80538 focus note A Microcredit Crisis Averted: The Case of Bangladesh B angladesh’s microfinance industry, which grew large in the 1990s, continued to expand well into the new century, adding 15–28 percent active loan products, and completed other house cleaning. The number of borrowers has plateaued ever since. borrowers annually from 2004 to 2007. Then in late The slight contraction of branches and staff and the 2007 microfinance institutions (MFIs) began to worry leveling off of customer numbers in turn affected the that continued rapid growth could have negative loan portfolios. Figure 3 shows their combined portfolio, consequences. In 2007, Shafiqual Haque Choudhury, and distinguishes the microcredit loans—to members the founder and president of ASA, one of the largest of village groups—from small enterprise loans (SEL) MFIs, remarked, “Excessive lending into a saturated to individual businesses. Interestingly, SEL rose from market could cause a ‘train crash’ that might cause 10 percent of portfolio in 2003 to 30 percent by 2012, great sector-wide damage and burden borrowers with the biggest shift happening during 2007–2008 with debts they did not need.� 1 just as microcredit lending slowed. Microcredit portfolios grew again in 2011–2012 as a result of a Bangladesh’s microfinance was on the verge of a sharp increase in loan sizes, a step-change that is sharp change in direction. The country’s big four expected to level off, with loan sizes in the future MFIs—ASA, BRAC, Buro, and Grameen Bank, which stabilizing in line with inflation. constituted two-thirds of microfinance supply for the past decade—in aggregate stopped adding branches The story behind the numbers and staff around 2008 (Figure 1). The change in course happened without notice or wider public discussion, Why was growth so fast up to 2008? Why did the and before microfinance crises in other countries, expansion of branches and staff suddenly stop in 2008? such as Nicaragua, Morocco, and India, came to light. Why did the number of borrowers level off? How has the market adjusted? Was a crisis brewing but then averted? Soon, the aggregate number of borrowers also stopped growing (Figure 2). The active borrower totals contracted This Focus Note explores these questions. To see modestly as the sector pulled back, closed some ancillary the full picture, we will describe not just how the Figure 1. Staff and branch numbers, big four MFIs MFI Staff and Branches 90 10 80 9 70 8 Branches Thousands 7 Staff Thousands 60 6 No. 87 50 July 2013 5 40 4 30 3 Greg Chen and Stuart 20 2 Rutherford 10 1 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Staff Branches 1 Stuart Rutherford interview with Choudhury in 2007. 2 Figure 2. Active borrowers, big four MFIs AcƟve Borrowers 20 18 16 14 12 Millions 10 8 6 4 2 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 (Note: Aggregates borrower figures from all four MFIs and therefore double- or multiply-counts borrowers with loans from more than one MFI. Estimates indicate that borrowers with loans from more than one MFI are just over 30 percent nationally.) MFIs behaved, but also compare the accounts given numbers provided by the four largest MFIs.2 Several by MFI leaders with the views of their clients. Our senior managers of each MFI were interviewed at aim is to describe the evolution of microfinance in length, and the MFIs also provided key internal policy Bangladesh over the past decade and to draw lessons documents. To add a demand-side perspective, 43 from the sharp change in direction that began in 2008 rural households were interviewed in depth during as Bangladesh averted a crisis. the first quarter of 2013 (summaries of each interview are available at http://www.cgap.org/publications/ Our information comes from three sources. A household-interviews-bangladesh-2013.) Combined decade’s worth of financial and performance data these three sources of evidence provide a robust and were taken from audits and augmented by other balanced picture. Figure 3. Gross loan portfolio, big four MFIs (constant 2012 $US) 3,000 2,500 2,000 $US Millions Small Enterprise 1,500 Loans 1,000 Microcredit 500 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2 The period 2003–2011 from audits; 2012 from provisional data provided by the MFIs. 3 The paper is arranged in six sections. The term microcredit originated in Bangladesh and refers to the provision of small, one-year loans primarily to rural I. An update describing how modern microfinance women organized in groups. These loans are repaid operates in Bangladesh. in frequent, regular installments, and there has always II. Explanations of why growth was so aggressive been a savings element. Today, just as they always have from 2002 to 2007. done, loan officers visit villages regularly to collect loan III. The story of how MFIs sensed a looming crisis repayments and savings deposits. But these outward and how they reacted between 2008 and 2010. continuities belie underlying fundamental changes. Any IV. A description of the kind of crisis averted by obligation for group members to repay loans for each MFIs and their clients. other (joint liability) has disappeared; it’s up to the loan V. A summary of the explanations of how the crisis officer to collect. Loan and savings payments are still was avoided. collected at a set time in each village but increasingly VI. Identification of five emerging trends in groups do not meet as a body. Some members send microfinance in Bangladesh following the marked their repayments through a neighbor or simply wait for changes in direction of the 2008–2010 period. the loan officer to appear at their doorstep. Savings, which were once fixed in value and compulsory, have I. Microcredit in Bangladesh become largely voluntary and more flexible. The story of microcredit can be fully understood only These changes reflect the long experience of clients within the context of modern Bangladesh, where who are deeply accustomed to the rules and rhythms economic and social conditions have been improving. of microcredit, and who can chose among several While income levels remain low, there has been steady competing MFIs. These factors have reduced the growth of 5 to 6 percent for the past decade that raised importance of meetings and group cohesion that incomes to $1,700 (per capita gross domestic product were once central to microcredit. [GDP] on purchasing power parity [PPP] basis) by 2012. Many rural households enjoy new income streams from The 43 household interviews conducted as background family members working overseas, or in factory jobs in for this paper were selected randomly and not the rapidly expanding ready-made garments industry. because they were known to be MFI clients (Box 1). Strong development fundamentals have positioned The interviews reveal how widely available microcredit Bangladesh as one of the few low-income countries has become. Twenty-seven of the 43 households were on target to achieve its Millennium Development actively borrowing, and held among them 51 MFI loan Goals (MDGs) (The Economist 2012). Roads have accounts. Sixteen households were borrowing from been much improved, and telecommunications have more than one MFI. Any of the 43 households could undergone a mobile-phone revolution. There is little have taken a loan from an MFI at any time. doubt that microfinance’s growth has been aided by the improving conditions, rising incomes, and The client interviews are a reminder that Bangladeshi expanding markets, especially in rural areas. We also microcredit functions in an environment where clients believe that microcredit has made a large contribution are well experienced with a service that has been to the improving conditions, especially as a means widely available for over a generation. Many current of engaging women more fully in the economic and microcredit users had mothers or even grandmothers political life of the country. As S. N. Kairy, the chief who were MFI borrowers. The interviews also reveal financial officer of BRAC aptly summarized, “The lives how many borrowers carefully weigh the tension that of rural people in Bangladesh are totally different than comes from having to meet the regular, frequent 20 years ago.� 3 repayments required of microcredit. 3 Unless otherwise noted, all quotations are from interviews with the authors conducted for this paper between January and May 2013. 