry is to come up with more satisfyingindicators of depth of outreach. We invite readersto send us suggestions for other variables we mightconsider in defining our peer groups that capturethe spirit of maximizing outreach.In our first Feature Article, “Reassessing theFinancial Viability of Village Banking …”, GaryWoller examines the prejudice that village bankingcannot be financially viable. In cooperation with the1 The definitions for all of the indicators presented in TheMicroBanking Bulletin can be found on page 39.SEEP Poverty Lending Working Group, heconsiders evidence from nine leading villagebanking institutions that participate in TheMicroBanking Bulletin and agreed to make theirdata public. He shows that village bankingprograms can perform as well as MFIs that useother lending techniques. Woller also finds thatdifferent programs achieve their overall resultsthrough distinct combinations of productivity andincome variables. He concludes with a call todevelop standards by which village bankingprograms can be compared to each other on theirown terms. To that end, the Bulletin provides acomparison of performance by lending methodology(see Tables A and B).Kathleen Stack and Didier Thys present theexperience of Freedom From Hunger, an NGO thatsupports programs that combine microcredit withlow-cost, high-impact education sessions innutrition, health and better business. Its innovativeapproach seeks to reduce delivery costs bypartnering with credit unions in a number ofcountries. This approach allows the very poor to beserved through an institution that already has amarket presence, which reduces delivery costs andenhances financial performance. This articleillustrates a creative way to break old paradigms inthe quest to fulfill the dual objectives of outreachand sustainability.In the Commentary section, we are pleased topresent an interview with David Gibbons ofCASHPOR, who discusses the network’s evolutionover the past five years. David highlights some ofthe unique features of working with poverty focusedprograms in Asia: the fact that the poor tend to workas agricultural laborers, the general resistance ofpolicy makers to allow MFIs to charge ‘high enough’interest rates, the regulatory barriers to savingsmobilization, and access to funding that would allowmore programs to scale up. In anotherCommentary, Beth Rhyne takes issue with DaveRichardson over the potential of MFIs to engage ineffective “self-governance”. Dave’s article on creditunions appeared in the last issue of the Bulletin.This issue includes four Case Studies. GeethaNagarajan, of the Bulletin’s Editorial Staff, preparedcases of two Asian MFIs. We also have two casestudies written by persons close to the institutions,which we have titled “In Their Own Words” toindicate this distinction.MICROBANKING BULLETIN, SEPTEMBER 2000 1

FINCA Uganda is one of the most matureprograms supported by FINCA international. Ina fascinating narrative, Mike McCord highlightsthe decisions that affected the evolution ofFINCA Uganda’s financial performance over thepast five years and its search to meet its dualobjectives.(1999 ALB: US$60; Depth: 19.4 percent) Pro Mujer in Bolivia started as an NGO thatbrought women into economic activity for thefirst time through a training program. Over timeit added a credit component. Nancy Natilsondescribes the challenges of operating in ancompetitive environment, yet maintaining anunwavering commitment to serving the poor.(1999 ALB: US$116; Depth: 11.5 percent) BURO Tangail, in Bangladesh, has made astrong attempt to incorporate deposit servicesfor poor clients. Geetha Nagarajan highlightsBURO’s fight to improve sustainability whileserving its relatively poor clients (92 percent liveon less than US$1 per day). BURO facesparticular challenges related to the additionalexpense of offering deposit services.(1999 ALB: US$57; Depth: 16.0 percent) Although it serves a heterogeneous clientele,BASIX extends its services into rural India.Nagarajan points to the unique challengesfaced by a private organization that seeks tocompete with subsidized credit to the poor.BASIX has had to face a reduced yield on itsportfolio brought on by increased delinquencyand higher costs associated with its ruralexpansion.(1999 ALB: US$208; Depth: 56.0 percent)In the Bulletin Highlights, Craig Churchill analyzesthe characteristics of financially sustainable MFIs.He finds that there is NOT a strong relationshipbetween loan balances and self-sufficiency, andthat some of the most profitable MFIs serve thepoorest clients. He then focuses on how low-endMFIs achieve financial self-sufficiency and identifiesimportant regional differences in their strategies.The Highlights section also includes an initial tasteof longitudinal analysis—how the performance ofMFIs changes over time.Programs targeting relatively poorer clients reachsustainability through a variety of means. We try toillustrate these distinct paths by creating a relativelylarge number of peer groups. In this issue, weadded one new peer group, dividing EasternEurope into High-end and Broad. Otherwise, thepeer groups remained relatively stable.As usual, the back of the Bulletin contains statisticaltables showing results both for peer groups that wehave created, and for several other indicators: ageof program, scale of operations, lendingmethodology, degree of financial intermediation,and target market.This Bulletin includes data from 114 MFIs in 46countries, an increase of ten institutions from IssueNo. 4. In each issue, new programs are added,while others drop out. Over half of the participatingprograms have updated their information within thepast year. We only include programs that havecontributed data in the past two years. Thisrepresents tighter criteria from previous issues ofthe Bulletin that included information from threeyears. So, while we have gained 17 new programsfor this issue, we lost 7 MFIs that did not updatetheir data recently.These regular changes in the composition of eachpeer group complicate the interpretation of datafrom one Bulletin to the next. Readers should becautioned NOT to draw inferences about theevolution of indicators across issues of the Bulletin.To address this issue, the next MicroBankingBulletin will include some information that showsthe evolution of key indicators for institutions fromwhich we have a series of data over a number ofyears. Obviously, these represent a smaller subsetof the entire group of programs that are currentlysending us data, but this longitudinal analysis willstill provide important insights into how MFIsdevelop over time.Finally, I would like to recognize the tremendousefforts of Jennifer McDonald who has recently leftthe Bulletin staff to work for a microfinanceinstitution in Africa. Jennifer is known to many ofyou as the heart and soul of the Bulletin during itsfirst three years. Her contributions are far toonumerous to list. Suffice it to say that herdedication, attention to detail, and commitment toquality set a standard that her successors mustwork hard to maintain. Thank you Jennifer.We hope that you will enjoy this issue and welcomeany suggestions for ways in which we can make theBulletin more relevant to your work.Robert Christen2 MICROBANKING BULLETIN, SEPTEMBER 2000

FEATURE ARTICLESCompartamosCompartamos’ management believes that the only way to make a significant dent in poverty is to achievemassive scale, and the best way to achieve massive scale is to be profitable. As a result, Compartamos hasrapidly expanded from 17,500 borrowers and a loan portfolio of US$552,000 in 1995 to 48,835 borrowers in1999 with nearly US$6.5 million outstanding. Compartamos has funded much of its growth through its ownequity, which it has built largely by charging high interest rates for its loans. The fact that Compartamosoperates in a monopoly environment has allowed it to pursue this strategy.At the same time, however, Compartamos has a high administrative cost structure that is in part a function of itssmall average loan size. Still, management believes that its administrative cost structure is too high, and hasimplemented a number of initiatives to bring costs in line, to raise staff productivity, and to reach out to newmarkets. These initiatives include a new management information system to track portfolio quality and a newsolidarity group loan product in urban areas. In conjunction with its new solidarity group loan product, it hasdesigned a paperless loan processing system. Compartamos’ management believes that as a result of theseand other initiatives, it will be able to drive its administrative costs down to 45 percent of its average loanportfolio. In 1999, Compartamos became a regulated financial institution.Efficiency comparisons made across lendingmethodologies (and target markets) often imply alack of understanding of these straightforwardprinciples. For example, Todd Farrington, in theFebruary 2000 Issue of the Bulletin, argues thatefficiency is a function of loan or portfolio size. 5 Theevidence for this assertion was the disparity in theadministrative expense ratios between MFIstargeting the low-end clients with those targetingthe middle to upper end of the market. The truth,however, is that a VBI with an average loan balanceof US$85 can be just as if not more efficient than anILI with an average loan balance of US$1,300. Thekey is whether the VBI is delivering its US$85 loansin a cost-effective manner given its availableresources, regardless of what the ILI is doing. 6How efficient are our nine VBIs relative to the fullsample of VBIs and relative to each other? Relativeto the full sample of VBIs, our nine institutions areless efficient on all four indicators. Compared toeach other, they are characterized by their lack ofconsistency—efficient in some areas and inefficientin others. Only Kafo Jiginew and Pro Mujer rateconsistently high, while only FINCA Kyrgyzstanrates consistently low. If one were to ask whetherour nine VBIs were “efficient,” the appropriateresponse would be, “According to which indicator?”The Return on PortfolioThe return on portfolio refers to an MFI’s financialreturn on its average lending portfolio. The returnon portfolio can be measured using at least two5 In his article “Efficiency in Microfinance Institutions,” Farringtonwrote that, “below a certain portfolio size it is difficult for an MFIto be efficient” (p. 18), and that “in Latin America, institutions withaverage loans below US$200 have difficulty achievingacceptable efficiency levels” (p. 19).6 Or, put another way, the key is whether the VBI is producing onor near its “production possibility frontier.”indicators: the portfolio yield and the interestspread. The portfolio yield, a proxy for the effectiveinterest rate, is equal to total interest incomedivided by the average loan portfolio. The interestspread is the difference between the portfolio yieldand the administrative expense ratio. It tells us theextent to which an MFI is pricing its products tocover its administrative costs.As seen in Figure 4, VBIs have earned a portfolioyield moderately higher than SGIs and substantiallyhigher than ILIs. Faced with a high cost structurerelative to the other two lending methodologies,VBIs appear to compensate by squeezing higherlevels of productivity out of their employees—thereby driving down their average cost perborrower—and by charging high interest rates.Nonetheless, because ILIs have a loweradministrative cost structure, they can charge lowerinterest rates and still earn a substantial spreadover costs, while SGIs earned negative spreads.Due to a high portfolio yield, VBIs have earned aspread, though small, over their higher costs.In contrast, our nine VBIs have earned a portfolioyield far in excess of those earned by the otherthree groups and an interest spread slightly lessthan the ILIs but considerably larger than the fullsample of VBIs and SGIs. Compared to the fullsample of VBIs, our nine VBIs appear to havecharged even higher interest rates, both in absoluteterms and relative to their administrative costs.This apparently has allowed them to achieve muchhigher levels of operational and financial selfsufficiencycompared to the other VBIs, despitehaving slightly higher administrative cost structures.6 MICROBANKING BULLETIN, SEPTEMBER 2000

FEATURE ARTICLESFigure 4: Comparison of the Return on Portfolioacross Lending MethodologiesPortfolioYield(%)InterestSpread(%)Lending MethodologyIndividual 36.5 15.6Solidarity Group 41.2 -4.5Village Banking 54.5 5.4Nine VBIs 70.0 14.9AGAPE 61.6 9.3Compartamos 111.4 48.8FINCA Kyrgyzstan a 107.7 18.8FINCA Nicaragua 75.1 18.3FINCA Uganda 84.4 -1.0CRECER 42.0 5.6Kafo Jiginew (VB product) 32.7 -0.2Pro Mujer Bolivia 39.4 1.6World Relief Honduras 75.3 33.3Source: The MicroBanking Bulletin, September 2000.a1998 figures.Factors Driving Relative Self-sufficiencyamong Our Nine VBIsSome insight into the factors driving the selfsufficiencyof our nine VBIs can be found in Figure5, which lists the nine institutions from the highestto the lowest on financial self-sufficiency and ranksthem on nine performance indicators thought to bedeterminants of financial self-sufficiency.Scanning across the rows in Figure 5, no patternemerges. None of the institutions performeduniformly well or poorly. While somewhere theremight exist a self-sufficient VBI that combines lowcosts, high productivity, and a high return onportfolio, it does not exist among our nine VBIs.Additional insight to this question can be gained byrunning a series of bivariate correlations betweenfinancial self-sufficiency and the same nineindicators. As seen in the bottom row of Figure 5,three indicators have large and statisticallysignificant correlation coefficients with financial selfsufficiency:portfolio yield, interest spread, and thenumber of borrowers. Of course, these are onlysimple correlations and do not imply causation;nonetheless, the strength of the correlations relativeto those of the other indicators suggests arelationship with financial self-sufficiency that isrelatively more robust. The conclusion, based onthis small, handpicked sample, is that for financialself-sufficiency, both the interest rate and scaleappear to matter most.Implications for Best PracticesThe (not surprising) implication for best practices isthat to achieve financial self-sufficiency, VBIsshould charge high interest rates at an “adequate”spread over costs and scale up. More surprisingare the weak correlations between financial selfsufficiencyand the other indicators in Figure 5. Allnine have taken different paths toward selfsufficiency,although each (with the exception ofKafo Jiginew) does appear to have compensatedfor relatively high administrative cost structures bycharging high interest rates.Although interest rate policies appear to have beenintegral to their success, it is necessary to questionthe long-term viability of this strategy. VBIs cancharge high rates because there is an excessdemand for loans. When competition and thesupply of loans increase, the equilibrium marketprice will fall. (It is probably no coincidence that theportfolio yields for Pro Mujer and CRECER areamong the lowest of the nine institutions. Bolivia isone of the more competitive microfinance marketsin the world.) Moreover, consumer preferences andother determinants of market demand change overtime. Therefore, it is probably not wise to base aninstitution’s long-term viability on the assumptionthat it can indefinitely charge monopolistic-typeinterest rates—although it is perhaps an effectivestrategy in the short to medium term.This returns us to consideration of other factorswhen discussing best practices, particularly that ofinstitutional efficiency. Despite the low correlationsof efficiency variables with financial self-sufficiencyamong our nine VBIs, policies and innovations(such as increased use of information systems andother technology) that drive down costs, increaseproductivity, and enhance the attractiveness ofproducts and services will be more important thaninterest rates over the long-run in determining aVBI’s financial viability. If the history of thecommercial banking industry is any indication, VBIshave barely scratched the surface in these areas.Improvements in efficiency will also free VBIs tocharge lower interest rates and still maintain anappropriate spread over costs (as the ILIs appear tohave done).This last point is especially important for povertylenders concerned about both depth and breadth ofoutreach. Charging very high interest rates mayreduce the demand for loans among the very poorwhose enterprises do not yield a rate of returnexceeding the interest rate (although perhaps stillyielding a moderate to high rate of return) andamong those segments of the low-income selfemployedwho have lower-cost borrowingalternatives. Once freed to charge lower interestrates, VBIs can reach downward, outward, andupward to all segments of their target markets.MICROBANKING BULLETIN, SEPTEMBER 2000 7

FEATURE ARTICLESFigure 5: Comparison of Factors Driving Self-sufficiency among Nine Village Banking InstitutionsInstitutionFinancial Selfsufficiency(%)Portfolio Yield(%)Interest Spread (%)Administrative Expense/ Average Loan Portfolio(%)Salary Expense /Average Loan Portfolio(%)Cost per Borrower (US$)Staff Productivity(no.)Depth(%)Borrowers(000’s)Salary Structure(multiple of GNP/ capita)1. Compartamos 143.7 111.4 (1) 48.8 (1) 62.6 (7) 36.1 (5) 63 (8) 207 (3) 3.4 (7) 48.8 (1) 1.9 (1)2. FINCA Kyrgyzstan a 108.0 107.7 (2) 18.2 (4) 89.5 (9) 57.2 (9) 67 (9) 78 (9) 17.7 (3) 9.9 (8) 5.4 (7)3. World ReliefHonduras107.1 75.3 (4) 33.3 (2) 42.0 (4) 30.9 (4) 35 (2) 134 (8) 11.5 (5) 18.7 (4) 4.2 (4)4. Pro Mujer Bolivia 100.1 39.4 (8) 1.6 (7) 37.8 (3) 21.0 (3) 47 (4) 180 (4) 11.5 (5) 18.9 (3) 5.1 (6)5. FINCA Nicaragua 99.0 75.1(5) 18.3 (3) 56.8 (6) 39.0 (6) 42 (3) 169 (6) 18.1 (2) 13.7 (6) 11.0 (8)6. CRECER 92.3 42.0 (7) 5.6 (6) 36.4 (2) 17.4 (2) 48 (5) 168 (7) 16.4 (4) 14.6 (5) 4.3 (5)7. AGAPE 88.5 61.6 (6) 9.3 (5) 52.3 (5) 39.7 (7) 57 (7) 175 (5) 3.2 (8) 4.9 (9) 2.2 (3)8. FINCA Uganda 87.7 84.4 (3) -1.0 (9) 85.4 (8) 47.8 (8) 49 (6) 221 (2) 19.4 (1) 20.8 (2) 16.1 (9)9. Kafo Jiginew(VB product)72.0 32.7 (9) -0.2 (8) 32.9 (1) 14.0 (1) 16 (1) 242 (1) 6.5 (6) 11.1 (7) 2.0 (2)Correlation Coefficient -- .73* .89* .28 .31 .63 -.27 -.20 .81* -.20Source: The MicroBanking Bulletin, September 2000. Note: Figures in parentheses denote rank order by factor.* Statistically significant at 5 percent level.a 1998 figures.Performance StandardsAfter so many years and so much talk about bestpractices, we are still a ways from defining bestpractices for village banking institutions. Theredoes appear to be broad consensus that bestpractice, at a minimum, translates into operationaland financial self-sufficiency. Beyond that,however, we can offer little more than broadrecommendations to keep costs low, improveefficiency, charge an adequate spread over costs,and scale up. It is argued here that we have beeninhibited in our search by a conception of bestpractices that pays too little attention to crucialdifferences in institutional characteristics and that,as a result, attempts to apply standards that are notalways relevant.What is needed is to identify a set of peer groupclassifications along a continuum of institutionaltypes and then develop performance standardsrelevant for each peer group, in addition toperformance standards that apply more or lessequally to all MFIs. Ideally this process will beaccompanied by a complementary process ofqualitative learning to provide necessary contextbehind the numbers. Used in isolation, theusefulness of performance indicators for informingbest practices among village banking institutions islimited, but used in conjunction with peer groupsand with the contextual information, their usefulnesscan be significant. The work of the Bulletin inidentifying peer groups and providing contextualinformation behind the numbers marks a significantstep in moving this process forward.All of the above being said, what insights do thenine VBIs examined for this article give us aboutspecific performance standards for village bankinginstitutions? This question is hard to answer giventhe limitations of the data set. Nonetheless, thesense from the data is that those VBIs that canpush administrative expenses significantly below 40percent of average loan portfolio, push the cost perborrower down into the US$30s, increase staffproductivity beyond 200, and earn an interestspread in the 15 to 20 percent range (andparticularly those who achieve a combination ofthese) will be expanding the frontiers of bestpractices in village banking.Gary Woller is an associate professor at the RomneyInstitute of Public Management in the Marriott School atBrigham Young University. During the last year, he wason sabbatical at FINCA International in Washington DC,where he served as acting Research Director. Gary canbe reached at gmwoller@hotmail.com. He thanksSEEP’s Poverty Lending Working Group for theirassistance with this article, and the nine village bankinginstitution for their willingness to share their performancedata.8 MICROBANKING BULLETIN, SEPTEMBER 2000

FEATURE ARTICLESA Business Model for Going Down Market:Combining Village Banking and Credit UnionsKathleen Stack and Didier ThysFew microfinance institutions consider going downmarket (serving poorer clients) as a good businessopportunity. Instead, they broaden services toreach a better-off clientele, or they say financialviability must be compromised to reach the poorest,or they reason that microfinance cannot help thevery poor, who need different types of socialservices. Even those who argue that breadth anddepth of outreach are compatible (with a fewnotable exceptions) have yet to demonstrate thatlarge numbers of the truly poor can be reachedviably.But what if there was a strategy that could reachvast numbers of the poorest and make goodbusiness sense? This paper presents a businessmodel for profitably serving the poorest segments ofthe microfinance market, along with someencouraging preliminary results.Recent Trends in Reaching the PoorestThe trend in the microfinance industry is towardcommercialization to ensure long-termsustainability. Unfortunately, this trend is coupledwith movement away from serving the very poor.This is because, in the search for financial viability,institutions are going up market. They have either:1) increased average loan sizes by growing withtheir customers; and/or 2) developed new, largeloanproducts for new markets.Using average loan size as a proxy for depth ofoutreach, microfinance success stories, such asBRI and BancoSol, are not serving particularly poorclients. The average loan size of each of theseinstitutions is more than US$500. In fact,BancoSol’s average outstanding balance has beencreeping up over the years so that as of 1998, itstood at US$914. 7Some might say that these large, financiallysustainable institutions are still reaching asignificant number of poor borrowers. But a studyof five Bolivian MFIs found that this is not so. 8 Only3 percent of BancoSol’s borrowers—900 people—7 The data are from a recent unpublished study on cooperativesand microfinance, Jeffrey Ashe et al. USAID, 1998.8 Navajas, Sergio, Mark Schreiner, Richard L. Meyer, ClaudioGonzalez-Vega and Jorge Rodriquez-Meza, “Microcredit and thePoorest of the Poor: Theory and Evidence from Bolivia", WorldDevelopment 28(2): 333-346 (1999).were considered to be among the poorest andindigent, and the total number of very poor reachedby all five institutions in the study was 2,600.The StrategyFreedom from Hunger, a US-based microfinancesupport organization, designed Credit withEducation to serve poorer segments of the marketsustainably. This integrated financial andeducational product is intended to be financiallyviable while reducing the causes of chronic hungerand malnutrition in rural households. It combines:1) group-based lending and savings services(village banking) for poor women with 2) low-cost,high-impact education sessions in nutrition, healthand better business.To achieve large-scale outreach with a financialproduct designed for very poor people, Freedomfrom Hunger has partnered with credit unions inseveral countries. We decided to collaborate withcredit unions for the following reasons: Outreach: Credit unions supply more lendingand savings services than any other type offinancial institution with the exception of banks.They are widespread, particularly in rural areas,and often have regional and national creditunion networks that can promote efficientproduct dissemination. Mission and Ownership: Credit unions werecreated to improve the welfare of their membersand communities through financial services. Asmember-owned institutions, credit unionsmaintain a commitment to both social andfinancial goals. Financing: They have the capacity to financenew products from internally generatedresources (savings). The savings-firstapproach of the credit unions is an attractivelong-term, self-financing feature. Multiple Products: Credit unions offer multipletypes of loan products. This product menumitigates against the risk of portfolioconcentration that occurs with MFIs that offeronly one or two types of loans. Also, as a creditunion client, a poor member can graduate froma group-based product to other credit unionproducts when and as the need arises.MICROBANKING BULLETIN, SEPTEMBER 2000 9

FEATURE ARTICLES Economies of Scope: Credit unions are in agood position to deliver a poverty-focusedproduct sustainably because back office andoverhead costs can be shared with otherfinancial services.To implement Credit with Education, credit unionsagree to finance the loan portfolio with membersavings. In exchange, Freedom from Hungerfinances the technical assistance and start-upcosts. Teams composed of field agents and asupervisor, employed either by the credit unions ora federation, form retail units to deliver Credit withEducation. All services are provided by the samefield agent to joint liability groups of clients in theircommunities. Loan cycles are 16 to 24 weeks withweekly, bi-weekly or monthly installment paymentsof loan principal, interest and savings to a groupsavings account at the credit union.From Freedom from Hunger’s perspective, thisstrategy is beginning to bear fruit. There arecurrently eight individual credit unions and six creditunion federations offering the Credit with Educationproduct. Credit union clients now represent 62percent of Freedom from Hunger’s total outreach,or over 85,000 women, with an average outstandingloan per borrower of US$75.Recent studies demonstrate the achievement ofboth depth of outreach and impact. In a povertyassessment carried out in two credit unionfederations in Mali, wealth ranking showed that 28percent and 57 percent of Credit with Educationclients fell into the bottom two wealth quartilesdescribed as food insecure—persons experiencingeither chronic or seasonal periods of hunger. 9 Thisindicates that, in Mali in 1998, Credit with Educationwas reaching 10,461 very poor credit union clients.Three-year impact studies carried out in Ghana andBolivia provide evidence that the children of Creditwith Education clients enjoyed significantlyimproved nutrition compared with a control groupwhen quality education accompanied financialservices. Improvements in women’s income, healthand nutrition practices, and empowerment werealso demonstrated. 109 Preliminary results of wealth ranking exercise conducted inMarch 2000 by Freedom from Hunger.10 “Impact of Credit With Education on Mothers’ and Their YoungChildren’s Nutrition: Lower Pra Rural Bank Credit with EducationProgram in Ghana,” Barbara MkNelly and Christopher Dunford.Freedom from Hunger Research Paper No. 4, (March 1998); and“Impact of Credit with Education on Mothers and their YoungChildren's Nutrition: CRECER Credit with Education Program inBolivia", Barbara MkNelly and Christopher Dunford. Freedomfrom Hunger Research Paper No. 5, (December 1999).From the perspective of cost effectiveness, it ismuch less expensive to work with credit unionpartners than to create new institutions or work withnascent MFIs. Freedom from Hunger spent US$6.4million in direct grants and technical assistance tocreate a capacity for reaching 30,000 womenthrough two specialized MFIs—or US$211 perborrower. Yet it only cost US$700,000 to create acapacity for reaching 36,000 women with two creditunion federations—or US$20 per borrower. Thecredit unions also reached this level of outreach inhalf the time and with a greater level of financialself-sufficiency than the specialized MFIs.But while the rationale for Freedom from Hunger towork with credit unions is strong, why would creditunion managers adopt a poverty-focused productgiven their need to run profitable organizations?The social reasons certainly play a role in thedecision-making process, but Freedom fromHunger could not make much headway if the socialinclination was not also accompanied by a solidbusiness rationale for going down market.The Poor Are Good BusinessBased on Freedom from Hunger’s earlyexperiences with the credit union/Credit withEducation model, there are numerous businessreasons for credit unions to offer a poverty-focusedloan product.Credit with Education allows credit unions topenetrate new markets. Most credit unionsrequire clients to come to a primary service centerto conduct their transactions, while Credit withEducation is delivered through a mobile bankingsystem. This decentralized model not only offersthe opportunity to provide services to poor clients inisolated communities, but it also helps the creditunion to test if there is sufficient demand to openoffices in those areas. Conversely, in regionswhere a credit union is struggling to sustain primaryservice centers, mobile banking systems can be alow-cost alternative.Risk mitigation and cash flow management havebeen two strong business reasons for adoptingCredit with Education. Rural credit unions, like KafoJiginew in the southern Mali cotton belt, primarilylend for agricultural purposes. Concentration ofportfolio in seasonal lending activities carries a highrisk and results in cash flow shortages. Credit withEducation, which represents 10 percent of KafoJiginew’s portfolio, has helped alleviate theseconstraints. The very poor take short-term workingcapital loans for commercial activities. Short loancycles and timely repayments placed regularly intocredit union accounts promote cash flow smoothing.10 MICROBANKING BULLETIN, SEPTEMBER 2000

