71311 FOCUS NOTE A New Look at Microfinance Apexes T his paper reviews recent experience with apex facilities (defined in Box 1) that support institutions delivering retail financial services to Box 1. What is an apex? An apex is a second-tier (or “wholesale�) fund that channels public resources to multiple retail financial poor and low-income clients.1 CGAP published its providers—typically lenders—in a single country. first study of such apexes 10 years ago (Levy 2002a Apexes provide mainly local currency loans, but may and 2002b). That study did not reach categorical also offer loan guarantees, equity investment, grants for operational cost support, and technical assistance. conclusions about apex effectiveness, but it did raise some serious concerns about the extent to which An apex is not always a standalone institution. It may apexes supported the development of sustainable be housed within a larger organization, such as a development bank. microfinance. Since then, the number of apexes and their total funding have grown a great deal. Apex facilities microfinance—for instance, lending to exporters have become increasingly popular with host-country or small and medium enterprises (SMEs). However, governments, as well as with development finance since this paper is limited to microfinance; “apex� institutions (DFIs) and multilateral agencies, such as here generally refers only to microfinance apexes. the European Commission (EC), the Inter-American Development Bank (IDB), the International Fund This paper does not cover federations of retail for Agricultural Development (IFAD), Germany’s cooperatives or credit unions, even though they Kreditanstalt für Wiederaufbau (KfW), and the World sometimes play a refinancing role. Such federations Bank. A 2008 CGAP mapping exercise identified often provide management or supervision in 76 apexes in 46 countries. In 2009 the largest 15 addition to finance and, therefore, tend to involve of these disbursed US$1.5 billion. By comparison, distinctive operational issues. Finally, some people total disbursements of all cross-border funders call associations of microfinance retailers “apexes,� (including bilateral and multilateral donors, DFIs, but these have no refinancing role and thus fall and microfinance investment vehicles [MIVs]) are outside the scope of this paper. roughly US$3 billion annually (El-Zoghbi, Gähwiler, and Lauer 2011). Not only has the number of Principal Sources apexes grown, but more individual apexes have accumulated enough years of experience to permit Data collection for this paper involved four efforts: some assessment of their usefulness. • The 2008 mapping referred to above, which Clearly, it is time for a fresh look. identified 76 microfinance apexes. • A literature review, including both global and Terminology country-specific studies, as well as reports from No. 80 apex institutions themselves (see Annex 4, June 2012 This paper uses the shorthand term “microfinance� Bibliography). to refer to financial services for low-income people. 2 • A 2010 quantitative survey of 23 apexes, Sarah Forster, Apex facilities exist for other sectors besides representing about 90 percent of all apex funding Eric Duflos, and Richard Rosenberg 1 These “microfinance institutions� assume a variety of organizational forms, including, for instance, not-for-profit nongovernment organizations, commercial banks, finance companies, postal savings banks, and financial cooperatives. 2 The financial services used by low-income people include credit, savings, transfers, and insurance. Today, many prefer to refer to such services in terms of “inclusive finance� or “access to finance� rather than the more common “microfinance,� arguing that the latter term has tended to connote microcredit rather than a full range of services, and specialized microfinance institutions rather than the full range of financial providers. Nevertheless, this paper uses “microfinance� for the sake of conciseness. Note that microfinance apexes have focused much more on supporting credit than on supporting other financial services, and that most institutions receiving apex funding specialize in microcredit. 2 in terms of their loan portfolios. This survey initial findings at a one-day workshop and provided collected information about the apex’s mission, useful inputs for this report.3 portfolio size, performance management, funding sources, and other characteristics using a Structure of This Paper questionnaire and follow-up telephone calls. • Case study analysis of six apexes representing This paper is organized into three parts: different regions and institutional forms (see Table • The first section offers a global picture of microfinance 1). These six apexes represent about 60 percent apex operations, including their size, trends, and of the total global apex portfolio surveyed. They characteristics. It draws primarily on the 2010 survey were judged to offer lessons of global relevance, of 23 major apexes, but also includes some results but the sample is not representative: it is biased from the broader mapping done in 2008. toward larger apexes and those more able to • The second section highlights key issues from provide information. The research team conducted across the six case studies. interviews with 73 stakeholders, including apex • The third section offers conclusions and management, microfinance institutions (MFIs), recommendations about maximizing the chances donors, and investors, as well as analysis of apex for apex success. It draws not only from the and Microfinance Information Exchange (MIX) specific research conducted for this paper, but data on the performance of the MFIs funded. The also from CGAP’s 17 years of experience with research team also conducted field visits to study (and publications regarding) the operations of the operations of Bancoldex, Palli Karma-Sahayak development funding agencies. Foundation (PKSF), Social Fund for Development (SFD), and Small Industries Development Bank of Like CGAP’s previous study of apexes, this report India (SIDBI). A list of those interviewed, which does not attempt to produce categorical conclusions includes apex managers, MFIs, and donors, is about the usefulness of microfinance apexes provided in Annex 1. worldwide. There is great diversity of experience At the outset of the project, CGAP members who among the 76 apexes we identified, and few of fund apexes or have an interest in them formed a them make enough information public to support working group that provided early input into the judgments about their success or failure. Rather, this research design. In March 2011, this group, along paper tries to distil lessons about how to improve with managers from MISFA and SIDBI, reviewed the apex performance. Table 1. Six case-study apexes Name of Apex Country Gross Loan Active MFI Clients Portfolio 2009 2009* (USD million) Small Industries Development Bank of India 847 19,620,000 India (SIDBI) Bancoldex Colombia 462 278,000 Palli Karma-Sahayak Foundation (PKSF) Bangladesh 430 8,260,000 Microfinance Investment Support Facility Afghanistan 69 250,000 for Afghanistan (MISFA) Social Fund for Development (SFD) Egypt 68 218,000 Banque Malienne de Solidarité (BMS) Mali 12 190,000 *Active clients of investee institutions, without implying that the apex is financing all of these clients. 3 The workshop was attended by AFD, AsDB, DFID, EC, GiZ, IFAD, IFC, KfW, UNCDF, and the World Bank as well as managers from SIDBI and MISFA. 3 Table 2. Top 15 apexes 2009, by Globally, apex funding is significant, growing, outstanding portforlio and concentrated. The largest 15 apexes had a SIDBI India $847,080,000 total gross loan portfolio (GLP) of over US$3 billion Nafinsa Mexico $491,195,601 in 2009. Although apexes exist in all regions, they Bancoldex Colombia $462,000,000 are most prevalent in Latin America and South Asia PKSF Bangladesh $429,601,508 (Duflos and El-Zoghbi 2010). PPAF Pakistan $137,705,613 BTS Tunisia $117,908,000 These apexes disbursed close to US$1.5 billion to Fedecredito El Salvador $102,215,808 MFIs in 2009, an increase of 25 percent over 2008. PCFC Philippines $74,743,843 Three quarters of the total portfolio was held by MISFA Afghanistan $69,348,000 the four largest apexes. Almost three quarters of SFD Egypt Egypt $67,889,186 total disbursements came from three apexes: SIDBI, FFSA Kazakhstan $66,344,711 Bancoldex, and PKSF, which are discussed in this JAIDA Morocco $47,500,000 paper as case studies. Funda-Pro Bolivia $39,610,405 RMDC Nepal $27,523,071 Apexes use a wide variety of organizational BfP (FNI) Nicaragua $24,404,795 forms, ranging from small divisions of government $3,005,070,541 development banks to large standalone foundations. Source: CGAP 2010 survey. Bancoldex, BMS, and SIDBI are all specialist units within development banks that have a broader focus on SMEs and industrial development. PKSF was 1. The Global Landscape: Apex established as a foundation specializing in microcredit, Characteristics and Trends though it is now supporting wider livelihoods efforts. SFD is a foundation that funds community-level CGAP’s 2008 mapping identified 76 microfinance infrastructure projects as well as microfinance and apexes. A 2010 survey collected information from 23 small enterprise development. MISFA began as part of the largest of these, representing the vast majority of a government ministry, but always had operational of apex portfolio and disbursements worldwide. independence and a nongovernment board majority. Numerical analysis in this section is based mainly on Table 3 illustrates the range of institutional types with the 15 largest apexes. examples from around the world Table 3. Institutional Type Type Ownership Examples Development Bank Government SIDBI (India), Bancoldex (Colombia), RDB (Cambodia), CFN (Ecuador) Commercial Bank Private BMS (Mali), Bank Andara (Indonesia) Standalone Fund (could be legally Government PCFC (Philippines), DAMU (Kazakhstan), PKSF constituted as a nonprofit foundation or (Bangladesh), NAFINSA (Mexico), SFD (Yemen), a company) PPAF (Pakistan) Quasi- MISFA (Afghanistan), SFD (Egypt) government Jaida (Morocco), Funda-Pro (Bolivia), MITAF (Sierra Leone), MDF (Mongolia), FFSA (Kazakhstan), Private RMDC (Nepal) Donor Project Implementing Unit LID (Bosnia) and MISFA (Afghanistan) were initially Government donor projects before legal entities were created 4 Table 4. Instruments by active portfolio (15 return in terms of strengthened MFI capacity, largest apexes) particularly in early stage markets (see Section 3). Instrument USD million Percentage (2009) Apexes lend to a large number of diverse Debt 2,894 97 institutions. The 15 largest apexes were lending to Equity 49 2 1,650 retail microfinance providers in 2009, across a Guarantees 18 1 wide variety of institutional types, including NGOs, Grants 16 1 cooperatives, microfinance banks, and commercial Total 2,980 100 finance companies. Among these are very large Source: CGAP survey 2010. (Percentages do not add to institutions (e.g., Bandhan, with nearly 3.5 million 100 percent because of rounding.) borrowers, funded by SIDBI in India) and very small institutions (e.g., the portfolio of SFD in Egypt). There tends to be some correlation between an apex’s institutional form and its culture. For example, Apexes’ microfinance operations are apexes housed within development banks often take overwhelmingly focused on MFIs’ credit a closer-to-commercial