OP 12-I.qxd 8/6/07 6:55 PM Page 1 40961 OccasionalPaper NO. 12 AUGUST 2007 SUSTAINABILITY OF SELF-HELP GROUPS IN INDIA: TWO ANALYSES Introduction and Summary The self-help group (SHG) model is the dominant form of microfinance in India. SHGs have grown explosively in recent years. It is reported that by March 2006, 2.23 Authors of Part I of this Occasional Paper are Jennifer million SHGs were reaching about 33 million members.1 Such outreach appears to Isern, lead microfinance represent a major breakthrough in a country where 50 million households live in specialist, CGAP; L. B. Prakash, poverty, with very limited access to financial services. executive director, Akshara; Anuradha Pillai, research Although the term self-help group is used in different countries to describe a vari- assistant, CGAP; and Syed ety of financial and nonfinancial associations, in India it refers to a group of 10­20 Hashemi, senior microfinance poor women who band together for financial services--beginning with periodic, specialist, CGAP. compulsory savings and then mainly loans--and sometimes social services as well. Robert Peck Christen, founder, SHGs are managed by their members, with varying degrees of external support. Boulder Institute of SHGs are formed with the assistance of self-help promotion institutions (SHPIs), Microfinance, and Gautam Ivatury, microfinance specialist, which include nongovernmental organizations (NGOs), government agencies, banks, CGAP, wrote Part II of this cooperatives, and microfinance institutions. In addition to helping with group forma- Occasional Paper. tion, SHPIs provide training, monitoring, and other support services. Occasionally, Richard Rosenberg, senior promoters give SHGs initial seed capital to lend, but more typically, groups begin by policy advisor, CGAP, wrote the saving and lending out their members' own resources. Most, but by no means all, Introduction. SHGs eventually borrow from an external source, usually a bank. This bank linkage The authors are grateful to the is the most distinctive characteristic of the Indian SHG model. SHG programs that provided The massive outreach of SHGs has generated interest in the model's sustainability their time and information for and replicability in India and elsewhere. Although SHGs have been widely studied this study. Richard Rosenberg, CGAP; Jeanette Thomas, (see the bibliography for examples), relatively little information has been published CGAP; and Vijay Mahajan, on their financial performance. BASIX/India, provided helpful This Occasional Paper reports on two separate studies of SHG programs con- comments on an earlier version of the paper. ducted by CGAP staff and partners. In Part I, Jennifer Isern, L. B. Prakash, Anuradha Pillai, and Syed Hashemi review SHGs developed by five different SHPIs, represent- © 2007, Consultative Group to ing the main approaches to SHG promotion in India. The study looks primarily at the Assist the Poor financial viability of these SHG programs. 1Data from NABARD: http://nabard.org/pdf/stmt1.pdf. Building financial systems for the poor OP 12-I.qxd 8/6/07 6:55 PM Page 2 Table 1. SHG Programs Included in the Two Studies Part Self-help promoting institution Location Part I Panagal Mandal Mahila Samakhya (PMMS), supported by Andhra Pradesh United Nations Development Programme (UNDP) South Asia Poverty Alleviation Program (SAPAP) of Government of Andhra Pradesh Sakhi Samiti promoted by Professional Assistance for Rajasthan Development Action (PRADAN) Professional Assistance for Development Action (PRADAN) Jharkand Chitradurga Gramin Bank (CGB) Karnataka People's Action for National Integration (PANI) Uttar Pradesh Part II Oriental Bank of Commerce (OBC) Uttaranchal Saravodaya Nanofinance Ltd. Tamil Nadu Dhan Foundation Tamil Nadu Microcredit Foundation of India (MFI) Tamil Nadu The study reported in Part II was done by strong programs, rather than weak ones, are most Robert Peck Christen and Gautam Ivatury for a relevant to assessing the potential of the SHG leading commercial bank in India. This study pro- movement. poses a methodology for designing SHG programs The reason for this approach can be illustrated to ensure their sustainability. In essence, this by looking at the evolution of conventional micro- approach asks (1) whether and how SHG pro- finance institutions. Most of the world's microfi- grams provide essential support services, and (2) nance institutions are probably unsustainable, for whether those services, and the SHGs they sup- reasons that include poor loan collection and inter- port, can be self-financing, without requiring con- est revenue that cannot cover operating costs. But tinuing infusion of external subsidies. The study not surprisingly, sustainable microfinance institu- demonstrates the methodology by analyzing the tions have grown much faster than unsustainable operational structure and financial performance of ones, and they now dominate the field. For four leading SHG programs. This study was not instance, of all the world's microcredit clients who originally intended for publication, but in CGAP's were reached by nongovernment microfinance view there continues to be a need to improve the institutions in 2005, about two-thirds of them got sustainability of SHG programs, and the method- their services from institutions that were already ology proposed is likely to interest a wider audi- financially sustainable.2 One strong microfinance ence. institution or SHG program that thrives and grows