FEATURE ARTICLESBy redirecting financial services to the rural poor,credit unions can improve utilization andcirculation of idle assets. In some regions, creditunions have a limited capability to transformsavings into loans and operating income. This isdue to the structural nature of their other loanproducts. Without a poverty-focused product, theyare oriented to markets with limited demand andoften carry high transaction costs for the borrower.When Freedom from Hunger began working withthe Réseau des Caisses Populaires in BurkinaFaso in 1993, the loan-to-savings ratio, or thesavings transformation rate, was 22 percent. In1998, the transformation rate was 150 percent. 11This vast improvement was partly due to Credit withEducation, which offered a product that fit thedemands of the rapidly growing market niche of therural poor. It allowed credit unions to transformsurplus urban savings into loans for a rural market.These new clients generate regular income for thecredit unions through interest payments on theseredirected savings while achieving the socialmission of the credit union to serve the community.Working with the very poor provides opportunities toachieve important efficiencies. In many markets,there is still limited, if any, competition for ruralmicrofinance providers. This lack of competitionpartly explains the willingness of the very poor toorganize groups of 25 or 30 members to obtainservices. Efficiency in terms of the ratio ofborrowers to field agent is high, averaging 285 forcredit unions in West Africa and the Philippines.The administrative expense per borrower for KafoJiginew’s Credit with Education product was justUS$16 in 1999.Staff costs can also be lowered while maintainingthe same quality of services by hiring literatewomen from the community. Group-based povertylending requires staff with the capacity primarily tobuild trust and strong relationships, as well asrecord and track simple financial transactions. In arecent staff assessment of Kafo Jiginew’s Creditwith Education product, women field agentsselected from the groups' management committeesperformed as well or better than field agents fromthe city with a high school education. The averagestaff cost per borrower in the village promotersystem is US$5.96 compared with US$8.20 for onethat uses high school graduates. 1211 “Study Design to Assess the Institutional Impacts of Credit withEducation on Credit Unions in Mali, West Africa,” Freedom fromHunger, 1998.12 “Evaluation du système de dotation du personnel de la ligne duproduit du Crédit avec Education de Kafo Jiginew,” Freedomfrom Hunger, May 2000.Extension of financial and educational servicesdeep into rural communities promotes credit unionmembership growth and diversification. Womenlearn to use financial services successfully andmany “graduate” to individual membership byopening their own savings accounts and takinglarger loans. In Burkina Faso and Mali, some groupmembers have been elected to boards of creditunions and/or hired as field staff, strengthening theparticipation of women in the credit union’smanagement. The percentage of women membersat Kafo Jiginew has increased from 1 percent to 20percent since Credit with Education was introducedfour years ago.The most important business rationale for servingthe very poor may be portfolio quality. Manycredit unions have a history of poorly performingloans and high delinquency rates. This is often dueto the business risk associated with the loan, thecredit risk associated with the client, or thebreakdown of selection and collection practicesassociated with the service delivery system. Creditwith Education provides a reliable screening anddelivery system to serve a low-risk market. Fear ofsocial castigation and strong rural communityinterdependence make poor clients terrific loanrepayers.The Case of the PhilippinesThis business approach to deepening outreach isexemplified by the preliminary accomplishments ofcredit unions in the Philippines that now offer Creditwith Education thanks to the Credit UnionEmpowering and Strengthening (CUES) Project ofthe World Council of Credit Unions (WOCCU).The first "Savings and Credit with Education"(SCWE) loans, as they are known in thePhilippines, were made in August 1998. From 1998through the first quarter of 2000, the program grewfrom four cooperatives and less than 2,000 clientsto eight cooperatives and over 13,000 clients. TheSCWE product line for six of the eight cooperativeswas operationally self-sufficient by that time.In terms of product share, SCWE is now one of sixor seven loan products offered by these eightcooperatives. At the end of 1999, SCWEaccounted for 33 percent of the total clients and 8percent of their outstanding portfolio.WOCCU's strong business-minded approach forworking with financial cooperatives set the contextwithin which Credit with Education was integrated.By implementing a more accurate measurement forportfolio quality, for example, the CUES programhelped to market the SCWE product. CUESMICROBANKING BULLETIN, SEPTEMBER 2000 11

FEATURE ARTICLEScreated an environment in which credit unionmanagers were forced to confront the poor qualityof their loan portfolios and the heavy costs ofprovisioning for the accumulated bad debt. Theyrealized that at least one loan product, SCWE, hada portfolio at risk of zero percent. This product wasalso being marketed to new members who were nottainted by the previous loan practices of theinstitution and with whom a more disciplined set ofexpectations could be developed. This led the firstfour credit union managers to expand their supportfor SCWE and brought in a second "batch" ofcooperatives whose managers cited lowdelinquency as the primary reason for their interest.Figure 1 shows selected performance indicators forthe eight cooperatives that offer SCWE, the fourthat do not offer the product, as well as data for all12 cooperatives. 13 A comparison between thosecooperatives that do and do not offer SCWE (for1998 and 1999) generally indicates that the povertyproduct is not a drag on overall financialperformance. As the product matures, it will likelyhave an increasingly positive effect.Figure 1: Profitability and Costs: CUESCooperatives in the PhilippinesOperating Self-sufficiency (%)1996 1997 1998 19991. Cooperatives w/ SCWE (8) 109 108 116 1262. Cooperatives w/o SCWE (4) 111 112 110 1193. All CUES Cooperatives (12) 110 110 113 122Portfolio Yield (%)1. Cooperatives w/ SCWE (8) 25.3 25.5 25.3 30.42. Cooperatives w/o SCWE (4) 26.8 30.4 28.9 34.13. All CUES Cooperatives (12) 25.8 27.1 26.5 31.6Return on Assets (%)1. Cooperatives w/ SCWE (8) 2.4 3.4 3.3 4.42. Cooperatives w/o SCWE (4) 2.3 2.6 2.5 4.13. All CUES Cooperatives (12) 2.4 3.2 3.0 4.3Operating Expenses / Average Total Assets (%)1. Cooperatives w/ SCWE (8) 10.6 10.7 9.1 9.72. Cooperatives w/o SCWE (4) 8.2 8.6 8.0 9.83. All CUES Cooperatives (12) 9.8 10.0 8.7 9.7Source: Freedom from Hunger.For the eight cooperatives using SCWE, operationalself-sufficiency improved in both years when theproduct was in place. The cooperatives usingSCWE had a lower portfolio yield than the averagetrend for all twelve cooperatives. The gap betweenthe yield for cooperatives using SCWE andcooperatives not using SCWE increased sharplyfrom 1996 to 1997, but decreased in 1998 andstabilized in 1999 after the introduction of SCWE.This gap indicates that although the non-SCWEcooperatives earned a greater return on theirportfolio to begin with and made more significantgains in improving their rate of return until 1997,that trend does not seem to hold following theintroduction of the new product.The return on assets ratio shows that thecooperatives using SCWE generally earned a betterreturn than the non-SCWE cooperatives. Thecooperatives using SCWE also lowered theiroperating expense ratio after introducing thepoverty product.All of these general improvements are the directresult of the improved overall managementperformance of the cooperatives, which wassubstantially supported by the management andtechnical assistance provided by CUES. What isinteresting to note about this data is that thoseinstitutions integrating a loan product for poorwomen were by no means disadvantaged or sloweddown in improving and growing their operationsrelative to those that did not.Specifically promoting deep outreach through acredit-led, group-based lending product did notimpede performance. It did not inhibit cooperativesfrom growing in terms of assets or savings. It didnot increase external dependency on borrowedfunds. It did not slow down the growth in the valueof outstanding loans. It did not increasedelinquency. It did not decrease operational selfsufficiency.In fact, it added an additionaldimension to the cooperatives by opening up a newand very large market. There are ample reasons tobelieve extending financial services to the very poormay not be bad for credit union business. With timeand larger scale efforts, it might even turn out to bepretty good business!Kathleen Stack is Senior Vice President for Research andInnovation at Freedom from Hunger. She is one of theco-creators of the Credit with Education methodology anddeveloped the strategy for linking Credit with Education tocredit unions. She can be contacted atkstack@freefromhunger.org.Didier Thys is Vice-President for Practitioner Services atFreedom from Hunger. Practitioner Services is atechnical assistance and training unit specialized inteaching MFIs how to deliver Credit with Educationservices. Further information is available fromdthys@freefromhunger.org.13 Note that the first SCWE loans were actually made in August of1998. The years 1996 and 1997 reflect no SCWE-relatedactivities.12 MICROBANKING BULLETIN, SEPTEMBER 2000

COMMENTARYCOMMENTARYServing the Poorest, SustainablyAn Interview with David Gibbons, CASHPORAbout CASHPORMicroBanking Bulletin (MBB): How didCASHPOR get started, what is its mission, andwhat services does it provide?David Gibbons (DG): CASHPOR was establishedby the Grameen Bank (GB) and 6 of its replicatorsin September 1991 as a vehicle to shareexperiences and to learn from each other. Over theyears, two main activities have evolved: 1) provisionof management training, surprise audits andtechnical assistance to its members (which nowinclude 20 MFIs in 9 countries in Asia); and 2)establishing GB-type start-ups in places where theyare very much needed and nobody has come forthlocally. Some examples include CFTS Ltd. inMirzapur, India and Project Naroman, which is juststarting in East Timor.CASHPOR wants to achieve a significant reductionof poverty throughout Asia by providing financialservices to poor households. Our mission is toprovide financial services to large numbers of poorwomen throughout Asia in a timely, honest, efficientand financially sustainable manner. As a result of adecision of the board, CASHPOR is currentlyconcentrating on the Philippines and India, but wecontinue to assist members in 7 other countries.CASHPOR no longer takes in direct members,unless there is no national network of GB-typeMFIs. National networks have been promoted inPhilippines, India and Nepal. GB-type MFIs in thosecountries join through their national network, whichCASHPOR assists directly.MBB: What is the relationship between CASHPORand the Grameen Bank?DG: Grameen Bank is a founding member ofCASHPOR. Periodically we call upon them toprovide experienced resource persons to assistother members in need. All members have agreednot to charge professional fees when called upon toassist another CASHPOR member.All CASHPOR members have adapted the GBapproach to their own economic, social, politicaland cultural contexts. We share the same vision asthe Grameen Bank, and like GB we deal exclusivelywith poor households. But in our operations, wetend to give more emphasis to attaining andmaintaining financial sustainability and to reachingthe poorest households. I think Prof. Yunus wouldagree that Grameen was reluctant to charge ahigher interest rate than the commercial banks inBangladesh; and that Grameen doesn’t reachenough of the poorest rural households.It is interesting to see some of the adaptations thatare emerging from CASHPOR members. One ofour MFIs in India, for example, was experiencingportfolio quality problems so it developed aninterest rebate. Each client who repaid on time foran entire quarter received a Rs. 10 refund, and ifthey paid on time for four quarters in a row, theyreceived an additional Rs. 10, or Rs. 50 for theyear. I was amazed at how well that worked inimproving the repayment rate.MBB: For the best CASHPOR members, whatfactors have contributed to their success?DG: The keys to their success have been: 1)capable, honest, visionary leadership; 2) highpriority for increasing their institutional capacity andkeenness to adopt promising new microfinancemanagement tools; 3) enough funding to continueto grow; and 4) CASHPOR’s guidance andassistance. Conversely, our weaker members lackgood leadership and/or are not sufficientlycommercially oriented (see Figure 1).MBB: What are the major changes that have takenplace to CASHPOR’s approach to microfinanceduring the past five years?DG: The most important change was recognitionthat we could not realize our vision unless we gavemore importance to attaining and maintainingfinancial self-sufficiency. To do this we had toimprove our financial management, reporting andanalysis.Most of the CASHPOR CEOs came from an NGObackground with a strong concern for social andeconomic development. Only later did they realizethat, for microfinance to be done well, it has to bedone with a commercial orientation, which meantthat we had to upgrade the financial managementcapacity of our members. CGAP deserves a lot ofcredit for bringing this to our attention.MICROBANKING BULLETIN, SEPTEMBER 2000 13

COMMENTARYFigure 1: Selected Performance Indicators from the CASHPOR Network (Dec. 1999)OrganizationCountryActiveBorrowersOutstandingPortfolio(US$)Admin. Expenses /Average LoanPortfolio (%)OperationalSelf-sufficiency(%)Portfolio atRisk(%)AIM * Malaysia 39,310 20,669,149 n.a. n.a. 1.2GB Biratnagar ** Nepal 39,048 3,086,544 n.a. n.a. 1.7ARDPAS China 35,546 2,503,630 n.a. n.a. 4.1SHARE India 29,490 2,609,845 27.5 98.4 0.0CARD Philippines 28,531 3,740,792 30.1 103.0 0.0CSD Nepal 26,817 1,285,234 21.5 69.6 0.1Nirdhan Nepal 21,644 1,371,559 29.7 69.3 0.6Dungganon Philippines 20,316 1,411,372 46.9 101.8 1.4TSPI (Kabuhayan) Philippines 9,694 886,045 n.a. n.a. 3.6GB Dhangadhi * Nepal 10,798 1,342,696 n.a. n.a. n.a.TYM Vietnam 10,058 823,939 18.8 100.8 0.1ASA India 6,738 286,631 45.3 60.9 0.5FPC* China 7,816 704,029 n.a. n.a. 0.6CEP Fund Vietnam 6,701 468,794 n.a. n.a. 3.2ASHI Philippines 6,627 468,263 57.6 48.1 3.3YUM * Malaysia 5,864 715,265 n.a. n.a. 20.8MKEJ * Indonesia 5,762 55,347 n.a. n.a. 0.4CFTS India 4,006 111,660 n.a. n.a. 10.8Nirdhan WB India 3,006 120,466 n.a. n.a. 24.0BSS India 98 3,157 n.a. n.a. 95.0Source: CASHPOR.Portfolio at Risk is amount of loans outstanding with at least one payment overdue above 90 days.* AIM figures are for end of May, 1999; ** GB, Biratnagar figures are for end of June, 1999.We also recognized that we needed to improve thetimeliness and quality of our reporting and increasethe computerization of our MIS. This change wascritical to improve the quality of our ratio analysis.Finally we have realized the need to improve ourbusiness planning and we have adopted CGAP’sMicrofin software for this purpose.Microfinance in AsiaMBB: How is microfinance in Asia different fromother parts of the world?DG: For the most part, Asian MFIs (excluding BRIand a few others like it) focus on using microfinanceas a tool for poverty reduction. I believe that in otherregions, mainly Latin America, microfinance isfocused mainly on promoting microenterprises. Tothe degree that microenterprises are not run by thepoor, microfinance in those countries would notdirectly reach the poor and the poorest households.Our focus on the poor and the poorest (bottom halfof the poor) results in smaller loan sizes, lowerinterest rates, better loan portfolio quality, anddifferent loan activities.This last issue is difficult for people who haven’tworked in Asia to understand. In using USAID’sAIMS tools for measuring impact, for example, wehave had to extensively customize them for theAsian context. The tools basically excludeagricultural activities from the definition of amicroenterprise; yet among CASHPOR members,the majority of first loans are used for animalhusbandry. In rural Asia, the primary route out ofpoverty is through agriculture.The nature of poverty is different in Asia. Ourclientele is fundamentally different from themicroentrepreneurs served by most Latin AmericanMFIs. An overwhelming proportion of our firstborrowers are landless agricultural laborers. Theylive well below the poverty line. Over half of whatlittle household income they have comes fromwages earned by working on someone else’s farm.To escape poverty, they need to work forthemselves and gradually increase the amount ofland that they operate as small farms. First loansare often for animal husbandry because borrowerswant to choose an income generating activity thatallows them to keep their other jobs. Then overtime, the household can get access to agriculturalland (through a combination of shareholding, leaseand purchase), reduce its reliance on income fromagricultural labor, and come out of poverty.MBB: What are the challenges facing MFIs in Asiain their efforts to create sustainable institutions?14 MICROBANKING BULLETIN, SEPTEMBER 2000

COMMENTARYDG: The most important challenge is developing afinancing strategy that would allow for optimumgrowth of outreach to the poor. None of ourmembers have been able to attract sufficientfunding to enable them to grow at optimum levels—and I don’t just mean grant funding. Certainly thereis a need for subsidized funds during the start-upphase, but once MFIs develop sufficient capacity,then they can move toward soft loans, and theneventually to loans at commercial rates.Ultimately, the key constraint will be the MFI’sability to access equity funding, since that willdetermine how much institutions can leverage.Now that several CASHPOR members haveachieved operational self-sufficiency, we arebeginning to help them identify equity investors,both locally and internationally. While the expectedreturn on investments in MFIs may not be sufficientto attract venture capitalists, well-run institutionscan generate levels of performance that will attractsocially responsible investors.There is also a crucial need to create anappropriate regulatory environment in Asia, whichwill allow for the commercialization of microfinance.In this regard, the biggest issue is the minimumcapital requirement to create regulated financialinstitutions, which is too stiff for NGOs to meet. Weare making some headway on improvingmicrofinance regulations in India, but it is neededelsewhere, especially China and Vietnam, whereCASHPOR members could really take off if theregulatory environment was more accommodating.But it isn’t all about money. There is also the needfor continual upgrading of management to have thecapacity to absorb increasing amounts of funds.Many Asian MFIs need help improving otheressential systems as well, like computerized MISand planning with Microfin, so as to givemanagement the tools they need.MBB: The Asian peer groups in the Bulletin aredramatically more efficient than their counterparts inLatin America. How do Asian MFIs achieve suchimpressive efficiency ratios?DG: The efficiency of Asian MFIs is based on theirsignificantly lower personnel costs and significantlyhigher loan portfolio quality. One of the reasons forthe higher repayment rates is precisely becausethey are serving a poorer market. While it mayseem counterintuitive that poorer people are betterrisks, our experience has generally been that if youreach down far enough, you’ll find a market that willbecome extremely faithful with its repayments. Thepoorest clients value the services more than theirbetter-off peers, and they are extremely keen tomaintain access to their line of credit. As shown inFigure 1, most CASHPOR members have excellentloan portfolio quality problems. But when they doexperience problems, it is usually with their moreaffluent clients.On a related note, MFIs in Asia also have a lowercost of funds due to the widespread availability ofsubsidized funds. I am not placing a valuejudgment on the availability of subsidized money; itis just a reality in some Asian countries. Over time,I am afraid that it will distort the market. In themeantime, it is necessary to take a pragmaticapproach. As long as these sources are available,such as 6.5 percent loans from NABARD in Indiawhen the market rate is 13 to 14 percent, we wouldbe silly not take advantage of them. But we alsoneed to prepare ourselves for the day when thesesources dry up.MBB: Another Bulletin observation is that AsianMFIs are generally less profitable than their LatinAmerican counterparts because of low portfolioyields. What are the barriers to charging higherinterest rates in Asia?DG: The barriers to charging appropriate interestrates are mainly political and exist primarily in SouthAsia, China and Vietnam. The increasingimportance of populist politics in these countriesmakes governments want to say they are limitinginterest rates to the poor. The result of course is alimiting of their access to microfinance. Some ofour MFIs are making headway, however, inshowing to their governments that significantnumbers of rural poor women are prepared to payappropriate interest rates in order to get continuingaccess to microfinance. I feel we need a couplemore years to demonstrate this with sufficientnumbers; but I am confident the governments willbe convinced by that time. The Microcredit Summitcampaign is playing a vital role here.Sustainability and the PoorestMBB: Why do you believe it is necessary to targetand exclusively serve the poor, instead of managinga diversified portfolio whereby larger loans cancross-subsidize the smallest loans?DG: If I can reach and maintain institutionalfinancial sustainability within a reasonable time (say5 years) while reaching significant numbers of thepoorest, the people with the greatest need for thisopportunity, why should I waste my time and otherresources on the non-poor? I have seen thishappen with MFIs in the Philippines that havemultiple product lines for different segments of themarket. When push comes to shove, managementMICROBANKING BULLETIN, SEPTEMBER 2000 15

COMMENTARYfocuses its attention on the bigger loans because ifthey start going bad, the institution is in trouble.Secondly, I am not aware of any cross-subsidizingMFIs that have reached large numbers, say morethan 50,000, of the poor, not to mention the pooresthouseholds. The cross-subsidization argumentoften appears to be a ruse.MBB: Is there is a trade-off between serving thepoorest clients and achieving self-sufficiency? Howdeep can an MFI go and also be sustainable?DG: To the bottom, but attainment of institutionalfinancial sustainability will take longer - say up to 5years. When I say the bottom, I am referringlandless households in rural areas whose irregularmonthly income is well below the poverty line. Youcan tell these people are on the bottom because thewife regularly works as a paid laborer. Most womenwon’t participate in paid agricultural work unlessthey absolutely have to.MBB: What are the most appropriate ways ofattaining self-sufficiency for programs that aretargeted at the very poor?DG: There are four issues that spring to mind.First, MFIs need to maximize outreach to the poorand poorest so that economies of scale can beenjoyed to the fullest. Second, they should offerlarger subsequent loans based on good loanperformance. Third, they need to fine-tune theirfinancial products to meet the real needs of thepoor and poorest. And last, but certainly not least,they need to charge appropriate interest rates.MBB: Besides small loan sizes, are there specificconditions involved in serving the poorest thathamper the attainment of self-sufficiency?DG: Much more motivation work is required, as thepoorest do not believe that microfinance is for themand they are afraid of getting into debt. It takes timeto open their eyes to the opportunity beingprovided, and this increases total staff costs.Demonstration effect from neighboring poor andpoorest households is essential to convince manyothers to participate. Thus it takes more time forthem to form groups.Savings Products for the PoorMBB: How would you compare the relativedevelopment benefits of savings versus creditproducts?DG: MFIs working with the poorest have to walkequally on both legs to succeed. Unfortunately,regulatory constraints make it difficult to providesavings services in many countries. Many of ourmembers are mobilizing deposits illegally, so theycan’t let savings amounts reach levels that wouldattract too much attention.To overcome this problem, we are encouraging ourmembers to establish a regulated financialinstitution that can legally offer savings services.CARD Bank, SHARE, and Nirdhan Utthan Bankhave made this exciting leap. But the CASHPOR“transformation” model is different from theapproach elsewhere because the NGO still has animportant role to play. The NGO continues to beinvolved in group formation and after the first loanthe NGO hands over borrowers to the bank. Thiscontinued link is necessary because CASHPORmembers don’t want to lose their social orientation.The integrated NGO/bank approach is designed toprovide clients with a wider array of services, toincrease the institution’s outreach by leveragingequity capital, and to retain its commitment to thepoorest.MBB: Do you think MFIs can offer voluntarysavings while requiring forced savings as collateral?DG: This is a difficult question. My personalpreference—although I doubt many of myCASHPOR colleagues would agree with me—would be to do away with forced savings andemphasize voluntary savings. But it isn’t that easy.In the Grameen methodology, collectiveresponsibility is very important. And voluntarysavings and collective responsibility don’t mix verywell.In the group, everyone knows how much each othersaves. So someone who is having repaymentproblems will cast a covetous eye toward thevoluntary savings of another—or the loan officerwho is striving for perfect repayments mayencourage this to occur. This possibility woulddiscourage voluntary savings from taking place. Atthe same time, collective responsibility is a keyelement in reaching the poorest. Because of theirirregular income flows, the only way the poorestclients feel comfortable accepting the risks of takinga loan is through collective responsibility. CARD isexperimenting with confidential, individual, voluntarysavings transactions immediately after their Centermeetings. Let’s see if it works.David Gibbons (gibbons@pc.jaring.my) is the ExecutiveTrustee of CASHPOR Inc., the Asian regional network ofGrameen Bank-type microfinance programs for the poor.He is also the Executive Chairman of CFTS in India, anexperimental GB-type fast-track microfinance program inMirzapur District. In addition, he is the author of books,training manuals, and articles on sustainablemicrofinance for the poor and poorest.16 MICROBANKING BULLETIN, SEPTEMBER 2000