4 Box 1: Snapshots of Rural Households and Microcredit To incorporate a client-side perspective on modern The households’ relationships with MFIs are as follows: microcredit 43 rural households were interviewed in depth in the first quarter of 2013. Twenty-five are Households randomly surveyed 43 from central Bangladesh, which is more commercially Households with MFI membership/ 41 (95%) active and highly competitive for microcredit; the savings ever rest are from southwestern Bangladesh, where MFI Households with MFI membership/ 36 (84%) competition and economic activities are moderate. savings at the time of the interview Summaries of the interviews are available at http:// Households with a loan from 11 (23%) www.cgap.org/publications/household-interviews- one MFI bangladesh-2013 Households with loans from 11 (23%) The number of households interviewed was small, two MFIs and the data obtained cannot represent Bangladeshi Households with loans from three 5 (7%) rural households generally, but the interviews do or more MFIs capture the feel and flavor of microfinance in this era. The individual households were selected by walking The households’ use of financial services into random villages to find respondents. Homes that All 36 current MFI users held savings at their MFIs, were obviously wealthy were bypassed, but other averaging about $160 each. Three households held than that interviews were held with any household savings of more than $500 (one had more than that was available and open to talk. The interviewers $1,000). These larger savings are mainly held in the described themselves as researchers on rural very popular commitment savings plans. Some held livelihoods and did not mention microfinance until MFI savings only because they were required to. Few well into the interview. Afterwards, the respondents’ households reported anything more than trivial levels verbal consent to use the interview and photos was of savings in cash at home. Five households were obtained. Normally only one interview was taken in using private saving clubs. any one village; however, exceptions were made in two cases where there were particularly interesting Three households held loans from formal banks, and and relevant respondents. five held policies with formal insurance companies (others had policies that had lapsed). Fifteen The households—basic demographics households held interest-bearing loans from private Average household size was 5.4, adults had an average non-MFI sources, many others had done so in the of 3.8 years of schooling, and almost all school-age recent past, and many of these loans were large children up to age 15 were in school. Nine households relative to MFI loans. Several households were were headed by women: six of them were widows; using MFI and private loans in combination (using three were deserted by husbands. Respondents were MFI loans to pay off, bit-by-bit, large private loans, both Muslim and Hindus. or using private loans to cover MFI repayments in times of stress). Almost everyone participated in the The households had an average of just under two active neighborhood-level exchange of interest-free loans— income earners each. Most employment was in daily for day-to-day expenses (including paying MFI dues) and casual wage labor, small shops and businesses, or for small-to-medium scale expenditures. Some had farming and livestock rearing, and rickshaw driving. borrowed large sums from relatives for purposes such Most households had more than one source of income, as sending people overseas to work. Most borrowing though some were occasional or seasonal. and repayment is in cash, but some farmers use When asked to rank themselves 24 (58 percent) traditional “paddy loans�—borrowing in cash to buy thought they were “middle income,� 16 (37 inputs and repaying in paddy. Pawning jewelry and percent) described themselves as “poor,� and 3 mortgaging land were also reported. The interviews (7 percent) thought of themselves as “very poor.� confirmed what we are coming to know well: financial Thirteen households had no land whatsoever. Of lives are complex, varied, and rich in the range of the remaining 30, nine had their own crop land instruments used. and 21 had homestead-land only (of these, several Three striking themes from the interviews farmed land that they leased or share-cropped). Ten households (23 percent) were getting remittance Experience leads to more careful use of MFIs. income from someone overseas or from a large city; Many households reported that they had reduced several more were trying to arrange for someone to the number of MFI accounts they held, or the work abroad. number of loans, or both, saying that experience had 5 taught them to moderate their borrowing. MFI staff, take  loans.� In his branch, less than 70 percent of several of whom we also interviewed, unanimously members are borrowing. reported that their members have become steadily MFI borrowing can be stressful. Many interviews “more conscious�—meaning more judicious in taking described the tension involved in MFI loan repayments, and using loans, and are better able to manage and this was one of their main reasons for taking repayments. fewer or smaller loans. This stress is aggravated  by Strong demand for savings services. The percentage the tenacious loan collection methods employed by of members who want to use MFIs to save rather most MFIs—especially afternoon or evening visits than borrow is growing. One regional MFI in central by motorcycle by loan officers working in pairs, who Bangladesh told us that “fifteen years ago members refuse to leave a late-payer’s home until at least a begged us for more loans: now we beg them to token payment is made. Beyond Microcredit commercial banks rather than MFIs. Because of Within Bangladeshi microfinance, microcredit the massive growth of mobile phone subscriptions, remains dominant. But new services are developing these services are increasingly available to the quickly. Some years ago MFIs became distribution same market segments served by microcredit networks for inward foreign remittances. The demand and SEL. for savings, as indicated in Box 1, is growing, and MFIs are responding with a broader range of flexible Supply—highly concentrated in four providers short-term and disciplined long-term plans. On There are well over 600 formal microfinance the credit side, the largest addition has been SEL, providers in Bangladesh licensed by the regulator, as Figure 3 revealed. In the 1990s, MFIs began to but the largest four dominate the market. Table 1 prioritize small enterprises as a way to promote off- presents data on each individually and collectively farm businesses and rural value chain linkages that as a proportion of the entire market. The big four could generate larger income and more employment. make up one-half to three-quarters of microfinance Small enterprises are seen as a way to diversify local supply. There are many regionally based MFIs that economic development beyond promoting the clients mention in field discussions, but these large smaller activities at the household level that are the four organizations have wide name recognition across targets of microcredit (Wood and Sharif 1998). Bangladesh. Another new market segment arose in the 1990s These largest four have many similarities, in­ from a realization that the very poorest households cluding a deep grounding in promoting national were not being reached by microcredit—the “bottom development following the country’s independence 20%.� To reach the destitute, wider livelihoods struggle. Each has a strong nonprofit orientation,5 support strategies were developed. There are a and they all provide a similar group-based number of ultra-poor programs, but the best known microcredit product as their primary service. At is BRAC’s Targeting the Ultra Poor (TUP) approach the same time, each has distinct features and replicated in several other countries. TUP couples 4 backgrounds: access to savings with livelihoods development and safety nets. • ASA was founded by Shafiqual Haque Choudhury and a small band of colleagues as a nongovernment One of the newest services in the microfinance organization (NGO) in 1978 promoting the social industry is mobile phone banking offered by and political empowerment of the poor. In 1991 4 CGAP and the Ford Foundation have partnered to test adaptations to the BRAC approach in eight other countries and 10 pilot sites. 5 Technically, Grameen Bank is a shareholding entity, but member–shareholders own a large majority, and it has typically had a return on assets lowest among the four. 6 Table 1. Largest Four MFIs and the Microfinance Market (2012) 4 MFIs as % of ASA BRACa BURO GRAMEEN All MFIsb Loan Portfolio ($US Millions) 636 750 110 997 65 Savings Deposits ($US Millions) 262 311 42 1,628 86 Branches 3,025 2,120 625 2,567 49 Active Borrowers (Millions) 4.2 4.4 1.0 6.7 72 Microfinance Staff 20,969 17,700 5,447 22,621 53 Year of Launch of Microcredit 1991 1974 1991 1976 — BRAC’s figures include only full-time microfinance staff, though shared services staff (e.g., accounting and finance) would increase the full- a.  time equivalent head count. Percentages based on large four MFI figures for 2011 as a proportion of industry totals from Bangladesh Microfinance Statistics, 2011, b.  