approach (e.g., Bancoldex services. Although apexes fund a diverse range of will lend to MFIs only if they are profitable). They institutions, they usually finance the expansion of may have in-house staff with the financial expertise access to credit for microentrepreneurs. Apexes’ and skills to carry out financial due diligence and founding policy objectives have typically included monitoring. They also have the financial strength that poverty alleviation, enterprise development, and comes from the capitalization of the bank. Nonbank job creation, all of which microcredit was expected apexes typically take a more holistic, socially to help support.4 Until recently, policy makers have oriented approach. They are more likely to have a focused little attention on possible apex roles with broader social mission and to fund higher risk, small, regard to other financial services. To expand their unregulated nongovernment organizations (NGOs) lending, MFIs need major funding, which matches or community-based MFIs. Different apex types may well with the debt finance that apexes offer. Savings, appeal to different funders’ missions and priorities. insurance, and transfers do not imply a similar funding need, though some grant funding can be Apexes mostly provide local currency loans. Among useful in the development of these services. the 15 largest apexes, almost all disbursements were loans, and over 90 percent of these loans were in local Some apexes are beginning to focus on the broader currency. A few apexes, such as SIDBI and SFD in financial inclusion agenda. For instance, Bancoldex Yemen, also provided quasi-equity and equity funding (Colombia) administers Banca de las Oportunidades to MFIs. The average loan maturity ranged between (BdO), which provides subsidies for new product one-and-a-half and four years, with an average of three development and new business models, including years. In 2009 the interest rate of apex loans to MFIs savings and branchless banking. MISFA is supporting ranged from 0 percent to 12 percent per annum, with development of microsavings in Afghanistan. an average of 8.1 percent (unweighted). Apexes assess the performance of their client MFIs. Capacity-building grants represent a minor share All the surveyed apexes report ongoing measurement of the disbursements of apex institutions. Only six of their client MFIs’ portfolio quality; 89 percent of the 23 apexes surveyed provide grant funding for measure their client MFIs’ operational sustainability, operational support and technical assistance for MFIs. and 84 percent measure operating efficiency. Although such grants represent less than 1 percent of Some apexes use CAMEL (Capital adequacy, total funding, the case study apexes and experiences Asset quality, Management quality, Earnings, elsewhere suggest that they can have a significant Liquidity management) performance standards. 4 For more recent assessments of the impact of microcredit, see Roodman (2012), Rosenberg (2009), and Bauchet et al. (2011). 5 Table 5. Number of MFIs served by 15 agencies. In some exceptional cases, apexes are apexes with largest portfolios, 2009 funded by commercial banks or by NGOs and Apex No. of MFIs private investors. Bank Andara (Indonesia) has a 1 SIDBI 146 consortium of investors led by Mercy Corps; Jaida 2 Nafinsa 24 (Morocco) has a mix of public and private investors. 3 Bancoldex 79 4 PKSF 192 Some apexes, particularly those that are subsidiaries 5 PPAF N/A of banks or those that are registered as banks 6 BTS 288 in their own right, borrow from banks. BMS, for 7 Fedecredito 55 example, borrows from Banque Ouest Africaine 8 PCFC 156 de Développement, the regional development 9 MISFA 16 bank for West Africa. As apexes grow their 10 SFD Egypt 446 lending operations, retained earnings become an 11 FFSA N/A increasingly important source of funds. 12 JAIDA 5 13 Funda-Pro 28 Most apexes are funded with subsidized resources 14 RMDC 208 (grants, or loans at well below market interest rates). 15 BfP (FNI) 7 At the same time, apexes are typically expected to TOTAL 1650 fund their operations by earnings from lending out Source: CGAP 2010 survey. or investing these resources. Apexes strive for their own sustainability. Twenty Anecdotal observations suggest that the quality and out of the 23 apexes say they regard their own effectiveness of this monitoring vary widely. financial sustainability as a goal, but the levels of current sustainability probably vary widely.5 Even At a time of increasing concern about the social though apexes are usually set up to address a performance of MFIs, some apexes are emphasizing temporary shortage of funding for a nascent measurement of social bottom line. For example, microfinance sector, apexes are no different from two-thirds of apexes track rural outreach and any public organization in that incumbent staff have the percentage of female clients, and some have a strong incentive to maintain their jobs. In the commissioned impact studies. course of its research, CGAP did not learn of any active apex that was contemplating a wrap-up of its Public funding is the main source of apex operations.6 Some microfinance apexes eventually resources. Typically, domestic finance ministries are shift their operations toward other objectives. the main source of host government funding, and international multilateral agencies (e.g., World Bank, 2. Highlights from Case Studies Asian Development Bank, European Commission, Inter-American Development Bank), bilateral donors This section reports observations based on the six (e.g., DFID, USAID), and DFIs (e.g., KfW, IFC) are apexes CGAP studied in detail in 2010–11:7 the main sources of international funds. Two-thirds • Bancoldex (Colombia) of apexes are funded by a mix of domestic and • Banque Malienne de Solidarité (BMS) (Mali) international public agencies. About a quarter • Microfinance Investment Support Facility for of them are financed exclusively by international Afghanistan (MISFA) 5 Reliable information on apex financial performance is often difficult to obtain. 6 LID in Bosnia–Herzegovina wrapped up its operations in 2007. 7 For most of the case study apexes, microfinance is only a part of their activities. Microfinance is not the primary activity of Bancoldex, BMS, SIDBI, or SFD. PKSF has been focused on microfinance, but that is changing, and the apex is now working on flood restoration and recovery, health services, an ultra-poor program, and various micro- and small enterprise programs, as part of a more holistic, community- based development approach. MISFA is now looking at lines of credit for SMEs. 6 • Palli-Kharma-Sahayak Foundation (PKSF) contrast, was created in 1990, when Bangladesh (Bangladesh) already had well-established MFIs, such as Grameen • Small Industries Development Bank of India and BRAC.8 Most of PKSF’s funding supported the (SIDBI) expansion of established MFIs, though the apex also • Social Fund for Development (SFD) (Egypt) funded selected early stage MFIs. Some characteristics of these apexes are presented The 2002 CGAP apex study suggested that apexes in Annex 2. In addition to the six case study apexes, were unlikely to be successful in countries that this section draws on the experience of the local did not already have a critical mass of competent initiatives departments (LID) apex in Bosnia– existing MFIs. The experiences of LID and possibly Herzegovina. MISFA provide counter-examples that argue otherwise. Both of these apexes have catalyzed the The case study apexes operate in a range of development of MFIs in a previously undeveloped diverse countries, some with dynamically shifting market, in particular by bringing in external microfinance contexts. At the time of this writing, expertise. LID wrapped up operations in 2007, by important events continued to unfold in India which time it had supported the development of following a microfinance crisis in the State of Andhra eight strong, profitable MFIs in Bosnia–Herzogovina Pradesh. Because the impacts of this crisis continue that were able to attract private, commercial funding to reverberate across Indian microfinance, it may from both local banks and MIVs as well as from the be too soon to be definitive about what the apex DFI-backed European Fund for South East Europe there might do differently. And in Afghanistan, (EFSE). Arguably, these MFIs ended up attracting the sector is retrenching after earlier growth, due too much funding, to the point of generating an partly to internal weaknesses of many MFIs there, over-lending crisis. SFD, on the other hand, has not but perhaps most influenced by a deteriorating played much of a role in catalyzing development of security situation. Such situations underscore the sustainable MFIs in Egypt. uniqueness of different contexts and also caution against generalizing too much about apexes. Apexes supporting start-up or young MFIs in undeveloped markets will need a different strategy Apexes operate at different market stages. than apexes operating in mature markets with Apexes have been established in both early stage established profitable MFIs. Early stage MFIs are and more advanced markets. In all market contexts, more likely to require intensive supervision and hand- there needs to be a clear underlying demand for holding, so it may be best for an apex in a start-up microfinance from low-income people. However, market to invest in a relatively small number of MFIs, the stage of MFI development and the level of MFI and to be prepared to deploy extensive technical demand for wholesale finance can vary widely. assistance to help build institutional capacity. Such apexes should expect to encounter significant At the time of MISFA’s creation, there were only a MFI performance problems, and they need to be few multisectoral NGOs running small microcredit disciplined in ending support to nonperforming schemes in Afghanistan. MISFA was created to jump- MFIs unless there are strong reasons to expect a start the development of microfinance by investing turnaround. in start-up MFIs, including NGOs, credit unions, and a microfinance bank. LID in Bosnia–Herzogovina In some cases, apexes have been established was another apex focused on start-ups in a country immediately after a conflict amid wider national with very little pre-existing microfinance. PKSF, by reconstruction efforts. These apexes typically need 8 PKSF did not fund Grameen Bank. 7 to build a microfinance sector from scratch, as was BMS has contributed to the sector’s development the case with MISFA. At the same time, such apexes in Mali, but its track record is mixed. A considerable are also often faced with lingering uncertainties amount of its support went to organizations that about security. In such circumstances, the need proved to have management problems or that did for sustainable MFIs and good risk management not stand up well in the face of problems in the is no less important, but at the same time certain cotton industry. contingencies or buffers may need to be considered in advance. For the apex and the MFI, this may In Afghanistan, it is too early to appraise MISFA’s require holding greater reserves and more soft success. It was successful in jump starting funding as a cushion against rapid shifts in the microfinance quickly post-conflict in 2003–2005, security situation. but more recently most of its initial investees have run into trouble, due to security problems and In more