COMMENTARYSix out of Seven Ain’t Bad (Credit Unions, Continued)Elisabeth RhyneYa gotta love Dave Richardson’s in-your-face articlein the last MicroBanking Bulletin (Issue No. 4). Yagotta love the way those credit union folks carry onlike they’re the ones who really inventedmicrofinance. Trouble is, in many ways, they did.Although I hesitate to endorse anything as religioussounding as doctrine, Dave’s “Seven Doctrines ofSuccess” are as good a summary of the creditunion approach as you’ll find. And he’s right thatthe microfinance “industry” has been very slow torecognize most of those points. It still hasn’t fullydigested some of them. The credit unions havesome strong messages for microfinance in severalof Dave’s doctrines, especially savings, efficiency,financial standards, and diversification of clientele.Microfinance practitioners should find ways to adoptsome of the credit unions’ approaches in theseareas, otherwise their institutions may lack thestaying power to survive when conditions change orcompetition heats up.However, I would like to challenge one of the sevenvirtues, self-governance. Dave, who is normally ahard-nosed realist, becomes all soft and mushywhen it comes to this one. He’s right thatmicrofinance has no perfect solution to ownershipand governance. Each available model is subjectto its own set of weaknesses, and self-governanceis no exception. Self-governing financial societies, especiallysmaller ones, are subject to capture byinfluential or highly motivated members whodirect policies toward their interests. Borrowerdomination led credit unions across the world toadopt anti-saver policies for years, therebylimiting their growth. Capture by a small groupof borrowers has caused many a credit unionsimply to collapse from what in another type offinancial institution would be regarded asinsider lending. Self-governance among multiple smallinstitutions is inefficient. While creatingindependent governing structures at thegrassroots level does seem to be a usefulapproach for developing financial services inareas banks can’t reach, it has turned out to bedifficult, labor intensive, and costly to bring lotsof tiny institutions to the point where they canbe reliable suppliers of quality services. Thereseem to be economies of scale formanagement structures: one managerialstructure with many outlets may be moreefficient than many separate superstructures. When credit unions grow to become majorsuppliers, the essence of self-governancebegins to disappear, making credit unions littledifferent from other types of financialinstitutions. For the clients of large-scale creditunions, being a member involves little morethan paying an initial membership contributionand checking an annual ballot. I suspect thatself-governance figures far below convenience,price, and products in the decision of mostclients on whether to bank with a credit union oranother institution. For large institutions,competent and professional managementmatters more than the particular type ofownership and governance.For these reasons, I’d prefer to think of selfgovernanceas one option in the overallmicrofinance toolkit, appropriate under certaincircumstances, but not the only solution.Why, after all, should there be a special linkbetween self-governance and financial services?As societies modernize, corporate formsincreasingly replace collective and participatoryforms. All the worse for the 21 st century, perhaps.But the move away from such forms is caused bymarket forces driving toward efficiency, and theseare very strong forces, indeed. The defenders ofparticipatory forms of action need to choose theirbattles carefully, searching for areas wherecollective participation and self-governance areworth going to the barricades for. The credit unionsin their fervor may believe that financial services area prime location for this battle to be fought. Iremain agnostic.Former director of USAID’s Office of Microenterprise,Elisabeth Rhyne is now an independent consultant inMozambique and a member of the Editorial Board of theMicroBanking Bulletin. She can be reached aterhyne@worldbank.org.MICROBANKING BULLETIN, SEPTEMBER 2000 17

CASE STUDIESCASESTUDIESIn Their Own Words: FINCA UgandaMichael J. McCordFrom as early as 1995, when FINCA Uganda onlyhad 1,200 clients, there was no question that itsobjective was to become a profitable microfinanceinstitution serving the women’s micro-businessmarket. The only questions were when and howprofitable? The two keys to achieving this goalwere always seen as 1) volume and 2) cost control(including efficiency).VolumeRecognizing that volume was a key, and thatwithout capacity there could be no volume—at leastnot with the requisite high portfolio quality—muchwork was spent on building staff and creating astrong foundation for growth.As seen in Figure 1, FINCA Uganda experiencedbetter than 100 percent growth in the number ofborrowers each year from 1996 through 1998. Thenumber of staff grew at a slower rate than thenumber of clients, which improved productivity.However, the loan portfolio also grew more slowly,which suppressed income. The relatively slowportfolio growth was a direct result of the averageloan outstanding actually falling rather significantlyin 1998 and not fully recovering in 1999 (let alonegrowing). Thus, all the wheels were spinning, butthe progress towards self-sufficiency was slow.Figure 1: FINCA Uganda - Growth1995 1996 1997 1998 1999Borrowers (no.) 1,289 3,324 8,473 17,228 20,769Total Portfolio (‘000 US$) 107 210 539 850 1,246Assets (‘000 US$) 321 404 1,031 1,485 2,066Staff (no.) 14 25 61 98 94Avg. Loan Balance (US$) 83 63 64 49 60Source: The MicroBanking Bulletin.The limited increase in average loan balance wasthe result of several factors, but five in particular:1) Rapid growth meant that the portfolio consistedof a large portion of new clients with smallerloans.2) As much as 50 percent of clients chose loanssmaller than the maximum available to them dueto limited business activity.3) Relatively high dropout rates caused significantwork in replacing clients and further increasedthe percentage of new clients (and their smallloans) in the portfolio.4) We decided to maintain the initial loan size at the1995 level, while offering faster growth after thefirst cycle. However, faster growth opportunitieswere not implemented until May 1999.5) FINCA maintains a commitment to serving thepoor.Additionally, in the midst of increasing competitionin Ugandan microfinance, it was necessary to makeseveral adjustments to the loan product to maintainits marketability. For various reasons, theseadjustments were not made quickly enough, whichretarded growth for much of the 1998/99 fiscal year.Several long awaited improvements wereimplemented beginning in May 1999, which haveled to dramatic positive results for fiscal 1999/2000(ends 31 st August). Product improvements includea group and client rating system, a significantlylower savings requirement (allowing client loans to“grow” faster), and other flexibilities that haveimproved market acceptance.In addition, Microcare, a health care financingscheme, was accessed to provide a mechanism forclients to buy a full range of health care servicesthrough hospitals. This scheme provides FINCAUganda with a competitive advantage among MFIs,by offering clients this risk management product tocomplement existing services such as voluntarysavings and group accident policy (with deathcoverage for the client and her family members).Given that any new product in the microfinanceindustry provides exclusivity for only six to twelvemonths, FINCA Uganda is actively working on itsnext round of adjustments and innovations.EfficiencyFINCA Uganda was slow to address efficiencyissues in the midst of its rapid growth (see Figure2). It was difficult to balance the need for efficiencywith the volumes of new (less productive)employees required for growth, especially whenmuch of that growth was into new areas.MICROBANKING BULLETIN, SEPTEMBER 2000 19

CASE STUDIESUntil mid-1997, FINCA Uganda primarily followed astrategy of geographically concentric growth. Thisinvolved moving consistently further from the centerin a controlled manner with the intention ofreasonable market saturation, rather than jumpingfrom city to city. This approach was reasonablyefficient and showed positive results.Figure 2: FINCA Uganda - Efficiency1995 1996 1997 1998 1999Total Admin. Expenses /Avg. Loan Portfolio (%) 92.5 68.1 71.9 81.6 85.4Salary Expenses / Avg.Loan Portfolio (%) 44.3 27.2 39.0 49.0 47.8Average Salary(multiple of GNP/ capita) 10.5 5.6 7.2 9.5 17.2Staff Productivity (no.) 92 133 139 176 221Source: The MicroBanking Bulletin.To comply with donor requirements for geographicgrowth, in 1997/98 the concentric strategy wasshelved and two branch offices were opened inoutlying areas with limited growth potential. Thestartup of a new branch is expensive, and althoughthe objective was to break even in 18 months,growth was slower than projected, extending thebreakeven period and resulting in reduced overallefficiency for the institution.Additionally, during this time (1997/98) new seniorlevel staff were brought on to manage currentgrowth and to prepare for future growth. This extralayer of management also added to the“inefficiencies”, though it was intended to preparebetter for the future.We also experienced significant wage inflation in aneffort to retain staff in an increasingly competitivemarket. This environment pushed, and continues topush, salaries in the Ugandan microfinanceindustry. This situation is unlikely to change,though FINCA Uganda has introduced incentiveschemes to create a better relationship betweenearnings of staff and earnings of the company.Productivity, as measured by borrowers per staffmember, has significantly improved throughreduced reporting requirements, expanded fieldtime, and higher productivity targets for field staff.Initially this effort was slow because the systemsand procedures needed adjustment to assist fieldstaff in reducing their time with their groups.Among the company’s annual objectives for fiscal1999/2000 were a reduction of total administrativeexpenses to portfolio to below 50 percent, andincrease borrowers per field staff to at least 400.To help accomplish these targets, the CreditDirector was provided a significant incentive, paidand reassessed on a quarterly basis, for achievingincrements of these objectives, while maintaining atop quality portfolio.FINCA Uganda is also trying to improve its backoffice efficiency. Until May 1999, the informationsystem consisted of part manual, part Excelspreadsheet accounting, loan tracking and reportingsystem. It was replaced by an electronic integratedaccounting, loan tracking, and reporting system inearly 2000. Once stability is reached with the newsystem, it is likely to improve efficiency significantly.Figure 3: FINCA Uganda – Profitability, Yieldand Sustainability1995 1996 1997 1998 1999AROA (%) -20.2 -10.4 -13.9 -11.4 -7.0AROE (%) -21.8 -10.9 -14.5 -11.6 -8.1Portfolio Yield (%) 53.1 58.5 53.5 68.9 84.4Real Yield (%) 41.0 47.9 43.4 60.8 73.3Operating Self-sufficiency (%) 55 84 73 82 95Financial Self-sufficiency (%) 51 71 67 77 88Source: The MicroBanking Bulletin.An additional mechanism to improve structuralefficiency (while improving yield) was implementedin May 1999. Previously, loans carried a 3 percentper month nominal rate, with a 1 percent affiliationfee, a fee for stationery, and a voluntary 1 percentfee for a group accident policy. These fees,collected at disbursement meetings, createdsignificant front and back office inefficiencies.Starting in May 1999, these fees were rolled into anew nominal interest rate of 4 percent per month.This reduced transactional inefficiencies, andprovided a small addition to earnings, which wasnot fully experienced until October 1999, althoughone can see partial results in Figure 3.Clearly it has taken too long for FINCA Uganda tosettle into consistent sustainability, both operationaland financial. However, the company is nowpositioned with a solid foundation—through capablestaff and management, a more flexible product, andan entrepreneurial executive—to reach and surpassthat elusive goal. 1997 and 1998 were importantinvestment and learning years. Figure 3 showssignificant improvement in 1999, with a hiddenspringboard for the achievement of full financialself-sufficiency in the very near future.Michael McCord was FINCA Uganda’s Chief Executivefrom 1995 to 2000, and FINCA’s Africa Regional Directorfrom 1998 through 2000. He is currently an independentconsultant acting as Chief Technical Advisor forMicroSave-Africa, and working with CGAP’s MFI ExternalAudit Capacity Building Project. He can be reached at(mmccord@cbu.edu, microinsurancecentre@hotmail.com).20 MICROBANKING BULLETIN, SEPTEMBER 2000

CASE STUDIESIn Their Own Words: Pro Mujer, BoliviaNancy Natilson1999 was a challenging year for many Bolivianmicrofinance institutions, including Pro Mujer. Thestellar performance of Pro Mujer Bolivia in 1998was not as stellar in 1999. Pro Mujer Boliviareached operational and financial sustainability of,respectively, 148 and 119 percent in 1998, provingto itself and others that village banking programscan be impressively sustainable. In 1999, however,these ratios were reduced to 110 and 100 percent,respectively—still sustainable, but only at abreakeven level (Figure 1).Figure 1: Sustainability and Profitability of ProMujer (1997-1999)1997 1998 1999AROA (%) -4.4 6.0 0.0AROE (%) -5.0 6.8 0.0Operating Self-sufficiency (%) 90 148 110Financial Self-sufficiency (%) 87 119 100Source: The MicroBanking Bulletin.What accounts for these increased pressures onPro Mujer’s bottom line, and how are weconfronting these challenges?Growth IssuesAlthough the number of borrowers has grownsteadily over the last couple of years (17 percent in1998 and 13 percent in 1999), our portfoliodecreased 6 percent in 1998 and remained flat in1999 (see Figure 2). This combination of trends isdue to a substantial reduction in average loan size.Reasons for this phenomenon are many, including:(a) economic crisis, (b) increased competition, (c)low client retention rates, and (d) continuedcommitment to serve the very poor.Figure 2: Outreach of Pro Mujer (1997-1999)1997 1998 1999Borrowers (no.) 14,226 16,669 18,919Total Portfolio (‘000US$) 2,336 2,200 2,197Staff (no.) 68 89 105Avg. Loan Balance (US$) 164 132 116Depth (%) 16.9 13.1 11.5Source: The MicroBanking Bulletin.Bolivia has been experiencing an economic crisissince 1998. In these conditions, the poorestsegments of the population – which is Pro Mujer’starget market – are the most vulnerable. While ourclients are struggling to maintain quality of life,increased debt is not necessarily the answer.Bolivia is one of the world’s most competitivemicrofinance markets with over-indebtedness fromoversupply. Traditional financial institutions thathave entered the microfinance sector are usingmethodologies unfamiliar to them and causingmarket distortions. Pro Mujer applauds the resultsof competition that have lowered interest rates andhave forced MFI’s to become more efficient. But,the other side of this competition story is increaseddelinquency. Pro Mujer has had some portfolioquality problems, but it is difficult to know how ourexperience compares to other Bolivian MFIs sincedata on write-offs and refinanced loans are notoften shared.Client retention remains a challenge. To this end,Pro Mujer analyzes client desertion and uses theinformation to make changes to communal bankmethodology and to pilot new products. Some ofthese improvements include increasing theefficiency of group meetings, raising the maximumloan size, and revising the saving requirements.We are also piloting individual loans designed toretain our clients whose credit needs have grownbeyond what communal bank loans can offer. Ourinitial experiences with this product are interesting.We were surprised to discover that many of ourexperienced clients who qualified for individualloans were reluctant to leave their communal banksand the required saving program, both of whichthey consider supportive and vital to their successas microentrepreneurs.Finally, Pro Mujer is proud of its commitment toserving the very poor. Reduction in average loansize bothers us only in that it negatively affects ourproductivity ratios and challenges our efforts to besustainable.Pro Mujer challenges the assumption among villagebanking practitioners that repeat clients willautomatically want larger loans. If theirmicroenterprises are not growing (which is often thecase), we do not encourage our clients to increasetheir loan size so that Pro Mujer’s institutionalsustainability can improve. Our target market isclients whose primary access to credit has beenexpensive moneylenders and who do not havecollateral to offer as a guarantee. The communalbank methodology offers them a good introductionto credit; we remain committed to providing pre-MICROBANKING BULLETIN, SEPTEMBER 2000 21

CASE STUDIEScredit training and quality service to this marketsegment. Therefore, we have only made ournewly-introduced individual loans available toclients who have progressed through our communalbanks, so as not to be tempted to go “up-market”and leave the very poor behind.Efficiency IssuesComparing Pro Mujer’s 1999 institutional efficiencyindicators with eight other leading village bankinginstitutions, we proudly rank on the edge of themost efficient third (see Woller’s article above). Butwe are constantly striving to improve our efficiencyby spending less and becoming more productive.Administrative costs as a percentage of averageloan portfolio rose from 1998 to 1999, as have costsper borrower. Notwithstanding the fact that ProMujer is an integrated program and hence ouradministrative costs are inflated (even though weattempt to segregate the direct costs of non-creditprograms from the costs of the credit program),efforts are being made to reduce costs at all levelsin order to compensate for the lack of portfoliogrowth. Also, costs were unusually high in 1999due to development of a new managementinformation system, whose benefits will be realizedin the years to come.Figure 3: Efficiency and Productivity for ProMujer (1997-1999)1997 1998 1999Total Administrative Costs /Avg. Loan Portfolio (%) 35.1 30.4 37.8Salary Expenses / Avg.Loan Portfolio (%) 22.5 19.2 21.0Cost per Borrower (US$)50 45 47Average Salary(multiple of GNP/ capita) 5.7 4.8 4.4Staff Productivity (no.)209 187 180Other Admin Costs / Avg.Loan Portfolio(%) 10.0 7.8 9.4Source: The MicroBanking Bulletin.Productivity, as measured by the ratio of borrowersto total staff, has declined as well, but productivityper loan officer shows a slight improvement in 1999over the previous year. The profitability marginsare being squeezed because the average loan sizeis smaller and costs are rising. Another challengewe face is the declining average number ofborrowers in each village bank, which is linked tothe retention rate mentioned earlier, as well as tocompetition and over-indebtedness.Strategic DirectionsAs Pro Mujer in Bolivia positions itself for the future,we have chosen a path that does not conform toindustry trends.We remain as committed to an integrated approach,combining microfinance with communal banktraining, business development services (bothtraining and technical assistance), health andhuman development training, and the provision ofbasic health services.We firmly believe that microfinance is only part ofthe solution to alleviating poverty; Pro Mujer’smission clearly addresses the economic and socialneeds of poor women. We are committed tosupporting focal centers in Bolivia, which havebecome community centers for our clients and theirfamilies. Although this approach incurs substantialcosts and deviates from the “traditional” villagebanking methodology of minimal infrastructure, itprovides a venue for invaluable training and healthservices.There are implications for organizational structure inour strategic decisions. Although we voluntarilysubmit our financial information to theSuperintendency of Banks in Bolivia, Pro Mujer hasdecided not to become a regulated financialinstitution because we believe that our mission isbest served by being a foundation. Nevertheless,we are investigating strategic alliances that wouldmake new funding options available to us.Finally, Pro Mujer’s sustainability must be achievedin harmony with our clients’ sustainability. Weintend to pass on efficiency savings to our clients,once we are sustainable on a more permanentbasis. Pro Mujer does not believe that institutionalsustainability is an end in itself, as sometimesseems the case when discussing best practices;instead, it is a means to enable our clients toachieve sustainability in their lives, so they can bemore fulfilled as individuals and as members of theirfamilies and communities.In summary, Pro Mujer Bolivia is experiencingincreased challenges to maintain and improve itssustainability. Some of the reasons are due toexternal factors beyond our control like competitionand oversupply of microcredit. But many of thereasons result from conscious decisions to serveour target market in ways we believe best meetsour clients’ needs.Nancy Natilson is Financial Advisor for ProMujer inBolivia. She also consults on financial management ofmicrofinance projects in developing countries. She can bereached at natilson@compuserve.com.22 MICROBANKING BULLETIN, SEPTEMBER 2000

CASE STUDIESBURO, Tangail, Bangladesh:Reaching the Poor with Savings and CreditGeetha NagarajanBangladesh Unemployed RehabilitationOrganization (BURO) in Tangail (BT) serves over67,000 low-end clients and provides them withdiverse types of credit and voluntary savingsproducts in a sustainable way. Since its inceptionin 1991, it has been a pioneer in offering depositservices. BT’s experiences are especiallyimpressive since several other MFIs in Bangladeshhave failed or scaled down deposit services due tohigh costs of operation.Through the critical input from external advisors andextensive feedback from customer focus groups,BT has developed six types of loans and threesavings products, summarized in Figure 1. Thedistinct features of BT’s products include flexibilityand variety, and unbundling of loans from savingsproducts. To be a member of BT, one does nothave to borrow; saving will suffice.OutreachSince 1995, BT has grown significantly in size,client coverage, and products offered, whilefocusing its attention at the low-end of the market.Until 1998, BT followed a horizontal growth strategyby moving into new areas to increase outreach. Inthe face of increasing competition, however, BTthen resorted to deepening its coverage of existingmarkets (Figure 2). New products were developedto meet the demand for financial services.The floods in 1998, indeed, put BT’s products,especially flexible savings to test. While severalcenters suffered from loan delinquencies, overallthe organization survived the disaster in fairly goodshape. In fact, after a brief period of decline, theaverage savings balance increased dramatically.Almost all of BT’s depositors save less than US$1 aweek, which suggests that BT is servinghouseholds with very limited disposable income.Figure 1: Summary of BT’s Savings and Credit ProductsSavingsGeneral SavingsContractualSavingsOne-time FixedDepositsLoansGeneral LoanSupplementaryLoanLine of CreditSanitation LoanBusiness orProject LoanDisaster LoanThis open access savings facility allows members to deposit more than the expected 10 taka (20 cents) everyweek and to access their savings any number of times without penalty. There is no ceiling on the amount ofdeposits or withdrawals. Annual interest rate is 7.5 percent.With these time deposits, members and associate members agree to save a fixed amount every week ormonth for a fixed period of time – 3, 5 or 10 years. The annual compound interest rates vary from 9 to 14%based on the length of the contract. Early withdrawals incur a penalty and failure to deposit threeconsecutive installments will result in transfer of funds to general savings.These are certificates of deposit for a large amount (minimum US$100) deposited at the beginning of thecontract period for a fixed term up to 5 years. These deposits earn 9 to 12 percent interest per annum, basedon the length of the contract.This product is available to all members and is repayable in fifty consecutive equal installments. The loandisbursed ranges from US$20 to US$1,200.These are available to general loan borrowers, six months after paying their dues, for an amount up to half ofthe original general loan.A line of credit is available to long-term clients with established businesses. The current ceiling is US$2,000for a period of two to three years, renewable each year.This is for building sanitary facilities or for drilling tube wells. The ceiling is US$50.This new product is offered to members who can use larger loans to create jobs in the community. Loansizes vary from US$2,000 to US$5,000 for a period of two years.This special loan facility is available during disasters for members who have lost their businesses and areunable to repay loan installments. Current ceiling is US$40 repayable in two years.All loans are charged an annual interest rate of 20 percent except the disaster loan at 5 percent. Except for line of credit andbusiness loans, a member can have up to two loans at a time. However, disaster loans are available to every member regardlessof the type and number of loans outstanding at the time of disaster. All are collateral free but group based loans.MICROBANKING BULLETIN, SEPTEMBER 2000 23

CASE STUDIESFigure 2: Outreach of BURO, Tangail1995 1996 1997 1998 1999No. of Branches 20 30 40 41 41No. of Staff 198 312 424 448 513No. of Borrowers(’000)11.8 12.7 25.7 55.7 62.4No. of Savers (‘000) 0 0 45.0 71.9 71.8Outstanding Portfolio(‘000 US$)Avg. Loan Balance(US$)504 707 1,446 3,606 3,33931 47 50 65 57Depth (%) 13 18 14 18 16Total SavingsBalance* (‘000 US$)Avg. SavingsBalance* (US$)0 0 295.5 562.8 885.50 0 6.6 7.8 12.3No. of Loan Products 3 4 5 6 6No. of SavingsProducts*1 1 2 3 3* These represent all deposits that have open access byallowing unlimited deposit and withdrawal.By offering flexibility, BT has increased averagesavings balances, although not at first. When BTtransitioned to completely liquid accounts in early1998, it experienced a heavy withdrawal of savingsas customers tested whether they really couldaccess their money. In addition, the worst floodsoccurred in mid-1998 increasing the deposits’outflow. After the initial frenzy of withdrawals,however, balances soon began to increase throughincreased volume of deposits and reduced volumeof withdrawals. The net growth in yearly depositsbetween 1998 and 1999 was 108 percent (seeFigure 3) and the average savings balanceincreased by more than 50 percent. This trend,though short, suggests a demand for, and depositorconfidence in, the savings instruments offered byBT.It is to be noted that while growth in the averagesavings balance was accompanied by an increasein number of depositors in 1997-98, it occurredthrough larger deposits in 1998-99. Besidesconducting an educational marketing campaign, BTalso increased deposits by creating a category of“associate members” who want to save and notaccess loans, and by opening a “savings only”branch in an urban area.Another indication that BT serves particularly poorpeople is the small loan balances maintained by theborrowers. In 1999, average loan balance relativeto GNP per capita was 16 percent. About 55percent of the active borrowers carried a loanbalance of less than US$100, accounting for 42percent of portfolio outstanding. Approximately 92percent of BT’s clients live below the poverty line(earning less than US$1 per day).Million Taka706050403020100Figure 3: Yearly Deposits andWithdrawals, BT 1995-991995 1996 1997 1998 1999YearsYr.depositsProfitability and SustainabilityYr.withdrawThe outreach strategy since 1996, comprised of amassive expansion amidst competition and newproduct launches, has impacted BT’s financialperformance.Figure 4: Profitability and Sustainability of BT,1995-1999, and Peers (in percent)OperatingIncomeOperatingExpenses1995 1996 1997 1998 1999Peers:SouthAsia27.0 31.9 23.1 25.5 23.4 14.936.1 30.3 33.1 27.2 25.6 20.7AROA -9.1 1.6 -9.9 -1.6 -2.2 -5.6Portfolio Yield 41.7 43.7 34.4 29.4 31.5 21.6Operating SelfsufficiencyFinancial Selfsufficiency82.5 110.1 72.0 103.8 114.8 92.574.8 105.2 69.7 94.0 91.6 70.7While operating expenses relative to average totalassets steadily declined from 36 percent in 1995 to26 percent in 1999, operating income relative toaverage total assets fell from 27 percent in 1995 to23 percent in 1999. This decline was causedprimarily by a drop in portfolio yield. In response tocompetition, BT lowered its annual interest rate in1997 on several loan products from 25 to 20percent. As a result, portfolio yield declined from 42percent in 1995 to 32 percent in 1999.BT also experienced an increase in delinquency inearly 1999 due to the 1998 floods. Portfolio at riskover 90 days was 2.2 percent in 1999 comparedwith less than 1 percent in previous years. Inresponse to portfolio quality problems, BT tightened24 MICROBANKING BULLETIN, SEPTEMBER 2000