Credit and Development Forum & Institute of Microfinance it turned single-mindedly to becoming the world’s II. Infectious Growth most efficient MFI. In 2006 ASA launched ASA International, which now owns and operates By 2002 microcredit in Bangladesh was already two affiliates in eight countries and is in the process of decades old, very large, and well-tested. Bangladesh establishing several more. had shown not only how to establish microcredit but • BRAC is one of the world’s largest NGOs and was how to scale it up. A number of set-backs had been recognized by the Global Journal in 2013 as the top overcome. In particular, disastrous flooding in 1998 NGO globally. BRAC was founded by F. H. Abed in and several years of injudicious ratcheting up of loan 1972 with a mission to fight poverty through large- sizes had combined to damage the portfolio of the scale basic health and education interventions, in flagship program, Grameen Bank (Yunus 2002). But addition to microfinance. In 2002 BRAC launched Grameen bounced back by 2002 with Grameen II, a operations internationally, and today operates redesign of its products that brought some welcome affiliates in 10 countries. flexibility to its lending. ASA had grown to become • Buro Bangladesh was founded in 1991 by a one of the largest MFIs in Bangladesh with among small group of people led by Zakir Hossain who the lowest costs anywhere. BRAC already had a large had fought in Bangladesh’s liberation war. It is the smallest of the four MFIs featured here, but microcredit program and was expanding its SEL has recently surpassed 1 million client members. portfolio aggressively, alongside its large-scale health It started as a regional player but now works at a and education work. Buro was moving beyond its national scale. Buro has a reputation for innovating home district of Tangail to establish a wider national flexible pro-poor savings services. presence. Microcredit was in good shape, and a • Grameen Bank pioneered modern group-based euphoric period of even faster growth set in between microcredit in the late 1970s. In 1983 a unique 2002 and 2007. ordinance established it as specialized bank for the poor. It was initially capitalized from To grow, MFIs had to raise new government sources, and later primarily from kinds of commercial capital shares of borrower–clients, or members. The Bank and its founder Muhammad Yunus won the In the 1980s and 1990s international donors helped Nobel Peace Prize in 2006. In 2010 Yunus fell into launch microcredit, often by funding poverty projects dispute with the Bangladesh Government and with grants. In 1990 the Government of Bangladesh resigned from the post of managing director in set up Palli Karma-Sahayak Foundation (PKSF), an May 2011. apex designed to wholesale funds to microcredit, 7 Figure 4. Sources of funding, all four MFIs (constant 2012 $US) 4,500 4,000 3,500 Other 3,000 $US Millions 2,500 Deposits 2,000 Borrowings 1,500 1,000 Equity 500 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Note: PKSF’s loans to MFIs amounted to about $400 million in 2011, PKSF Annual Report 2011. and later the World Bank and other donors provided Grameen Bank has member–shareholders, none it with financing of over $400 million (Forster, Duflos, of the four MFIs has ever sold equity to outsiders, and Rosenberg 2012). PKSF loans fueled the growth so management has never had to respond to the of many MFIs in the 1990s. Among the large four, expectations or timelines of external investors. Thus, ASA and BRAC borrowed from PKSF heavily while while Bangladeshi microcredit was able to grow large Grameen began with loans from other sources, such by tapping into commercial borrowings and savings, as IFAD. the four large MFIs retained a nimbleness that helped them to adapt quickly and avoid a crisis. But in the new century the growth ambitions of MFIs outpaced what PKSF or donors alone could But, why did MFI management provide. The four MFIs began to cultivate new choose to grow so fast? sources of funding. Grameen II mobilized ample savings not just from borrowers but also from Even though conditions were conducive and ample the general public, and today the Bank holds funding was found, there was no imperative for MFIs savings worth more than one-and-a-half times to continue to grow or to grow at a fast rate. MFI its loan portfolio. ASA demonstrated how cost- boards and managers controlled their own destiny, effectiveness can generate large surpluses, and so why did they decide to grow so fast from 2002 by 2007 it was financing its loans almost entirely to 2007? from retained earnings and deposits. BRAC also mobilized deposits from clients but began to Ambitious, intensely competitive institutions command loans from domestic commercial banks Many of Bangladesh’s MFI leaders came to the fore in almost equal amounts. The funding vehicles following a bloody war of independence in 1971 and became more sophisticated: BRAC, for example, a famine in 1974 that left their country in chaos. Their opened a securitization window in mid-2006. Buro sense that “something must be done for my country� arranged a syndicated loan from a bank consortium went beyond do-gooder sentiment and drove an to fund its national growth strategy. ambition to make a contribution at the national level. As F. H. Abed, the founder of BRAC, says, “Small may With more capital raised from private savers and be beautiful, but large is necessary.� The imprint of from bank loans, Bangladesh microfinance has the founders’ ambitions remains deeply embedded become increasingly commercial. However, although in their respective organizations. 8 Ambition is a large part of what makes these Bank won the Nobel Peace Prize, and in 2007 ASA organizations so formidable and influential, and was recognized by Forbes Magazine as the top it fosters competition among them that has only MFI globally. Such attention naturally spurred the intensified as each institution has grown. Sheer size— Bangladeshi MFIs to become even larger. especially the number of clients—soon became the popular measure of inter-MFI rivalry. Buro, once a New microcredit regulation regional player, became determined to establish Regulation also played an important, if unintended, a national presence and measure itself against the role in accelerating growth. In 2006 Bangladesh’s giants. ASA was a late starter, having begun with parliament passed the Microcredit Regulatory Act, an ideological preference for social over financial creating for the first time an independent authority empowerment, but once it decided, in 1991, that to oversee all microcredit operators except Grameen microcredit was the most effective way to help Bank.6 When the new Microcredit Regulatory millions of poor people, it grew fast. By 2002 ASA Authority (MRA) began floating possible regulations was snapping at the heels of BRAC and Grameen, in 2007, it proposed that MFIs would need MRA and during the 2002–2007 “growth spurt,� it caught permission for each new MFI branch. The idea was up with them, at least in terms of the number of its later discarded, but by then it had provoked a reaction clients and branches. by MFIs who rushed to set up branches in advance of receiving their initial licenses. BRAC was particularly MFI management knew that the industry was already aggressive, leaping from 1,500 to 2,900 branches large and that there would eventually be a ceiling during the second half of 2007 just before it received to microcredit expansion: that sooner or later the its license in late November that year. Medium and market would peak and growth would slow. But that small NGOs also set up additional branches in this only intensified their efforts to claim market share period, and some NGOs turned to microcredit for before saturation was reached. It became a “race to the first time, in part to ensure themselves a source of the edge.� As Mosharrof Hossain, director of finance income as a licensed provider. From the end of 2007 at Buro Bangladesh explained, “We continued to to the end of 2008 MRA licensed nearly 400 MFIs and expand our operations rapidly from 2005 through would license more than 200 in subsequent years. 2010 to achieve our goal to be a nationwide microfinance organization. We knew that overlapping III. Reaching the Limits (multiple borrowing) was increasing but we felt that of Microcredit despite this that we were experienced and could handle things.