mature markets, the focus will be more on internal weaknesses. It did fund one MFI (BRAC/ meeting the external financing needs of growing, Afghanistan) that achieved wide outreach early on creditworthy MFIs in a way that does not crowd and First Microfinance Bank Afghanistan, which has out private investment. This may include structuring performed well financially. subordinated loans or equity investments to leverage other sources of funding, such as deposits SFD has funded over 500 Egyptian MFIs, most of and local commercial bank loans. Apexes in mature them community-based NGOs. However, very few markets have the option to deploy most of their of SFD’s investees have achieved strong, sustainable portfolio with relatively low-risk MFIs. Bancoldex, growth. With the support from the World Bank, SFD for example, lends only to profitable MFIs. PKSF has now shifted its policy to concentrate on 10–15 financed quite a few small and start-up MFIs, but high-potential MFIs. given the large number of MFIs in Bangladesh, it was able to be very selective—for a while the apex The experiences of SFD, BMS to an extent, and was accepting only one out of every 10 applications perhaps the early stage of MISFA raise the question it received. Most of SIDBI’s funding goes to larger of whether good results are less likely when the MFIs, but it also has a portfolio of smaller, early apex focuses too much on disbursement and early stage partners and sees part of its role as expanding outreach, rather than on the quality of partner the frontiers of microfinance by taking more risk MFIs. The apex’s funders can be a source of undue than commercial banks. An important question disbursement pressure. Current MISFA management these apexes regularly confront is how to reposition estimates that over-rapid disbursement, combined what they do when markets are better developed— with insufficient focus on capacity building, resulted though this is not always easy for large, well-funded in large losses in defaulted portfolio and shutdown organizations to do. costs, as well as grants to institutions that didn’t survive. Most of the case study apexes have accelerated the growth of sustainable microfinance. PKSF, As maturing markets grow, the importance of LID, SIDBI, and Bancoldex have funded sustainable the apexes’ lending as a source of funds for MFIs that significantly expanded outreach in their microfinance development tends to decline—which respective markets. In all four cases, other funding is, of course, a desirable outcome. SIDBI has seen sources were severely limited, especially during the very strong growth by its top MFIs, which include early years of operation. Of course, other factors large, high-growth MFIs, such as Bandhan, Equitas, contributed to growth as well, but interviewees and Ujjivan. According to SIDBI’s management, its (including MFI managers) were nearly unanimous share of debt funding for the largest MFIs and for that these apexes played an important role in the sector as a whole has declined from nearly 100 accelerating the growth of sustainable microfinance percent of lending in its initial years of operations in the countries concerned. to 10 percent by early 2010, as these profitable 8 MFIs gained access to commercial bank loans.9 In banks.10 In the case of MISFA, there is virtually no Bangladesh, too, the largest, most mature MFIs, commercial bank lending available to MFIs. Only such as ASA and BRAC, have increased their PKSF is charging rates well below a demonstrable commercial funding, including deposits, so that by commercial market. 2009 PKSF was financing only about 11 percent of microcredit in Bangladesh (Sinha 2011). However, the pricing of apex loans is an imperfect indicator of whether the apex is crowding out Have the case study apexes crowded out commercial funding. Even where an apex charges commercial funding? The case for apexes is commercial interest rates, it may still offer an strongest in environments where other funding is in advantage over bank loans, in terms of longer loans short supply—for instance, where commercial banks or lower collateral requirements. That said, some do not yet have confidence in MFIs’ creditworthiness, apexes have an explicit objective of promoting or where an unfavorable regulatory environment commercial funding. SIDBI has a policy of providing keeps MFIs from taking deposits or foreign no more than 15–20 percent of larger MFIs’ funding investment. But what happens when those funding needs. To facilitate commercial funding, it has sources start to become available? created a Lender’s Forum, with over 40 domestic banks. SIDBI management says that it shares A common critique is that publically funded apexes information on MFI performance with these banks, crowd out private finance, as their subsidized money including SIDBI’s own credit ratings. Furthermore, reduces MFIs’ motivation to go after deposits, bank SIDBI structures some of its own finance to fill loans, private equity, and other nonsubsidized critical gaps in equity and subordinated quasi-equity funds. However, some of the case study experiences loans, which facilitates bank lending to the MFI.11 suggest that this need not always be the case. Bancoldex has a similar policy, with apex investment capped at 40 percent of an MFI’s portfolio. Recently, Four of the case study apexes are now charging rates PKSF has started to help more of its small partners comparable to what MFIs are paying commercial build relationships with commercial banks. Table 6. Interest rates and loan terms of apexes Interest rates charged to MFIs (2009) (%) Average loan term (2009) Apex name Apex Commercial Banks Apex Commercial Banks (Country) PKSF 7.0* 12.2 3 2.5 (Bangladesh) SIDBI (India) 12.0* 12.0 4 2.8 BMS (Mali) 8.1 8.4 3 2.5 Bancoldex 7.9 9.7 (Colombia) 3 2.8 MISFA 5.0 n/a (Afghanistan) 10 n/a SFD (Egypt) 10.0 10.8 4 2.5 Source: CGAP Apex Survey 2010, MIX Market *PKSF charges a lower interest rate of 4.5 percent to its small partner organizations, defined as those with fewer than 100,000 active clients. SIDBI charges lower rates to some MFIs working in remote areas. 9 The commercial banks—most of them government owned—were willing to lend to MFIs mainly because such lending satisfied the banks’ priority-sector lending obligations. In the wake of the crisis in Andhra Pradesh, bank loans have become harder to obtain. 10 In SIDBI’s early years its loans were well below market cost, but it has moved to market pricing more recently. Other apexes as well have subsidized rates in early years, on the grounds that (1) commercial borrowing is simply unavailable to MFIs, so there is no market to distort, and (2) MFIs’ early stage operating losses need to be financed. 11 External factors have also driven bank lending to MFIs in India. The government’s Priority Sector Lending policy sets targets that banks can partly meet by lending to MFIs. SIDBI has helped facilitate the market entry of banks but it cannot be said to have caused this trend. 9 In terms of actual results, unsubsidized commercial overheating can occur without apexes present, and funding has become predominant in India and in Bosnia–Herzegovina the overheating happened Bosnia–Herzegovina, making it hard to see a serious well after LID’s exit. Nevertheless, apexes are usually crowding-out problem with those apexes. 12 In recipients of subsidized funds, and therefore, they Bangladesh, although some MFIs still choose lower bear a significant leadership responsibility for the cost PKSF money instead of available commercial microfinance sector they fund. Apexes ought to be sources, the sector as a whole has moved heavily especially attuned to the risks of overheating, and toward unsubsidized funding. Some of the largest they should even take measures to raise awareness as MFIs view the conditions of PKSF’s loans as overly these risks intensify. At the same time, it is important restrictive, especially PKSF’s interest rate cap. So to acknowledge that apexes that support the entry large MFIs in Bangladesh have avoided apex loans of private commercial funders over time see their even though the interest rates are low. In Mali, the relative influence over the microfinance industry microfinance sector still depends on subsidies, but decline. This perhaps highlights the importance of regular commercial banks are increasingly lending the role of apexes and others to promote industry to some MFIs, based on those MFIs’ track records infrastructure, such as credit bureaus, earlier on in paying their loans from BMS. 13 MFI managers before markets mature or become overheated.15 in Colombia and India cite a similar demonstration effect: commercial lenders become more interested Some of the case study apexes are trying to foster after they see an MFI successfully repaying its apex “responsible� finance. Many apexes are starting to loans. measure the social performance of their MFIs. All the case study apexes monitor basic social performance As with other observations about the case study indicators, such as the number of clients reached MFIs, their performance with respect to commercial and percentage of female borrowers, but they funding cannot be taken as representative of apexes are increasingly going further. MISFA is currently worldwide. But the experiences of some of them developing indicators for client poverty as well as for demonstrate clearly that it is possible to have a changes in poverty and well-being. BMS has funded vigorous apex without precluding an eventual shift a workshop on social performance measurement to unsubsidized resources. Obviously, the availability methods and indicators for its partner MFIs. of large apex funding can have a tendency to delay that shift, especially if the apex’s terms are more Especially after the Andhra Pradesh crisis, SIDBI favorable than what the commercial market offers. is raising the importance of responsible finance. It At the same time, apexes that step in and take early now requires all partner MFIs to implement a code stage risks can create a demonstration effect that of conduct that lays out a set of core values for accelerates the entry of private capital. microfinance, including ethical behavior, avoidance of over-indebtedness, transparency, and promotion As markets have matured, overheating is a serious of financial literacy. This commitment has been risk. When markets for any form of retail credit—not written into SIDBI’s loan agreements, which require just microcredit—become saturated, problems with all MFIs to undergo an independent assessment over-indebtedness are likely.14 With rapid growth of adherence to the Code of Conduct as well as in Bosnia–Herzegovina and in the regional market other measures to ensure transparency and ethical of Andhra Pradesh in India, supply expanded very behavior (see Annex 3, SIDBI and Responsible quickly, contributing to a rapid push by lenders Finance). Both SIDBI and LID also commissioned to make many loans in a short time span, often independent research into the impact of microcredit without care for consumer needs or interests. Such on borrowers. 12 In Bosnia–Herzegovina, it could be argued that the EU- and DFI-funded EFSE has competed with, if not crowded out, private funding. 13 BMS is itself a commercial bank, though its shareholders are MFIs. 14 See Schicks and Rosenberg (2011) for a discussion of this dynamic as well as of other issues associated with microcredit over-indebtedness. 