CASE STUDIESlending requirements and consequently disbursedfewer loans in 1999 than in 1998. BT alsoimplemented a grading system for branches andcenters based on their loan repaymentperformance. Branches and centers with poorportfolio quality were penalized by reduced accessto future loans. A few centers were permanentlyclosed.Subsidized funding helped to finance BT’sexpansion. It constituted about 16 and 26 percentof loan portfolio in 1998 and 1999, respectively, andclose to 42 percent of the total liabilities in bothyears, pushing adjustment costs higher. As aresult, the FSS ratio in 1999 lagged behind that of1998, despite continued improvement in operatingself-sufficiency.EfficiencyOver the past five years, BT has increased itsnumber of clients by nearly six fold, and hasintroduced 3 new loan and 2 new savings products.Product experimentation combined with fast growthgenerally does not favor an increase in efficiency.Nonetheless, Figure 5 indicates that BT hasimproved its efficiency: administrative costs relativeto average loan portfolio declined from 47 percentin 1995 to 25 percent in 1999, while cost perborrower fell from US$25 to US$14.Figure 5: Efficiency of BURO, Tangail1995 1996 1997 1998 1999Administrative Expenses /Avg.Loan Portfolio (%)47.4 36.4 44.6 28.3 25.0Avg. Loan / GNP perCapita (%)13 18 14 18 16Staff Productivity (no.)a. Loans59 41 61 124 122b. Savings-- -- -- 30 72Avg. Salary (multiple ofGNP/ capita)1.6 1.7 1.8 2.2 3.1Cost per Borrower (US$) 24.7 17.8 29.6 17.4 14.3Efficiency was enhanced considerably by theincrease in staff productivity, despite the fact thatfield staff handle both loans and depositmobilization. Equipping field staff with motorbikesand bicycles improved staff productivity, especiallyin reaching clients in remote areas. To avoidmisuse of vehicles and to minimize maintenancecosts, BT made a vehicle loan to the staff andprovides a fixed allowance for upkeep.To improve efficiency, BT has also had to overcomean increase in its salary structure. Managementbelieved that higher wages were necessary toretain trained staff, as well as to attract qualifiedpersonnel to handle the specialized requirements ofobtaining savings. This strategy paid off throughincreased deposits, reduced employee turnover,and success in recruiting qualified personnel to staffthe expanding training, internal audit and MIS units.BT has undertaken several additional steps toimprove efficiency that have not paid off yet.Training expenses are being reduced by requiringnew field workers to be apprentices for two weeksin the field before beginning training at the headoffice. By moving forward the timing of theapprenticeship, BT hopes to cut the cost of trainingrecruits who then resign after exposure to the field.The organization recently introduced businessloans and time deposits that can cross-subsidizethe smaller loans and deposits. Despite BT’sthorough product development process, theseproducts have not yet had the desired effect. Thedemand for time deposits has been sluggish, sincefew clients could deposit large amounts for a longperiod of time. Only a few business loans havebeen issued because it is difficult to locate poorclients with good projects that merit large loans.To further improve efficiency, BT is experimentingwith increased branch automation throughcomputerization. This approach, however, is adouble-edged sword, since increased expensesdue to equipment costs and the expensiveemployees required to operate the computers mayoffset productivity gains. Inconsistent power supplyis another drawback.Lessons from BTBT’s experience shows that it is possible to servethe poor in a sustainable way by offering a varietyof credit and savings products. Challenges,however, remain for BT that may influence thefuture path of MFIs attempting to serve the poor.The BT case especially presents an interestingdilemma for MFIs planning on offering voluntarydeposit services.While demand for flexible savings products mayexist, it has been difficult to appropriately price thesavings products due to challenges in evaluatingtheir costs.Accounting for the time spent on mobilizing savingsby the field staff, who constituted about 75 percentof the total staff in 1999, the Bulletin estimates thatlending productivity would increase from 225 to 325clients per field staff if they were freed frommobilizing savings. This would bring BT’s total staffproductivity to 235 clients per employee. This levelof productivity may help to offset the costs ofMICROBANKING BULLETIN, SEPTEMBER 2000 25

CASE STUDIESproviding smaller loans. And by not offeringsavings, BT might be able to lower staff salaries.On the other hand, the estimated gains in staffproductivity on the credit side cannot be separatedfrom the costs of alternative ways of raising fundsfor lending. The increase in costs due to depositmobilization may indeed be compensated for by alower cost of funds for its loan operations. Savingsconstitute about 27 percent of the total loan portfolio(and 55 percent of the liabilities) at BT. If BT wereto mobilize these resources from commercialsources, then interest costs may increase by about3.2 percent, pushing operating costs higher by atleast another one percent.It is also important to mention that the maintenanceof liquid reserves to meet potential savingswithdrawals creates an additional cost, since itreduces the funds available for loans that cangenerate income. But the costs can likely be offsetif the average savings balance grows above thereserve requirement.It is also probable, although not yet documented,that BT enhances customer loyalty, and all theaccompanying cost savings, by offering openaccesssavings.Certainly this analysis is hypothetical and requires amuch more careful look to determine the costs andbenefits of providing voluntary savings. Themicrofinance industry currently lags behind in itsunderstanding of ways to evaluate the costs andbenefits of savings products. It is imperative for BT,and for the broader microfinance community, toexamine the unit costs of savings on a productbasis to evaluate gains to efficiency and profitability.Geetha Nagarajan is a member of the Bulletin’s editorialstaff. This article is based on her due diligence visit inFebruary 2000 and information submitted to the Bulletinby BURO, Tangail. The MicroBanking Bulletin thanks BTfor granting permission to publish its financial results.26 MICROBANKING BULLETIN, SEPTEMBER 2000

CASE STUDIESBASIX, India:Reaching Rural Clients with CreditGeetha NagarajanBASIX, a non-bank financial institution that startedin 1997, illustrates the challenges of a young MFI inpursuit of rural outreach in India where the formalfinancial sector actively provides subsidized credit.Outreach StrategyBASIX is not trying to serve the poorest of the poordirectly: loans less than US$100 represent only sixpercent of the outstanding portfolio. Its outreachstrategy is based on the assumption that largerloans to non-poor in rural areas will facilitate thecreation of employment for the poor. BASIX servesa broad target market of over 12,600 clients in ruralareas with an average loan balance of US$208,which is 55 percent of GNP per capita (see Figure1).Figure 1: Outreach of BASIX1997-98 1998-99 1999-00No. of branches 5 8 10No. of staff 36 74 79No. of active borrowers 1,127 9,044 12,626Loan portfoliooutstanding (‘000 US$)Average loan balance(US$)456 1,422 2,592404 157 208Depth (%) 109 42 56Loan outstanding foragriculture (% to total)-- 59 50BASIX serves a heterogeneous clientele whodemand a range of credit products, includingagriculture loans. To deliver services to this marketefficiently, BASIX recognized that diverse channelswere necessary. As a result, it has developed avariety of group and individual loan products for arange of purposes, as summarized in Figure 2.Loans for agricultural activities account for half ofthe portfolio, while non-farm loans account for 35percent.These products are delivered through a variety ofdifferent channels, including self-help groups,intermediaries such as trader organizations, agroprocessingfirms and NGOs, BASIX loan officersand customer service agents paid on commissions.Profitability and SustainabilityUsing commercial sources of funds (87 percent ofassets in March 2000), BASIX expanded rapidly inthree years, both in terms of the number of clientsand the menu of products. But it has paid the pricefor zealous growth in its financial performance.Figure 2: Loan Products Offered by BASIXProductCrop loans through selfhelpgroups (indirectgroup loans)Crop loans tointermediaries for onlending(indirectindividual loan)Agricultural investment(direct individual loan)Agri-allied activities(direct individual loan)AverageLoanDisbursed(US$)Farm LoansAnnualInterestRate (%)Avg. Term(month)283 21 97,236 24 12500 24 18274 24 18Non-farm loans (direct, individual loans)Microenterprises 290 24 12Growth enterprises 707 24 36Small enterprises 8,724 24 15Housing loans to repeatclients-- 18 36General purpose loans (no end-use restrictions)Self-help groups 2,767 15 to 21 12Individual loans to repeatclients276 24 12As of March 2000, BASIX has yet to consolidate itsfinancial position and become sustainable (seeFigure 3). Indeed, the performance has beenvolatile. Financial self-sufficiency dropped from 97percent in 1998-99 to 65 percent in 1999-00.Between 1998-99 and 1999-00, the operatingincome declined from 24 to 21 percent whileoperating expenses increased from 25 to 32percent.Figure 3: Financial Performance of BASIX1998-99 1999-00Adjusted return on assets (%) -0.8 -11.3Operational self-sufficiency (%) 105 77Financial self-sufficiency (%) 97 65Operating income / Avg. totalassets (%)24.3 20.7Portfolio yield (%) 29.6 24.0Operating expenses / Avg. totalassets (%)25.1 31.9MICROBANKING BULLETIN, SEPTEMBER 2000 27

CASE STUDIESThe large concentration of the portfolio inagricultural loans significantly affected theorganization’s income. Portfolio yield declined from30 percent in 1998-99 to 24 percent in 1999-00partly because portfolio at risk over 90 days climbedfrom 0.6 percent to 6.5 percent during the sameperiod. Rebates for on-time repayments intendedto improve portfolio quality further reduced the yield.BASIX attributes its portfolio quality problems to thefollowing: (i) prolonged drought affecting repaymenton farm loans; (ii) reduced efforts to recover loansdue to expansion and development of humanresources in new areas; and (iii) lateimplementation of a good MIS system to track loandelinquencies. Furthermore, the Government ofIndia required all financial institutions to reduce theircollection efforts until the drought was over.Interest Rate PolicyTo increase portfolio yield, BASIX has now revisedits interest rate policy. Nominal interest rates perannum range from 15 to 24 percent according tocost of funds, operating costs, risks andcompetition. For example, the rate for self-helpgroups was reduced from 20 to 15 percent since aline of credit was negotiated in 1998 with theNational Bank for Agriculture and RuralDevelopment, at 10.5 percent. In addition,competition from the formal sector for self-helpgroups with a cheaper rate compelled BASIX toreduce its margin. On the other hand, BASIX wasable to raise rates for other products from 21 to 24percent to cover costs.To reduce the long gestation period in realizingincome from farm loans, BASIX now requires thatthey be repaid in two installments. Indeed, manyagricultural clients have non-farm incomegeneratingactivities that allow them to pay farmloans in installments. If this change had not beenimplemented, the portfolio at risk in March 2000would have been much higher.EfficiencyThe complexity of the organization’s product menunecessitated a higher salary structure. BASIXrecruited several high-cost employees to support itsnew product development. As BASIX translates itsinitial learning into standard procedures, itestimates that loan officer productivity will increasefrom 250 to 400 clients.BASIX has kept its administrative expense ratio lowdespite high expansion costs and an increase instaff salaries, by maintaining large loan balancesand by enhancing staff productivity (Figure 4).Loan officers now carry a portable loan file thatcontains a short summary on each of their clientsthat is updated weekly for quick tracking andprocessing of loans. Computerization of borrowerrecords at the branch level is expected to furtherimprove efficiency.Figure 4: Efficiency of BASIXAdministrative expenses /Average loan portfolio (%)1997 1998 199914 19 16Depth (%) 109 42 56Staff productivity (no.) 31 123 160Average Salary(multiple of GNP/ capita)1.0 2.7 4.5Cost per borrower (US$) 27 34 30ChallengesThus far, the organization has attracted adequatefunds to finance its growth and experimentation.With external funding becoming tighter and costlier,however, BASIX may resort to alternative sourcesof funding such as deposit mobilization from thepublic. This may further increase the challenges fora young MFI.While cross-subsidization may be possible,increased competition and interest rate restrictionson farm loans may reduce margins and compelBASIX to withdraw some products. By usingintermediaries to deliver some loans, BASIX maybe missing opportunities to build long-termrelationships with clients. There is also a challengein selecting and training self-help groups to becomeefficient conduits. Increasingly, commercial banks,subsidized NGOs, and government programs areattracting self-help groups by providing loans atcheaper rates. Competition may challenge theretention and cohesion of BASIX groups.The organization regularly assesses customersatisfaction, which has shown that “ease of access”is its strength. As it grows, BASIX can achieveviability only by building on its strengths, learningfrom past experiments with products and deliverychannels, and continuing to attract commercialsources of funds.Geetha Nagarajan is a member of Bulletin Editorial Staff.This article is based on her due diligence visit in March2000 and information submitted to the Bulletin by BASIX.The MicroBanking Bulletin thanks BASIX for grantingpermission to publish its financial results.28 MICROBANKING BULLETIN, SEPTEMBER 2000

BULLETIN HIGHLIGHTS AND TABLESBULLETINHIGHLIGHTS AND TABLESBulletin HighlightsCraig F. ChurchillThe Bulletin database provides an excitingopportunity to answer some of the challengingquestions facing the microfinance industry. Inkeeping with the theme of this issue, this Highlightssection will explore the potential trade-off betweentarget market and self-sufficiency by looking closelyat the characteristics of financially self-sufficientmicrofinance institutions.As shown in Figure 1, the number of financially selfsufficient(FSS). 14 MFIs has increased over time,which reflects both the maturation of the industry aswell as our increasing success in encouragingorganizations to participate in the Bulletin.IssueFigure 1: Bulletin Participants over Time#1Oct1997#2July1998#3July1999#4Feb2000#5Sept2000# of All MFIs 28 72 86 104 114# of FSS MFIs 21 34 40 60 65% FSS 75 47 47 58 57Between the first and second issues, the Bulletinwaived the requirement that participating institutionshave a FSS ratio of at least 75 percent in an effortto broaden and deepen its coverage. That openedthe floodgates for a large increase in newparticipants, many of which were smaller, newer,and not sustainable. Some of these organizationshave improved over time, so that now three out ofevery five participants are financially self-sufficient.Of the financially self-sufficient MFIs, 24 institutionshave a FSS ratio of 110 percent or higher, whichroughly translates into an AROA above 4 percent.Characteristics of Financially self-Sufficient MFIsA closer look at the set of FSS MFIs revealsconsiderable variety. As shown in Figure 2, withthe exception of the very young programs inMENA/Central Asia, each peer group has at leastone FSS MFI, and 8 of the 14 groups have amajority that is financially self-sufficient. (For more14 The definitions for Bulletin ratios can be found in “The Index ofRatios and Tables” on page 39.details about the Bulletin peer groups, see “AnIntroduction to Peer Groups and Tables” followingthe Highlights section.)Figure 2: Self-sufficiency by Peer GroupPeer Group# of FSS # of non-MFIs FSS MFIs1. Latin America Large 10 12. Latin America Medium Broad 9 43. Latin America Medium Low-end 11 24. Latin America Small Low-end 1 45. Latin America Credit Unions 10 16. Asia Large 3 27. Asia-Pacific 5 08. South Asian 3 69. Africa Small 1 810. Africa Medium 2 411. Africa/MENA 4 212. MENA/Central Asia 0 613. Eastern Europe Broad 2 614. Eastern Europe High-end 4 3Total 65 49Figure 3 provides the characteristics of the top tenperforming Bulletin MFIs ranked by their financialself-sufficiency ratio.Figure 3: Characteristics of the Top Ten MFIsbased on Financial Self-sufficiency RatioRegion Target Market Methodology Size1. LA Broad Individual Medium2. LA High Individual Large3. Africa Broad Individual Medium4. LA Low Individual Medium5. Asia Low Solidarity Large6. LA Low Village Medium7. Asia Broad Solidarity Large8. LA Broad Solidarity Medium9. MENA Broad Individual Large10. LA Broad Individual LargeWith the exception of small programs andorganizations in Eastern Europe, all regions, targetmarkets, methodologies and sizes of institutions arerepresented on the Top Ten list. Although the topfour institutions all use an individual lendingMICROBANKING BULLETIN, SEPTEMBER 2000 29

BULLETIN HIGHLIGHTS AND TABLESmethodology, they are serving three different targetmarkets. In fact, three of the Top Ten institutionsserve the low-end market, which means that theiraverage loan balance is either below $150 or below20 percent of GNP per capita. Furthermore, theTop Ten list includes NGOs, regulated financialinstitutions and credit unions. Obviously there isnot one road to financial self-sufficiency.As shown in Figure 4, FSS MFIs are generallymedium sized, located in Latin America, serve abroad target market, and use individual lendingmethodology. Nevertheless, the commoncharacteristics of FSS MFIs highlighted in Figure 4are prevalent partly because there are moreinstitutions with those characteristics represented inthe Bulletin database.Figure 4: Characteristics of Financially Self-sufficient MFIs (FSS MFIs)Number of Financially Self-sufficient MFIs454035302520151050424136352219181210865 51LowBroadHighAfricaAsiaLAMENAE. EuropeSmallMediumLargeIndividualSolidarityMarket Region Size MethodologyVillageFigure 5: Comparison of FSS and Non-FSS CharacteristicsCharacteristic # of FSS MFIs # of non-FSS % FSS to TotalMFIsAgeMature (>6 years) 48 18 73Young (3 to 6 years) 8 16 33New (< 3 years) 9 15 38RegionAfrica 6 13 32Asia 12 10 55Eastern Europe 5 9 36Latin America 41 12 77Middle East North Africa 1 5 17Scale of OperationsLarge (portfolio > US$ 8 million) 18 4 82Medium (portfolio between US$ 1 million and US$ 8 million) 42 25 63Small (portfolio < $1 million) 5 20 20MethodologyIndividual 36 16 69Solidarity (groups of 3-9 borrowers) 19 21 48Village (groups with ≥ 10 borrowers) 10 12 45Target MarketLow-end (depth < 20% OR average loan balance < US$ 150) 22 28 44Broad (depth between 20% and 149%) 35 17 67High-end (depth ≥ 150%) 8 4 6730 MICROBANKING BULLETIN, SEPTEMBER 2000

BULLETIN HIGHLIGHTS AND TABLESFigure 5 highlights the diversity among FSS MFIs.It shows, for example, that while FSS MFIs fromLatin America vastly outnumber sustainableprograms in Asia (41 to 12), there is a much smallergap in the percentage of FSS MFIs to total MFIs inthose regions (77 percent to 55 percent).Two findings stand out from a simplistic analysis ofFigure 5. First, older MFIs have a distinctadvantage over their younger counterparts.Second, small MFIs have a real disadvantage intheir efforts to achieve financial self-sufficiency. It isalso interesting to note that nearly half of theprograms using group methodologies (solidarity andvillage) or serving low-end market are sustainable.Performance VolatilityMany institutions experience wide swings in theirfinancial self-sufficiency levels. If we had publishedthe Top Ten list in the last issue of the Bulletin, only5 of those listed in the prior issue would also appearon this year’s list. The Bulletin database currentlyconsists of 52 participants (not just FSS MFIs) forwhich we have both 1998 and 1999 data. Of the52, 22 experienced an increase in the FSS ratio ofmore than 10 percentage points, 14 decreased bymore than 10 points, and 16 basically stayed thesame (+/- 10 points).As shown by Figure 6, newer institutions tend toexperience more volatility than older MFIs, but theyhave not cornered the market on fluctuations. Moremature institutions operating in unstablemacroeconomic conditions or competitive marketsalso experience big swings in their FSS ratio.How Low Can They Go?Now to the crux of the matter: is there a trade-offbetween financial self-sufficiency and depth ofoutreach? Figure 7 lists financially self-sufficientMFIs ranked by their depth of outreach (averageoutstanding loan / GNP per capita) and averageoutstanding loan balance.The data in Figure 7 suggest that it is possible toprovide very small loans and be financially selfsufficient—atleast in Latin America and parts ofAsia. With one exception, MFIs in Africa andEastern Europe have achieved financial selfsufficiencyby serving a broad or high-end targetmarket.Figure 6: Percent Change in Financial Self-sufficiency from 1998 to 1999New and Young MFIsMature MFIs-100% -50% 0% 50% 100% 150% 200% 250%MICROBANKING BULLETIN, SEPTEMBER 2000 31

BULLETIN HIGHLIGHTS AND TABLESFigure 7: Financially Self-sufficient MFIs with the Smallest Loan SizesRanked by Depth IndicatorOrganization Region MethodologyScaleof Ops.Depth(%)Ranked by Average Loan BalanceOrganization Region MethodologyScaleof Ops.1. Compartamos LA Village Medium 3.3 1. EMT Asia Solidarity Medium 332. CEAPE/ PE LA Solidarity Medium 7.4 2. FINCA Malawi Africa Village Small 373. Pro Mujer Bolivia LA Village Medium 11.5 3. BRAC Asia Solidarity Large 534. WR Honduras LA Village Medium 11.5 4. BURO, Tangail Asia Solidarity Medium 575. FMM Popayán LA Individual Medium 12.0 5. ASA Asia Solidarity Large 666. EMT Asia Solidarity Medium 12.7 6. FINCA Nicaragua LA Village Medium 747. Enlace LA Solidarity Medium 12.8 7. WR Honduras LA Village Medium 858. Hublag Asia Solidarity Small 12.9 8. Pro Mujer Bolivia LA Village Medium 1169. RSPI Asia Solidarity Small 13.9 9. Compartamos LA Village Medium 12910. CMM Medellín LA Individual Medium 14.4 10. AKRSP Asia Village Medium 143Avg.LoanBal.(US$)How Low-end MFIs Achieve Financial SelfsufficiencyAn exploration of performance indicators by targetmarket reveals how the financially self-sufficientMFIs serving the poorest clients accomplished theirremarkable feat. Figure 8 shows that theadministrative expense ratio for low-end programsis approximately twice as high as that of MFIsserving other target markets. However, a look atthe salary structure and the cost per borrowerreveals that sustainable, low-end MFIs keep a lid onexpenses, so that their high administrative expenseratio is primarily a function of low loan sizes.Low-end(n=20)Broad(n=33)Figure 8: Efficiency Indicators of FSS MFIs byTarget Market aAverageSalary(multipleof GNP/capita)Cost perBorrower(US$)PortfolioYield (%)35.9 3.9 53 53.4 170AdminExp. /Avg.LP b (%)Productivityof Staff(No.)19.5 5.2 104 37.7 106High-end 16.3 9.1 294 29.5 55(n=6)a Data above were calculated by dropping top and bottompercentiles in each group.in average loan size and depth of outreach. Asshown in Figure 9, low-end MFIs have average loanbalances that are only 6 percent of those of highendprograms, and their depth indicator is only 4percent that of high-end programs, showing a muchhigher depth of outreach.Low-end FSS MFIs also target women moreeffectively than sustainable programs that providelarger loans. Eighty-three percent of the clients oflow-end FSS MFIs are women, compared to 50percent for broad MFIs and 32 percent for high-endinstitutions.Figure 9: Financial Performance of FSS MFIs byTarget Market aLow-end(n=20)Broad(n=33)High-end(n=6)Avg.LoanBalance(US$)Depth(%)OSS(%)FSS(%)AROA(%)164 14 129 106 1.6751 67 125 111 2.52,629 340 141 117 2.3aData above were calculated by dropping top and bottompercentiles in each group.Besides keeping costs down, to compensate fortheir small loan sizes, financially self-sufficient, lowendMFIs also charge much higher interest ratesand have higher productivity ratios. While thesecompensating efforts do not result in the samelevels of financial performance as MFIs servingbroad and high-end clientele (i.e., low-endprograms have slightly lower FSS and AROAratios), the results of the low-end FSS MFIs arequite extraordinary given the enormous disparities32 MICROBANKING BULLETIN, SEPTEMBER 2000

BULLETIN HIGHLIGHTS AND TABLESRegional Analysis of DepthSuccess in profitability serving the low-end marketvaries significantly by region. An analysis of theentire Bulletin database (not just the financially selfsufficientsubset), summarized in Figure 10,provides some indication of key regionaldifferences. The first finding is that there are notany participating MFIs from Eastern Europe thatserve the low-end market.In Latin America, participating low-end MFIsachieve almost the same levels of self-sufficiencyon average as institutions serving a broad market.This accomplishment appears to be largely theresult of a 19-point spread between portfolio yieldand the administrative expense ratio.On average, low-end MFIs in Africa are reallystruggling. Unlike other regions, the low-endinstitutions in Africa have higher costs per borrowerand a higher salary structure than MFIs serving abroad market. Although they try to make up fortheir high costs with increased productivity andhigher yields, the astronomical administrativeexpense ratio of 76 percent is too much to bear. Tobe fair, the low-end African institutions are thesmallest (average loan portfolio of US$1.1 million)of the subsets listed in Figure 10 and, except for theEastern European programs, have been inoperation for the least amount of time (average 4.3years).The low-end Asian MFIs are heterogeneous. Justover half are financially self-sufficient, includingsome that are very profitable. At the other extremeare programs that, despite low costs and highproductivity, are unprofitable primarily because theyare not generating a sufficient portfolio yield. Infact, only in Asia do the MFIs with the smallest loansizes charge the lowest interest rates.Change in Depth and FSS over TimeThe average age of sustainable, low-end MFIs is 11years, which is notably higher than the programsserving broad markets (8.6 years) and almost twicethe age of high-end programs (5.8 years). It isoften assumed that MFIs experience a gradualincrease in average loan size over time. But thefact that these MFIs are on average 11 years oldand are still serving the low-end market suggeststhat loan balances do not necessarily have to creepup as MFIs mature. Indeed, it is probable thatthese experienced MFIs become more adept inserving new clients who demand small loan sizesand hence maintain a small average loan balance.Figure 10: Performance Indicators by Region and Target Market aAverageLoanBalance(US$)Depth(%)FSS(%)PortfolioYield (%)AdminExpense/Avg. LoanPortfolio(%)Cost perBorrower(US$)AverageSalary(multipleof GNP/capita)Productivityof Staff(No.)Latin AmericaBroad (n=32) 956 63 103 40 22 141 4.9 100Low (n=19) 250 11 97 64 45 87 2.5 131Africa bBroad (n=8) 332 67 88 30 30 67 8.2 96Low (n=15) 125 21 63 49 76 72 9.0 153Asia cBroad (n=4) 258 40 109 39 16 37 3.6 147Low (n=15) 83 18 82 32 32 28 3.1 197Eastern EuropeHigh (n=6) 2,866 250 88 29 21 460 8.6 48Broad (n=8) 1,277 76 87 31 25 245 6.1 74a Data above were calculated by dropping top and bottom percentiles in each group.bAll MFIs in MENA are included in Africa.c All MFIs in CA are included in Asia.MICROBANKING BULLETIN, SEPTEMBER 2000 33