� No one told Bangladesh’s biggest MFIs to stop the expansion of microcredit abruptly in 2008. There were International encouragement no directions from regulators or government, nor The domestic ambitions of the MFIs fueled growth, recommendations from industry bodies. There were but international encouragement greatly magnified damaging cyclones in 2007 and 2009, but they were it. In 2002 Bangladesh’s major MFIs were the less severe than others in Bangladesh’s history, and world’s biggest, and among the most efficient. their effect on microcredit was not serious enough Their managers regularly showcased their work at to be part of decisions to slow growth. The dispute international events. Their scale and efficiency were between Yunus and the government was still some praised, promoted, and tracked by organizations years off and only came to the surface in 2011. Global such as the Microcredit Summit, CGAP, and MIX. encouragement was undiminished—internationally The United Nations declared 2005 the Year of in 2008 growth remained the watchword, and this Microcredit. In 2006 Muhammad Yunus and Grameen was the era that saw a rapid influx of new investment 6 This did not include Grameen Bank, which fell under its own ordinance. 9 vehicles for microfinance. Yet, Bangladesh microcredit began to slow well before microfinance Box 2. The Size of Bangladesh’s Microcredit Market problems in Bosnia, Morocco, Nicaragua, or India The local microfinance industry body, the Credit emerged. The Bangladesh slowdown also preceded and Development Forum (CDF), collected data the international financial crisis that began in late from some 612 MFIs in 2008, and reported a 2008. total of 24 million borrowers.a After adjusting for estimates of multiple borrowing (clients with loans from more than one MFI), there were about Moreover, the slow-down was not coordinated among 17 million unique individuals borrowing, from about the big players, nor did they act at exactly the same 15 million households. By this estimate nearly half time. For example, in mid-2007, just as ASA started of Bangladesh’s 33 million households had loans from MFIs. Since low-income households are the to back-pedal, BRAC began the biggest branch primary target, the level of coverage among that expansion in its history. But by mid-2008 these late subgroup must have been even greater. In terms spasms of growth had ended, and aggregate branch of the percentage of targeted borrowers reached, microcredit must surely have been very close to and staff numbers stopped growing. Only Buro, which full saturation. Sanjay Sinha of MCRIL offers a very had taken on large bank debt and needed to deploy similar analysis based on 2009 data.b it quickly, bucked the trend, and did not slow growth until 2010. Microfinance Statistics 2009, Institute of Microfinance and a.  Credit and Development Forum, Dhaka Bangladesh Microfinance Review, August 2011, BRAC b.  Development Institute MFIs slowed their own growth in reaction to two main problems: Our random interviews of 43 low-income rural • First, they began to sense the negative conse­ households in early 2013, when the number of quences of market saturation of the core microcredit borrowers was similar to 2008, showed that every market. household that wanted a microcredit loan either had • Second, they became more aware of the man­ one or could have easily have obtained one. Of the 43, agement problems created by the rapid growth of only two households had never been MFI clients. Thirty- 2002 to 2007. six households held MFI memberships (and therefore MFI savings) at the time of the interviews. Of these, 27 A crowded market were borrowing and among them had 51 loans. Market saturation occurs when the provision of Differing views on the benefits and risks of a service reaches the limits of a targeted client multiple borrowing segment’s effective demand, but this concept is Multiple borrowing—borrowing simultaneously from difficult to measure. In Bangladesh in 2007, MFIs two or more MFIs by a single borrower—had long had still not reached some villages in remote areas, been noticed in Bangladesh, where it is commonly and many of the poorest households were left out of referred to as “overlapping.� In 2001 Wright, microcredit. SEL lending and the ultra-poor programs Christen, and Matin noted that “[c]lients are now able were still well short of the total demands of these to choose between as many as five or more MFIs in market segments. But the supply by MFIs to their many villages, and multiple membership of MFIs has core targeted client segment of basic microcredit— risen to unprecedented levels.� loans to women from low-income households in rural areas—had come very close to saturation. Nearly Since that 2001 report the industry had expanded every target rural household who wanted a loan massively, and by 2007 multiple borrowing almost already had one or more than one. Nearly every certainly had increased, with some estimates indicating household had been a member or borrower at some that just over 30 percent of borrowers had loans time in the recent past. from more than one MFI (Khalily and Faruqee 2011). 10 Shafiqual Haque Choudhury, president of ASA, certainly might go back on their promises to do the same. thought so. Alerted by reports from his staff, he made The forum also discussed setting up a credit bureau, an incognito visit to some ASA clients and found in his but this, too, failed to gain momentum at the time. words a “horrifying situation�: some of them held loans The meetings were convened regularly for a year from three, four, or even five MFIs. Choudhury took a or more but were eventually quietly abandoned. straightforward view, “if borrowers take too many loans Competitive rivalry dampened any potential they will run into repayment problems, and would stop collective response. repaying me or someone else.� Instead, each MFI slowed growth in its own way. . . Not everyone shared Choudhury’s fears. Shameran ASA was the first big MFI to take unilateral action. Abed, associate director of BRAC’s microfinance In mid-2007 it halted the growth of its branch programs notes, “BRAC was sensitive to multiple network. The halt was ostensibly to allow time for borrowing, but we were also confident that our new computerization to be embedded; but it was experience equipped us to manage this reasonably never reversed once that goal had been achieved. well.� Some in the industry noted that multiple ASA’s branches peaked at 3,334 in 2007 and have borrowing has merits, arguing that it allows declined slowly to 3,015 by the end of 2012. This borrowers to better manage cash flows and it decision was driven by ASA’s worries about multiple spreads repayment risk among the MFIs. Research borrowing, but reinforced by a rise in delinquency ASA shows that households need to form more than one experienced late in 2007. The circular that ASA sent to lump sum each year, and therefore access to loans its field staff announcing the cessation of any further more than once a year can be helpful. Mosharoff branch openings also instructed branch managers Hossain of Buro notes, “We’ve long been aware to discourage multiple borrowing by cutting the of multiple borrowing, and have sent senior number of clients per worker and to slow down client management teams out to see how many passbooks recruitment, “even if this means that total member our borrowers keep. We determined that in most numbers go down.� To provide added security against cases it was not risky and therefore carried on to potentially delinquent loans, clients were to be told reach our targets up through 2010.� that they must open an additional commitment savings account if they borrowed more than $200. Despite Collective action failed. . . this decisive early action, ASA still had some doubts. ASA, however, remained anxious. In late 2007 Enamul Haque, then ASA’s executive vice-president of ASA, BRAC, Buro, and Grameen Bank set up an operations, recalls, “I felt that we were losing our edge informal discussion forum attended by senior when growth slowed and we could regain client and managers.7 The main agenda was how to manage staff enthusiasm by resuming growth.� This view has multiple borrowing. ASA, motivated by its fear almost won out as ASA has reconsidered its strategy that over-supply could lead to a repayment crisis, periodically, but cooler heads have prevailed and the argued for a division of territory, and suggested slowdown has remained in place. reciprocal closure of branches in over-served areas. This earned a tepid response. The others did not Just as ASA put an end to the opening of new share ASA’s level of concern. Moreover, Grameen’s branches, BRAC was entering its final sharp growth Mohammad Shahjahan says it has always been the spurt in advance of receiving its MRA license in late Bank’s philosophy that “clients are better off if they 2007. But shortly thereafter BRAC ran up against a have a choice between competing MFIs.