The nine SHG programs that were analyzed in to massive size is more important than a dozen both studies (see Table 1) are not a representative that remain small or perish. sample of Indian SHG programs. Knowledgeable Some of the more interesting findings that observers indicate that many SHG programs are emerge from these nine programs are listed below. weak and unsustainable. The nine programs stud- It is reasonable to suppose that they may have ied here were chosen because most of them were some broader relevance as well. reputedly stronger than average. The intent was to 2Based on data from the Microfinance Information eXchange (MIX) get a picture of the potential of the SHG model and the MicroCredit Summit, provided by Adrian Gonzalez, MIX re- when it is well executed. The authors believe that searcher for CGAP projects. 2 OP 12-I.qxd 8/6/07 6:55 PM Page 3 I Sustainability. Many well-executed SHG Most Indian SHGs are externally funded by banks programs are achieving financial sustainabil- or by promoter organizations who borrow from ity, even when all promotion and support banks. Indian commercial banks, most of which are costs are included, though this cannot be government owned, began lending to SHGs generalized for the entire SHG movement. because of government-imposed, priority-sector I Support services. Well-run SHPIs seem to lending quotas. Elsewhere in the world, it is hard be able to provide an adequate package of to find success stories among community-managed external support services to establish SHGs loan funds that are externally funded. Rather, and then keep them functional, and the almost all of the successful programs are savings SHGs seem to be willing and able to pay the based.3 cost of that support. What accounts for the success of externally funded community finance in India? One is I Loan collection. In well-run SHG pro- grams, there is usually very little default on tempted to speculate that the difference might be repayment of external loans to the SHGs that, in India, external funding comes in the form from banks or promoters. In addition, SHGs of a direct or indirect loan from a commercial bank in some of the programs studied are quite that is serious about getting its money back. On successful in eventual collection of their the other hand, Indian banks have experienced internal loans even though those loans had high default rates on other priority-sector lending been subject to unusually high levels of late in the past. payment. If the banking relationship is important to the success of the Indian SHG model, this may have I Cost levels. Generally, the SHG programs implications for the replicability of the model in studied here compare favorably with other other countries. It remains to be seen how many microfinance models in terms of administra- other governments will impose such directed tive costs. However, this does not factor in credit requirements, and whether commercial the time members have to spend at meetings banks will be willing to lend to community groups and the risks group members are subject to. without them. I Savings. Though all SHGs require periodic Note that the performance information on pro- fixed amounts of savings from members, the grams discussed in this paper is fairly old. We do model is credit driven. Clients of the nine not view this as a major concern, because the point programs studied in this paper joined an of the paper is not to give a current overview of the SHG mainly to get loans, not to save. For Indian SHG movement, but rather to investigate some programs, there is little voluntary sav- the question of whether and how SHG programs ing beyond the minimums required to qual- can be run sustainably. ify for loans. I Reaching the poor. Most of the programs in 3A recent CGAP study (Murray and Rosenberg 2006) found that mem- the studies reach very poor and marginalized bers of community-managed loan funds were much less disciplined in clients. handling outside money than they were in handling their own money ac- cumulated through savings. I Elite capture. The researchers found little evidence of elite capture, which has been a problem with some other forms of commu- nity-based and member-managed finance. 3 OP 12-I.qxd 8/6/07 6:55 PM Page 4 Part I. Do India's Self-help Groups programs; some even participate in local govern- Provide Value for Money? ment elections. Because Indian law requires that larger organizations be formally registered, SHGs What Is an SHG? have no more than 20 members. Most groups have SHGs are autonomous collectives that deliver 10 to 20 members. Members typically are poor small loans to their members. SHGs are run by women who have similar socioeconomic back- their members, who choose their own leaders and grounds and who are from the same locality. bank account signatories. They decide on compul- Members join to get loans and other services, such sory savings amounts, interest rates, lending as livelihood support--where the group offers norms, and distribution of surpluses. Accounts are skills training, promotes new markets, provides maintained by literate members, hired bookkeep- access to assets in addition to credit, and provides ers, or staff of promoting institutions. access to irrigation. Most SHGs are part of a federation that helps Most SHGs are formed with assistance from a them with governance and financial monitoring promoting institution--such as the Small and promotes new groups where needed. Industries Development Bank of