BULLETIN HIGHLIGHTS AND TABLESFigure 11: Percent Change in Average Loan Balance for 50 MFIs (1998 to 1999)Low endBroad/High end-80% -60% -40% -20% 0% 20% 40% 60% 80% 100% 120%To look at this issue in more detail, we refer againto the set of MFIs with both 1998 and 1999information (MFIs which are not all financially selfsufficient).Of the 50 institutions with loan balancedata for both years, roughly half (23) experienced adecrease in average loan balance from 1998 to1999 and the other half (25) experienced anincrease (2 stayed the same). Their percentchange is depicted in Figure 11.A closer look at the characteristics of theseinstitutions reveals that a larger number of MFIsserving broad or high-end markets increased theiraverage loan balance from 1998 to 1999. Also,except for one MFI, the MFIs that experienced thelargest decreases in their average loan balanceswere all serving the low-end market. One possibleexplanation is that perhaps the low-end MFIs weregrowing faster, which means that their portfoliosconsisted of a large percentage of new clients withsmaller loans. This explanation only works for acouple of low-end MFIs, and in general there doesnot appear to be a strong correlation betweenchange in average loan balance and the change inthe number of clients.While it might be assumed that a decline in averageloan balance will negatively affect self-sufficiency,this is not necessarily the case. Figure 12 showsthe change in FSS for low-end MFIs that increasedtheir depth of outreach. Out of the 13 programsthat fall within this category, 6 increased their FSSratio from 1998 to 1999. This issue deservesfurther analysis, but it reinforces the idea that thereis not necessarily a trade-off between depth ofoutreach and profitability.34 MICROBANKING BULLETIN, SEPTEMBER 2000

BULLETIN HIGHLIGHTS AND TABLESFigure 12: Percent Change in Financial Self-sufficiency for Low-end Programs withDeclining Average Loan Balances (1998 to 1999)50%40%30%20%10%0%-10%-20%-30%-40%-50%-60%Average Loan Balance (ALB) ChangeFinancial Self-sufficiency (FSS) ChangeTotal OutreachOn an unrelated note, the Bulletin has receivedseveral requests for total outreach by participatingMFIs. Figure 13 summarizes some maincharacteristics of Bulletin participants organized intopeer groups. Since our database consists only of114 institutions, and there are several thousandorganizations providing microfinance, thesenumbers do not indicate the size of the market. Infact, one of the largest MFIs in the world, thePeer GroupGrameen Bank, does not participate in the Bulletin,so there are at least another 2.5 million clients thatare missing from these totals.Perhaps the most interesting observation is that fiveAsian MFIs—ASA, BRAC, BRI, BAAC and BankDagang Bali—account for 89 percent of theborrowers and outstanding portfolio.Figure 13: Total Outreach of Bulletin MFIs by Peer GroupNumber ofBorrowersOutstanding LoanPortfolio ($)Total Assets ($)Number ofVoluntarySaversTotal VoluntaryDeposits ($)1. LA Large 312,043 304,473,502 375,583,369 78,685 123,487,6982. LA Medium Broad 88,286 41,969,977 59,514,550 12,047 4,858,4353. LA Medium Low-end 170,857 31,192,284 44,095,964 0 04. LA Small Low-end 14,761 2,126,482 3,115,225 0 05. LA Credit Unions 59,203 44,025,192 67,061,499 268,908 38,351,2406. Asia Large 11,004,567 4,773,596,634 8,071,536,010 24,916,688 5,381,822,5787. Asia-Pacific 135,419 19,390,124 24,352,476 5,881 9,0588. South Asian 261,722 18,753,452 32,607,076 444,797 1,723,5389. Africa Small 80,899 7,085,562 13,462,370 14,431 647,14710. Africa Medium 142,781 21,837,009 32,966,902 171,722 12,485,38911. Africa/MENA 58,437 46,760,601 94,429,058 209,308 28,387,45612. MENA/Central Asia 37,766 5,233,867 11,220,128 0 013. E. Europe Broad 21,190 24,784,255 28,314,020 0 014. E. Europe High-end 7,844 22,533,261 30,118,157 142 175,478Total 12,395,775 5,363,762,206 8,888,376,807 26,122,659 5,591,948,017MICROBANKING BULLETIN, SEPTEMBER 2000 35

BULLETIN HIGHLIGHTS AND TABLESSetting up the Peer GroupsAn Introduction to the Peer Groups and TablesThe MicroBanking Standards Project is designed tocreate performance benchmarks against whichmanagers and directors of microfinance institutionscan compare their own performance.The microfinance industry consists of a wide rangeof institutions, with very different characteristics.For the reference points to be useful, an MFI needsto be compared to similar institutions.The MicroBanking Bulletin addresses this issue withits peer group framework. Peer groups are sets ofprograms that have similar characteristics—similarenough that their managers find utility in comparingtheir results with those of other organizations intheir peer group. They are based on threeindicators:1) Region: With regulatory environments, interestrate policies, and macroeconomic conditionsvarying widely around the world, microfinancediffers by region.2) Scale of Operations: Microfinance institutionschange and develop as the scale of theiroperations grows. We classify MFIs as small,medium or large according to the size of theirportfolio, so that MFIs are compared with othersat a similar stage of growth and outreach.3) Target Market: We classify institutions into threecategories—low-end, broad, and high-end—according to the range of clients that they serve.The target market is measured by the ratio oftheir average outstanding loan per borrower toGNP per capita.The quantitative criteria used to determine eachinstitution’s peer group are summarized in Figure 1below. For each peer group, we provide averagecharacteristics and a wide range of performanceratios (see Tables 1 to 4 on pages 40 to 46). Thesenumbers represent performance benchmarks orstandards for MFIs based on their region, scale,and target market. For MFIs interested incomparisons by other characteristics, Tables A andB (pages 48 to 51) include selected ratios by age,scale, lending methodology, level of retail financialintermediation, and target market.New Peer GroupsThis issue of The MicroBanking Bulletin includesone new peer group and a redefinition of another.The Eastern European peer group has now beenseparated into two, Broad and High-end, whichsignificantly improves the homogeneity of thesegroups. In the last issue of the Bulletin, we tried toadd a fourth characteristic to our peer groupdefinitions: financial intermediary. However, onlyone region (Latin America) had enoughintermediaries to justify a separate peer group, andnearly all of those MFIs were credit unions. So, forthe purposes of clarity, the peer group LatinAmerican Intermediaries is now called LatinAmerican Credit Unions.Peer Group Composition and Data QualityThe members of each peer group are listed inFigure 2 on the following page and more detailedinformation about each institution can be found inAppendix II. Since the Bulletin relies primarily onself-reported data, we have rated the quality of thatinformation based on the degree to which we haveindependent verification of its reliability. The dataquality rating is NOT a rating of the institution’sperformance.Statistical IssuesIn the statistical tables that follow, the averages foreach peer group are calculated on the basis of thevalues between the 2 nd and 99 th percentiles, whichusually means that the top and bottom values foreach indicator are dropped. For the entire sampleof MFIs, the top and bottom deciles were excluded.These exclusions were done to reduce the effect ofoutliers. For more details, see Appendix I.Figure 1: Peer Group Criteria1. Region Latin America Asia Africa Africa/ MENA MENA/ CentralAsiaEasternEurope2. Scale of Operations Small Medium LargeTotal Loan Portfolio (US$) < 1,000,000 1,000,000 to 7,999,999 ≥ 8,000,0003. Target Market Low-end Broad High-endAverage Loan Balance /GNP per capita< 20% OR Avg. LoanBalance < US$15020% to 149% ≥ 150%36 MICROBANKING BULLETIN, SEPTEMBER 2000

BULLETIN HIGHLIGHTS AND TABLESFigure 2: A Guide to the Peer GroupsPEER GROUPNDATA QUALITY RATING(No. of MFIs with eachrating)AAA A BPARTICIPATING INSTITUTIONS *1. LA LargeSize: LargeTarget: Broad/High-end2. LA Medium BroadSize: MediumTarget: Broad3. LA Medium Low-endSize: MediumTarget: Low-end4. LA SmallSize: SmallTarget: Low-end5. LA Credit UnionsSize: AllTarget: Broad6. Asian LargeSize: LargeTarget: Low-end/Broad7. Asia-PacificSize: AllTarget: Low-end/Broad8. South AsianSize: Small/MediumTarget: Low-end/Broad9. African SmallSize: SmallTarget: Low-end10. African MediumSize: MediumTarget: Low-end11. Africa/MENASize: Large/MediumTarget: Broad/High-end12. MENA/CASize: Small/MediumTarget: Low-end13. Eastern Europe High-endSize: AllTarget: High-end14. Eastern Europe BroadSize: AllTarget: Broad11 5 5 1 Agrocapital, BancoADEMI, BancoSol, Calpiá, CM Arequipa,FIE, Finamérica, FWWB Cali, Los Andes, Mibanco,PRODEM13 3 6 4 ACODEP, ACTUAR, ADOPEM, ADRI, Banco la PequeñaEmpresa, CHISPA, EMPRENDER, Enlace, FAMA,FONDECO, FUNADEH, ProEmpresa, Sartawi13 2 5 6 CAM, CEAPE Pernambuco, CMM Medellín, Compartamos,Contigo, Crecer, FED, FINCA Honduras, FINCA Nicaragua,FMM Popayán, Portosol, ProMujer Bolivia, WR Honduras5 2 1 2 AGAPE, Banco do Povo de Juiz de Fora, FINCA Ecuador,FINCA México, Vivacred11 11 0 0 15 de Abril, 23 de Julio, ACREDICOM, Chuimequená,COOSAJO, ECOSABA, Moyután, Oscus, Sagrario,Tonantel, Tulcán5 3 2 0 ASA, BAAC, Bank Dagang Bali, BRAC, BRI5 0 5 0 ACLEDA, EMT, Hublag, RSPI, TSPI9 3 5 1 AKRSP, BASIX, BURO Tangail, CDS, FWWB India,KASHF, Nirdhan, SEEDS, SHARE9 2 4 3 FAULU, FINCA Malawi, FINCA Uganda, Foccas, RFF, SAT,SEF, UWFT, WAGES6 1 4 1 Kafo Jiginew, Nyésigiso, PAMÉCAS, PRIDE Vita (Guinea),PRIDE Tanzania, PRIDE Uganda6 2 1 3 ABA, ACEP, CERUDEB, Citi S&L, PADME, UNRWA6 1 4 1 Al Amana, Al Majmoua, Constanta, FATEN, FINCA Kyrgyzstan,Microfund for Women7 0 7 0 AMK, FEFAD, MEB, Moznosti, Network LeasingCorporation, SUNRISE, WVB8 0 4 4 BOSPO, Fundusz Mikro, Inicjatywa Mikro, LOK, MC-SEA,Mikrofin, Nachala, NOAAll MFIs 114 35 53 26† The MicroBanking Bulletin uses the following ratings system to classify information received from MFIs:AAA The information is supported by an in-depth financial analysis conducted by an independent entity in the lastthree yearsA The MBB questionnaire plus audited financial statements, annual reports and other independent evaluationsB The MBB questionnaire or audited financial statements without additional documentationLA = Latin America MENA = Middle East/North Africa CA = Central Asia* The institutions in italics and bold are new to the Bulletin. A short description of all institutions can be found in Appendix II.MICROBANKING BULLETIN, SEPTEMBER 2000 37

BULLETIN HIGHLIGHTS AND TABLESIndex of Ratios and TablesINDICATORS AND RATIOS DEFINITIONS TABLEOUTREACH AND INSTITUTIONAL INDICATORSTOTAL ASSETS Total assets (US$) 1,ANUMBER OFFICES Head office and branch offices (number) 1,ANUMBER STAFF Head office and branch staff (number) 1AGE OF INSTITUTION Years functioning as a MFI (years) 1NO OF PRESENT BORROWERS Borrowers with outstanding loans (number) 1,APERCENT WOMEN BORROWERS Women borrowers to total borrowers (%) 1,AMACROECONOMIC INDICATORSGNP PER CAPITA (CURRENT PRICES) GNP per capita (US$) 4GDP GROWTH RATE Annual average, 1990-1998 (%) 4INFLATION RATE Inflation rate (%) 4DEPOSIT RATE Deposit rate (%) 4FINANCIAL DEEPENING M3 / GDP (%) 4PROFITABILITYUNADJUSTED RETURN ON ASSETS Unadjusted net operating income / average total assets (%) 2ADJUSTED RETURN ON ASSETS (AROA) Adjusted net operating income / average total assets (%) 2,BADJUSTED RETURN ON EQUITY (AROE) Adjusted net operating income / average equity (%) 2OPERATIONAL SELF-SUFFICIENCY (OSS) Unadjusted operating income / unadjusted operating expense (%) 2,BFINANCIAL SELF-SUFFICIENCY (FSS) Adjusted operating income / adjusted operating expense (%) 2,BPROFIT MARGIN Adjusted net operating income / adjusted operating income (%) 2INCOME & EXPENSESOPERATING INCOME Adjusted operating income / average total assets (%) 2OPERATING EXPENSE RATIOAdjusted operating expenses (administrative, interest, adjustment and loan loss3provision expenses) / average total assets (%)NET INTEREST MARGINAdjusted net interest margin (operating income less interest and fee expense, inflation 2expense, subsidy expense and exchange rate expense) / average total assets (%)INTEREST EXPENSE Adjusted interest and fee expense, exchange rate expense / average total assets (%) 3ADJUSTMENT EXPENSE Inflation and subsidy adjustment expense / average total assets (%) 3LOAN LOSS PROVISION EXPENSE Adjusted loan loss provision expense / average total assets (%) 3SALARY EXPENSE – ASSETS Adjusted staff salary and benefits expense / average total assets (%) 3SALARY EXPENSE – PORTFOLIO Adjusted staff salary and benefits expense / average loan portfolio (%) 3,BOTHER ADMINISTRATIVE EXPENSE – Adjusted administrative expenses other than staff salary and benefits /3ASSETSaverage total assets (%)TOTAL ADMINISTRATIVE EXPENSE Adjusted total administrative expense (personnel, office supplies, deprecation,3,Brent, utilities, transportation, and others) / average loan portfolio (%)PORTFOLIO YIELD Adjusted total interest and fee income from loan portfolio / average loan portfolio (%) 2,BREAL INTEREST YIELD (Portfolio yield – inflation rate) / (1+inflation rate) (%) 2AVERAGE SALARY Adjusted average staff salary / GNP per capita (multiple of GNP per capita) 3,BCOST PER BORROWER Adjusted total administrative expenses / average number of borrowers (US$) 3,BSTAFF PRODUCTIVITY Present borrowers per staff member (number) 3,BPORTFOLIO INDICATORSPORTFOLIO AT RISK > 90 DAYS Outstanding balance of loans overdue > 90 days / total loan portfolio (%) 3,BTOTAL LOAN PORTFOLIO Portfolio outstanding (US$) 1,AAVG. LOAN BALANCE Total loan portfolio / present borrowers (US$) 1,ADEPTH Average loan balance / GNP per capita (%) 1,3,BCAPITAL AND LIABILITY STRUCTURE“MARKET” BASED FUNDING All liabilities with “market” cost / average loan portfolio (%) 1,ACAPITAL / ASSETS Adjusted total equity / adjusted total assets (%) 1,ANote: The tables listed by number are for the Peer Group comparisons. Those listed by letter are for the Additional Analyses, whichcompare MFIs based on the following five categories: Age, Lending Methodology, Level of Financial Intermediation, Target Market, andScale of Operations.MICROBANKING BULLETIN, SEPTEMBER 2000 39

BULLETIN HIGHLIGHTS AND TABLESTABLE 1. INSTITUTIONAL CHARACTERISTICS AND OUTREACH INDICATORSPEER GROUPAGE OFFICES STAFFTOTAL ASSETS(years) (no.) (no.) (US$)CAPITAL/ ASSETStotalcapital /totalassets (%)ALL MFIs avg 7 11 88 5,681,315 51.6stdv 4 8 64 5,809,443 22.9N 88 86 89 92 92Fully Sustainable MFIs avg 9* 50* 336* 14,514,649* 46.7stdv 5 154 1,181 27,006,912 22.7N 53 57 60 63 631. LA Large Target: Broad/ High-end avg 11* 19* 206* 29,559,763* 27.2*Agrocapital, Banco ADEMI, BancoSol, Calpiá, CM Arequipa, FIE, Finamérica, stdv 3 13 88 11,101,790 15.1FWWB Cali, Los Andes, Mibanco, PRODEM N 9 9 9 9 92. LA Medium Broad avg 8 9 76 4,352,691 45.1ACODEP, ACTUAR, ADOPEM, ADRI, Banco Pequena Empresa, CHISPA, stdv 4 5 54 2,072,391 18.9EMPRENDER, Enlace, FAMA, FONDECO, FUNADEH, ProEmpresa, Sartawi N 11 11 11 11 113. LA Medium Low-end avg 10 7 75 3,302,021 67.3*CAM, CEAPE Pernambuco, CMM Medellín, Compartamos, CONTIGO, CRECER, stdv 3 5 42 1,276,273 15.9FED, FINCA HO, FINCA NI, FMM Popayan, Portosol, ProMujer, World Relief HO N 11 11 11 11 114. LA Small Target: Low-End avg 5 1* 26 569,688 59.5AGAPE, Banco do Povo de Juiz de Fora, FINCA Ecuador, FINCA México, stdv 4 1 7 21,784 24.9Vivacred N 3 3 3 3 35. LA Credit Unions Size: All Target: Broad avg ---- 4* 54 5,974,420 37.115 de Abril, 23 de Julio, ACREDICOM, Chuimequená, COOSAJO, ECOSABA, stdv ----- 2 19 2,139,870 8.7Moyutan, Oscus, Sagrario, Tonantel, Tulcán N ----- 9 9 9 96. Asian Large Target: Low-end/ Broad avg 25* 1,021* 8,775* 1,082,504,622* 17.3*ASA, BAAC, Bank Dagang Bali, BRAC, BRI stdv 4 480 4,205 1,646,058,228 16.1N 3 3 3 3 37. Asia-Pacific Size: All Target: Low-End/Broad avg 10 14 137 2,490,015 58.6ACLEDA, EMT, Hublag, RSPI, TSPI stdv 2 6 78 1,544,946 10.7N 3 3 3 3 38. South Asian Size: Small/Medium Target: Low-End/Broad avg 8 20* 132 3,413,312 59.8AKRSP, BASIX, Buro Tangail, CDS, FWWB India, KASHF, Nirdhan, SEEDS, stdv 5 17 137 2,487,937 19.3SHARE N 7 7 7 7 79. African Small Target: Low-End avg 5 8 57 1,413,772 70.7*FAULU, FINCA Malawi, FINCA Uganda, FOCCAS, RFF, SAT, SEF, UWFT, stdv 1 6 21 366,326 20.6WAGES N 7 7 7 7 710. African Medium Target: Low-End avg 6 31* 102 5,177,794 34.2Kafo Jiginew, Nyésigiso, PAMÉCAS, PRIDE Tanzania, PRIDE Uganda, stdv 3 14 25 2,496,809 13.5PRIDE Vita N 4 4 4 4 411. Africa/MENA Size: Large/Med. Target: Broad/High-end avg 7 9 118 15,172,381* 51.8ABA, ACEP, CERUDEB, Citi S&L, PADME, UNRWA stdv 2 5 82 10,695,939 29.4N 4 4 4 4 412. MENA/CA Size: Small/Medium Target: Low-End avg 3* 11 64 1,865,050 98.1*Al Amana, Al Majmoua, Constanta, FATEN, FINCA Kyrgyzstan, stdv 2 6 33 479,805 1.8Microfund for Women N 4 4 4 4 413. Eastern Europe High-end Size: All avg 3* 4* 30* 4,098,846 42.2AMK, FEFAD, MEB, Moznosti, Network Leasing Corporation, SUNRISE, WVB stdv 1 1 12 1,495,401 26.8N 5 5 5 5 514. Eastern Europe Broad Size: All avg 2* 8 25* 2,616,054 62.5BOSPO, Fundusz Mikro, Inicjatywa Mikro, LOK, MC-SEA, MIKROFIN, stdv 1 5 8 1,117,429 33.3Nachala, NOA N 6 6 6 6 6Note: Standard deviations and sample sizes are listed below the peer group averages. The averages are calculated on the basis of the values between theninth and second deciles for all MFIs and between second and the 99th percentiles for each peer group; therefore, sample sizes vary across indicators. Groupaverages different from average for all MFIs at 5 percent significance level are marked with an asterisk (*). Additional statistical information is available atwww.calmeadow.com. Abbreviations: LA= Latin America; MENA=Middle East/North Africa; CA=Central Asia.40 MICROBANKING BULLETIN, SEPTEMBER 2000

BULLETIN HIGHLIGHTS AND TABLES“MARKET” BASEDFUNDINGTOTAL LOANPORTFOLIOPRESENTBORROWERSAVG. LOANBALANCEDEPTH% WOMENBORROWERS“market” priced liabilities /avg loan portfolio(%)(US$)(no.)loan portfolio /present borrowers(US$)avg. loan balance/GNP per capita(%)women borrowers /total borrowers(no.)31.3 3,881,619 10,574 581 48 60.534.6 3,974,994 9,132 535 37 20.792 92 90 90 90 7265.0* 11,050,522* 73,920 803* 80* 56.358.4 22,036,367 341,448 854 109 22.854 63 61 61 61 4981.6* 23,728,939* 26,050* 1,107* 75* 52.125.1 9,491,877 10,879 740 46 109 9 9 9 9 748.8 3,090,453 6,404 859 60 45.8*21.5 1,174,271 5,609 953 30 15.111 11 11 11 11 1022.0 2,167,713 11,093 232* 12* 79.5*23.7 910,300 6,310 183 5 1911 11 11 11 11 100.0 393,679 3,145 295 7 87.50.0 65,936 1,991 359 7 53 3 3 3 3 291.7* 3,804,955 4,884 990* 63 42.3*19.9 1,363,368 2,660 500 31 3.19 9 9 9 9 9126.0* 352,532,708* 2,046,752* 394 29 58.9118.4 430,311,722 835,243 402 9 48.03 3 3 3 3 218.5 1,730,343 25,087* 168 15 83.017.7 969,448 27,610 16 3 7.23 3 3 3 3 315.3 1,859,430 22,413* 82* 22 79.0*14.0 1,565,927 20,253 38 9 26.57 7 7 7 7 78.5 791,069* 8,161 108* 25 91.8*12.8 198,009 3,613 30 13 10.77 7 7 7 7 757.8 3,274,217 22,942* 144 48 58.053.7 1,485,344 9,699 21 17 7.84 4 4 4 4 369.5* 8,153,933* 8,840 778 127* 44.687.6 4,003,912 3,584 291 98 19.44 4 4 4 4 40.0 958,843 5,284 153 12* 97.5*0.0 296,888 2,150 69 4 5.04 4 4 4 4 43.7 3,308,772 1,067* 2,932* 343* 34.5*4.1 938,526 338 457 205 3.05 5 5 5 5 48.3 2,305,720 1,995* 1,277* 76 44.214.0 933,350 994 276 33 5.06 6 6 6 6 6MICROBANKING BULLETIN, SEPTEMBER 2000 41

BULLETIN HIGHLIGHTS AND TABLESTABLE 2. OVERALL FINANCIAL PERFORMANCE AND OPERATING INCOMEUNADJUSTEDRETURN ONASSETSADJUSTEDRETURN ONASSETSADJUSTEDRETURN ONEQUITYPEER GROUP net operating income /avg. total assets(%)adj. net operatingincome / avg. total assets(%)adj. net operatingincome / avg. equity(%)ALL MFIs avg 1.5 -3.2 -5.1stdv 5.8 5.9 13.4N 92 92 91Fully Sustainable MFIs avg 6.2* 2.5* 7.6*stdv 5.3 4.8 13.9N 63 63 621. LA Large Target: Broad/ High-end avg 3.6 2.6* 13.9*Agrocapital, Banco ADEMI, BancoSol, Calpiá, CM Arequipa, FIE, Finamérica, stdv 3.5 2.0 13.2FWWB Cali, Los Andes, Mibanco, PRODEM N 9 9 92. LA Medium Broad avg 1.9 -2.3 -15.0ACODEP, ACTUAR, ADOPEM, ADRI, Banco Pequena Empresa, CHISPA, stdv 7.6 6.8 46.4EMPRENDER, Enlace, FAMA, FONDECO, FUNADEH, ProEmpresa, Sartawi N 11 11 113. LA Medium Low-end avg 7.7* 0.8* 1.4CAM, CEAPE Pernambuco, CMM Medellín, Compartamos, CONTIGO, CRECER, stdv 7.8 8.3 12.0FED, FINCA HO, FINCA NI, FMM Popayan, Portosol, ProMujer, World Relief HO N 11 11 114. LA Small Target: Low-End avg 1.2 -9.2 -21.0*AGAPE, Banco do Povo de Juiz de Fora, FINCA Ecuador, FINCA México, stdv 2.4 5.3 15.7Vivacred N 3 3 35. LA Credit Unions Size: All Target: Broad avg 7.4* 1.5* 5.8*15 de Abril, 23 de Julio, ACREDICOM, Chuimequená, COOSAJO, ECOSABA, stdv 4.1 2.2 7.1Moyutan, Oscus, Sagrario, Tonantel, Tulcán N 9 9 96. Asian Large Target: Low-end/ Broad avg 3.8 1.1 -6.7ASA, BAAC, Bank Dagang Bali, BRAC, BRI stdv 3.8 4.1 35.5N 3 3 37. Asia-Pacific Size: All Target: Low-End/Broad avg 5.0 1.3 1.5ACLEDA, EMT, Hublag, RSPI, TSPI stdv 3.0 3.6 4.6N 3 3 38. South Asian Size: Small/Medium Target: Low-End/Broad avg -1.2 -5.6 -7.7AKRSP, BASIX, Buro Tangail, CDS, FWWB India, KASHF, Nirdhan, SEEDS, stdv 3.6 4.2 8.6SHARE N 7 7 79. African Small Target: Low-End avg -10.4* -14.4* -19.3*FAULU, FINCA Malawi, FINCA Uganda, FOCCAS, RFF, SAT, SEF, UWFT, stdv 8.8 8.9 9.7WAGES N 7 7 710. African Medium Target: Low-End avg -11.3* -13.0* -27.4*Kafo Jiginew, Nyésigiso, PAMÉCAS, PRIDE Tanzania, PRIDE Uganda, stdv 12.7 14.4 14.8PRIDE Vita N 4 4 411. Africa/MENA Size: Large/Med. Target: Broad/High-end avg 3.3 1.0 7.9ABA, ACEP, CERUDEB, Citi S&L, PADME, UNRWA stdv 4.7 6.9 8.7N 4 4 412. MENA/CA Size: Small/Medium Target: Low-End avg -12.3* -17.0* -20.6*Al Amana, Al Majmoua, Constanta, FATEN, FINCA Kyrgyzstan, stdv 6.5 4.2 3.2Microfund for Women N 4 4 413. Eastern Europe High-end Size: All avg 1.5 -2.2 -4.2AMK, FEFAD, MEB, Moznosti, Network Leasing Corporation, SUNRISE, WVB stdv 4.0 2.6 4.0N 5 5 514. Eastern Europe Broad Size: All avg 0.9 -3.8 -8.7BOSPO, Fundusz Mikro, Inicjatywa Mikro, LOK, MC-SEA, MIKROFIN, Nachala, stdv 1.6 1.1 6.4NOA N 6 6 6Note: Standard deviations and sample sizes are listed below the peer group averages. The averages are calculated on the basis of the values between theninth and second deciles for all MFIs, and between second and the 99th percentiles for each peer group; therefore, sample sizes vary across indicators. Groupaverages different from average for all MFIs at 5 percent significance level are marked with an asterisk (*). Additional statistical information is available atwww.calmeadow.com. Abbreviations: LA= Latin America; MENA=Middle East/North Africa; CA=Central Asia.42 MICROBANKING BULLETIN, SEPTEMBER 2000