� There clear sign of market saturation: word filtered back was also a collective action dilemma—the fear by from the field in early 2008 that many of the newly each MFI that if it were to close branches its rivals opened branches couldn’t recruit enough clients 7 This included key lieutenants from among the senior ranks of each MFI: Sushil Ray of ASA, Shabbir Choudhury of BRAC, Mosharrof Hossain of Buro, and Dipal Chandra Barua of Grameen Bank. 11 to achieve the minimum loan portfolio to reach ensure that poor people got access to credit, so that breakeven. Kairy, BRAC’s chief financial officer, “when we saw this ambition being fulfilled through recalls, “We were pushing to meet disbursement our own work and that of other MFIs, we were happy targets to achieve the necessary portfolio size, but to shift from fast growth to improving quality.� some areas were simply not viable.� Buro, the smallest of the four, continued to grow until BRAC’s growth spurt happened just as BRAC was 2010. Its ambition was to achieve a national presence, expanding its international operations from one to eight but it was also obligated to grow due to the large countries, stretching senior management very thin. bank loans it had taken. Buro’s slowdown came by Within management ranks there were also increasingly 2010, once its size targets had been achieved and in divided views on the health of its microcredit operations. reaction to deteriorating performance. Some noted that those high growth figures temporarily masked underlying problems, and this concern gained Management problems in traction when delinquency rose markedly in 2007 and the wake of growth 2008. By mid-2008 BRAC slowed growth but it was only by late 2009 that it made significant changes. While market saturation was an important reason Senior management installed a new and younger to slow growth, many of the problems that arose team to manage BRAC’s microfinance operations and were caused not so much by saturation as by the usher in a series of reforms. The reforms had many torrent of growth from 2002 to 2007. As those elements, but importantly BRAC scaled back from a problems emerged, the MFIs devoted much time to peak of 2,900 branches in 2008 to fewer than 2,200 by solving them, creating a second and more sustained the end of 2012. phase of moderation. In particular, MFIs focused on re-establishing credit discipline, fixing staffing For ASA, slowing growth also meant dropping challenges, improving internal controls, and repairing a product. To boost its client numbers ASA had their finances. launched guardian loans in 2005—loans to the husbands or fathers of its female members, putting Re-establishing preloan credit discipline a second loan into borrowers’ households. It found Credit discipline begins with attracting suitable these loans hard to collect, and began to phase out borrowers. Bangladesh microcredit had always been the scheme in 2009, contributing to the slight decline focused on poverty alleviation, preferring to take on in active borrower numbers. clients that matched a target poverty profile. This approach worked well enough when microcredit was By the end of 2008 Grameen Bank also slowed new in short supply and growth was moderate. But in the branch openings, and its branch network has remained fierce competition of the years leading up to 2008, almost unchanged since then. Its managers did not the pressure on staff to achieve recruitment and share the same level of concern as those at ASA and disbursement targets over-rode any doubts about BRAC. Grameen believed that multiple borrowing was whether they should take on this client or give a big moderate, with high levels confined only to a few loan to that one. Growth targets over-rode good districts. Its branch growth had been steadier than judgment. BRAC’s in part driven by Grameen’s savings products, which provided the resources to open new branches The first reactions to trouble focused on delinquency according to the steady rise in deposit levels. But management. ASA hired a senior ex-police officer like BRAC, Grameen saw reduced demand for group who called on his old colleagues in the districts to membership and microcredit, and began to relax help staff bring pressure on bad payers, and he the recruitment targets it gave its staff. Mohammad developed an intimidating-looking “contract� that Shahjahan, Grameen’s acting managing director, overdue borrowers were asked to sign. All the MFIs points out that the Bank’s mission was always to appealed more strongly for help from community 12 leaders. BRAC formed a dedicated collection team to Improving internal controls follow up on serious overdue payments. ASA emptied Grameen Bank also focused on improving its lending its executive floor in its Dhaka headquarters, sending and recovery processes. Passbooks, for example, have all the directors to live in the field for months at a fewer errors than they did four or five years ago. Similar time, to lead the fight against delinquency. improvements have been made by the other big MFIs. Buro says its passbooks used to be “messy�; now it These (ex post) measures to follow up on delinquency has them reconciled in public, at the weekly meetings. were only moderately effective, and it was fixing BRAC was dismayed to find that an uncomfortably (ex ante) preloan checks on borrowers that repaired common practice of disbursing “ghost loans� had portfolio quality over time. From 2008 ASA told loan grown up among target-driven staff. Loans were officers not to recruit clients who they knew to have “disbursed� to false names, so that their proceeds loans from other MFIs, something they had a turned could be used to make up the overdue amounts of a blind eye to in the past. BRAC developed new pre- other clients. This practice manipulated repayment and loan documentation that required staff to more disbursement targets. Shameran Abed of BRAC notes thoroughly investigate the livelihoods of would-be that “when we went to investigate we found that ghost borrowers. Critically, the number of branches and loans amounted to less than 2% of portfolio, but more branch targets were reduced, making it possible for troubling than the amount of money involved were staff to use more discretion in picking new borrowers. the signs of the pressures staff were under to meet targets, and our own lax oversight. Until we changed Shoring up staff capacity and compensation. . . our attitude we were at risk of this problem growing The pace of growth also meant that MFIs found it worse.� BRAC tackled this problem by requiring harder and more expensive to recruit staff. In  late branch accountants, who reported up a different 2007 BRAC grew its staff numbers by almost chain of command than loan officers, to process loan 50  percent and found it difficult to fill mid-level disbursement paperwork and verify customers. At the positions with qualified people. ASA, which had just same time it doubled its monitoring unit and filled added an extra staff member at each branch to run its many vacant positions in its internal audit unit. It also new computers, also felt compelled to raise salaries, closed a loophole that let staff withdraw savings to and its staff expense ratio surged. Investments in make up for late loan repayments without having to get staff and training added to operating expenses. (See the consent of the client. Figure 5.) These weakened financial performance in the short run, but were critical investments to re- Buro found that some loans nominally disbursed establish the quality of operations. to group members as individuals ended up in the Figure 5. Operating expenses as % of gross loan portfolio, all four MFIs 11.50% 11.00% 10.50% 10.00% 9.50% 9.00% 8.50% 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 13 Figure 6. Return on Assets, big four MFIs 4.50% 4.00% 3.50% ASA 3.00% 2.50% BRAC 2.00% Buro 1.50% 1.00% Grameen 0.50% 0.00% -0.50% 2004 2005 2006 2007 2008 2009 2010 2011 2012 Note: Grameen Bank’s loan interest rate has long been the lowest of the big four hands of just one or two members, often from elite reserves. Attitudinally, they prepared themselves for households. It stopped this practice by staggering the downturn. For instance, the managers of BRAC’s disbursements and instituting pre- and post-loan microfinance program gave advance warning to checks on borrowers. senior leadership of a coming period of lower growth and poorer performance. Importantly, management Bracing for a financial hit of all the MFIs understood that a short dip in MFIs in Bangladesh are not backed by deep- performance would be necessary to get them back pocketed investors who can pump in cash on short onto stable footing over the longer term. In the end notice, and therefore keep a very sharp eye on their their performance did suffer, especially in 2008 and finances. As they discovered and then set about 2009, but by 2011 and 2012 the overall performance correcting problems, MFIs braced themselves for in the sector had been largely restored to solid a period of worsening performance. Each of the footing.9 four organizations experienced a rise in its own delinquency measures.8 For ASA it happened by end IV. What Kind of Crisis 2007; BRAC and Grameen Bank were affected during Was Averted? 