India (SIDBI), Federations and other promoters also link individ- Rashtriya Mahila Kosh (RMK), Housing and ual groups to external financial and social resources. Urban Development Cooperation (HUDCO), SHGs collect periodic savings and make loans Housing Development Finance Corporation to their members. Savings are usually compulsory, (HDFC), and Friends of Women's World Banking with the amount and frequency of savings collec- (FWWB). Most of them receive funding from tions decided by the group. Loans are funded by commercial banks (almost all government owned) savings, revenues (interest, fees, penalties), and or their promoting institutions. loans from banks and other external sources. The SHG-Bank Linkage Program, launched in Initial loans to members are usually funded by 1992 by the National Bank for Agriculture and savings. Such loans are typically small and used for Rural Development (NABARD), stimulated the consumption or to repay existing debt borrowed at development of many SHGs nationwide. NABARD higher interest rates from other sources. These loans is a government-owned apex refinance (wholesale range from 100 to 2,000 rupees (US$2.5­US$45), loan) institution with a combination of promo- with a maximum loan period of six months and a tional, supervisory, and refinance functions for bullet repayment scheme (that is, all principal is retail institutions--rural branches of commercial repaid in a single installment at the end of the banks, regional rural banks, and cooperative banks. loan, with interest paid monthly or with the prin- SHGs were initially promoted mainly by NGOs cipal repayment). Over time, SHGs mobilize more such as MYRADA and PRADAN. Since the middle savings, retain earnings, and often borrow external of the 1990s, when the model was scaled up, pro- funds, enabling larger loans for consumption and motional work was largely done by specialized gov- business purposes. These loans range from 1,000 to ernment agencies, such as the District Poverty 20,000 rupees (US$23­US$450) and are repaid in Initiatives or the Velugu project in Andhra Pradesh monthly installments over one to three years. and the Kudumbshree project in Kerala, the In addition to financial services, some SHGs Women's Development Corporations in the states provide health care services (such as polio vaccina- of Tamil Nadu and Maharashtra, the Women and tions and family planning information), social Child Development departments, and the District empowerment activities (such as literacy training), Rural Development Agencies (DRDAs) in most food-for-work opportunities, and mid-day meal others states. Although the more specialized agen- 4 OP 12-I.qxd 8/6/07 6:55 PM Page 5 Table I-1. Models of Support for Bank-Linked Indian SHGs Type of support Model 1 Model 2 Model 3 Promotion NGO or government agency NGO or government agency Bank promotes the group promotes the group promotes the group Financing Bank lends directly to NGO or government agency Bank lends directly to the group the group obtains funds from bank and lends to the group cies have, by and large, established SHGs of fair these two models because the costs of group for- quality, the departments and the DRDAs pursued mation and support are borne by SHPIs. a numbers approach that produced SHGs of indif- In the third model, a bank acts as an SHPI-- ferent quality. In Andhra Pradesh, for example, the forming SHGs, training them, and then lending to government uses gas connections and revolving them. As of March 2005, 21 percent of bank- loan funds as incentives to encourage women to linked SHGs had been promoted by banks. form SHGs. Members joined the groups to cap- NABARD's program is designed to integrate ture these benefits. But after achieving their short- informal savings and credit groups with the main- term goals, most of these groups stopped func- stream banking system. Under the program, tioning. SHGs in which members have not been NABARD refinances bank loans to SHGs--that is, "bribed" by quick-and-easy subsidies have proven it provides financing to banks at a below-market more durable. The incentives for such groups are interest rate (currently 6%), though banks continue more conducive to member participation and to carry the risk for their loans. group solidarity, both of which are crucial to a By March 2006, NABARD's program had lent group's sustainability. 114 billion rupees (US$2.8 billion to 545 banks through 44,362 branches, half of which was still Models Linked to Banks outstanding). These banks in turn extended loans Most--though not all--Indian SHGs eventually to 2.23 million SHGs that served an estimated 33 get loans from commercial banks. Three models million poor women over 13 years.4 have emerged for SHG­bank linkages. In the Banks establish links with groups that have dominant model, used by NABARD's SHG­Bank maintained regular savings relationships with them, Linkage Program, a bank lends directly to a group usually after six months. Banks then lend to the after evaluating the group's operations, maturity, group without collateral, relying on self-monitoring and capacity to absorb credit. The groups lend the and group peer pressure for repayment. Banks typ- proceeds to their members. An SHPI--an NGO ically initiate lending to SHGs with a loans-to- or government agency--remains involved with the savings ratio of 1:1 or 2:1, then gradually increase group, but is not part of its funding chain. As of this ratio to 4:1. March 2005, 72 percent of bank-linked SHGs