BULLETIN HIGHLIGHTS AND TABLESOPERATIONALSELF-SUFFICIENCYoperating income /interest, loan loss andadministrative expense(%)FINANCIAL SELF-SUFFICIENCYadj. operating income /interest, adjustment,loan loss andadministrative exp. (%)OPERATINGINCOMEadj. operatingincome / avg. totalassets(%)PROFITMARGINadj. net operatingincome / adj.operating income(%)NETINTERESTMARGINadj. net interestmargin /avg. total assets(%)PORTFOLIOYIELDinterest incomefrom portfolio /avg. loan portfolio(%)REALYIELD(portfolio yield –inflation rate) /(1 + inflation rate)(%)106.6 91.8 29.3 -13.5 19.5 39.9 28.721.6 17.4 9.5 25.1 7.6 12.7 11.792 92 92 92 92 92 92129.7* 112.4* 32.5 8.1* 22.1 42.4 29.631.4 24.0 12.1 14.2 9.8 18.4 15.963 63 63 63 63 63 63112.3 108.4* 30.7 7.5* 20.2 37.1 28.810.4 6.3 4.5 5.5 2.4 7.8 3.89 9 9 9 9 9 9106.4 94.7 40.4* -9.0 27.5* 51.7* 35.722.2 16.0 9.8 22.7 8.1 14.8 12.011 11 11 11 11 11 11123.1* 102.4 41.8* 0.1 29.1* 59.6* 45.2*22.9 16.7 5.9 15.5 4.8 21.0 8.211 11 11 11 11 11 11103.4 84.8 47.3* -18.4 29.4* 66.6* 47.6*6.4 6.4 5.8 9.4 1.1 10.9 6.43 3 3 3 3 3 3134.6* 109.9* 27.7 8.2* 10.9* 30.7* 7.3*14.2 10.6 8.6 9.2 2.0 9.0 12.99 9 9 9 9 9 9121.7 105.0 23.1 2.9 8.9* 24.4* 11.7*21.9 17.5 4.0 17.2 4.6 1.1 5.53 3 3 3 3 3 3116.9 103.7 33.6 2.9 27.5 46.2 38.57.9 10.2 5.9 9.4 7.3 2.1 5.63 3 3 3 3 3 392.5 70.7* 14.9* -54.5* 8.6* 21.6* 10.8*22.3 20.9 5.5 52.6 5.1 8.5 7.17 7 7 7 7 7 773.0* 66.3* 32.2 -57.4* 25.1 60.9* 46.6*16.6 15.0 12.7 33.7 11.1 13.6 9.87 7 7 7 7 7 775.6* 73.9* 21.8 -43.3* 19.7 35.2 31.318.6 19.0 7.2 42.4 9.6 12.2 7.74 4 4 4 4 4 4134.1* 119.6* 20.3 7.9 13.8 28.7 22.341.9 43.4 5.3 32.5 6.8 7.7 2.14 4 4 4 4 4 457.2* 51.2* 18.1* -99.2* 15.2 39.5 30.811.7 8.1 4.9 33.8 5.1 5.6 5.84 4 4 4 4 4 4105.7 89.1 19.8* -14.1 12.2* 27.5* 12.6*20.1 12.0 4.8 17.3 4.0 4.7 1.95 5 5 5 5 5 5103.5 87.2 27.4 -14.8 20.1 30.9 19.2*6.1 3.2 3.3 4.2 2.5 2.9 4.36 6 6 6 6 6 6MICROBANKING BULLETIN, SEPTEMBER 2000 43

BULLETIN HIGHLIGHTS AND TABLESTABLE 3. OPERATING EXPENSES AND PORTFOLIO MANAGEMENT INDICATORSPEER GROUPOPERATINGEXPENSEinterest, adj.,loan loss &admin. exp./avg. total assets(%)INTERESTEXPENSEinterest exp. /avg. totalassets(%)ADJ.EXP.adj. exp./ avg.totalassets(%)LOAN LOSSPROV. EXP.loan lossprovision exp. /avg. totalassets(%)SALARYEXP./ TAstaff exp. /avg. totalassets(%)ALL MFIs avg 33.1 4.5 3.8 2.3 11.2stdv 11.2 3.5 2.7 1.4 6.0N 92 90 92 92 92Fully Sustainable MFIs avg 30.5 6.2* 3.7 2.3 9.4stdv 12.6 4.8 3.9 1.6 5.7N 63 63 63 63 631. LA Large Target: Broad/ High-end avg 28.9 8.9* 0.8* 3.2 7.4Agrocapital, Banco ADEMI, BancoSol, Calpiá, CM Arequipa, FIE, Finamérica, stdv 6.9 4.2 0.7 1.3 2.8FWWB Cali, Los Andes, Mibanco, PRODEM N 9 9 9 9 92. LA Medium Broad avg 42.0* 7.4* 3.6 4.6* 12.6ACODEP, ACTUAR, ADOPEM, ADRI, Banco Pequena Empresa, CHISPA, stdv 9.8 4.7 2.2 2.7 4.7EMPRENDER, Enlace, FAMA, FONDECO, FUNADEH, ProEmpresa, Sartawi N 11 11 11 11 113. LA Medium Low-end avg 42.5* 4.7 7.4* 2.5 15.1*CAM, CEAPE Pernambuco, CMM Medellín, Compartamos, CONTIGO, CRECER, stdv 8.3 3.2 3.3 1.4 6.2FED, FINCA HO, FINCA NI, FMM Popayan, Portosol, ProMujer, World Relief HO N 11 11 11 11 114. LA Small Target: Low-End avg 59.4* 5.2 9.2* 3.3 26.0*AGAPE, Banco do Povo de Juiz de Fora, FINCA Ecuador, FINCA México, stdv 6.7 4.7 1.9 0.4 5.6Vivacred N 3 3 3 3 35. LA Credit Unions Size: All Target: Broad avg 25.8 9.6* 5.1 1.3* 4.7*15 de Abril, 23 de Julio, ACREDICOM, Chuimequená, COOSAJO, ECOSABA, stdv 11.2 2.0 6.3 0.6 1.4Moyutan, Oscus, Sagrario, Tonantel, Tulcán N 9 9 9 9 96. Asian Large Target: Low-end/ Broad avg 21.4 9.6* 2.9 2.0 3.4*ASA, BAAC, Bank Dagang Bali, BRAC, BRI stdv 4.3 8.1 1.6 1.0 2.7N 3 3 3 3 37. Asia-Pacific Size: All Target: Low-End/Broad avg 31.4 2.9 2.7 3.7 12.1ACLEDA, EMT, Hublag, RSPI, TSPI stdv 2.1 0.3 0.7 0.8 0.7N 3 3 3 3 38. South Asian Size: Small/Medium Target: Low-End/Broad avg 20.7* 2.1 4.3 1.2* 5.9*AKRSP, BASIX, Buro Tangail, CDS, FWWB India, KASHF, Nirdhan, SEEDS, stdv 6.3 1.6 1.6 0.9 3.8SHARE N 7 7 7 7 79. African Small Target: Low-End avg 48.6* 1.9 4.5 1.9 20.7*FAULU, FINCA Malawi, FINCA Uganda, FOCCAS, RFF, SAT, SEF, UWFT, stdv 14.0 1.9 2.1 1.5 8.1WAGES N 7 7 7 7 710. African Medium Target: Low-End avg 35.7 1.6 1.2 1.6 15.6Kafo Jiginew, Nyésigiso, PAMÉCAS, PRIDE Tanzania, PRIDE Uganda, stdv 19.5 0.6 1.5 0.2 11.2PRIDE Vita N 4 4 4 4 411. Africa/MENA Size: Large/Med. Target: Broad/High-end avg 16.2* 1.7 0.6* 1.6 6.5ABA, ACEP, CERUDEB, Citi S&L, PADME, UNRWA stdv 6.6 1.6 0.7 1.4 1.7N 4 4 4 4 412. MENA/CA Size: Small/Medium Target: Low-End avg 39.7 0.0* 2.5 1.1 21.0*Al Amana, Al Majmoua, Constanta, FATEN, FINCA Kyrgyzstan, stdv 15.0 0.1 1.2 1.0 8.3Microfund for Women N 4 4 4 4 413. Eastern Europe High-end Size: All avg 21.4* 2.0 4.3 2.3 7.0AMK, FEFAD, MEB, Moznosti, Network Leasing Corporation, SUNRISE, WVB stdv 5.6 1.9 2.4 1.0 3.2N 5 5 5 5 514. Eastern Europe Broad Size: All avg 30.4 2.0 5.4 2.6 13.4BOSPO, Fundusz Mikro, Inicjatywa Mikro, LOK, MC-SEA, MIKROFIN, stdv 4.0 1.8 1.5 1.7 2.1Nachala, NOA N 6 6 6 6 6Note: Standard deviations and sample sizes are listed below the peer group averages. The averages are calculated on the basis of the values between the ninthand second deciles for all MFIs, and between second and the 99th percentiles for each peer group; therefore, sample sizes vary across indicators. Group averagesdifferent from average for all MFIs at 5 percent significance level are marked with an asterisk (*). Additional statistical information is available atwww.calmeadow.com. Abbreviations: LA= Latin America; MENA=Middle East/North Africa; CA=Central Asia.44 MICROBANKING BULLETIN, SEPTEMBER 2000

BULLETIN HIGHLIGHTS AND TABLESOTHER ADMIN.EXP./ TAother admin. exp. /avg. total assets(%)ADMINEXP. / LPtotal admin.exp. / avg.loan portfolio(%)SALARYEXP. / LPstaff exp. /avg. loanportfolio(%)PORTFOLIOAT RISKoutstandingbalance overdue> 90 days /total loan portfolio(%)DEPTHavg. loanbalance /GNP percapita (%)AVERAGESALARYavg. staff salary/ GNP per capita(multiple ofGNP/ capita)STAFFPRODUCTIVITYpresent borrowers /number of staff(no.)COST PERBORROWERtotal admin. exp/ avg. number ofborrowers(US$)8.9 31.0 17.2 2.0 48.3 5.1 111 1503.8 16.8 10.7 1.4 37.0 3.1 49 18492 92 92 77 90 89 89 908.2 24.9* 13.4* 2.3 79.7* 5.5 127 1224.8 14.2 8.7 1.6 108.5 4.1 84 11763 63 63 54 61 60 60 487.5 17.6* 9.0* 2.3 74.7* 6.2 124 1832.5 3.3 3.5 1.1 46.0 3.0 32 939 9 9 9 9 9 9 911.7* 32.8 17.0 3.5* 60.1 5.6 72* 1665.2 10.4 6.0 1.7 30.4 3.5 41 13711 11 11 10 11 11 11 710.8 38.5 22.1 2.1 12.4* 2.7* 140 511.9 9.0 8.2 1.9 4.7 1.6 48 1411 11 11 10 11 11 11 912.1 57.5* 39.9* 0.7 7.4 1.9 98 1442.3 13.5 12.0 0.9 6.8 0.1 49 1273 3 3 2 3 3 3 34.5* 13.8* 6.9* 2.2 63.3 2.8* 79 881.5 4.7 2.1 0.8 31.1 0.7 19 169 9 9 9 9 9 9 41.5* 9.3* 6.1 0.7 29.0 2.9 221* 120.3 3.9 3.8 0.3 9.4 1.4 105 93 3 3 2 3 3 3 39.2 33.3 17.2 4.2* 14.8 3.4 108 530.9 2.9 1.7 2.0 2.6 2.2 55 43 3 3 3 3 3 3 35.7* 20.0 10.3 1.1 22.1 2.9 229* 192.8 8.5 7.0 1.3 9.1 1.0 235 107 7 7 6 7 7 7 719.0* 84.2* 45.6* 1.4 25.0 10.3* 153* 716.4 15.8 16.5 1.6 13.0 4.7 57 397 7 7 7 7 7 7 715.1* 48.0 24.8 1.1 47.6 13.1* 178* 536.3 24.8 16.4 1.4 17.0 9.0 115 274 4 4 4 4 4 4 24.1* 17.0 10.9 2.1 126.7* 9.8* 87 1232.4 8.9 6.6 1.6 97.9 3.4 28 1024 4 4 4 4 4 4 410.8 66.4* 43.1* 0.6 11.9* 3.6 86 561.4 8.5 1.8 0.8 3.5 1.2 12 164 4 4 3 4 4 4 36.4 19.2 9.5 0.4* 342.7* 7.8 44* 400*2.1 5.9 3.4 0.4 205.2 2.8 19 2745 5 5 5 5 5 5 58.3 25.1 15.7 0.8* 76.1 6.1 74 2451.4 2.7 2.7 0.9 33.2 3.3 19 876 6 6 6 6 6 6 5MICROBANKING BULLETIN, SEPTEMBER 2000 45

BULLETIN HIGHLIGHTS AND TABLESTABLE 4. MACROECONOMIC INDICATORSGNP PERCAPITAGDP GROWTHRATE, ANNUALAVG. 1990-98INFLATIONRATEDEPOSITRATEFINANCIALDEEPENING(M3 / GDP)PEER GROUP (US$) (%) (%) (%) (%)ALL MFIs avg 1,247 4.2 8.2 14.2 36.0stdv 755 1.0 4.8 6.8 12.2N 92 82 92 91 75Fully Sustainable MFIs avg 1,366 4.1 11.0 16.5 37.9stdv 1,046 1.4 14.0 12.1 15.5N 65 61 65 64 551. LA Large Target: Broad/ High-end avg 1,666 4.8 5.9 16.1 38.6Agrocapital, Banco ADEMI, BancoSol, Calpiá, CM Arequipa, FIE, Finamérica, stdv 677 0.8 6.1 6.4 10.4FWWB Cali, Los Andes, Mibanco, PRODEM N 11 11 11 11 112. LA Medium Broad avg 1,896* 4.2 11.6 17.0 39.0ACODEP, ACTUAR, ADOPEM, ADRI, Banco Pequena Empresa, CHISPA, stdv 2,012 1.0 13.2 10.5 12.0EMPRENDER, Enlace, FAMA, FONDECO, FUNADEH, ProEmpresa, Sartawi N 13 13 13 13 133. LA Medium Low-end avg 2,332* 3.9 10.9 19.9* 38.3CAM, CEAPE Pernambuco, CMM Medellín, Compartamos, CONTIGO, CRECER, stdv 1,661 1.4 13.5 11.9 12.3FED, FINCA HO, FINCA NI, FMM Popayan, Portosol, ProMujer, World Relief HO N 13 13 13 13 134. LA Small Target: Low-End avg 3,400* 3.2* 13.8* 28.0* 29.2AGAPE, Banco do Povo de Juiz de Fora, FINCA Ecuador, FINCA México, stdv 1,354 0.6 13.7 10.3 3.3Vivacred N 5 5 5 5 55. LA Credit Unions Size: All Target: Broad avg 1,553 3.6* 25.7* 26.0* 38.715 de Abril, 23 de Julio, ACREDICOM, Chuimequená, COOSAJO, ECOSABA, stdv 31 0.6 25.4 22.0 0.0Moyutan, Oscus, Sagrario, Tonantel, Tulcán N 11 11 11 11 56. Asian Large Target: Low-end/ Broad avg 1,056 6.5* 19.9* 19.1 53.5*ASA, BAAC, Bank Dagang Bali, BRAC, BRI stdv 993 0.9 21.9 13.1 24.5N 5 5 5 5 57. Asia-Pacific Size: All Target: Low-End/Broad avg 794 4.0 6.5 9.4 43.3ACLEDA, EMT, Hublag, RSPI, TSPI stdv 491 1.0 2.5 2.5 29.2N 5 5 5 5 58. South Asian Size: Small/Medium Target: Low-End/Broad avg 422* 5.6* 9.5 11.0 46.7*AKRSP, BASIX, Buro Tangail, CDS, FWWB India, KASHF, Nirdhan, SEEDS, stdv 162 0.7 2.6 3.4 7.5SHARE N 9 8 9 9 89. African Small Target: Low-End avg 957 4.8 10.9 15.3 22.3*FAULU, FINCA Malawi, FINCA Uganda, FOCCAS, RFF, SAT, SEF, UWFT, stdv 1,278 2.7 13.3 10.5 16.4WAGES N 9 9 9 9 910. African Medium Target: Low-End avg 353* 4.3 2.7* 6.9* 18.1*Kafo Jiginew, Nyésigiso, PAMÉCAS, PRIDE Tanzania, PRIDE Uganda, stdv 146 1.7 3.7 4.0 6.1PRIDE Vita N 6 6 6 6 611. Africa/MENA Size: Large/Med. Target: Broad/High-end avg 742 4.6 5.5 8.9 26.9ABA, ACEP, CERUDEB, Citi S&L, PADME, UNRWA stdv 540 1.7 4.9 7.5 30.2N 6 5 6 6 512. MENA/CA Size: Small/Medium Target: Low-End avg 1,443 -2.0* 11.0 25.9* 107.0*Al Amana, Al Majmoua, Constanta, FATEN, FINCA Kyrgyzstan, stdv 1,004 10.5 12.7 23.9 41.5Microfund for Women N 6 5 6 5 313. Eastern Europe High-end Size: All avg 978 1.8* 11.9 13.9 36.8AMK, FEFAD, MEB, Moznosti, Network Leasing Corporation, SUNRISE, WVB stdv 279 0.1 6.9 1.5 31.1N 7 2 7 6 214. Eastern Europe Broad Size: All avg 2,221* 0.8* 9.0 10.5 34.4BOSPO, Fundusz Mikro, Inicjatywa Mikro, LOK, MC-SEA, MIKROFIN, stdv 1,529 3.4 7.0 4.4 1.1Nachala, NOA N 8 4 8 8 2Note: Standard deviations and sample sizes are listed below the peer group averages. The averages are calculated using all observations for all MFIs and on the basis of thevalues between the second and the 99th percentiles for each peer group; therefore, sample sizes vary across indicators. Group averages different from average for all MFIs at 5percent significance level are marked with an asterisk (*). Additional statistical information is available at www.calmeadow.com. Abbreviations: LA= Latin America;MENA=Middle East/North Africa; CA=Central Asia.46 MICROBANKING BULLETIN, SEPTEMBER 2000

BULLETIN HIGHLIGHTS AND TABLESAdditional Analysis TablesTables A and B provide data on selectedperformance indicators for groups of institutionsbased on the following five characteristics:1) Age of the MFI: Microfinance institutionsdevelop as they mature. The Bulletin classifiesMFIs into three categories (new, young, andmature) based on the difference between theyear they started their microfinance operationsand the year for which the institutions havesubmitted data.2) Scale of operation: MFIs are classified assmall, medium and large according to the sizeof their loan portfolio to facilitate comparisons ofinstitutions with similar outreach.3) Lending Methodology: Performance may varyby the methodology used by the institution todeliver loan products. The Bulletin classifiesMFIs based on the primary methodology theyused as determined by the number of loansoutstanding.4) Level of Financial Intermediation: Thisclassification is based on the ratio of totalvoluntary time and passbook deposits to totalassets. This ratio indicates the MFI’s ability tomobilize retail savings and fund its loan portfoliothrough deposits.5) Target Market: The Bulletin classifiesinstitutions into three categories—low-end,broad, and high-end—according to the range ofclients they serve based on average loanoutstanding.The quantitative criteria used to categorize thesecharacteristics are summarized in the table below.A list of institutions that fall into these categories islocated immediately following Table B.These Additional Analysis Tables provide anothermeans of creating performance benchmarksbesides the peer groups. Two of thesecharacteristics—scale of operation and targetmarket—are also factors in determining peer groupcomposition. The purpose of the AdditionalAnalysis Tables is to look at these characteristicssingularly, rather than within context of the peergroups.The inclusion of these additional tables is the resultof feedback we have received from readers ofprevious issues of the Bulletin. We would greatlyappreciate additional suggestions on how we canmake these tables more useful in the future.Age of the MFIScale of operationsLending MethodologyLevel of Retail Financial IntermediationTarget MarketNew: 1 to 2 yearsYoung: 3 to 6 yearsMature: over 6 yearsLarge: portfolio > US$ 8 millionMedium: portfolio US$ 1 to 8 millionSmall: portfolio < US$ 1 millionIndividualSolidarity Group: group of 3 to 9 borrowersVillage Banking: groups with ≥ 10 borrowersFinancial Intermediary: passbook and time deposits ≥20 percent of total assetsOther: passbook and time deposits < 20 percent of totalassetsLow-end: depth < 20% OR average loan size < US$150Broad: depth between 20% and 149%High-end: depth ≥ 150%MICROBANKING BULLETIN, SEPTEMBER 2000 47

BULLETIN HIGHLIGHTS AND TABLESTABLE A: INSTITUTIONAL CHARACTERISTICS AND OUTREACH INDICATORSTOTAL ASSETS CAPITAL / ASSETS “MARKET” BASEDFUNDINGall liabilities with “market”CRITERIAtotal capital / total assets cost / avg loan portfolio(US$)(%)(%)AGE New (1 - 2 years) avg 2,356,811* 55.8 13.1*stdv 1,260,124 30.3 19.1N 22 22 22Young (3 - 6 years) avg 4,622,207 57.0 29.0stdv 4,726,330 32.8 70.3N 22 22 22Mature (> 6 years) avg 61,781,141 48.4 57.1*stdv 372,003,749 24.4 47.6N 64 64 64SCALE OF Large (Portfolio > US$ 8 million) avg 190,543,109* 32.0* 88.5*OPERATIONS stdv 658,415,764 21.1 72.9N 20 20 20Medium (Portfolio US$ 1 to 8 million) avg 4,276,559 54.3 37.4stdv 2,391,563 25.1 39.5N 65 65 65Small (Portfolio < US$ 1 million) avg 1,226,378* 61.8 14.3*stdv 755,236 30.6 26.8N 23 23 23METHOD- Individual avg 70,133,026 43.4 60.0*OLOGY stdv 420,616,841 26.0 60.6N 50 50 50Solidarity Groups avg 10,792,057* 51.7 33.6(groups of 3 to 9 borrowers) stdv 20,727,044 25.7 38.8N 38 38 38Village Banking avg 2,571,521* 72.7* 12.8*(groups with ≥ 10 borrowers) stdv 1,892,150 22.9 19.3N 20 20 20RETAIL Financial Intermediaries avg 139,042,446* 23.6* 126.2*FINANCIAL (passbook and time deposits stdv 593,003,373 16.5 64.3INTER- ≥ 20% of total assets) N 25 25 25MEDIARYOther avg 5,890,092 60.3* 20.4*(passbook and time deposits stdv 10,683,267 25.6 24.7< 20% of total assets) N 85 85 85TARGET Low-end avg 5,102,014 63.4* 21.8GROUP (depth < 20% OR avg. loan balance stdv 12,552,299 26.7 31.5< US$ 150) N 48 48 48Broad avg 69,897,939 43.3* 59.6*(depth between 20% and 149%) stdv 420,629,069 24.6 46.7N 50 50 50High-end avg 14,190,423* 37.6 41.2(depth ≥ 150%) stdv 15,930,427 26.1 77.3N 10 10 10Note: Standard deviations and sample sizes are listed below the group averages. The averages are calculated on the basis of the values between the second andthe 99th percentiles for each group; therefore, sample sizes vary across indicators. Group averages different from average for all MFIs at 5 percent significancelevel are marked with an asterisk (*). Additional statistical information is available at www.calmeadow.com.48 MICROBANKING BULLETIN, SEPTEMBER 2000

BULLETIN HIGHLIGHTS AND TABLESBRANCHOFFICES% WOMENBORROWERSTOTAL LOANPORTFOLIOPRESENTBORROWERSAVG. LOANBALANCE(no.) (%) (US$) (no.)total loan portfolio / no.present clients(US$)7* 56.1 1,830,619* 3,123* 2637 24.4 1,060,564 3,259 28622 17 22 22 1111 74.8* 3,065,045 9,839 188*9 26.9 3,606,613 7,298 14422 20 22 22 19100* 61.2 24,273,835 113,788* 331*364 24.9 106,900,749 461,188 26962 58 64 63 49172* 49.8 72,728,502* 328,385* 539398 21.7 185,215,265 788,818 29720 16 20 20 1312 59.2 2,966,943 11,406 298*17 24.1 1,531,921 13,477 23362 58 65 64 457* 84.2* 623,572* 5,235* 112*7 21.2 252,366 3,933 5523 21 23 23 2141 44.4* 23,852,227 57,255 1,341*219 14.2 119,291,547 348,873 1,06850 41 50 50 5034* 72.4* 8,137,669* 45,171 222*99 22.5 17,339,930 174,107 18038 34 38 38 3335 86.7* 1,672,098* 13,879 109*117 23.1 1,440,229 11,584 5518 20 20 20 2079* 45.6* 46,881,657* 114,547* 891*308 11.5 167,898,436 491,801 69025 22 25 25 2535 69.0* 4,350,521 24,939 707129 26.7 8,730,900 118,148 95582 74 85 84 8456* 84.3* 3,578,968 39,040 156*172 20.3 10,565,846 156,958 14745 41 48 47 4744 50.4* 24,756,209 61,133 481218 17.7 119,498,358 348,471 25150 47 50 50 3211 33.6* 8,930,351* 4,221* 2,741*10 3.6 9,137,688 4,536 1,12110 7 10 10 10MICROBANKING BULLETIN, SEPTEMBER 2000 49