15 Bangladesh microcredit as a whole has not yet encountered an obvious over-indebtedness crisis, but because the market appears to be close to saturation and delinquency is high by world standards, there is reasonable cause for concern (Sinha 2011). 10 Most of the case study apexes have provided of new branches and “corresponsales no bancarios� capacity-building support. The main activity of (branchless banking agents). PKSF and SIDBI are now an apex is to provide wholesale finance to retail supporting new product development, including MFIs; however, apexes that have an objective of SME loans, agricultural lending, and insurance developing the institutional capacity of early stage products. MFIs also provide grants. For example, MISFA, PKSF, SIDBI, BMS, and LID have provided time- Most of the apexes have not done much to bound grants that help young MFIs with training support development of market infrastructure— and technical assistance, fixed assets, and operating for instance, consulting capacity, credit bureaus, deficits. Such grant funding is typically a small rating agencies, or trade associations. SIDBI is an percentage of portfolio (see Table 3), but funders exception. It has helped finance trade associations, and apex managers regard it as an important commissioned M-CRIL to undertake MFI ratings activity.16 (before the ratings market was well-developed),17 trained auditors and technical support providers, and SIDBI is recognized as having made a significant most recently promoted the development of a client contribution to the institutional development of the protection Code of Conduct assessment tool by a microfinance sector in India through a sector-wide private company. Additionally, SIDBI is supporting strategy for capacity-building. During its first phase the establishment of a platform to consolidate MFI (1999–2005), SIDBI provided a comprehensive performance reporting at the national level. package of support to partner MFIs based on an annual Capacity Building Needs Assessment The case study apexes have sometimes influenced funded by DFID and IFAD. SIDBI has helped start a policies and regulation. Apexes often have the generation of young and promising MFIs across India. advantage of having one foot in the practice of According to management, it has also instituted microfinance through the MFIs they fund and one innovative capacity-building programs, such as the foot in the policy sphere through their governance. Young Professionals program and support to the The apexes studied have all participated in the Institute of Management to develop microfinance development of microfinance policy, law, and modules. In Bosnia–Herzegovina, MFI managers and regulation. SFD was a key player in the development World Bank staff say that LID technical assistance of Egypt’s Microfinance Strategy. SIDBI provided was crucial to the development of the sector. input into the Microfinance Bill, and it has been consulted as the Indian government shapes its The case study apexes have not done much to spur response to the crisis triggered by events in Andhra innovation. Not surprisingly, it is individual MFIs Pradesh. MISFA is advising the Afghan Central rather than the apexes that have historically driven Bank on how to monitor and supervise MFIs, and it development of new products and systems. More contributed to a draft law for deposit-taking MFIs. recently, some of the case study apexes are trying LID also lobbied—unsuccessfully—for MFI deposit- to provide more support for innovation. Bancoldex, taking regulation in Bosnia–Herzegovina. for example, now administers BdO, which is the Colombian government’s policy to promote access Given some recent crises in microcredit and the to financial services by, for example, subsidizing the prevailing political economy of many countries, development of new products (e.g., microinsurance interest rate caps are on the agenda of many and savings for recipients of conditional cash governments. Apex institutions can sometimes transfers) and extending the reach of the financial affect policy in this area. Bancoldex was influential sector (including MFIs) by subsidizing the opening in encouraging the Government of Colombia to 16 Sometimes the apex finances third-party technical assistance through a grant or a contract. Alternatively, the apex’s staff may provide the assistance. Either way, such technical support represents a nonreimbursable contribution. 17 M-CRIL subsequently developed its rating work more widely and became the largest specialized microfinance rating firm in the world. 11 increase an interest rate cap. PKSF, on the other rather than independent microfinance wholesale hand, has a cap on final borrowing rates for its institutions in need of investment to deliver on a partner MFIs of 12.5 percent flat and has influenced unified strategy. the government to set a general cap of 27 percent effective.18 3. Recommendations for Apexes and Their Funders In some countries (e.g., Bangladesh), questions have been raised as to whether apexes can play a Before offering specific recommendations, two role in supervising MFIs as they transition to formal, general notes are in order. First, the evidence base regulated institutions. If an apex were to prudentially for this third section is broader than the evidence supervise MFIs it has invested in, there could be a base for the first two. Apexes are funding agencies conflict between the supervisor’s duty to protect that support microfinance. The same is true for the public and the investor’s desire to collect on its the international development agencies who are investment. 19 But apexes can certainly encourage members of the CGAP consortium. Apexes work transparency and prepare MFIs to become licensed in a single country, while CGAP’s members work institutions subject to government supervision. They in multiple countries, but many of the same issues can, for example, teach MFIs to track and report on and lessons learned apply equally to both types of indicators required by supervisory authorities. They funding operations. Accordingly, this section draws can also advise central banks on the characteristics not only on the apex research conducted for this of MFIs and appropriate supervision regimes, as paper but also on CGAP’s 17 years of experience MISFA is doing in Afghanistan. around issues of funder effectiveness. One useful summary of that experience is the “Pink Book� Donor coordination is a challenge. Apexes and (CGAP 2006); but many other CGAP publications their funders could do more to pursue consistency address issues that apex managers confront of policy, coordination of funding, and uniformity in regularly. performance standards and reporting requirements. Funders are more effective when they share Second, this paper does not offer conclusions information and collaborate on (rather than about whether setting up an apex is likely to be duplicate) programs.20 a useful tool in any given market. As mentioned earlier, CGAP research does not support general When an apex is receiving funds from multiple conclusions about how often apexes are likely to donors, there is a tendency to design and maintain succeed. Rather, the recommendations address separate programs. SFD has found it difficult to situations where an apex already exists, or the coordinate its multiple donors, which included the decision to create one has already been taken. This World Bank, UNDP, EC, the Kuwait Government, section offers suggestions about how to make those and the Japanese development agency JICA. One apexes as effective as possible. donor required highly subsidized loans. Another insisted that MFIs could receive second loans 3.1. Apex Setup only after those applying for a first loan were served. Through inconsistent policies and funding Begin with a thorough market study. Apex requirements, some donors pushed SFD toward managers and their funders need to have a thorough good practice microfinance, while others pulled understanding of the microfinance marketplace. In it back. Unfortunately, donors sometimes treat some circumstances it may be appropriate to begin apexes as their own project implementation units by assessing the demand for microfinance among 18 For a discussion of concerns about interest rate caps, see Helms and Reille (2004). 19 For instance, the public interest might call for closure of a troubled deposit-taking MFI, but a supervising apex might be reluctant to take this step if the MFI could repay its loan from the apex only by continued operations. There would be less of a conflict if the apex provides nonprudential supervision (e.g., consumer protection). 20 CGAP 2005–2007 (Country-Level Effectiveness and Accountability Reviews). The cases of Sri Lanka, Nicaragua, and to some extent Pakistan show how little coordination took place among funders in supporting an apex and the negative market consequences. 12 investors, and technical support that start-ups will Box 2. Donor coordination issues in require. MISFA MISFA has played an effective role in coordinating donor support for microfinance in Afghanistan, though The study should also assess which other sources challenges remain. MISFA was created following a of local and international funding are available war for control of Afghanistan, and as donors funded to MFIs, including commercial funding, to verify the post-conflict reconstruction efforts, individual the nature of the funding gap the apex is aiming countries’ donor agencies tended to want to allocate their funding to areas where their national forces were to address. Such a study might also include an deployed, or to promote alternative livelihoods to overview of the policy and regulatory environment opium production, or a myriad other individual donor for microfinance and what work is needed in this priorities. These varied needs made it challenging to pursue a coherent national policy. domain. Ideally, the study would be carried out as a joint exercise between the apex and all MISFA was successful, however, in persuading a range of funders to channel most of their microfinance interested funders.21 resources into a World-Bank-administered trust fund, which was then programmed into MISFA with a more Protect political independence. Promising MFI coherent single strategy. This not only permitted partners are often scarcer than originally expected, a single national strategy, but it also meant that managers of MISFA and their investee MFIs faced a and building a portfolio of such partners can be single set of conditions and reporting requirements. hindered if funding decisions are distorted by Had this not been the case, MISFA management extraneous political considerations. Political thinks it would have had to spend much more time on independence should be bolstered, not just by donor requirements and much less on its main task of promoting microfinance. expressions of intent, but by governance structures built into the apex’s constitutive documents. In Donors have sometimes insisted, with MISFA’s acquiescence, that contributions be earmarked for MISFA, for example, it was agreed that a majority specific activities rather than be used to support of the board would be independent persons who MISFA’s core fund. These projects in general have have no formal role in the Afghan government. In been difficult to administer and have often been unsustainable. Indonesia, Bank Andara has been set up as a fully private apex with