had SHGs normally borrow from banks at an annual been financed through this model. interest rate of 8­12 percent and lend to their In the second model, the promoting institution members at 24 percent, although in some cases it also plays a funding role. A bank lends to the pro- has come down to 18 percent. Groups retain the moter, which then on-lends the funds to its SHGs. interest rate differential as earnings. Over several As of March 2005, 7 percent of bank-linked SHGs had been financed using this approach. Banks like 4 http://nabard.org/pdf/stmt1.pdf 5 OP 12-I.qxd 8/6/07 6:55 PM Page 6 years, this revenue typically surpasses member sav- (formerly known as the Velugu project) and ings as a source of funds. More than 95 percent of funded by the World Bank. the bank loans to SHGs backed by the NABARD I Sakhi Samiti, an SHG federation in Rajasthan, program are repaid.5 The NABARD program is initially promoted by Professional Assistance nationwide, but it is especially active in the south- for Development Action (PRADAN), an ern states of Andhra, Tamil Nadu, and Karnataka, NGO. In 1999, the federation assumed which are home to more than half of all bank- responsibility for SHG promotion. linked SHGs. I A PRADAN program in the Lohardaga dis- Most of the banks that lend to SHGs in India trict of Jharkhand, formed in 1992, that began are government owned, though a few private promoting SHGs for savings and credit in banks participate as well. What motivates the 1996. banks to lend to such unconventional borrowers? I Chitradurga Gramin Bank (CGB), a rural The dominant factor is government-mandated bank in Karnataka selected by NABARD in lending targets of 40 percent of total bank credit 1992 as one of nine regional rural banks to to borrowers from priority sectors, including agri- promote SHGs. culture, microfinance, small industry, housing, and I People's Action for National Integration education. Of this, 10 percent must be to the (PANI), a leading NGO in Uttar Pradesh that "economically weaker sections." These targets are espouses a Gandhian ideology of integrated monitored by bank senior managers and govern- human development. ment officials, who are answerable to Members of Parliament. But some banks are engaging in SHG Table I-2 summarizes the inception, goals, and and other microfinance operations because they services of these five groups. Table I-3 provides a think this market may be profitable. It is an open statistical overview of the five programs and the question whether banks would have much interest 150 SHGs covered by the study. in doing business with SHGs if there were no gov- ernmental lending targets. Methodology A two-stage sampling process was used to select Study Findings SHGs for detailed review. First, villages were cho- The analysis in Part I is based on data from five sen through probability proportionate to size sam- well-established institutions that represent the pling, which weighted samples based on the popu- main models for promoting Indian SHGs, cover- lations of different villages. This approach ensured ing diverse regions: appropriate representation of small, medium-size, and large villages. Next, 30 SHGs were randomly selected from each institution's list of those that I Panagal Mandal Mahila Samakhya (PMSS), an SHG federation in Andhra Pradesh, was pro- had been active for at least three years in the sam- moted as part of a United Nations ple villages. Development Programme (UNDP) South Data for the 150 SHGs in the sample were col- Asia Poverty Alleviation Programme (SAPAP) lected between May 2003 and April 2004 and cov- initiative that ended in 2000. PMSS is cur- ered all the members of the sample groups. rently supported by Indira Kranti Pathakam Analysis focused on the source and volume of SHG funds, loan portfolio quality, profitability, operat- ing costs, efficiency, growth, outreach, and per- 5 Here and elsewhere in Part I, "repayment rate" is calculated as ceived life changes among members resulting from (amount repaid minus prepayments of loans) / repayment due during the period. 6 OP 12-I.qxd 8/6/07 Table I-2. Overview of Five Selected Institutions That Promote SHGs in India 6:55 Legal form of SHG Institution promoter When and how SHG promotion began Objective Services to SHG members/clients PM PMMS Panagal Mandal,a Mutually aided 1995: Five-year program to organize poor Empower poor women to overcome social, · Training of SHG members and bookkeepers Mahabubnagar Society, cooperative society women for social change and livelihood economic, cultural, and psychological barriers · Linkages to banks and other external funders Page Andhra Pradesh Three tiers: SHGs, village generation began as part of the UNDP-SAPAP through self-managed institutions of the poor. · Conflict resolution clusters, and mandal project. UNDP initially contracted NGO partners · Market linkages 7 federation to form SHGs; later the NGOs withdrew and · Reproductive and child health services SHGs were formed directly by project staff. · Social initiativesb Sakhi Samiti Public society 1987: SHGs formed during a severe drought to Improve women's access to government · Bookkeeping (Kishangarh Bas block, Two tiers: SHGs and buy fodder on credit. PRADAN promoted the schemes and enable them to meet · Internal audits Rajasthan) federation initial SHGs, but the federation handles emergency financial and/or credit needs. · Linkages to banks and other external funders promotion now. · Conflict resolution · Reproductive and