BULLETIN HIGHLIGHTS AND TABLESTABLE B: FINANCIAL PERFORMANCE AND EFFICIENCY INDICATORSCRITERIAADJUSTEDRETURN ONASSETSadj. net operatingincome / avg. totalassets(%)OPERATIONALSELF-SUFFICIENCYoperating income /interest, loan loss &admin expense(%)FINANCIALSELF-SUFFICIENCYadj. operating income /interest, adjustment, loanloss & admin expense (%)PORTFOLIOYIELDinterest incomefrom portfolio / avg.loan portfolio(%)AGE New (1 - 2 years) avg -8.4* 89.5* 77.7* 35.8stdv 9.2 26.5 20.2 10.0N 22 22 22 22Young (3 - 6 years) avg -7.3* 94.2* 81.0* 45.2stdv 8.8 28.4 21.3 21.2N 22 22 22 22Mature (> 6 years) avg -0.6* 120.7* 102.9* 42.6stdv 8.0 35.6 29.5 19.9N 64 64 64 64SCALE OF Large (Portfolio > US$ 8 avg 2.6* 126.5* 114.3* 34.2OPERATIONS million) stdv 4.2 36.4 25.1 11.6N 20 20 20 20Medium (Portfolio US$ 1 to 8 avg -3.3 111.4 93.9 41.7million) stdv 9.6 32.5 25.4 20.0N 65 65 65 65Small (Portfolio < US$ 1 avg -10.4* 85.2* 71.5* 48.1*million) stdv 8.4 24.6 18.8 17.9N 23 23 23 23METHOD- Individual avg -0.3* 121.8* 104.8* 36.5OLOGY stdv 6.8 33.7 29.9 15.3N 50 50 50 50Solidarity Groups avg -7.7* 92.6* 81.9* 41.2(groups of 3 to 9 borrowers) stdv 11.6 31.1 26.2 13.2N 38 38 38 38Village Banking avg -5.7 103.4 85.7 54.5*(groups with ≥ 10 borrowers) stdv 7.0 27.5 18.7 26.5N 20 20 20 20RETAIL Financial Intermediaries avg 0.0* 114.8 104.3* 34.5FINANCIAL (passbook and time deposits stdv 5.0 26.4 23.4 11.6INTER- ≥ 20% of total assets) N 25 25 25 25MEDIARYOther avg -5.3 106.6 89.2 44.1(passbook and time deposits stdv 10.6 37.5 28.0 21.2< 20% of total assets) N 85 85 85 85TARGET Low-end avg -7.6* 99.6 82.4* 50.8*GROUP (depth < 20% OR avg. loan stdv 11.2 37.7 26.4 24.1Balance < US$150) N 48 48 48 48Broad avg -1.1* 113.6 99.7* 36.6(depth between 20% and 149%) stdv 6.3 25.4 22.0 12.1N 50 50 50 50High-end avg -0.1 121.2 103.7 27.7*(depth ≥ 150%) stdv 4.9 42.2 32.8 5.3N 10 10 10 10Note: Standard deviations and sample sizes are listed below the group averages. The averages are calculated on the basis of the values between the second and the99th percentiles for each group; therefore, sample sizes vary across indicators. Group averages different from average for all MFIs at 5 percent significance level aremarked with an asterisk (*). Additional statistical information is available at www.calmeadow.com.50 MICROBANKING BULLETIN, SEPTEMBER 2000

BULLETIN HIGHLIGHTS AND TABLESADMINEXPENSE/ LPtotal admin exp. /avg. loan portfolio(%)SALARYEXPENSE/ LPPORTFOLIOAT RISKDEPTHAVERAGESALARYSTAFFPRODUCTIVITYCOST PERBORROWERavg. loan avg. staff salary /staff exp. / avg. outstanding balance balance/ GNP GNP per capita present borrowers / total admin. exp. /loan portfolio overdue > 90 days / total per capita (multiple of GNP/ no. of staffavg. no. of(%)loan portfolio (%)(%)capita)(no.)borrowers (US$)40.3* 22.6 1.0* 93.1* 6.2 79* 290*29.1 18.6 1.5 90.1 3.8 45 24822 22 22 22 22 22 1744.7* 24.9* 1.6 57.1 7.1* 122 11425.3 15.2 1.7 76.8 5.4 55 10822 22 19 22 22 22 2128.1 15.6 2.7* 49.8 4.8 142* 88*20.3 13.0 2.4 50.7 4.0 119 7864 64 57 63 62 62 4916.7* 8.9* 2.0 95.6* 6.1 132 1428.8 5.5 1.1 91.5 4.6 69 11820 20 18 20 20 20 1931.3 17.3 2.3 72.8* 5.9 118 15318.3 11.4 2.9 102.7 4.7 88 18765 65 59 64 63 63 4759.1* 34.1* 2.1 20.8* 4.4 116 8734.0 21.6 2.0 15.3 3.9 66 8623 23 21 23 23 23 2120.9* 10.2* 2.2 108.9* 4.8 96 21711.7 6.5 2.0 122.4 4.0 73 20550 50 49 50 50 50 3945.7* 26.4* 2.2 37.9 6.7* 121 9632.7 20.4 2.3 32.0 5.5 68 6838 38 33 38 37 37 3149.1* 29.7* 1.6 16.3* 5.5 190* 40*24.0 14.9 1.6 12.3 3.6 144 2020 20 16 20 20 20 1719.3* 9.3* 2.2 75.7* 5.1 116 11813.7 6.4 1.1 56.8 4.4 84 9025 25 23 25 25 25 1739.4* 22.5* 2.2 62.6 5.8 126 14227.7 17.6 2.8 99.7 4.8 101 16985 85 76 84 84 84 7151.3* 29.7* 2.1 16.0* 4.8 162* 63*30.8 19.9 2.7 9.8 4.8 119 6048 48 40 47 47 47 4022.9* 12.2* 2.5 65.4* 5.8 105 14212.4 7.4 1.9 32.2 4.3 64 10750 50 48 50 50 50 3817.1* 9.0* 0.7* 307.1* 7.9* 47* 374*8.3 6.2 0.7 147.4 3.8 21 21610 10 10 10 10 10 9MICROBANKING BULLETIN, SEPTEMBER 2000 51

BULLETIN HIGHLIGHTS AND TABLESComposition of Additional Analysis GroupingsAGENew(1 - 2years)AlAmanaAMKBanco do PovoBanPeqEmpresaBASIXBOSPOCONSTANTAFEFADKASHFLOKMC-SEAMEBMFWMIKROFINMoznostiNachalaNOAPAMÉCASPortosolPRIDE-UgandaSATSUNRISEVivacredWVBYoung(3 - 6years)ACLEDAAgrocapitalAl MajmouaCEAPE/PECERUDEBCitiS&LEnlaceFATENFAULUFINCA EcuadorFINCA KyrgyzstanFINCA MalawiFundusz MikroFOCCASFONDECOInicjatywa MikroNirdhanNetwork Leasing Corp.PADMEPRIDE-TanzaniaProMujer BoliviaRFFSHAREWAGESMature(> 6years)15 de Abril23 de JulioABAACEPACODEPACREDICOMACTUARADOPEMADRIAGAPEAKRSPASABAACBanADEMIBancoSolBDBBRACBRIBURO, TangailCALPIÁCAMCDSCHISPAChuimequenáCM ArequipaCMMMEDCompartamosCONTIGOCOOSAJOCRECERECOSABAEMPRENDEREMTFAMAFEDFIEFinaméricaFINCA HondurasFINCA MexicoFINCA NicaraguaFINCA UgandaFMMPopFUNADEHFWWBCaliFWWBIndiaHublagKafo JiginewLosAndesMibancoMOYUTANNyésigisoOscusPRIDE VitaPRODEMProEmpresaRSPISagrarioSartawiSEEDSSEFTONANTELTSPITulcánUNRWAUWFTWR HondurasSCALE OF OPERATIONSLarge(Portfolio >US$ 8million)ABA\ACEPACLEDAAgrocapitalASABAACBanADEMIBancoSolBDBBRACBRICALPIÁCERUDEBCM ArequipaCOOSAJOFIEFinaméricaFundusz MikroFWWBCaliLosAndesMibancoPRODEMMedium(PortfolioUS$ 1 to 8million)15 de Abril23 de JulioACODEPACREDICOMACTUARADOPEMADRIAKRSPAlAmanaAMKBanPeqEmpresaBASIXBOSPOBURO, TangailCAMCEAPE/PECHISPAChuimequenáCMMMEDCompartamosCONTIGOCRECERECOSABAEMPRENDEREMTEnlaceFAMAFATENFEDFEFADFINCA HondurasFINCA KyrgyzstanFINCA NicaraguaFINCA UgandaFMMPopFONDECOFUNADEHKafo JiginewLOKMC-SEAMEBMIKROFINMoznostiNachalaNetwork Leasing Corp.NOANyésigisoOscusPADMEPAMÉCASPortosolPRIDE-TanzaniaPRIDE-UgandaPRIDE VitaProEmpresaProMujer BoliviaSagrarioSartawiSEEDSSHARESUNRISETONANTELTSPITulcánUNRWAWRHondurasWVBSmall(Portfolio

BULLETIN HIGHLIGHTS AND TABLESComposition of Additional Analysis Groupings, ctd.RETAIL FINANCIAL INTERMEDIATIONRetailFinancialIntermediary(passbook andtime deposits ≥20% of totalassets)15 de Abril23 de JulioACREDICOMBAACBanADEMIBancosolBDBBRIBURO, TangailCERUDEBChuimequenáCitiS&LCM ArequipaCOOSAJOECOSABAEnlaceFIEFinaméricaKafo JiginewLosAndesMOYUTANNyésigisoOscusPAMÉCASSagrarioTONANTELTulcánUWFTOther(passbook andtime deposits

APPENDICESAPPENDICESAppendix I: Notes to Statistical SectionThe MicroBanking Standards Project, of which TheMicroBanking Bulletin is a major output, is open toall MFIs that are willing to disclose financial datathat meet a simple quality test. Participating MFIstypically have three characteristics: 1) they arewilling to be transparent by submitting theirperformance data to an independent agency; 2)they display a strong social orientation by providingfinancial services to low-income persons; and 3)they are able to answer all the questions needed forour analysis.The one hundred and fourteen institutions thatprovided data for this issue represent a largeproportion of the world’s leading microfinanceinstitutions. They have provided data generally bycompleting a detailed questionnaire, supplementedin most cases by additional information. Allparticipating MFIs receive a customized reportcomparing their results with those of the peergroups.presented, or for consequences resulting from theiruse. We employ a system to make tentativedistinctions about the quality of data presented tous and include only information for which we have areasonable level of comfort. However, we cannotexclude the possibility of a programmisrepresenting its results.The most delicate areas of potential distortion are:(1) unreported subsidies and (2) misrepresentedloan portfolio quality. There can also beinaccuracies in reporting the costs of financialservices in multipurpose institutions that alsoprovide non-financial services, in part because ofdifficulties in assigning overhead costs. These risksare highest for younger institutions, and forinstitutions with a record of optimistic disclosure. Ifwe have grounds for caution about the reliability ofan MFI’s disclosure, we will not include itsinformation in a peer group unless it has beenexternally validated by a third-party.Data Quality IssuesThe Bulletin classifies information from participatinginstitutions according to the degree to which wehave independent verification of its reliability. AAAratedinformation has been independentlygenerated through a detailed financial analysis byan independent third party, such as a CAMELevaluation, a CGAP appraisal, or assessments byreputed rating agencies. A-rated information isbacked by accompanying documentation, such asaudited financial statements, annual reports, andindependent program evaluations that provide areasonable degree of confidence for ouradjustments. B-rated information is from MFIs thathave limited themselves to completing ourquestionnaire. These ratings signify confidencelevels on the reliability of the information; they areNOT intended as a rating of the financialperformance of the MFIs.The criteria used in constructing the StatisticalTables are important for understanding andinterpreting the information presented. Given thevoluntary nature and origin of the data,CALMEADOW, the Editorial Board, and CGAP cannotaccept responsibility for the validity of the resultsAdjustments to Financial DataThe Bulletin adjusts the financial data it receives toensure comparable results. The financialstatements of each organization are converted tothe standard chart of accounts used by the Bulletin.This chart of accounts is simpler than that used bymost MFIs, so the conversion consists mainly ofconsolidation into fewer, more general accounts.Then three adjustments are applied to produce acommon treatment for the effect of: a) inflation, b)subsidies, and c) loan loss provisioning and writeoff.In the statistical tables the reader can compareadjusted and unadjusted results.InflationThe Bulletin reports the net effect of inflation bycalculating increases in expenses and incomes dueto inflation. Inflation causes a decrease in the realvalue of equity. This “cost of funds” is obtained bymultiplying the prior year-end equity balance by thecurrent-year inflation rate. 15 Fixed asset accounts,on the other hand, are revalued upward by the15 Inflation data are obtained from line 64x of the InternationalFinancial Statistics, International Monetary Fund, various years.MICROBANKING BULLETIN, SEPTEMBER 2000 55

APPENDICEScurrent year’s inflation rate, which results in inflationadjustment income, offsetting to some degree theexpense generated by adjusting equity. 16 On thebalance sheet, this inflation adjustment results in areordering of equity accounts: profits areredistributed between real profit and the nominalprofits required to maintain the real value of equity.MFIs that borrow from banks or mobilize savingshave an actual interest expense, which is anoperating cost. In comparison, similar MFIs thatlend only their equity have no interest expense andtherefore have lower operating costs. If an MFIfocuses on sustainability and the maintenance of itscapital/asset ratio, it must increase the size of itsequity in nominal terms to continue to make thesame value of loans in real (inflation-adjusted)terms. Inflation increases the cost of tangible itemsover time, so that a borrower needs more money topurchase them. MFIs that want to maintain theirsupport to clients must therefore offer larger loans.Employees’ salaries go up with inflation, so theaverage loan balance and portfolio must increase tocompensate, assuming no increase in interestmargin. Therefore, a program that funds its loanswith its equity must maintain the real value of thatequity, and pass along the cost of doing so to theclient. This expectation implies MFIs should “pay”interest rates that include the inflation-adjustmentexpense as a cost of funds, even if this cost is notactually paid to anyone outside the institution.Some countries with high or volatile levels ofinflation require businesses to use inflation-basedaccounting on their audited financial statements.We use this same technique in the Bulletin. Ofcourse, we understand that in countries where highor volatile inflation is a new experience, MFIs mayfind it difficult to pass on the full cost of inflation toclients. We are not recommending policy; rather,we are trying to provide a common analyticalframework that compares real financialperformance meaningfully.SubsidiesWe adjust participating organizations’ financialstatements for the effect of subsidies byrepresenting the MFI as it would look on anunsubsidized basis. We do not intend to suggestwhether MFIs should or should not be subsidized.Rather, this adjustment permits the Bulletin to seehow each MFI would look without subsidies forcomparative purposes. Most of the participatingMFIs indicate a desire to grow beyond the16 In fact, an institution that holds fixed assets equal to its equityavoids the cost of inflation that affects MFIs, which hold much oftheir equity in financial form.limitations imposed by subsidized funding. Thesubsidy adjustment permits an MFI to judgewhether it is on track toward such an outcome. Afocus on sustainable expansion suggests thatsubsidies should be used to enhance financialreturns. The subsidy adjustment simply indicatesthe extent to which the subsidy is being passed onto clients through lower interest rates or whether itis building the MFI’s capital base for furtherexpansion.The Bulletin adjusts for three types of subsidies: (1)a cost-of-funds subsidy from loans at below-marketrates, (2) current-year cash donations to fundportfolio and cover expenses, and (3) in-kindsubsidies, such as rent-free office space or theservices of personnel who are not paid by the MFIand thus not reflected on its income statement.Additionally, for multipurpose institutions, TheMicroBanking Bulletin attempts to isolate theperformance of the financial services program,removing the effect of any cross subsidization.The cost-of-funds adjustment reflects the impact ofsoft loans on the financial performance of theinstitution. The Bulletin calculates the differencebetween what the MFI actually paid in interest on itssubsidized liabilities and the deposit rate for eachcountry. 17 This difference represents the value ofthe subsidy, which we treat as an additionalfinancial expense. We apply this subsidy to thoseloans to the MFI that are priced at less than 75percent of prevailing market (deposit) rates. Thedecreased profit is offset by generating an“accumulated subsidy adjustment” account on thebalance sheet.If the MFI passes on the interest rate subsidy to itsclients through a lower final rate of interest, thisadjustment may result in an operating loss. If theMFI does not pass on this subsidy, but instead usesit to increase its equity base, the adjustmentindicates the amount of the institution’s profits thatwere attributable to the subsidy rather thanoperations.Loan Loss ProvisioningFinally, we apply standardized policies for loan lossprovisioning and write-off. MFIs vary tremendously17 Data for shadow interest rates are obtained from line 60l of theInternational Financial Statistics, IMF, various years. Thedeposit rate is used because it is a published benchmark in mostcountries. Sound arguments can be made for use of differentshadow interest rates. NGOs that wish to borrow from bankswould face interest significantly higher than the deposit rate. Alicensed MFI, on the other hand, might mobilize savings at alower financial cost than the deposit rate, but reserverequirements and administrative costs would drive up the actualcost of such liabilities.56 MICROBANKING BULLETIN, SEPTEMBER 2000

APPENDICESin accounting for loan delinquency. Some count theentire loan balance as overdue the day a paymentis missed. Others do not consider a loan delinquentuntil its full term has expired. Some MFIs write offbad debt within one year of the initial delinquency,while others never write off bad loans, thus carryingforward a hard-core default that they have littlechance of ever recovering.We classify as “at risk” any loan with a paymentover 90 days late. We provision 50 percent of theoutstanding balance for loans between 90 and 180days late, and 100 percent for loans over 180 dayslate. Wherever we have adequate information, weadjust to assure that all loans are fully written offwithin one year of their becoming delinquent.(Note: We apply these provisioning and write-offpolicies for ease of use and uniformity. We do notrecommend that all MFIs use exactly the samepolicies.) In most cases, these adjustments are notvery precise. Nevertheless, most participating MFIshave high-quality loan portfolios, so loan lossprovision expense is not an important contributor totheir overall cost structure. If we felt that a programdid not fairly represent its general level ofdelinquency, and we were unable to adjust itaccordingly, we would simply exclude it from thepeer group.Financial Statement Adjustments and their EffectsAdjustment Effect on Financial Statements Type of Institution Most Affectedby AdjustmentInflation adjustment of equityReclassification of certain long termliabilities into equity, and subsequentinflation adjustmentSubsidy adjustment: Interest savingson subsidized liabilities involving atleast a 25 percent discount in relationto market based loans to the sameinstitution or, in the absence of suchloans, the deposit rateSubsidy adjustment: Current-yearcash donations to cover operatingexpensesSubsidy adjustment: In kind donationof goods or services (e.g., line staffpaid for by technical assistanceproviders)Loan loss provision and write-offadjustment: Applying policies whichmay be more aggressive than the MFIemploys on its own booksIncreases financial expense accountson profit and loss statement, to somedegree offset by inflation incomeaccount for revaluation of fixed assets.Generates inflation adjustment accountin equity section of balance sheet withnet balance of inflation adjustments.Decreases concessionary loan accountand increases equity account;increases inflation adjustment on profitand loss statement and balance sheet.Increases financial expense on profitand loss statement. Increases subsidyadjustment account on balance sheet.Reduces operating income on profit andloss statement (if the MFI recordsdonations as operating income).Increases subsidy adjustment accounton balance sheet.Increases expense on profit and lossstatement, increases subsidyadjustment account on balance sheet.Increases loan loss provision expenseon profit and loss statement. Onbalance sheet, increases in loan lossreserve and/or write-offs are accountedfor by equal reductions in loan lossreserve and portfolio.NGOs funded more by equity thanby liabilities will be hard hit,especially in high-inflationcountries.NGOs that have long-term lowinterest“loans” from internationalagencies that function more asdonations than loans.Banks or NGOs that use large linesof credit from governments orinternational agencies at highlysubsidized rates.NGOs during their start-up phase.This adjustment is relatively lessimportant for mature institutionsincluded in this edition.NGOs during their start-up phase.Less important for matureinstitutions included in this edition.MFIs that allow bad loans toaccumulate within their portfolio.This common problem tends tohave a limited effect on leadingMFIs because their loan losses arelow, even after adjustment.MICROBANKING BULLETIN, SEPTEMBER 2000 57

APPENDICESStatistical IssuesThe Bulletin reports the means and standarddeviations of the performance indicators for eachpeer group. At this stage, peer groups are stillsmall and the observations in each peer groupshow a high variation. Outliers distort the results ofsome of the peer group averages. Consequently,the reader should be cautious about the interpretivepower of these data. Over time, as more MFIsprovide data, we will be in a better position togenerate deeper and more sophisticated types ofanalyses of the data at our disposal, and will have ahigher degree of comfort with the statisticalsignificance of the differences between the meansof the distinct peer groups.To ensure that the averages reported represent thegroup as accurately as possible, we have excludedoutliers for each of the indicators. Statistics for thecategory All MFIs were calculated by deletingobservations in the first and last deciles for eachindicator. In other words, the values between the11th and 89th percentiles were used for theanalysis. For the FSS sample and peer groupcalculations, the first and last percentileobservations were excluded for each indicatorexcept macroeconomic indicators. The averagesare calculated on the basis of the values betweenthe 2nd and the 99th percentiles for each group. Ineffect, for each indicator we rank the MFIs in thegroup and eliminate the top and bottom values. Inmost cases, this exclusion eliminates twoobservations for each peer group: the institutionwith the highest and the lowest value on eachindicator. In cases where indicators containobservations with tied values for highest and lowestvalues, more than two observations are deleted.For this reason, we have reported the sample sizefor each group and indicator on the tables. Wherethe sample size is reduced to n=1, we have notreported the result so as to maintain confidentiality.This method helps to prevent outliers fromdominating group results, and smoothes the data byminimizing data dispersion.We have carried out statistical tests to determinethe impact of outliers where they exist, and toquantify the results in terms of how well theyrepresent the peer groups. Where large differencesexist between the means of different peer groups orgroups sorted by selection criteria, we have verifiedtheir statistical significance using t-tests. Thesetests compare the mean of the group to the mean ofall MFIs in the sample, taking into account factorslike the number of observations and the dispersionof the sample. The test statistic is then comparedto a standard critical level (using 5 percent as thesignificance level) to decide whether the differencebetween the group and the sample as a whole isstatistically significant. In other words, they allowus to decide whether the difference we see isrobust, by considering it in the context of howcohesive and how large the group is.58 MICROBANKING BULLETIN, SEPTEMBER 2000

APPENDICESAppendix II: Description of Participating MFIsACRONYM NAME, LOCATION DATE15 de Abril Cooperativa 15 de Abril,Ecuador23 de Julio Cooperativa 23 de Julio,EcuadorABAACEPACLEDAACODEPACREDICOMACTUARADOPEMADRIAGAPEAgrocapAKRSPAl MajmouaAl AmanaAlexandria BusinessAssociation,EgyptAgence de Crédit pourl’Enterprise Privée,SenegalAssociation of CambodianLocal EconomicDevelopment Agencies,CambodiaAsociación de Consultorespara el Desarrollo de laPequeña, Mediana yMicroempresa,NicaraguaACREDICOM,GuatemalaCorporación Acción por elTolima - ACTUARFamiempresas,ColombiaAsociación Dominicanapara el Desarrollo de laMujer,Dominican RepublicAsociación para elDesarrollo Rural Integrado,Costa RicaAsociación General paraAsesorar PequeñasEmpresas,ColombiaFundación Agrocapital,BoliviaAga Khan Rural SupportProgramme,PakistanLebanese Association forDevelopment -- AlMajmoua,LebanonAssociation Al Amana,MoroccoDATAQUALITYRATINGDESCRIPTION OF MICROFINANCE PROGRAM06/99 AAA 15 de Abril is a credit union in Ecuador that has participated inWOCCU’s technical assistance program since in 1995. 15 de Abriloffers both credit and voluntary savings services to members.06/99 AAA 23 de Julio participates in WOCCU’s technical assistance program inEcuador. It is a credit union offering credit and savings services tomembers.12/99 AAA ABA provides credit to small and microenterprises using anindividual lending methodology. It is an NGO founded in 1988 andbased primarily in urban areas. The credit program began in 1990.12/99 B ACEP began as an NGO in a provincial town in 1987 and hasexpanded to operate in other urban areas in Senegal. It hasconverted to a credit union.12/99 AAA ACLEDA was started in 1993 as an NGO. It provides small andmicro loans to enterprises and trains entrepreneurs in smallbusiness management. Both group and individual loans are made.12/98 B Founded in 1989, ACODEP serves small and micronterprisesprimarily in Managua and other urban areas of Nicaragua. It iscurrently negotiating a voluntary supervision agreement with theSuperintendent of Banks in Nicaragua.09/99 AAA ACREDICOM is a member of the FENACOAC credit union system inGuatemala, and participated in WOCCU’s technical assistanceprogram. It primarily lends for agriculture and to a lesser extentmicroenterprise activities, and mobilizes savings from members.12/99 B ACTUAR Tolima was founded in 1986. It is an NGO offering loansto microenterprises in Tolima and surrounding areas, and is affiliatedwith ACCION International and Cooperativa Emprender in Colombia.12/99 A ADOPEM, an affiliate of Women’s World Banking, is an NGOdedicated to credit for women microentrepreneurs. It has been inoperation since 1982.08/99 A ADRI is an NGO offering loans to small and microenterprises inCosta Rica. Founded in 1986, it also offers training and businessdevelopment services to its clients.12/99 A Founded in 1975, AGAPE operates principally in Barranquilla,offering microcredit through a mixture of methodologies includingvillage banking, solidarity groups and individual loans. It is anaffiliate of Opportunity International.12/99 AAA Fundación Agrocapital focuses its services on agriculture and agroindustry,working mainly in rural and small urban areas of Bolivia. Itis an NGO founded in 1992, and offers a mixture of microloans andlonger term mortgage loans.12/98 A AKRSP is a multi-service NGO that works in the “Roof of the World”region of northern Pakistan. Its credit program began in 1983,offering loans through its network of village organizations.12/98 A Al Majmoua is a Lebanese NGO, offering village banking-typeservices in both urban and rural areas. The program beganoperations in 1994 as a project of Save the Children. Ownershipwas transferred to the Lebanese institution in 1998.12/98 AAA Al Amana offers solidarity group loans through a wide network ofbranches in urban areas of Morocco. Founded in 1997, it is anaffiliate of Pride Vita.MICROBANKING BULLETIN, SEPTEMBER 2000 59