the backing of Mercy Corps, IFC, Source: Communication from Dale Lampe, MISFA director of operations. KfW, Hivos-Triodos Fund, and Cordaid. low-income households, but this has seldom proved Name a qualified board. Typically, the majority of a limiting factor for microfinance development, the directors should be independent, and the board given the large unmet demand for financial services should have expertise in areas such as banking and in most countries. Typically, the most important finance, law, accounting, economic development, element of a market study is assessment of the and ideally the needs of end clients. It may be useful existing microfinance providers that are candidates to structure an independent investment committee to receive funding from the apex. Apex designers reporting to the board, to further insulate funding have often omitted this step, but it need not be decisions from political considerations. particularly burdensome. In a day or two spent with an MFI, a first-rate microfinance operations Be clear that the apex’s objective is to support expert can usually get a meaningful impression of development of sustainable retailers. Retailers the quality of management and operations, and become “sustainable� when they can collect their the probability that the MFI will be able to reach loans effectively and they are profitable enough sustainable scale. In an early stage market where to continue expanding their operations without start-up of new MFIs will be needed, the study further subsidies, by drawing on commercial bank should look at whether there are realistic prospects loans, foreign and domestic investment, deposits, to recruit the foreign or domestic managers, and other sources of market funding. 21 For examples, see CGAP Country Level Effectiveness and Accountability Reviews (2005–2007). 13 For most microfinance apexes, financial services Establish clear expectations about the apex’s are seen as serving the ultimate policy goal of financial performance. There is considerable income generation and poverty alleviation. The discussion about whether apexes should or can apex wants to see quality services extended to as be sustainable. The answer will depend on one’s many low-income people as possible. But in the definition of sustainability. If “sustainable� means short to medium term, the focus should be on “unsubsidized,� the proposition that apexes should building sustainable retailers, more than on numbers be sustainable becomes an odd one. It is hard to see of clients. If retailers are sustainable, expansion how they can fill a gap in the commercial market—i.e., of outreach has a high probability of eventually finance retailers whose risk is too high or whose return taking care of itself. Conversely, expansion without is too low for commercial lenders—without using sustainability is likely to be only temporary. subsidized resources, whether grants or soft loans. Build in leverage of commercial finance as a stated If, on the other hand, “sustainable� means “able to objective, making it clear that the apex’s role is operate without continuing infusions of subsidy,� to substitute for commercial finance when it’s not then any apex, no matter how inefficient, can be available, and not to crowd out commercial finance made sustainable with a big enough initial subsidy. once it does become available. In practice, many apexes can generate surpluses, covering their administrative costs with the interest Don’t create large disbursement pressure in early rate differential between the low cost of their years—especially in early stage markets. It takes funds and the interest income from lending out or time for apex staff to learn the business and to investing those funds. Of course, this will depend on assemble a portfolio of strong partners. The partners the apex being able to collect a high percentage of themselves are often well advised to concentrate its loans to MFIs. on piloting and refining their systems, rather than expanding rapidly in their first few years. If an Some apexes are financed by continuing fiscal apex has too much money that it has to disburse, infusions from the government. When this happens, it may be forced into funding unpromising MFIs or it is usually not the result of the initial plan, but pressuring its partners for inappropriate growth. It rather because the apex is lending a lot of money may be better to fund the apex in stages, starting to institutions that fail to repay. out small and adding resources as they become necessary. Plan for what will happen to the apex once “success� has been achieved—that is, once the Establish regular, transparent public reporting country has a good supply of strong retailers with on the performance of the apex and its investee adequate commercial funding available. Should the MFIs. Most people agree that when MFIs regularly apex close operations at that point, or shift to other disclose their financial and social results, it not work? If the latter, what work? Without this sort of only serves the interests of their funders, but also explicit planning, the apex’s evolution at that point tends to improve the performance of the reporting is more likely to be driven by the staff’s incentive MFIs. The same principle applies to the apex. The to maintain their positions, rather than by pure financial reports of both the apex and its partner development objectives. investees should comply with generally accepted accounting principles as codified in the country or— 3.2. Early Stage Markets22 even better—with International Financial Reporting Standards. MIX can help develop reporting systems Expect to have to do a lot of hand-holding and for the MFIs. capacity building. In markets where the MFIs are 22 While this section focuses on early stage microfinance markets, the recommendations are also relevant for apexes supporting start-ups in more advanced markets. 14 mainly new and/or weak, there will be performance the MFI. Second, technical assistance will obviously problems. In such settings, the expertise of apex be more effective when it is demand-driven, when management and staff is particularly important. it is based on specific issues identified through an If it is not possible to recruit people who already assessment of the MFI’s needs, and when the source have experience in microfinance operations, then of technical assistance is not tied by the funder. apex personnel should receive intensive training MISFA, LID, and SIDBI have conducted on-site and technical assistance in the early years and be institutional assessments with their MFIs. Third, it supported by international experts, if need be. (This may be necessary to develop the firms and technical does not imply that staff expertise is unimportant in assistance providers as part of the institution- later stage markets with more advanced MFIs.) building process. For example, SIDBI supported the growth of microfinance consulting firms and rating The apex should have appropriate instruments agencies in India by contracting them for specific (typically grants), as well as adequate resources and assignments. staffing to provide high-quality capacity-building support to partner MFIs. In early years, apex staff Consider time-limited grants to retailers to cover are relatively inexperienced and less likely to be start-up losses. As experience with microcredit able to provide strong technical assistance, so the grows, young MFIs are reaching profitability more focus should be on recruiting third-party experts, quickly than used to be the case (Gonzalez and relying as much as possible on those who have a Rosenberg 2006). Still, new MFIs typically need demonstrated track record of producing sustainable a few years before income catches up with costs microfinance, whether locally or in a foreign setting. and the resulting losses need to be financed. Some apexes have provided operating grants to cover Managing an effective capacity-building program is these losses in the early years. a resource-intensive undertaking, whether done by apexes or independently, and it requires dedicated Other apexes have chosen to finance start-up costs staff and resources. First, apexes should separate by initially lending to new MFIs at subsidized rates. its capacity building and grants function from its This is a reasonable approach. Some would argue, wholesale lending. This kind of firewall lowers the though, that it is better to use grants (an equity risk that the apex will be liable for the performance of infusion) in combination with full-price loans. This latter approach defrays the start-up losses, but it also ensures that an MFI from the very beginning Box 3. An apex that liquidated itself gets used to liabilities at the cost that it will face as The World Bank-funded LID project in Bosnia– Herzegovina is the only apex known to CGAP that it funds its long-term expansion. planned an exit strategy early on. MFIs understood clearly that the project would end, and that they When an apex finances start-up or young needed to become financially sustainable and seek MFIs, it should be ready to take risks. But two other funding sources. Apex and World Bank staff assisted this process by introducing MFIs to investors, qualifications are called for (1) no investee partner such as Triodos and Blue Orchard, and facilitating should be selected without a thorough appraisal by relationships with local commercial banks. The portion a microfinance operations expert, and (2) the apex of the original loan capital that had been funded by must be willing and able to drop partners who do bilateral donors was converted into capital grants for MFIs once they met strict standards related to not perform. Both of these points are discussed management capability, governance, outreach, and further in Section 3.4. financial sustainability. These grants gave the MFIs a capital base with which to leverage commercial funding. The remaining outstanding loans were Prioritize building early market infrastructure and transferred from the apex to the Ministry of Finance, reporting. One of the benefits of having apexes and the apex’s lending ceased. At the end of eight coordinate a small growing industry is that much years of apex operations, there were eight profitable can be done to structure the industry early on. This MFIs with outreach across the country. may involve establishing standards around reporting 15 and definitional issues. It may be useful to establish often lower than many people expect (Anand standards for audits and governance. It might be too and Rosenberg 2008). When competitive markets early for a credit bureau to be useful or viable, but approach saturation—that is, when supply begins to it can be important nevertheless to begin planning catch up with demand—problems with client over- for a credit bureau in preparation for the later time indebtedness are almost inevitable. Apexes need when it will be needed. Credit bureaus often take to be alert for this dynamic, and should consider several years to build. As a major source of funds, collaborating with MFIs and regulators to develop apexes in such industries may have the clout to an early warning system and encourage responsible ensure that this infrastructure gets attended to early practices.24 on. Consider providing more support in the form 3.3. Advanced Markets of equity or quasi-equity. A more advanced microfinance market implies stronger retailers who Actively