child health services · Social services PRADAN Unregistered 1996: PRADAN formed SHGs to ensure Promote and strengthen women's SHGs as a · Internal audits (Lohardaga District, Two tiers: SHGs and sustainability of irrigation activities in Lohardaga viable financial intermediary, create compact · Computerized bookkeeping Jharkand)c federation district. PRADAN has since withdrawn; groups clusters for outreach, and follow up by · Linkages to banks and other external funders are now self-managed by members. consolidating enterprise-based livelihood · Facilitation of SHG clusters activities. CGB Regional rural bank; no 1992: Five SHGs promoted by MYRADA linked Inculcate savings habits among the poor, · Training of SHG members (Chitradurga District, federation as part of the SHG-Bank Linkage Program. facilitate access to bank credit through an · Monitoring of SHG operations Karnataka) Groups are self-managed; CGB provides credit effective credit delivery system, and build linkages and mentoring. mutual trust between CGB and the rural poor. PANI Unregistered 1995: Formed SHGs for free bonded laborers to Empower the target community so that they · Training of SHG members (Faizabad District, Uttar Two tiers: SHGs and help liquidate their debt. may help themselves. · Bookkeeping Pradesh) federation · Internal audits · Linkages to banks and other external funders Notes: aA mandal is an administrative unit below the district consisting of 40­50 villages. In Andhra Pradesh, development blocks are divided into mandals, while the rest of the country used the taluk as the administrative block. bSocial initiatives include promoting the rights of lower caste people, advocating against child marriages, promoting schooling for female child laborers, and promoting livelihoods for disabled and desti- tute women. cPRADAN promoted its first SHGs in 1987 in the Kishangarh Bas block of Alwar district, Rajasthan. They were later federated in 1997 as Sakhi Samiti. 7 OP 12-I.qxd 8/6/07 6:55 PM Page 8 Table I-3. Statistical Overview of 150 Indian SHGs, by Promoting Institution Sakhi Indicator PMMS Samiti PRADAN CGB PANI Average Average age of SHGs (years) 6 6 5 4 5 5 Total number of members in sample SHGs 429 389 533 475 378 441 Average number of members per SHG 14 13 18 16 13 15 SHG members with outstanding loans (percent) 98 88 76 89 90 88 Average SHG outstanding loan portfolio (US$) 2,301 1,846 509 1,845 441 1,388 Average SHG outstanding bank loan (US$) 162a 903 84 1,177 372 740 Average SHG savings (US$) 580 881 573 1,101 166 660 Average savings per member (US$)b 41 68 32 69 13 45 Number of SHG promotion staff 28c 5 10 NA 79 31 Note: The survey covered 30 groups for each institution. Dates for financial information are for the year ending March 2003 for PMSS, Sakhi Samiti, and PRADAN and for the year ending March 2004 for CGB and PANI. Loan analysis was based on annual information: PMSS, 31 March 2003; Sakhi Samiti, 31 October 2003; PRADAN, 30 November 2003; CGB, 28 February 2004; and PANI, 31 March 2004. For all five SHG programs, financial information for the periods ending March 2001, March 2002, March 2003, and March 2004 was converted at Rs. 46.577 to US$1, Rs. 48.733 to US$1, Rs. 47.65 to US$1, and Rs. 44.125 to US$1, respectively. aAll PMSS SHGs had borrowed from their federation. The federation borrows from banks and then on-lends the proceeds, together with the federation's own funds, to member SHGs. bEstimated as average savings per SHG/average members per SHG. cStaff consists of 5 PMSS staff and 23 village organization staff. SHG participation. Members were interviewed for bank, such as CGB, which serves clients who live both personal information and information about near its offices. To determine how effectively the their SHG. Financial and meeting records were SHG model reaches these populations, the study reviewed, verified, and entered into spreadsheets analyzed the locations of the 150 SHGs in the to create detailed financial statements, loan port- sample and the economic and demographic pro- folio reports, and attendance records for each files of their members (see Table I-4). Most of the group. Data were cross-checked using loan and groups' members live far from paved roads, bank saving ledgers, member passbooks, and minutes of branches, and health centers, although there is meetings, and verified with members during indi- variation among the programs. CGB, for example, vidual and group interviews. The CGAP research is based in a district capital, and its SHGs are closer team assessed the five institutions, focusing on to paved roads and health centers. PRADAN, on their costs of forming SHGs, provision of ongoing the other hand, aims to work with more disadvan- support, and links to external sources of credit. taged groups and so operates in one of India's least developed states (Jharkhand), where settlements Outreach are more remote. Most Indian SHG programs reach out to vulnera- Accordingly, of the five programs studied, ble and marginalized people who own little or no PRADAN had the deepest outreach: almost all land, are predominantly illiterate, and who lack SHG members are tribal people or members of access to formal sources of financing. The depth of scheduled castes. PRADAN's clients are also the programs' outreach to such populations depends poorest: 85 percent had no homestead land or only on program design and on their promoters' vision, marginal nonagricultural landholdings, and almost leadership, and commitment. An