APPENDICESACRONYM NAME, LOCATION DATEAMKASABAACBanADEMIBanco doPovo de Juizde ForaBancoSolBancoPeqEmp(BPE)Bank DagangBASIXBOSPOBRACBRIBURO,TangailCalpiáCAMAMK Posusje,Bosnia and HerzegovinaAssociation for SocialAdvancement,BangladeshBank for Agriculture andAgricultural Cooperatives,ThailandBanco de DesarrolloADEMI, S.A.,Dominican RepublicBanco do Povo de Juiz deFora,BrazilBanco Solidario, S.A.,BoliviaBanco de la PequeñaEmpresa, S.A.,Dominican RepublicBank Dagang Bali,IndonesiaBharatiya SamruddhiFinance Ltd.,IndiaBOSPO,Bosnia and HerzegovinaBangladesh RuralAdvancement Committee,BangladeshBank Rakyat Indonesia,Unit Desa System,IndonesiaBURO, Tangail,BangladeshFinanciera Calpiá, S.A.,El SalvadorCentro de Apoyo a laMicroempresa,El SalvadorDATAQUALITYRATINGDESCRIPTION OF MICROFINANCE PROGRAM12/99 B AMK is a limited liability company founded in 1997 to providemicrocredit to low income self-employed individuals in urban areas.It is financed by the Local Initiatives Department of the World Bankthat aims to improve access to credit to the poor to promoteeconomic reconstruction.12/99 AAA ASA is an NGO that offers credit services to the rural poor inBangladesh. The majority of its clients are landless women. It wasfounded in 1978 and shifted from an earlier, integrated developmentstrategy to its current focus on financial services in the early 1990s.It uses a village level group lending methodology.03/98 AAA BAAC is a government-owned agricultural bank that lends to smallfarmers and farmers’ cooperatives. Founded in 1966, its outreach inrural areas of Thailand is now estimated to cover more than 80% offarm families.12/99 A Banco ADEMI is a formal financial institution, which beganoperations in 1998. The bank is the successor to the NGO, ADEMI,which was involved in microcredit since 1982.06/99 AAA Banco do Povo de Juiz de Fora is an NGO operating in Juiz de Forain Brazil. It offers individual loans to microentrepreneurs and wasfounded in1997. It was formerly known as FAEP.12/99 A BancoSol is a licensed commercial bank devoted to microfinance,offering microenterprise credit and passbook savings. Its creditprogram focuses on group loans, and it operates primarily in urbanareas of Bolivia. It grew out of the NGO PRODEM and was spun offas a bank in 1992. It is an affiliate of ACCION International.12/98 AAA Banco de la Pequeña Empresa was created to serve bothmicroenterprises and small businesses, and has just completed itsfirst year of operations. It s a formal financial sector institution andholds a license to operate as a development bank. It is an affiliate ofACCION International.12/98 AAA Bank Dagang is a private commercial bank that offers savings andcredit facilities to primarily low-income clients in Bali. It was foundedin 1970.03/99 AAA BASIX was set up in 1996 to provide financial services to the ruralpoor, to promote self-employment, and to provide technicalassistance to microentrepreneurs and rural financial institutions.12/99 B BOSPO is a NGO founded in 1995 to provide microcredit to solidaritygroups made of low income women entrepreneurs in secondarycities of Tuzla. It is financed by the Local Initiatives Department ofthe World Bank that aims to improve access to credit to the poor topromote economic reconstruction.12/99 AAA BRAC is an NGO that started in 1972. It provides both financial andnon-financial services primarily in rural areas. The financial servicesinclude the provision of microloans and mobilization of savings.12/99 AAA BRI is a government-owned bank oriented towards rural areas,which has operated since 1897. The Unit Desa system is anextensive network of small banking units, which function as profitcenters and provide individual loans and savings services. Thesystem has existed in its current form since 1984.12/99 AAA Flexible voluntary open-savings, microloans and insurance servicesare provided by BURO Tangail since 1990. It is an NGO.12/98 AAA Financiera Calpiá began as an NGO, AMPES, and was convertedinto a finance company in 1995. It offers individual loans tomicroenterprises and small businesses and has started to mobilizesavings. It operates mainly in urban areas, although 25% of itsportfolio is now in rural areas.12/99 B FINCA’s affiliate in El Salvador, the CAM was founded in 1990 and isone of FINCA’s largest affiliates serving over 16,000 clients in all 15geographic departamentos in El Salvador.60 MICROBANKING BULLETIN, SEPTEMBER 2000

APPENDICESACRONYM NAME, LOCATION DATECDSCommunity DevelopmentSociety,IndiaDATAQUALITYRATINGDESCRIPTION OF MICROFINANCE PROGRAM03/99 A CDS offers microcredit and non-financial services in the Nagpurregion of India. It was founded in 1985 and is an affiliate ofOpportunity International.CEAPE/PECERUDEBCiti S&LCHISPAChuimequenáCM ArequipaCMM/MedCOMPARTCONSTANCONTIGOCOOSAJOCRECERECOSABAEmprenderEMTCentro de Apoio aosPequeños EmpreendimentosPernambuco,BrazilCentenary RuralDevelopment Bank,UgandaCiti Savings & Loans,GhanaFundación Chispa,NicaraguaCooperativa San MiguelChuimequená,GuatemalaCajas Municipales deArequipa,PeruCorporación Mundial de laMujer Medellín, Medellín,ColombiaAsociación ProgramaCompartamos, I.A.P.,MexicoConstanta,GeorgiaFundación CONTIGO,ChileCooperativa San JoséObrero,GuatemalaCRECER,BoliviaECOSABA,GuatemalaEmprender Buenos Aires,ArgentinaEnnathian MoulethanTchonnebat,Cambodia06/99 AAA CEAPE Pernambuco is an urban-based microenterprise creditprogram. A member of the FENAPE network in Brazil, and ofACCION International, it was founded in 1992.12/99 A CERUDEB was founded as a trust company in 1983, and obtainedits banking license in 1992. It received technical assistance fromIPC from 1993-98, and its current shareholders are the UgandaCatholic Secretariat, the Catholic Dioceses of Uganda, Hivos-TriodosFond and SIDI. CERUDEB provides credit and savings services inKampala and Uganda’s district towns.12/99 B Citi Savings is a private non-bank financial institution that operates inGreater Accra, Ghana. It lends to rotating savings and creditassociations (susu clubs) and informal savings collectors, andmobilizes savings from the public.06/98 AAA Founded in 1991, CHISPA works primarily in urban areas ofNicaragua. It is affiliated with the Mennonite Economic DevelopmentAssociation (MEDA).09/99 AAA San Miguel Chuimequená is a Guatemalan credit union. It is amember of the FENACOAC system and it participates in WOCCU’stechnical assistance program. It offers loans and savings services toits members.12/98 AAA The municipal savings and credit banks of Peru are owned by citygovernments. Arequipa is one of the largest and most successfulbanks of the national network, and offers pawn and microenterpriseloans as well as savings products.12/99 A CMM Medellín is affiliated to the Women’s World Banking network,and operates in Medellín and surrounding areas. It was founded in1985 and lends to both men and women.12/99 B Compartamos is the lending arm of Gente Nueva, a Mexican NGOthat was founded in 1985. The program uses a village bankingmethodology focusing on women, in rural and semi-urban areas ofMexico. It began lending in 1990.12/98 A Constanta was established in 1997 with a grant from UNHCR/Savethe Children as a local NGO to provide group loans to poor selfemployedwomen.12/99 A CONTIGO began lending operations in 1989, and offers creditservices to microentrepreneurs in communities in the south ofSantiago de Chile.09/99 AAA San José Obrero is a member of the FENACOAC credit unionfederation, and participated in WOCCU’s technical assistanceprogram in Guatemala. It offers loans and savings services to itsmembers.06/99 B CRECER is an NGO working primarily in rural areas of Bolivia. Itparticipates in Freedom from Hunger’s “Credit with Education”program, using a village banking methodology.09/99 AAA ECOSABA is a member of the FENACOAC credit union federation,and participated in WOCCU’s technical assistance program inGuatemala. It offers loans and savings services to its members.04/99 A Emprender, founded in 1992, is an ACCION affiliate that offersmicroenterprise credit in urban areas of Argentina. The majority of itslending is to solidarity groups.12/99 A EMT was founded in 1991 as a rural credit project run by the Frenchagency, GRET. It is in the process of transformation to anindependent Institution, and operates in rural areas in the south ofCambodia. It offers individual and solidarity group loans.MICROBANKING BULLETIN, SEPTEMBER 2000 61

APPENDICESACRONYM NAME, LOCATION DATEENLACEFAMAFATENFAULUFEDFEFADFIEFinaméricaFINCA ECFINCA HOFINCA KYFINCA MAFINCA MXFINCA NIFINCA UGFMM PopFOCCASFONDECOFUNADEHPrograma ENLACE, BancoSolidario,EcuadorFundación de Apoyo a laMicroempresa,NicaraguaPalestine for Credit andDevelopment,West Bank and GazaFood for the HungryInternational,UgandaFundación Ecuatoriana deDesarrollo,EcuadorFoundation for EnterpriseFinance and Development,AlbaniaFFP - Fomento aIniciativas Económicas,S.A.,BoliviaFinanciera América, S.A.,ColombiaFINCA Ecuador,EcuadorFINCA Honduras,HondurasFINCA Kyrgyzstan,KyrgyzstanFINCA Malawi,MalawiFINCA México,MexicoFINCA Nicaragua,NicaraguaFINCA Uganda,UgandaFundación Mundo MujerPopayán,ColombiaFoundation for Credit andCommunity Assistance,UgandaFondo de DesarrolloComunal,BoliviaFundación Nacional parael Desarrollo de Honduras,HondurasDATAQUALITYRATINGDESCRIPTION OF MICROFINANCE PROGRAM09/99 B ENLACE is the microfinance division of Banco Solidario in Ecuador.The program was founded in 1995, and Banco Solidario receivestechnical assistance from ACCION International. ENLACE offersboth credit and savings services to microentrepreneurs. It alsoadministers a pawn-lending product.12/99 A FAMA operates mainly in urban areas of Nicaragua, providingmicroenterprise credit. It was founded in 1991 and is affiliated withACCION.12/99 A FATEN was initiated as a Save the Children affiliate in 1995 andspun-off as an independent NGO in 1999. It provides microcredit topoor women entrepreneurs using group methodology.12/99 B Founded in 1995 as an affiliate of Food for the Hungry International,FAULU provides group based credit and voluntary deposit servicesto small and microentrepreneurs in urban and semi-urban areas.12/99 A Founded over 30 years ago, FED has an extensive branch networkthroughout Ecuador providing individual microloans. It is an affiliateof ACCION International.12/98 A Operating mainly in urban areas of Albania, FEFAD offers smallbusiness loans. It was founded in 1995 as an initiative of theAlbanian and German governments, and receives technicalassistance from IPC.12/99 A FFP - FIE is a for-profit financial institution offering individual loans tomicroenterprises in urban areas of Bolivia. It began lending in 1988as an NGO, and began operating as a “Private Financial Fund” in1998 under regulation by the Bolivian Superintendency of Banks.12/99 AAA Finamérica is a regulated finance company operating in Bogotá andsurrounding areas. Its predecessors were the NGO Actuar Bogotá,founded in 1988, the NGO Corposol, and the financiera Finansol. Itis an affiliate of ACCION International.12/98 B FINCA Ecuador was founded in 1994 and provides village bankingservices to low-income families in three regions of the country:Pichincha, Guayas, and Imbabura.12/98 B FINCA Honduras is one of the largest FINCA affiliates in terms ofportfolio size. It was founded in 1989 and operates in 13 of the 18departamentos of Honduras.12/99 A Founded in 1995, FINCA Kyrgyzstan is operating in five of the sixoblasts of Kyrgyzstan and offers both village banking and individualloan products to 10,000 clients.08/99 A FINCA Malawi works with women in the country’s southern region,and has been in operation since 1994.12/98 B FINCA Mexico currently operates village banking groups in the stateof Morelos. It was founded in 1989.06/99 A FINCA’s Nicaraguan affiliate began lending in 1992, and has sinceexpanded to have branch offices in several urban areas inNicaragua.12/99 AAA One of FINCA’s largest programs, FINCA Uganda has been inoperation since 1992. The program offers village banking services towomen in Kampala, Jinja and Lira.12/99 B FMM Popayán is a Women’s World Banking affiliate working in thestate of Cauca in Colombia. It began lending to microenterprises in1985.12/98 B FOCCAS, an affiliate of Freedom from Hunger, operates a villagebanking-style program in Uganda’s district towns and villages. It isbased on a credit with education model.12/99 A FONDECO is an NGO working primarily in rural areas in Bolivia. Itwas founded in 1995.12/97 AAA FUNADEH works with small and microenterprises in urban areas ofHonduras. It is an affiliate of ACCION International and was foundedin 1985.DATA62 MICROBANKING BULLETIN, SEPTEMBER 2000

ACRONYM NAME, LOCATION DATE QUALITYRATINGFunduszMikroFWWB CaliFWWBIndiaHUBLAGInicjatywaMikroKafo JiginewKASHFLos AndesLOKMC-SEAMEBMibancoMicrofundfor WomenMIKROFINMoyutánMoznostiNachalaNetworkLeasingFundusz Mikro,PolandFundación Women’s WorldBanking Cali,ColombiaFriends of WWB,IndiaHUBLAG DevelopmentFinance Programme,PhilippinesInicjatywa Mikro,PolandKafo Jiginew,MaliKashf Foundation,PakistanCaja de Ahorros y CréditosLos Andes,BoliviaLOK SarajevoBosnia and HerzegovinaMercy Corps – ScottishEuropean Aid,Bosnia and HerzegovinaMicroenterprise Bank,BosniaBanco de laMicroempresa,PeruMicrofund for Women,JordanMIKROFIN,Bosnia and HerzegovinaCooperativa Moyután,GuatemalaMoznosti,MacedoniaNachala,BulgariaNetwork LeasingCorporation Ltd.,PakistanDESCRIPTION OF MICROFINANCE PROGRAMAPPENDICES09/99 A Fundusz Mikro began operations in 1995, and now lends tomicroentrepreneurs across Poland through an extensive branchnetwork. It is a member of the MicroFinance Network.12/98 AAA FWWB Cali, an affiliate of Women’s World Banking, began lending in1982. It makes individual loans to urban microenterprises in Cali.03/98 A FWWB India lends to rural women through savings and creditgroups. It was founded in 1982.12/98 A The Hublag Development Finance Programme is the microlendingarm of the Gerry Roxas Foundation. It lends to microenterprises withboth individual and group lending methodologies, and beganoperations in 1987.12/99 A Inicjatywa Mikro lends to microenterprises mainly in urban areas ofPoland. It is affiliated with Opportunity International.12/99 B Kafo Jiginew is a federation of credit unions operating in rural areasin the south-central region of Mali. It was founded in 1987.03/00 A KASHF is a NGO founded in 1996 to provide microcredit to lowincome women entrepreneurs in rural and urban areas. It is anaffiliate of ASA, Bangladesh.12/99 A Caja Los Andes grew out of ProCrédito, an NGO that began lendingoperations in 1992. It was converted to a special finance company in1995. Los Andes operates in urban and some rural areas in Bolivia,providing individual loans and savings services.12/99 B LOK is a NGO founded in 1997 to provide individual credit to smallentrepreneurs in urban and rural areas. It is financed by the LocalInitiatives Department of the World Bank that aims to improve accessto credit to the poor to promote economic reconstruction.12/99 B MC SEA is a NGO that started its operation in 1997 and providesindividual credit to microenterprises in war affected areas. It isfinanced by the Local Initiatives Department of the World Bank thataims to improve access to credit to the poor to promote economicreconstruction.12/98 A The Microenterprise Bank was launched by IPC in 1997 to providefinancial services such as loans, money transfers and depositservices to micro and small enterprises in Bosnia-Herzegovina.12/99 A Mibanco is a commercial microfinance bank offering microenterprisecredit in Lima, and is affiliated with ACCION International. Formerlyoperated as an NGO under the name Acción Comunitaria del Perú,the institution was transformed into a bank in 1998.12/99 B This former Save the Children village banking program in Jordan wasfounded in 1994. It focuses primarily on Palestinian women fromsquatter communities.12/99 B MIKROFIN is an affiliate of CARE international and started itsoperations in 1997. It provides individual and group loans tomicroentrepreneurs in semi-urban areas. It is financed by the LocalInitiatives Department of the World Bank and CARE.09/99 AAA Moyután is a member of the FENACOAC credit union federation, andparticipated in WOCCU’s technical assistance program inGuatemala. It offers loans and savings services to its members.12/99 A Moznosti, an affiliate of Opportunity International, began lending in1996. It operates both in urban and rural areas of Macedonia, andlends to microenterprises and small businesses.12/99 B Nachala, an affiliate of Opportunity International, converted into acooperative in 1998. It operates both in urban and rural areas andmakes individual loans to microenterprises and small businesses forworking capital.06/99 A Network Leasing is a private for profit financial company which offersfinancial services to microentrepreneurs. It uses leasing, amethodology considered compatible with Islamic law, which forbidsborrowing on interest.MICROBANKING BULLETIN, SEPTEMBER 2000 63

APPENDICESACRONYM NAME, LOCATION DATENIRDHANNOANyésigisoOscusPADMEPAMÉCASPortosolPride VitaGuineaPRIDE TZPRIDE UGPRODEMProEmpresaProMujerRFFRSPISagrarioSartawiSATNirdhan Utthan,NepalNOA,CroatiaRéseau Nyésigiso,MaliCooperativa Oscus Ltda.,EcuadorAssociation pour laPromotion et l’Appui auDéveloppement desMicroEntreprises,BeninProgramme d’Appui auxMutuelles d’Epargne et deCrédit au Sénégal,SenegalPortosol,BrazilPride Finance Guinea,Republic of GuineaPromotion of RuralInitiatives andDevelopment Enterprises,TanzaniaPromotion of RuralInitiatives andDevelopment Enterprises,UgandaFundación para laPromoción y Desarrollo dela Microempresa,BoliviaEDYPME ProEmpresa,PeruProMujer,BoliviaRural Finance Facility,South AfricaRangtay Sa PagrangayInc.,PhilippinesCooperativa El Sagrario,Ltda.,EcuadorServicio Financiero Rural,Fundación Sartawi,BoliviaSinapi Aba Trust,GhanaDATAQUALITYRATINGDESCRIPTION OF MICROFINANCE PROGRAM06/99 A Nirdhan is an NGO founded in 1991. It is a Grameen replicateproviding credit and deposit services to the poor. Both compulsoryand voluntary deposits services are offered. The NGO hastransformed into Nirdhan Utthan Bank Limited in July 1999. It is amember of the CASHPOR network.12/98 B NOA, an affiliate of Opportunity International, was started in 1997 toprovide individual and group loans to self employed persons inagriculture and small businesses.12/99 A Established in 1990 as a credit union, Nyésigiso offers credit andsavings services to both men and women in urban and rural areas ofMali. It is a member of the Development International Desjardinsnetwork.06/99 AAA Oscus is a credit union in Ecuador, and it participates in WOCCU’stechnical assistance program. Oscus offers both credit and voluntarysavings services to members.06/99 AAA PADME is an NGO working in urban and peri-urban areas of Benin.It offers loans to small and microenterprises, and was founded in1993.12/99 A Pamécas was established as a credit union in 1996. It offers a widerange of savings and credit services, primarily to women, usingindividual, solidarity and village banking products in urban and periurbanSenegal. It is a member of the Development InternationalDesjardins network.06/99 AAA Portosol is an NGO operating in Porto Alegre in Brazil. It offersindividual loans to microentrepreneurs and was founded in1996.12/98 AAA Pride Vita (or Pride Finance) works primarily in urban and semiurbanareas of Guinea and was founded in 1991.12/99 A PRIDE offers microcredit in urban and semi-urban areas ofTanzania. It was founded in 1993.12/98 A PRIDE in Uganda was started in 1996. It provides microloans toborrowers organized as groups in urban and semi-urban areas ofUganda.12/99 B PRODEM began in 1986 as an NGO offering group loans to urbanmicroenterprises, and was the precursor to BancoSol. When itsurban portfolio was passed to BancoSol in 1992, it began to developa new clientele in rural areas in Bolivia.12/99 A ProEmpresa, formerly the IDESI network, is now operating as aformal financial institution in Peru.12/99 A ProMujer Bolivia was founded in 1991, to provide training and creditto predominantly women clients.03/99 AAA RFF is a non-profit organization offering microcredit in rural areas ofSouth Africa. The institution also operates a separate housing loanprogram for salaried employees. RFF’s microcredit program wasestablished in 1993.12/98 A RSPI, an Opportunity International partner, lends primarily to selfhelpgroups in the Cordillera and Iloco regions of the Philippines.06/99 AAA El Sagrario is a credit union in Ecuador, and participates inWOCCU’s technical assistance program, begun in 1995. It offersboth credit and voluntary savings services to members.12/98 A Fundación Sartawi offers group credit to producers and othermicroenterprises in rural areas of Bolivia. The credit program hasoperated in its current form since 1990.12/98 B The Sinapi Aba Trust is a member of Opportunity International, andoffers individual and group loans both in rural and urban areas ofGhana. It was founded in 1995.64 MICROBANKING BULLETIN, SEPTEMBER 2000

APPENDICESACRONYM NAME, LOCATION DATESEEDSSEFSHARESUNRISETonantelTSPITulcánUNRWAUWFTVivacredWAGESWR HondurasWVBSarvodaya EconomicEnterprises,Sri LankaSmall EnterpriseFoundation,South AfricaSociety for HelpingAwakening Rural poorthrough Education,IndiaSUNRISE SarajevoBosnia and HerzegovinaCooperativa Tonantel,GuatemalaTSPI DevelopmentCorporation,PhilippinesCooperativa Tulcán, Ltda.,EcuadorUnited Nations ReliefWorks Agency,GazaUganda Women’s FinanceTrust,UgandaVivacred,BrazilWomen and Associationsfor Gain both Economicand Social,TogoWorld Relief Honduras,HondurasWorld Vision,BosniaDATAQUALITYRATINGDESCRIPTION OF MICROFINANCE PROGRAM03/99 B SEEDS was established in 1987 to provide loans for employmentcreation and increasing standard of living, to mobilize depositsthrough compulsory and voluntary savings programs and to providelife and natural disaster insurances.06/99 AAA SEF is an NGO working in the Northern Province of South Africa. Itworks with a Grameen methodology to provide loans to rural women,and was founded in 1991.03/99 AAA SHARE lends to women in rural areas of Andhra Pradesh in India. Itis a member of the CASHPOR network.12/99 B SUNRISE is a NGO founded in 1997 to provide individual credit tostart-up and established micro enterprises. It is financed by theLocal Initiatives Department of the World Bank that aims to improveaccess to credit to the poor to promote economic reconstruction.09/99 AAA Tonantel is a member of the FENACOAC credit union federation,and participated in WOCCU’s technical assistance program inGuatemala. It offers loans and savings services to its members.06/99 A TSPI operates in urban and semi-urban areas of the Philippines,offering group loans to microenterprises. It was founded in 1981 andis affiliated to the Opportunity Network, the MicroFinance Networkand CASHPOR, among others.06/99 AAA Tulcán is a credit union in Ecuador, and participates in WOCCU’stechnical assistance program, begun in 1995. It offers both creditand voluntary savings services to members.12/99 B The Income Generation Program of UNRWA lends tomicroenterprises and small businesses in Gaza. It began operationsin 1991.12/99 A Uganda Women’s Finance Trust offers solidarity group and individualloans to women in Kampala and district towns of Uganda. It is anaffiliate of Women’s World Banking.06/99 AAA Vivacred is an NGO operating in Rio de Janeiro in Brazil. It offersindividual loans to microentrepreneurs, and was founded in 1997.12/99 A WAGES serves women in Lomé and surrounding areas, workingwith borrowers’ associations in a village-banking type methodology.It was founded in 1994.09/99 B World Relief, Honduras was founded in 1981 as a NGO. It is part ofCOVELO network and network of NGOs FODIPREH. It offers a mixof individual, solidarity and village banking loan products to women inurban and semi-urban areas in Honduras.09/99 A Founded in 1996 as an affiliate of World Vision, the NGO providesindividual and group loans to self-employed small andmicroentrepreneurs.Note: Sources for macroeconomic country data are the IMF, International Financial Statistics and the World Bank, World DevelopmentIndicators, unless otherwise indicated.MICROBANKING BULLETIN, SEPTEMBER 2000 65