leverage commercial finance. Advanced are becoming able to borrow from commercial markets typically have a critical mass of strong, sources. An MFI’s ability to take advantage of this profitable MFIs that can be creditworthy borrowers opportunity is limited by the strength of its equity from commercial banks, or candidates for other base. If an apex injects equity capital or makes commercial investment or a deposit-taking license. quasi-equity subordinated loans, the MFI can Apexes in these markets should consider the borrow more. following steps: Consider supporting development of savings • Setting an upper limit on the percentage of a services, other new products, and market large MFI’s assets that can be financed by the infrastructure. Low-income clients’ need for financial apex. services goes far beyond traditional microcredit • Making the raising of unsubsidized commercial products. These clients need deposit facilities, fund resources a performance target in the apex’s transfers, and insurance, as well as a wider range agreement with advanced MFIs. of credit products. An apex could provide financial • Offering equity, quasi-equity, or other support for such efforts (e.g., by funding technical subordinated lending to MFIs. This kind of assistance), and could also work with regulators to instrument allows another creditor—for instance, help develop an appropriate licensing regime for a bank—to collect its loan before the apex collects depository microfinance. (such subordination encourages commercial lending to the MFI by making it safer).23 More Development of a sound microfinance industry generally, apexes should focus on offering funding requires the evolution of market infrastructure, that is not available from other sources. including audit, consulting, and training capacity; • Requiring MFIs to report their financial industry networks; and eventually credit bureaus. performance using formats and standards that are Apexes can play a role in supporting such familiar and credible to commercial lenders. infrastructure. • Organizing a lenders’ forum or some other mechanism to share the apex’s knowledge of its Consider other roles. Eventually, advanced investees with potential commercial investors. microfinance providers will be able to access commercial sources for most or all of their funding Be alert for over-indebtedness as the microcredit needs. At this point, the apex may need to reconsider market approaches saturation. The demand for its mission. As a practical matter, few apexes will microcredit is not infinite. Effective demand is choose to terminate their operations (whether or 23 Use of these financial instruments may require augmenting the apex’s staff expertise. 24 For a full discussion of microcredit over-indebtedness and possible ways to deal with it, see Schicks and Rosenberg (2011). 16 not this would be a desirable outcome). An apex the form of minimum performance thresholds for key that has been funding mainly large, advanced MFIs indicators.25 A short list of performance indicators might refocus its support on younger, smaller MFIs is more likely to be effective than a long list. At operating in more difficult environments that are a minimum, the list should include cost recovery judged to be risky by private funders. An apex might (profitability) and loan collection. Core financial take on the role of “lender of last resort,� providing and outreach indicators are discussed in Rosenberg emergency liquidity in times of crisis. Or the apex (2009).26 might take on other development activities besides microfinance. Many of those other activities tend Some apexes may wish to contractually enforce to need grants rather than loans; in such cases, the performance against one or more social indicators. apex might face a sustainability problem when it no In this case, care should be taken to select an longer has interest income flowing in from loans to indicator that is both meaningful in terms of social MFIs. outcomes and unambiguously measurable. 3.4. Selection and supervision of MFIs Consequences for failure to meet a minimum threshold should be spelled out, probably including Be selective in picking retailers, based on the apex’s right to suspend or cancel future the likelihood that the investee will become disbursements. Performance-based contracts sustainable. For an apex that aims at developing of this sort will not be credible unless the apex a sustainable microfinance industry, the central demonstrates that it is willing to enforce them. question in choosing investee partners should be, Sometimes there will be extenuating circumstances “Is this MFI sustainable in terms of profitability and for a missed threshold. But when an MFI fails to loan collection, and if not, how high is the probability perform, the apex should not continue funding the that it will become so?� Evaluating the prospects MFI unless the MFI demonstrates a clear, specific, of a young or start-up MFI is as much an art as a and credible plan to remedy the problem. In any science. The evaluation is based on both current event, the apex—like any other development performance and the quality of management. funder—should not continue supporting an MFI Current performance can be quantified, for the that repeatedly fails to meet defined minimum most part. Appraising the quality of management, performance standards. on the other hand, is more complex, and often comes down to judgement. Such an appraisal should Focus supervision on the overall institutional be conducted by an expert who has extensive performance of the investee, not just on some experience working with strong MFI managers. component of its operations funded by the apex. There are useful guides for evaluating MFIs (e.g., If the apex’s objective is to develop a sustainable Isern, Abrams, and Brown [2008]), but they cannot microfinance market that provides high-quality substitute for the experienced judgement of services, achievement of that objective will depend someone who knows the business well. on how every retail institution as a whole performs. Unless the institution as a whole is sustainable, Use performance-based contracts that focus whatever specific activities may be earmarked for on key performance indicators. If the apex’s apex funding are unlikely to continue and expand investment is premised on the expectation that the over the long term. MFI will move toward, or maintain, sustainability, then this expectation should be quantified and Independently evaluate MFIs’ loan collection incorporated into the loan or grant agreement, in performance, using specialized testing. Loan 25 A minimum performance threshold is different from a planned target. The latter represents the performance that the MFI expects to achieve. By contrast, a minimum threshold represents the lowest acceptable performance; in other words, the apex is saying that it won’t think its investment has been justified if the MFI doesn’t reach this level. 26 The recommendation here is that there should be only a very few contractually required minimum performance thresholds. The list of indicators that the apex monitors (as opposed to contractually enforces) will be longer. 17 Require investees to adopt and implement codes Box 4. Core performance indicators of responsible finance and consumer protection. Most funders will want reports on a wide range of The SMART Campaign (www.smartcampaign.org) outreach and financial measurements, but there is a small set that most analysts would regard as core has developed a global set of client protection indicators. principles, along with tools and resources to 1. Breadth of outreach—number of active clients implement them. The Social Performance Task Force served (2011) has developed a set of standards for social 2. Loan repayment—percentage of portfolio that is late, or written off performance management. 3. Financial sustainability—net profit as a percentage of equity or assets Conduct regular reviews of the performance 4. Efficiency—administrative cost per dollar of loans outstanding or per active client of all apex investees. At least twice a year apex In recent years there has been increasing emphasis management should conduct a portfolio-wide on reporting social performance, not just financial review looking at the performance of each partner performance. MIX has adopted a set of 11 social MFI against key performance indicators. LID in performance indicators developed by the Social Performance Task Force, including measures such Bosnia–Herzegovina shared key performance as client poverty levels, client retention rate, client indicators among all of its investees and found that protection policies, and the range of products and the practice stimulated healthy competition among services.* the MFIs. *See http://www.themix.org/social-performance/Indicators Invest heavily in apex staff capacity, through collection is the biggest business risk facing recruitment and training. It is axiomatic that an providers of microloans, and a vast majority of MFI apex’s effectiveness in selecting and supervising failures are attributable to collection problems. partner MFIs depends on the skill and experience of Furthermore, the loan portfolio is the element of its staff and management. This is often a challenge for MFI financial statements that is most commonly new apexes, which tend to be created in environments subject to material misstatement. where there are not many experienced microfinance specialists available for recruitment. If possible, the Accordingly, keeping track of an investee’s loan apex should recruit microfinance experience, even portfolio quality is the apex’s most important risk- if that means hiring some foreign personnel at control challenge. Unfortunately, normal audit the beginning. In any event, the apex should have procedures provide little assurance as to portfolio a substantial training budget and use it. It would quality. 27 Meaningful assessment of loan quality be penny-wise and pound-foolish to skimp on staff requires specialized testing above and beyond a training because people are too busy carrying on the regular audit, whether this testing is performed apex’s day-to-day work. Apexes do not necessarily by the external auditor, by apex staff, or by some need large numbers of staff, but they are well served other expert evaluator. Suggested portfolio testing by having a highly capable team of top managers processes can be found in Christen and Flaming who know the microfinance business well. (2009). MicroSave provides training in another portfolio audit tool.28 3.5. Structuring funding to be useful to retailers Most apexes (and MFIs) do not do this kind of portfolio testing. Even though it adds to the cost Try to structure some lending as working capital of auditing, periodic independent portfolio testing loans. Some apexes require full repayment of an is strongly recommended. It probably would have MFIs’ loan before the MFI can borrow again. This identified the problems brewing in overheating kind of lending may be suitable for fixed-asset markets well before they exploded. acquisitions, but it does not match well against the 27 For suggestions on MFI audits, see CGAP (1998). 