NGO, such as 90 percent live in thatched huts or are squatters. PRADAN, whose mission is to help vulnerable CGB's clients were somewhat better off, with 70 groups, has deeper outreach than a regional rural percent living in concrete or brick houses and 40 8 OP 12-I.qxd 8/6/07 6:55 PM Page 9 Table I-4. Depth of Outreach of SHGs, by Promoting Institution Sakhi Indicator PMMS Samiti PRADAN CGB PANI Average Distance from paved road or highway (miles) 3.3 1.7 10.6 2.4 8.2 5.2 Members from scheduled castes and tribesa (percent) 21.9 58.4 96.6 35.2 68.0 56.0 Clients who are marginal farmers or have no homestead or agricultural land (percent) 58.5 83.3 85.6 60.6 96.6 76.9 Members who are illiterate or can only sign name (percent) 88.0 89.0 86.0 14.0 93.0 74.0 Members without a house or only a thatch hut (percent) 11.2 28.0 87.6 29.7 35.2 38.3 aScheduled castes and tribes are communities accorded special status by the Constitution of India. These communities were considered "outcastes" and were excluded from the four-caste system that was the social superstructure of Hindu society in India for thousands of years. These castes and tribes were relegated to the most menial labor, with no possibility of upward mobility, and degenerated into the country's most economically and socially backward communities. percent owning agricultural land. In addition, 86 Leadership percent of CGB clients are literate, compared with In about two-thirds of the SHGs, elections for just 7­14 percent in other programs. Annex tables leadership positions had been held at least once IA-1 through IA-6 provide further data for each since the group was created. In some cases, the program. same leaders were reelected, reportedly to avoid problems in changing bank account signatories. Institutional performance Decision-making The institutional sustainability of SHGs depends Most SHGs reported that decisions were made at on their management, systems (including external meetings attended by all (58%) or most members support), and membership. Overall, the five pro- (31%). But a quarter of the groups promoted by grams studied had good survival rates among their CGB and PANI reported that decisions were made SHGs, relatively low member dropout levels, and by a few members. consistent attendance and member participation. Membership and group stability In addition, field staff of the promoting institutions Over an average of five years before the study's rated most groups in the sample as above average to start, 15 percent of members dropped out of the exceptional in their management, systems, and 150 sample SHGs; 9 percent joined as new mem- membership, indicating that these groups were bers.6 The main reasons for dropping out were probably capable of managing themselves. death, marriage, or migration. Inability to meet Record-keeping. Because of low literacy rates savings requirements or attend weekly meetings among members, SHG records are maintained by was also cited in a few cases. NGO staff, a literate member of the group or its SHGs tend to last longer if their promoters pro- federation, or--most commonly--a literate person vide good organizational support and social mobi- in the village who is paid for the service. Records lization. For example, only 2 percent of SHGs pro- generally include data on attendance, savings, moted by PANI--which achieved deep outreach to loans, and member passbooks. The quality of records was rated good or acceptable in more than 6 In most cases, if someone wants to join an SHG, she has to contribute half of the sample SHGs and average in one-third to the group an amount equal to the accumulated savings per member of the sample. Rating criteria are defined in annex at that time. This deters prospective members from joining existing groups, and they often choose to motivate other people and start a new SHG. Table IA-7. 9 OP 12-I.qxd 8/6/07 6:55 PM Page 10 Table I-5. Age Distribution of SHGs, by Promoting Institution Program 3­5 years 5­7 years More than 7 years Average PMMS 11 6 13 6.0 Sakhi Samiti 14 8 8 5.9 PRADAN 26 4 0 4.5 CGB 23 7 0 4.2 PANI 20 8 2 4.5 Total 94 33 23 5.0 vulnerable and marginalized groups--disbanded groups promoted by the three other institutions. between April 2001 and March 2004. When the Members were more likely to have family members government had actively promoted SHGs (typi- make payments when meetings were held in the cally by offering immediate incentives for joining), morning, which is a more convenient time for higher numbers of recently launched groups disin- hired SHG bookkeepers than for members who tegrated, reflecting the absence of adequate time need to be at work. When group members or leaders for group organization and social mobilization. managed records, most meetings were held in the Table I-5 shows the ages of the 150 SHGs ana- evening. (Annex Table IA-8 provides further details.) lyzed for this study. Perceptions of promoting institutions The low rates of dropout and turnover convey a The research team asked field staff from the five strong message about the groups' utility. Members promoting institutions to rank the cohesion and would not continue to attend and invest in their overall functioning of SHGs based on their links to groups unless they found them helpful. external funding (such as from a bank or MFI), Meeting attendance repayment rates on internal loans, and adherence SHG meetings are held regularly, with 82­100 to group norms. Although all the programs rated percent of groups