2008. Buro’s difficulties surfaced most visibly later in 2011, after it had achieved a national presence. In hindsight, the signals of impending trouble were clear by 2007–2008. But signals that become The MFIs took several measures to handle the obvious later are often invisible or overlooked or financial downturn. All of them slowed the increase brushed aside at the time. Financial crises can take in loan size during 2009 and 2010, and pressured on unpredictable dynamics. A reaction by a market staff to increase deposits mobilized. Savings provide player may be misunderstood by others who may a buffer against default by allowing MFIs to adjust feel compelled to take action to protect themselves savings against loans that were overdue. MFIs also even at the expense of the wider public interest. upped their provisioning to strengthen their loan loss There can be a precipitous loss of confidence by 8 MFIs in Bangladesh each measure delinquency, but their systems for doing so are not uniform, making it difficult to offer comparable delinquency figures. 9 A review of multiple years of loan delinquency measures, loan loss provisioning, and write-offs tells a consistent story about a rise in problems around 2008 and improvements since then. 14 borrowers or savers leading to a chain reaction of in overall welfare. The case of household 005 falling repayments or rapid savings withdrawals. shows this: Bangladesh was fortunate to avoid a descent into an unpredictable or spiraling deterioration. The improvements in portfolio quality and operations in Mr. MM’s household was landless, and years ago they borrowed heavily in the private market to 2011 and 2012, combined with the more optimistic secure a small piece of homestead land, which views of managers today indicate that more serious is now their one big asset. Mr. MM’s wife joined problems were avoided. There was some damage two major MFIs and year after year they have been using MFI loans to repay the private lenders. from deteriorating performance for the four largest The MFI loans themselves are repaid through the MFIs but this would have been several times worse strenuous work of their son ML who commutes had the trends seen in 2008 been allowed to fester. two hours each day to the local town on his rickshaw, which he then rides for hire all day. Meanwhile, his father goes out begging to get MFIs may have avoided a crisis for themselves by their food. Mostly they eat twice a day, sometimes putting their own houses in order, but where does only once. But all three adults think their sacrifice that leave their clients? As the market expanded is worthwhile. rapidly in the high-growth period, were clients dangerously over-indebted? To explore this question, we turn to our household interviews. Just how pervasive is over-indebtedness and how often does it lead to serious harm? The Institute of Over-indebtedness is tricky to measure. In some Microfinance released a paper in 2011 that presents markets it can be seen in excessive “debt-service evidence that over-indebtedness was low even at ratios�—that is, the borrower spends too big a the height of the growth period (Khalily and Faruqee proportion of income repaying debt. But where 2011). The MFI managers we interviewed feel that households have irregular income streams, as in rural high levels of harm have been avoided. Shafiqual Bangladesh, this definition is not always helpful, as Choudhury of ASA, who had been among the most the case of Mrs. MA (interview 006) shows: vocal in warning of the dangers to clients of over- indebtedness now says that “the train crash was avoided.� On the whole, the evidence from our Mrs. MA’s MFI loan repayments each month exceed household interviews in early 2013 supports this her monthly income. However, she expects to be able to repay the loan because her son should soon conclusion. Twenty-seven of our respondents were reach Malaysia and start remitting money, and even holding MFI loans at the time of the interviews. if he is delayed she is confident of securing family Among them is one tragic case (interview 004) loans, secured against future remittances, to tide where a spiraling combination of business failure and her over. multiple private and MFI loans resulted in a serious long-term drop in standard of living: We therefore use “over-indebtedness� to refer to situations where clients have ongoing difficulty with Mrs. MB (interview 004) joined six MFIs and loan repayments, and to measure the seriousness borrowed from each of them to try to support of the over-indebtedness by the extent to which a husband whose business failed and who had it contributes to serious deteriorations in welfare, contracted loans from private moneylenders. In the end they sold virtually all their assets (including such as reduced consumption, poor health, loss of their land and home) to settle the debts. In the creditworthiness, or excessive stress (Schicks and village, neighbors were more aware of Mrs. MB’s Rosenberg 2011). Some clients accept repayment MFI debts than of her husband’s private ones, which tension as a sacrifice worth making for the benefits were contracted in secrecy in the market. Mrs. MB reports that villagers say things like �MB has been of borrowing, so not all sacrifices made for loan finished off by the MFIs.� repayments should be equated with a reduction 15 Mrs. MB was the only one of our respondents who industry will need to remain vigilant to steer past the had themselves suffered catastrophic loss of assets, potential pitfalls that lie ahead. though we heard hearsay reports of similar cases. Some households reported long periods of reduced V. How Was the Crisis Averted? food intake, illness, and high levels of stress. Two of them (005 and 012) mentioned that they are eating Senior management of the four large MFIs took less to find loan repayments, and one (002) was decisions independently of each other beginning contemplating selling the household’s rice stocks. in late 2007—especially decisions to simply stop Readers of the interviews may judge for themselves growing. There was no credit information bureau to whether they indicate an acceptable or unacceptable provide reliable feedback on borrower indebtedness, incidence of harm. nor were there any regulatory instructions to prompt them. While none of them was omniscient, and Our opinion is that cases of the most serious kind they were unable to agree on the nature, timing, of over-indebtedness, where there is a permanent or severity of the problems they saw, the MFIs were drop in household welfare, are not common now, right to anticipate serious trouble. They pulled back nor were they during the high growth period up and mended their systems. But why were the MFIs through 2008 when microcredit volumes were in Bangladesh—in contrast to those in some other similar to today. That said, the interviews reveal a markets—able to act in time to head off problems level of stress that the microfinance industry would before they worsened? do well to take more seriously, and to seek ways to reduce still further, perhaps by relaxing the mantra Using experience to see through the of zero tolerance on loan repayment schedules so “fog of growth� that repayments fit more comfortably with the often A well-known feature of Bangladesh microcredit irregular and unreliable incomes streams of poor is just how long the four largest MFIs have been households.10 The recent rise in nominal microcredit operating—in each case over 20 years and in the loan sizes, up from $108 in 2008 to $221 today, case of BRAC and Grameen Bank for more than 30. also deserves examination to ensure that the size of Amid the “fog of growth� that enveloped microcredit repayment obligations do not cause further stress, up through 2007 there were many conflicting nor accelerate incidences of over-indebtedness, nor signals. Having already lived through many small discourage poor but creditworthy households from difficulties and having survived some larger ones, borrowing. their managers—many of whom had been in place since the founding of their institutions—knew that This rise in loan sizes is a reminder that risk requires the underlying health of microcredit cannot be taken constant diligence. Bangladesh’s big MFIs averted a for granted. As S. N. Kairy of BRAC notes, “Growth crisis around 2008 but that doesn’t mean that they are in many ways made our numbers look better but we now immune from future ones. Analyses of Grameen’s sensed potentially deeper problems.