28 http://www.microsave.org/sites/default/files/toolkits/toolkits/Loan_Portfolio_Audit_Toolkit_Summary.pdf 18 ongoing working capital needs of a loan portfolio. resources. Reporting on sub-borrowers increases Low-income borrowers’ willingness to repay depends the cost of lending, which usually gets passed on to on the MFI offering prompt follow-on loans when borrowers in the form of higher interest rates. these clients’ previous loans are paid off. This implies a fairly steady funding requirement, which would be An apex should usually avoid requiring MFIs to better served by a line of credit, or an overlapping report on job or enterprise creation as a result series of concurrent loans, rather than by a single of their loans. It is very difficult to get reliable large loan that has to be paid down to zero before information about these variables and to determine further borrowing. On longer term loans, it may be whether they are attributable to the loan. useful to build in a grace period during which the MFI does not have to amortize capital. More generally, any reporting requirement that goes beyond standard financial reporting should Ensure that disbursements are timely and reliable. be justified by a specific need to know, and a clear If the MFI does not have dependable liquidity to fund notion of what decisions will be affected by such follow-on loans, it will quickly run into repayment reporting. problems. Liquidity management is more crucial for microlenders than for normal collateralized lenders. Be cautious about restrictions on product terms, lending methodologies, and target markets. As a Keep red tape to a minimum and reporting general matter, retailers tend to be better situated requirements reasonable. MFIs’ most common than wholesale funders when it comes to figuring complaints about apexes are that apex loans are out what services clients want, how to deliver them encumbered with unnecessary requirements and at low cost, and which markets are practical to serve. that the transaction costs of dealing with apexes are high. Even after taking into account the MFIs’ Apexes sometimes tie funding to particular natural bias (everyone would like money with no geographical areas or particular clients. This may strings attached) it seems that there is some merit occasionally be appropriate—the argument being to this complaint. All development funders, not made here is only that such requirements should be just apexes, have to work hard against their natural scrutinized carefully. When funders have wanted to inclination to proliferate requirements. provide direct support to a specific area or clientele, they have had better results when they find a retailer As a general rule, it is preferable to allow the apex’s who already wants to serve that market, rather than investment to be used as core funding for the MFI, twisting the arm of a retailer whose basic interests rather than to tie it to particular loans or activities. lie in other directions. In particular, apexes should usually avoid making MFIs report on individual subloans funded with Some apexes cap the interest rate that partner MFIs the apex’s money. Money is fungible, and it makes can charge to the ultimate borrowers. This practice little difference in terms of development objectives is controversial. It has the effect of discouraging which of the MFI’s funding sources are allocated service to some clients who are more expensive to a given loan, at least in an MFI whose credit to reach, for instance because of small loan size, all goes to a microclientele. The apex should be remote location, or low population density. Before concerned with seeing that the MFIs’ total microloan adopting such a policy, the apex should carefully portfolio increases by a given amount, rather than weigh the perceived benefits of the measure against with identifying particular borrowers with the apex’s the outreach limitations it may entail.29 29 See Helms and Reille (2004). 19 Lend at commercial interest rates. Most apexes’ lack of strategic coordination among development resources come from grants or well-below-market- funders and administrative burden on MFIs because rate loans. It would often be possible for them to of conflicting reporting requirements. Sometimes lend to retailers at below-market rates, and still to apexes can help to reduce this problem. have enough of an interest margin to fund their administrative costs. In the first place, if multiple international donors channel their resources through an apex, some However, the countries where microcredit is available strategic coordination results, at least if the donors to most of the qualified customers who want it are don’t attach idiosyncratic objectives and restrictions almost always countries where the retailers have to their contributions. Of course, this decision rests been able to tap commercial resources, including ultimately with the donors, but an apex’s board and bank loans and deposits. Young MFIs need to build management may be able to encourage coordination. business models capable of absorbing the full cost of such resources. When apexes lend at less-than- Second, even when the donors channel their funds commercial rates, they risk lowering the retailers’ directly rather than through the apex, there is incentive to move on to commercial borrowing, room for harmonization of reporting requirements, and to mobilize deposits (which not only fund so that an MFI with three funders does not have microloans but also are a valued client service in to prepare three separate reports. Funders tend and of themselves).30 to underestimate the time and resources tied up in MFI reporting. Arguably, a domestic apex with Sometimes, as in the case of SIDBI, charging strong leadership might be in a good position to commercial rates to retailers results in large profits enlist international donors in working out a unified for the apex. Some argue that these profits should reporting scheme. be passed on to the retailers in the form of lower rates. The question here is who should capture Microfinance apexes are a major source of funding the subsidy. For the reasons indicated above, the for low-income financial services, and they are best solution is usually to leave the profit in the likely to remain so for some time. They face most apex, where it can be used for other public-benefit of the same challenges that other donors and purposes. development agencies have faced for decades. Supporting financial access for the poor is a difficult Where possible, pursue donor harmonization, business. Apexes that pay attention to the lessons including unified reporting requirements. In many of international experience are much more likely to microfinance markets, resources are wasted due to be effective at that business. 30 The rationale for limited-term subsidized lending to new MFIs was discussed above in Section 3.2. 20 Annex 1: List of Interviews for Case Study Apexes Type of case study) Number of Key interviewees interviews conducted Bancoldex Field visit 11 Bancoldex: CEO, CFO, Head of MF, Head of BdO (Colombia) MFIs: WWB Cali, Finamérica, Contactar, Bancamía, Emprender Others: MF Network, Planet Finance BMS (Mali) Phone interviews 13 BMS: CEO, Head of MF department MFIs: Kafo Jiginew, Nyesigiso, Demesow, FADEL SA, Jigiyaso Ba, Soro Yiriwaso, Misileni Others: BIDC, Oikocredit, Grameen Foundation, Sidi MISFA Phone interviews 4 MISFA: Managing director (Afghanistan) MFIs: BRAC, FMFB Donors: DFID, WB PKSF Field visit 18 PKSF: Chairman, management team (Bangladesh) MFIs: BRAC, Shayida Foundation, Padakhep, POPI, CCDA, Uddipan, Buro Donors: DFID, WB SFD (Egypt) Field visit 11 SFD: Director of MF sector and his deputy MFIs: SBACD, DBACD, Lead, Tadamun, Mubadara, Matareyyah, Shabab Misr Donors: AfDB, JAICA SIDBI (India) Field visit 16 SIDBI: Chairman, management team in both Lucknow (HQ) and Kolkata branch MFIs: ASA, Bhandan Donors: DFID, KfW, WB Other: Sa-Dhan, MFIN, M-CRIL 21 Annex 2: Self-Reported Characteristics of Case Study Apexes Institutional Form and Governance Apex Institutional Form MF is primary Board Members (No. of members) activity Bancoldex Division of a No Government, member appointed by the Shareholders government-owned General Meeting, member appointed by the development bank Exporters Association Guild Council (5) BMS Division of a private No Representatives of the Government, Para public commercial bank organizations, one social investor, MFIs (12) with majority MFI ownership MISFA Not-for-profit, Yes Members representing the government of limited liability, Afghanistan (Ministry of Finance and Rural joint stock company Rehabilitation and Development), donors, private (nondividend sector, and academia (7) disbursing), government owned PKSF Government- Yes, but may MD of PKSF, representatives of Bangladesh Economic established not-for- change Association, academia, award-winning social worker, profit association Prof. Muhammad Yunus (7) SIDBI Division of No Government, State Bank of India, Bank of India, West government-owned Bengal Financial Corporation, IDBI Bank Ltd., Life development bank Insurance Corporation of India (11) SFD Semi/quasi- Yes Representatives of the government and private sector governmental headed by the prime minister organization Performance monitoring tools Performance Frequency Financial Indicators measurement tools Bancoldex Portfolio reports, Quarterly CAMEL (capital adequacy, asset quality, management financial statements quality, earnings, liquidity management) BMS Financial Quarterly Prudential ratios reported to central bank (capital statements adequacy, liquidity management, risk limitation, legal reserves) MISFA Portfolio Monthly 22 performance standards, including financial and reports, financial nonfinancial indicators (e.g., strategic and business statements, planning, capitalization, accounting treatment of institutional nonperforming loans and write offs, etc.) development action plans PKSF Portfolio reports, Monthly Financial performance indicators (e.g., debt-to-equity, financial statements capital adequacy, debt service coverage, liquidity, return on capital, adequacy of provisions) SFD Portfolio Monthly E.g., number of loans disbursed, number of active reports, financial borrowers, portfolio and repayment rate, OSS, etc. statements, simplified GIRAFE* tool SIDBI Institutional Monthly Credit score approach (e.g., profitability, capital assessment, adequacy, debt-service coverage, liquidity, PAR, loan external rating, loss provisions) portfolio reports, financial statements *An evaluation tool, similar to CAMEL, developed by Planet Finance. 