reporting regular meetings. But highly on cohesion, ratings were less impressive for meetings seem to become less frequent as groups overall functioning (Table I-6). (Annex Table IA-9 age. Attendance was consistent across the five provides rating criteria.) programs, with about 80 percent of members attending. Financial performance In a quarter of the SHGs, members sometimes Savings sent family members or others to make payments Savings in the 150 SHGs consisted entirely of the on their behalf. This practice was more common compulsory deposits that are a condition of mem- among the groups promoted by Sakhi Samiti and bership and loan access. The absence of large vol- PRADAN, which have poorer members than the untary savings balances probably indicates that Table I-6. Promoting Institutions' Ratings of SHGs (percentage of groups considered above average or exceptional) Sakhi Area PMMS Samiti PRADAN CGB PANI Average Group cohesion 77 71 80 90 84 80 Overall functioning 70 60 47 64 60 60 10 OP 12-I.qxd 8/6/07 6:55 PM Page 11 Table I-7. SHG Members with Outstanding Loans, by Promoting Institution (percent) Sakhi PMMS Samiti PRADAN CGB PANI Average 98 88 76 89 90 88 most members do not consider groups useful for Lohardarga, Jharkhand, where PRADAN works, storing surplus cash and use other savings mecha- SHG members reported that moneylenders went nisms. In addition, the complexity of managing out of business because of reduced demand for cash flows may limit SHGs from offering such their services. Overall, 14 percent of members of services to their members. the sample SHGs supported by PRADAN still bor- rowed from moneylenders. The average loan per Loans to members member was 6,750 rupees, where SHGs had access The average outstanding loan portfolio (loans to to outside capital from federations or banks, and members) was US$1,388 per SHG. Almost all the 1,450 rupees where there was no such support. groups surveyed charged members a monthly interest rate of 2 percent for loans.7 Only two of Loans to members accounted for 68­98 percent of the SHGs' assets. Although data on portfolio at the 150 groups charged 3 percent a month-- risk can be artificially improved by rescheduling though in the past (generally at the start of the loans, none of the SHGs in the sample did so. microfinance program), most SHGs in India None of the SHGs, however, maintained loan loss charged that amount. Loan terms, typically six reserves. Although most SHGs nominally required months to three years, with monthly repayments, monthly repayments, in practice, members repaid were not related to loan size. their loans flexibly based on available cash flows, At the time of the survey, 88 percent of SHG which did not necessarily correspond to monthly members had outstanding loans (Table I-7). loan installment schedules. Such repayment behav- About two-thirds of the loans (68%) were report- ior was especially common among members who edly for business use: agriculture (32%), animal depend on seasonal income, such as from agricul- husbandry (23%), and microenterprise (13%). The other third was reported as consumption loans.8 ture and animal husbandry. SHGs in the PMMS program had the best port- (Annex Table IA-10 provides further details.) folio quality (lowest portfolio at risk), possibly Loans were not concentrated in the hands of a few because of the program's more individualized loan members. For example, SHG leaders represented 16 terms (see Table I-8). Members of all five pro- percent of group members and 21 percent of loans grams said they consider SHG loans to be from outstanding. family and friends; as a result, these loans are Among the sample SHGs, 23 percent of mem- treated with less discipline than loans from banks bers still used moneylenders for credit. But among to SHGs. But members also reported that most SHGs promoted by NGOs, only 2 percent of SHG loans to members are eventually repaid, with members borrowed from moneylenders. In low eventual default rates. 7The only exception was for groups that receive funds from Rashtriya Because the SHGs in the sample did not provi- Mahila Kosh, a government program that limits the nominal annual in- sion for loan losses, the analysis adjusted their terest rate on member loans to 12 percent. financial statements to include loan loss allowances 8Distinctions between business and consumption microloans are inher- ently difficult to track. Other microcredit studies have found that actual based on rating standards developed by APMAS, loan use often differed from reported loan use. an Indian agency that specializes in assessment and 11 OP 12-I.qxd 8/6/07 6:55 PM Page 12 Table I-8. Average Outstanding Loan Portfolio and Portfolio at Risk for SHG Loans to Members, by Promoting Institution Sakhi Indicator PMMS Samiti PRADAN CGB PANI Average Average outstanding loan portfolio (US$) 2,301 1,846 509 1,845 441 1,388 Portfolio at risk > 30 days (percent) 8 18 23 53 25 25 Portfolio at risk > 90 days (percent) 7 18 21 49 25 24 Portfolio at risk > 365 days (percent) 1 13 7 19 16 11 Note: Portfolio at risk is calculated as the total outstanding balance of loans with any payments overdue by more than x days divided by the total outstanding loan portfolio. capacity building of SHGs and their federations (Table I-9).9 APMAS derived these standards from Table I-9. Adjustments Made to SHGs' Financial Statements for Loan Losses an empirical analysis of SHGs' success in eventu- ally collecting overdue loans. This provisioning