� Their antennae performance by David Roodman11 show that the value were up and sensing something amiss. of rescheduled loans in the bank’s portfolio has been rising in 2011 and 2012. This change may not pose a Motivations to act—in it for the long-term serious risk to Grameen or the wider industry, but it While long years of experience helped MFI does highlight how performance is always subject to management sense problems brewing, it was their dynamic change. Having averted a crisis before, the future expectations that compelled MFIs to act. 10 See Sinha (2013). 11 See his blog at http://international.cgdev.org/blog/grameen-bank-portfolio-continues-deteriorating 16 Each organization has staked its reputation on of difficulty with their loans, they can also net off microcredit, and their future viability depends on a 12 their savings against their debts and exit without healthy microfinance industry. The main stakeholders the embarrassment associated with default. The in MFIs are senior management, and for them there interviews include many descriptions of clients who is no exit strategy, no personal shares to cash out— used their savings in this way. many count microfinance as a large part of their life’s work. As F. H. Abed of BRAC notes, “Microfinance For the MFIs, savings is both a growing source of has made a big contribution to Bangladesh’s funding necessary to maintain national scale and a development and we expect to continue to provide vital device for maintaining strong credit discipline, microfinance in the years ahead.� offering a risk management tool that helps protect them. Savings provided extra resilience that helped Experienced clients helped, too the MFIs correct a deteriorating situation. This tool A striking feature of our 2013 household interviews was not available in some other national markets is the care with which potential borrowers approach where MFIs’ ability to raise savings was more the decision to take a loan. This comes out of long restricted, or banned outright. experience of working with MFIs and the basic microcredit product whose rules and rhythms they What contribution, if any, did the social goals of know well. Many of today’s MFI borrowers are from MFIs make to the change in course? households with three generations of exposure In acting as they did, the MFIs were influenced to microcredit. They are experienced in making by historical forces whose influence is large, if judgments about the obligations and rules of loan hard to measure. Bangladesh microfinance began agreements. Often this simply means not taking any soon after a brutal war of independence followed more loans: the interviews show many such cases. by a famine that left the country impoverished. Were the clients in Bangladesh less experienced, The belief that microfinance exists primarily to it is likely that many more of them would have draw people out of poverty is widely accepted, borrowed injudiciously, creating much higher levels and microfinance is seen as a “movement� rather of over-indebtedness. The experience of clients than as just an industry. This is not to say that likely muted some of the risks brewing in the rapid Bangladesh MFIs are purely socially driven. MFIs growth period. Zakir Hossain of Buro, commenting take pride in the surpluses they generate, and a ruefully on his MFI’s difficulties with recruiting and good year for many MFI managers often equates training good staff during the growth years, noted to a strong bottom line. The social commitment that “the borrowers were more experienced and to poverty reduction didn’t stop MFIs from knowledgeable than our staff.� undertaking reckless and potentially irresponsible growth up through 2007. Promoting savings helped as a buffer in the MFI–client relationship But those who work in Bangladesh microfinance Bangladesh microcredit has been coupled with argue that their nonprofit character makes a savings from the very beginning, but during the era difference. S. N. Kairy of BRAC remarks, “External of rapid growth from 2003 to 2007, savings products investors are not interested in the same development proliferated and take-up multiplied, with short-term, goals as NGOs.� Choudhury of ASA shares the open-access passbook savings and longer-term, same sentiment about his Bangladesh operations, disciplined commitment plans. This has proven to “Commercially driven private equity would have be a critical tool for managing credit risk for both made the situation worse leaving us unable to quickly clients and MFIs. For the clients their balances offer correct the problems that emerged.� Whether these not only a cushion in times of need, but in case beliefs hold true is debatable, but they underscore 12 Although BRAC is perhaps better known for its public health and primary education work. 17 a common Bangladeshi view about microfinance. must figure out a way to reach back out to those who It is true that while Bangladesh microfinance is require smaller loans.�13 increasingly commercially funded, management has not ceded governance control. We cannot say what Other changes are underway. Grameen Bank has would have happened had more outside investor already introduced the option to top-up loans to interests played a role. There are countries where their full disbursement value part way through external investors and commercialization have built the term, and reschedules delinquent loans into a responsible microfinance sectors. But it would be flexible repayment plan. BRAC has adopted the top- defying the obvious to ignore the development roots ups and is experimenting with loan rescheduling and founding visions in motivating MFIs to pause and and refinancing. Other product refinements under reconsider their growth trajectory. Efforts to remain consideration include loan terms other than the year- true to their professed beliefs in development helped long standard and more flexible loan repayment shape behavior during this critical change in direction schedules to better match customers’ cash flows. in Bangladesh. BRAC is shifting some loan repayments from weekly to monthly. Grameen has introduced a new “young VI. The Road Ahead: entrepreneur’s loan� aimed at the sons and daughters Five Trends for Bangladesh of Grameen members, youngsters who have enjoyed better education and have brighter horizons than The shifts that began in 2008 mark a turning point their mothers and need “something extra� to attract for microfinance in Bangladesh. As traditional them into the Grameen fold. microcredit, which until then had dominated the growth of the industry, began to plateau, other parts 2. Lending to small enterprise will remain a of it—notably small enterprise lending and savings— significant market segment have accelerated, and new possibilities in the shape of mobile services have emerged. As of 2013, five The most marked development since 2007 has been the trends stand out. step change in the value of SEL lending. Bangladeshi MFIs recognized that some of their group members 1. The core microcredit product is holding steady had the capacity to take and repay bigger-than-average and evolving incrementally loans that they would invest in (or repay by means of) larger standalone enterprises, usually retail stores, Microcredit will serve a stable rather than expanding small-scale manufacturing, construction and transport number of clients. Change will come from incremental services (interview 039), or intensive food production. product refinements. MFIs are already moving toward disbursing bigger loans. Loan sizes saw an Of the four big MFIs, BRAC is the only one that upward shift in 2011 and 2012 across all four of the has separated its SEL line of business (which it calls big MFIs. This was done in part to keep up with progoti, meaning progress) from its microcredit. inflation, but as Mosharoff of Buro notes, “The loan Progoti often operates out of newly established size increase was primarily driven by the introduction branches (usually in the market town area rather than of the interest rate cap of 27 percent by MRA in the village) and employs separate, specially trained 2011.� BRAC no longer gives loans of less than US$ staff. It lends mostly to men, as individuals, and loans 125 to microcredit borrowers. The loan size increase are repaid in monthly rather than weekly installments. is a cause of concern. F. H. Abed of BRAC notes, The other three MFIs have not separated their SEL “We are worried about leaving out poorer clients and to this extent and are more cautious. Nevertheless, 13 Traditionally, microcredit borrowers were supposed to own no more than half an acre of cultivable land or its equivalent in other assets. When we asked an MFI field worker what he looked for these days when recruiting members for the microcredit groups, he told us he preferred clients with “good assets�—such as an acre or more of farm land, plus homestead land big enough to grow trees and vegetables. 