22 Capacity-building activities of apexes Bancoldex Client-level capacity building (Programa EOCM Especial: credit linked to TA), TA for 6 MFIs, incentives to MFIs to open new branches BMS No large-scale capacity-building programs but has provided in-house TA to 7 small MFIs MISFA Intensive TA to turn around poor performing institutions PKSF Institutional development fund used to provide training and capacity-building support to all MFIs. Provides training both directly (in-house) and out sources to qualified consultants. SFD In the past, TA was provided particularly during the first 1–2 years of funding an NGO-MFI; however, SFD currently does not have funding for capacity-building support and is relying on other donors to fund this. SIDBI Capacity-building support to MFIs; support to MFI associations, rating agencies, auditors, technical consulting firms, training of trainers program, Young Professionals program, support to Institute of Management to develop microfinance modules 23 Annex 3: SIDBI and balance basis) being charged to the ultimate beneficiaries. Responsible Finance (e) to prepare a Board approved note on recovery SIDBI Lenders’ Forum—Additional practices that would be displayed in local language at each branch and to give an Terms of Sanction (March 2011) undertaking to take steps to ensure responsible and non-coercive loan recovery practices at the The borrower shall agree: field level (a) to furnish financial and operational data in (f) to develop a Board approved strategy to check the specified format to the India Microfinance multiple lending/over-indebtedness amongst Platform (IMFP) within reasonable timelines and clients by December 31, 2010 and implement it with accuracy. thereafter and also obtain annual affirmation of the strategy by its Board. (b) to undergo a third party Code of Conduct Assessment with a view to assessing the degree (g) to put in place an effective grievance redressal of adherence to the voluntary microfinance Code mechanism by December 31, 2010 – to be placed of Conduct through accredited agencies for the in the website of MFI and also displayed in the purpose. branch offices. (c) to undergo a Systems and Portfolio Audit (h) to take steps to ensure that some acceptable involving detailed examination of operational form of electronic, written or printed systems and procedures, funds utilization, acknowledgement of financial transactions is assessment of loan portfolio in respect of the left with the individual borrower or the group/its risk parameters, finance as well as planning and representative. control etc. by an external agency. (d) to ensure transparency and uniformity in (i) to furnish regularly, accurate and comprehensive calculating and reporting (to clients and in the data to Credit Bureaus like CIBIL and High Mark public domain) the effective cost (on reducing Credit with regard to beneficiaries. 24 Annex 4: Bibliography CGAP/Symbiotics. 2010. 2010 MIV Survey Report. http://www.cgap.org/gm/document-1.9.47373/ General Reading CGAP_2010_MIV_Survey_Report.pdf Anand, Malika, and Richard Rosenberg. 2008. “Are Chen, Gregory, Stephen Rasmussen, and Xavier Reille. We Over-Estimating Demand for Microloans?� Brief. 2010. “Growth and Vulnerabilities in Microfinance.� Washington, D.C.: CGAP. Focus Note 61. Washington, D.C.: CGAP. http:// www.cgap.org/gm/document-1.9.42393/FN61.pdf Bauchet, Jonathan, Cristobal Marshall, Laura Starita, Jeanette Thomas, and Anna Yalouris. 2011. Christen, Robert Peck, and Mark Flaming. 2009. “Due Latest Findings from Randomized Evaluations of Diligence Guidelines for the Review of Microcredit Microfinance. Forum 2. Washington, D.C.: AFI, J-PAL, Loan Portfolios: A Tiered Approach.� Technical and CGAP. http://www.cgap.org/p/site/c/template. Guide. Washington, D.C.: CGAP. http://www. rc/1.9.55766/ cgap.org/gm/document-1.9.36521/DueDiligence_ TechGuide_ENG.pdf Berger, Marguerite. 2003. “Case Study: Peru. The Second Story. Wholesale Finance in Latin America.� Duflos, Eric. 2008. Note on Apexes. CGAP internal http://idbdocs.iadb.org/wsdocs/getdocument. document. aspx?docnum=1481942 Duflos, Eric, and Mayada El-Zoghbi. 2010. “Apexes: ———. 2004. “Do Apex Models Make Sense for An Important Source of Local Funding.� Brief. Microfinance?� Microenterprise Development Review. Washington, D.C.: CGAP. http://www.cgap.org/p/ Vol. 6, No. 2. http://www.microfinancegateway. site/c/template.rc/1.9.43025/ org/gm/document-1.9.24955/21312_ 40.pdf El-Zoghbi, Mayada, Barbara Gähwiler, and Kate Lauer. 2011. “Cross-Border Funding of Microfinance.� Focus CGAP. 1998. External Audits of Microfinance Note 70. Washington, D.C.: CGAP, April. Institutions: A Handbook. Technical Guide. Washington, D.C.: CGAP. http://www.cgap.org/p/ Glisovic, Jasmina, Mayada El-Zogbi, and Sarah site/c/template.rc/1.26.2110/ Forster. 2010. “Advancing Savings Services: Resource Guide for Funders.� Technical Guide. ———. 2002. “Helping to Improve Donor Effectiveness Washington, D.C.: CGAP. http://www.cgap.org/gm/ in Microfinance—Apex Institutions in Microfinance.� document-1.9.55535/TG_AdvancingSavingsServices. Donor Brief. Washington, D.C.: CGAP. http://www. pdf cgap.org/gm/document-1.9.2427/DonorBrief_05.pdf Gonzalez, Adrian, and Richard Rosenberg. 2006. ———. 2006. Good Practice Guidelines for Funders “The State of Microcredit: Outreach, Profitability, of Microfinance. Microfinance Consensus Guidelines. and Poverty.� http://www.microfinancegateway.org/ Washington, D.C.: CGAP. http://www.cgap.org/p/ gm/document-1.9.26787/25.pdf site/c/template.rc/1.9.2746/ Gonzalez-Vega, Claudio. 1998. Microfinance Apex ———. 2009. “Role of Government during the Mechanisms: Review of the Evidence and Policy Global Crisis� Microfinance Blog, 12 December. Recommendations. CGAP-OSU Research Project on http://microfinance.cgap.org/2009/12/10/ Apex Mechanisms. http://aede.osu.edu/programs/ governments%E2%80%99-role-in-times-of-crisis- ruralfinance/PDF%20Docs/Publications%20List/ toward-a-new-paradigm/ Papers/98P20.pdf 25 Helms, Brigit, and Xavier Reille. 2004 “Interest ———. 2010. “Does Microcredit Really Help Poor Rate Ceilings and Microfinance: The Story so Far.� People.� Focus Note 59. Washington, D.C.: CGAP. Occasional Paper. Washington, D.C.: CGAP. http:// http://www.cgap.org/gm/document-1.9.41443/fn59. www.cgap.org/gm/document-1.9.2703/OP9.pdf pdf Isern, Jennifer, Julie Abrams, and Matthew Brown. Schicks, Jessica, and Richard Rosenberg. 2011. 2008. “Appraisal Guide for Microfinance Institutions.� “Too Much Microcredit? A Survey of the Evidence.� Technical Guide. Washington, D.C.: CGAP. Occasional Paper 19. Washington, D.C.: CGAP. http://www.cgap.org/gm/document-1.9.4394/ http://www.cgap.org/gm/document-1.9.55377/ MFIAppraisalTechnicalGuide.pdf OP19.pdf Levy, Fred. 2002a. “Apex Institutions in Microfinance.� Sinha, Sanjay. 2011. Bangladesh Microfinance Occasional Paper 6. Washington, D.C.: CGAP. Review 2011, BRAC Development Institute. http:// www.microfinancegateway.org/p/site/m/template. ———. 2002b. “Apex Institutions in Microfinance.� rc/1.9.55265?cid=PSD_MFGatewayBulletinEN_W_EXT Donor Brief 5. Washington, D.C.: CGAP, July. http:// www.cgap.org/p/site/c/template.rc/1.9.2702/ Social Performance Task Force. 2011. Universal Standards for Social Performance Management. Navajas, Sergio, and Mark Schreiner. 1998. “Apex http://sptf.info/sp-standards Organizations and the Growth of Microfinance in Bolivia.� Economics and Sociology, Occasional Paper World Bank. 2006. Microfinance in Southasia: No. 2500. http://www.microfinance.com/English/ Toward Financial Inclusion for the Poor. http://web. Papers/Bolivia_Apex.pdf worldbank.org/WBSITE/EXTERNAL/COUNTRIES/SO PKSF Newsletter. 2010. ENRICH: A Holistic Approach UTHASIAEXT/0,,contentMDK:21404284~pagePK:14 to Household-focused Poverty Eradication: A New 6736~piPK:146830~theSitePK:223547,00.html Initiative of PKSF. Dhaka, Bangladesh, August. http://www.pksf-bd.org/~pksfbd/images/stories/pdf/ Evaluations and Field Reports wessite%20document%20for%20newsletter.pdf Afghanistan: MISFA Richter, Patricia. 2004. “The Integration of the CGAP. 2008. Building a Microfinance Industry from Microfinance Sector into the Financial Sector Scratch. http://www.cgap.org/p/site/c/template. in Developing Countries: The Role That Apex rc/1.26.2213/ Mechanisms Play in Uganda.� http://www. microfinancegateway.org/p/site/m//template.rc/ Nagarajan, Geetha, Henry Knight, and Taara 1.9.25917 Chandani. 2006. Mid-Term Review of the Microfinance Roodman, David. 2012. Due Diligence: an Impertinent Sector and MISFA in Afghanistan. http://www.mrrd. Inquiry into Microfinance. Washington, D.C.: Center gov.af/MISFA/FULL%20REPORT%20-%20GN%20 for Global Development, Brookings Institution Press. -%20Oct06%20-%20FINAL.pdf Rosenberg, Richard. 2009. “Measuring Results of Bosnia–Hezegovina: LIDs Microfinance Institutions: Minimum Indicators That Donors and Investors Should Track.� Technical Goodwin-Groen, Ruth. 2003. “Avoiding Apex Guide. Washington, D.C.: CGAP. http://www. Pitfalls: LIDs of Bosnia and Herzegovina.� CGAP cgap.org/gm/document-1.9.36551/Indicators_ DIRECT donor brief. http://www.cgap.org/gm/ TechGuide.pdf document-1.9.2810/CaseStudy_06.pdf/ 26 Social Development Notes. 2001. Innovative Approaches Pakistan: PPAF to Microfinance in Post-Conflict Situations: Bosnia Local Initiatives Project. http://siteresources.worldbank. Duflos, Eric, Alexia Latortue, Rochus Mommartz, org/INTRANETSOCIALDEVELOPMENT/873351- Graham Perre, and Stefan Staschen. 2007. CGAP 1111664103368/20489583/sdn50-Bosnia.pdf Country-Level Effectiveness and Accountability Review with a Policy Diagnostic. http://www.cgap.org/gm/ India: FWWB document-1.9.2183/clear_pakistan_report.pdf Mohan, Brij, Mythili Narayan, and Vinod Jain. 2005. Sierra Leone: MITAF Mid-Term Review Report of FWWB USAID Project. http://pdf.usaid.gov/pdf_docs/PDACG427.pdf Duval, Ann. 2009. Development of a Sustainable Pro- Poor Financial Sector in Sierra Leone. MITAF Final Nepal: RMDC Evaluation. http://erc.undp.org/evaluationadmin/ downloaddocument.html?docid=3439 Man Shrestha, Shankar. Microfinance in Nepal: Experiences of RMDC as an Apex Microfinance Duval, Ann, and Franklin Bendu. 2006. Development Organization. http://www.pksf-bd.org/Seminar%20 of a sustainable pro-poor financial sector in Sierra Paper/RMDC-%20Nepal_Shankar%20Man%20 Leone. MITAF Mid-term Evaluation. http://www. shrestha.doc uncdf.org/english/microfinance/uploads/evaluations/ SIL-03-C01_midterm.pdf No. 80 June 2012 Please share this Focus Note with your colleagues or request extra copies of this paper or others in this series. CGAP welcomes your comments on this paper. All CGAP publications are available on the CGAP Web site at www.cgap.org. CGAP 1818 H Street, NW MSN P3-300 Washington, DC 20433 USA Tel: 202-473-9594 Fax: 202-522-3744 Email: cgap@worldbank.org © CGAP, 2012 Authors Sarah Forster, Eric Duflos, and Richard Rosenberg wrote this paper with contributions from Nina Holle and Mayada El-Zoghbi. This paper would not have been possible without the generous contributions of six apexes that agreed to be case studies: BMS (Mali), Bancoldex (Columbia), MISFA (Afghanistan), PKSF (Bangladesh), SFD (Egypt), and SIDBI (India). Likewise, the 23 apexes that completed our global survey provided valuable data and insights. The authors would also like to thank an informal working group of apex donors, made up of AFD (Philippe Serres), AsDB (Qifeng Zhang and Binh Nguyen), DFID (Claire Innes), EC (Alessandra Lustrati), GiZ (Johannes Majewski), IFAD (Michael Hamp), IFC (Deepak Khanna and Momina Aijazuddin) , JICA (Kazuto Tsuji), KfW (Matthias Adler), UNCDF (Henri Dommel and John Tucker) and the World Bank (Doug Pearce and Niraj Verma). This group provided early input into the research design and also provided feedback on the initial findings at a one-day workshop, in which managers from SIDBI and MISFA also participated. The conclusions of this very productive workshop contributed substantially to the paper. Finally, the authors thank Beatriz Marulanda and CGAP staff and research assistants who provided useful inputs and comments on the report. In particular, Mohammed Khaled in Egypt and Corinne Riquet in Mali collected much vital information, while Greg Chen, Jasmina Glisovic, Kate McKee, Steve Rasmussen, and Anna Martinez Zubizaretta provided input and helpful feedback on drafts. The suggested citation for this paper is as follows: Forster, Sarah, Eric Duflos, and Richard Rosenberg. 2012. “A New Look at Microfinance Apexes.� Focus Note 80. Washington, D.C.: CGAP, June.