18 SEL  is likely to remain an important and growing MFI inclination to push borrowing is only reinforced portion of microfinance operations. by the MRA regulator, which has regulations that restrict savings offerings.14 3. Client demand for savings services is increasing; but is inadequately met Savings had always been linked to the core microcredit product, but in recent years the service has become Studies consistently show that, even in the absence of increasingly flexible. Now, as we can see in the formal services, poor people seek ways to save and interviews, almost every MFI offers passbook (that will tolerate the risks of using informal tools to satisfy is, unrestricted withdrawal) savings and commitment this need. As incomes rise and poverty recedes, savings (in one form or another). Almost everyone we populations become more future-orientated, more spoke to knows what these instruments are, even if inclined to worry about providing for tomorrow as they don’t have one themselves. well as for today, and therefore ever more interested in saving. Bangladesh, with its reductions in poverty, It was the expansion of savings under Grameen II that is no exception. The household interviews suggest made a big difference. Grameen’s Pension Scheme, that the awareness of and desire for savings are high. a long-term commitment savings product that safely Even insurance agents now find it worth their while collects regular deposits and pays a good rate of interest to visit quite modest homes to sell policies (often for terms of five or 10 years, has proved attractive for 10-year endowment savings). accumulating large sums. Grameen Bank’s savings balances today are by far the largest of any of the MFIs Mrs. KA (interview 007) has been using MFIs since and provides evidence that demand could be much before she was deserted by her husband. When better met if other MFIs achieve a similar scale, but for he was still around, they managed MFI loans this to happen the regulation of MFI NGOs would need reasonably well from his income. Now she is on her to adapt. This is the main point of trend four. own she has found that saving is a better way to help her secure the future of her children. She has left two of her MFIs and is now in just one friendly 4. There is growing momentum to create a local one that doesn’t insist on her taking a loan. special category of deposit-taking MFIs Mr. ABS and his wife Mrs. MB (interview 016) head a modest farming family. Mrs. MB has two Licenses have now been extended to some 600 or insurance policies to help the couple as they age. One is a 10-year endowment savings plan, and she more MFIs, but the regulations remain based on pays 100 taka ($1.28) a month into it, sourced from a microcredit approach that prioritized credit and selling eggs and saving on the housekeeping. The de-emphasized savings.15 Regulators must protect other is a 15-year term life policy. This requires an depositors’ funds, so caution is warranted for a large annual payment of 5,077 taka ($65)—quite a large sum. To raise this sum on time each year Mrs. MB majority of the 600 licensed MFIs. But applying borrows from an MFI. restrictive rules to all MFIs indiscriminately misses the opportunity to offer more conducive regulations Many MFIs assume that their clients want to, and for a small number of MFIs that are strong enough should, borrow continuously. One of our interviewed to operate under a different regulatory approach. households (003) describes this attitude as “clumsy,� There is increasing discussion in Bangladesh of and several others (like Mrs. KA in the box above) creating a new class of microfinance deposit-taking would like to see MFIs become more willing to allow institution and extending licenses to perhaps five to their members to enjoy ongoing savings services 10 organizations. These MFIs could be given greater without being obliged to borrow. But the general flexibility to mobilize deposits, particularly if critical 14 Microcredit Regulatory Authority unofficial translation of regulations: www.mra.gov.bd 15 There are limits, for example, governing the ratio of deposits held to loans outstanding. 19 organizational standards are met. Views appear to be Global Journal, The. 2013. “Top 100 NGOs 2013.� swinging favorably in this direction: the chairman of http://theglobaljournal.net/group/top-100-ngos/ the MRA (the governor of Bangladesh’s Central Bank) remarked in April 2013 that MFIs should “begin to Institute of Microfinance and Credit and Development think about setting up MFI Banks.�16 Forum. 2009. Microfinance Statistics 2009. Dhaka: Institute of Microfinance and Credit and Development 5. Mobile phone payments services are developing Forum. rapidly Khalily, Baqui, and Rashid Faruqee. 2011. “Multiple Several new mobile-phone-based deposit and payment Borrowing by MFI Clients.� Policy Brief. Dhaka: services are growing rapidly across Bangladesh, Institute of Microfinance. including one from Dutch Bangla Bank and another, bKash, a service provided by a subsidiary of BRAC ———. 2011. “Multiple Memberships (Overlapping) Bank. 17 By the end of April 2013 these services have in Microcredit Market in Bangladesh.� Dhaka: more than 5 million accounts and 80,000 agents, and Institute of Microfinance. these numbers are growing fast. This has put the opportunity to open a basic mobile-phone-based PKSF (Palli Karma-Sahayak Foundation). 2011. Annual account or accessing an agent within reach of millions. Report 2011. Dhaka: PKSF. It is not yet clear what impact this might have on microfinance; but by reaching many of the same clients Schicks, Jessica, and Richard Rosenberg. 2011. “Too mobile phone based accounts and payments introduces Much Microcredit? A Survey of the Evidence on Over- a potentially dynamic new force on the scene. Indebtedness?� Occasional Paper 19. Washington, D.C.: CGAP, September. References Sinha, Sanjay. 2013. “A Challenge to Flat Earth Credit and Development Forum and Institute for Thinking in Microfinance.� Gurgaon, India: Micro- Microfinance. 2011. Bangladesh Microfinance Statistics, Credit Ratings International Limited, May. http://m-cril 2011. Dhaka, Bangladesh: The University Press Ltd. .com/article/A-challenge-to-flat-earth-thinking-in- microfinance.pdf Economist, The. 2012. “Out of the Basket.� November. http://www.economist.com/news/briefing/21565617- Wood, Geoffrey, and Iffath Sharif, eds. 1998. “Who bangladesh-has-dysfunctional-politics-and-stunted- Needs Credit? Poverty and Finance in Bangladesh.� private-sector-yet-it-has-been-surprisingly London: Zed Books. Forbes. 2007. “The 50 Top Microfinance Institutions.� Wright, Graham A. N., R. P. Christen, and I. Matin. 2001. Forbes, 20 December. http://www.forbes.com/2007 “Introducing Savings Services into ASA, a Microcredit /12/20/microfinance-philanthropy-credit-biz-cz Institution.� Small Enterprise Development, vol. 12, _ms_1220microfinance_table.html no. 3: 20–32. Forster, Sarah, Eric Duflos, and Richard Rosenberg. Yunus, Muhammad. 2002. “Grameen Bank II: 2012. “A New Look at Microfinance Apexes.� Focus Designed to Open New Possibilities.� Dhaka: Note 80. Washington, D.C.: CGAP, June. Grameen Bank. 16 Daily Star, 4 April 2013 17 BRAC Bank does not work in microfinance, but its primary promoter is the NGO BRAC. bKash the mobile banking service is a subsidiary of BRAC Bank. No. 87 July 2013 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, 2013 The authors of this Focus Note are Greg Chen and Stuart Rutherford. This Focus Note relied heavily on insights provided by the management of MFIs in Bangladesh. We are especially indebted to Shafiqual Haque Choudhury, Enamul Haque, and Sushil Roy of ASA; Shabbir Choudhury, Ishtiaq Mohiuddin, Shameran Abed, S. N. Kairy, and F. H. Abed of BRAC; Mosharrof Hossain and Zakir Hussain of Buro Bangladesh; and Mohammad Shahjahan and A. S. M. Mohiuddin of Grameen Bank. We benefited greatly from an external review by Syed M Hashemi and Elizabeth Rhyne and from Alexia Latortue and Stephen Rasmussen from CGAP. We are indebted to S. K. Sinha who assisted with the client interviews. Importantly, we are grateful to the 43 households who were open and willing to share their views to add a critical client perspective. The suggested citation for this Focus Note is as follows: Chen, Greg, and Stuart Rutherford. 2013. “A Microcredit Crisis Averted: The Case of Bangladesh.� Focus Note 87. Washington, D.C.: July. Print: ISBN 978-1-62696-016-9 epub: ISBN 978-1-62696-018-3 pdf: ISBN 978-1-62696-017-6 mobi: ISBN 978-1-62696-019-0 UKa from the British people