Number of Loan loss schedule is much less stringent than is normal for days of overdue allowance principal repayment (percent) microfinance institutions or banks--where, for 1­30 0.0 instance, loans more than a year overdue are usu- 31­60 0.0 ally provisioned 100 percent or written off. This is 61­90 2.5 91­180 5.0 because APMAS has found that SHGs eventually 181­365 12.5 collect a substantial portion of late loans. > 365 50.0 The levels of loan delinquency shown in Table Note: When adjusting financial statements, loan loss allowances were spread over two financial years, on the I-9 would be disastrous for most microcredit assumption that the delinquency had accumulated over that providers.10 Yet it appears that SHGs are surviving period or longer. despite this. This has to do with the fact that a sig- nificant part of the SHG loans were used for crop revenues from interest rate spreads, fees, and cultivation and livestock rearing, neither of which penalties for late loans. SHGs in the CGB program offer a monthly cashflow. Yet loan installments had the worst repayment record on member loans were often fixed at monthly intervals, often out of and, not surprisingly, the worst repayment rates on inexperience and sometimes out of a desire to keep external (bank) loans. a discipline of "repaying something in each meet- ing." Thus the high level of late repayments in Loans from external sources SHGs did not always translate into defaults. Three-quarters of the sample SHGs had borrowed Nonetheless, this is an area to monitor closely for from a bank or federation. In the case of PMMS, SHG programs, as it is for any lending operation. member SHGs borrowed from federations; SHGs Despite late repayments from their members, under the other programs borrowed directly from SHGs in the sample were generally able to repay banks. Outstanding loan balances from all external their loans to banks using member savings and sources among the five programs ranged from US$84 to US$1,177 per SHG, with an average 9 APMAS, a technical organization based in Hyderabad, Andhra loan balance of US$739 (Table I-10). Some of the Pradesh, has developed a rating methodology for SHG federations and programs achieved high repayment of their bank rated more than 200 of them. 10In most other forms of microcredit, loan collection tends to spin out loans, while others did not. Portfolio at risk at 30 of control if the portfolio at risk more than 30 days late stays above 10 and 365 days was excellent in three programs but percent for very long. poor in the other two. 12 OP 12-I.qxd 8/6/07 6:55 PM Page 13 Table I-10. Average Outstanding Loan Portfolio and Portfolio at Risk of Loans from External Sources to SHGs, by Promoting Institution Sakhi Indicator PMMS Samiti PRADAN CGB PANI Average Average outstanding loan portfolio (US$) 1,162 903 84 1,177 372 740 Portfolio at risk > 30 days (percent) 0 14 0 32 0 9 Portfolio at risk > 90 days (percent) 0 14 0 26 0 8 Portfolio at risk > 365 days (percent) 0 9 0 13 0 4 Costs of group promotion women for empowerment, social change, and Costs of promoting SHGs include the costs of livelihood generation. The program sets up three launching, providing training, and monitoring tiers (SHG, village cluster, and federation) and them for about three years, after which groups can provides an initial US$105 per SHG and per vil- usually function with considerably less--and less lage cluster to defray the costs of launching the costly--external support. Promoters' costs were group or cluster. The program also gives each group analyzed based on total spending over three years and cluster US$210 to meet ongoing expenses. divided by the number of SHGs promoted or sup- CGB had the cheapest promotional model. This ported during that time. The average cost was program focuses on literate, less poor women who US$259 per SHG, although costs varied widely live close to bank branches. Its promotion costs are among the five programs. The highest costs were mainly limited to the launch of SHGs, plus initial for multipurpose programs that promote empow- orientation and training, with the remaining costs erment of members, such as PMMS (US$443 per covering CGB's administration. CGB SHG mem- SHG) and PRADAN (US$361).11 The average bers have the highest socioeconomic backgrounds cost of the CGB program, which formed SHGs of all sample groups, which means that they had solely for onlending, was only US$50 per group. the least proportion of poor members. They had Several factors led to the higher costs for the weakest performance for on-time loan repay- PMMS, which aims to organize poor people and ment. CGB's groups were the least profitable (see Table I-12), at least before promotional costs are 11Costs for the PRADAN Lohardaga program included research and development for a pilot program. considered. Table I-11. Costs of Promoting SHGs, by Promoting Institution (US$) Sakhi Indicator PMMS Samiti PRADAN CGB PANI Average Number of groups promoted 300 135 123 360 785 341 Total cost of promotion and support per group 443 232 361 50 211 259 Average cost per group launch 201 139 93 30 111 115 Average cost of training and monitoring per group 241 94 268 20 100 145 Average promotion and support cost per group member 32 18 20 3 16 18 Note: Data are for costs incurred over the previous three years, on average. For additional details, see annex Table IA-11. 13 OP 12-I.qxd 8/6/07 6:55 PM Page 14 Table I-12. Financial Sustainability (Profitability) of SHGs, by Promoting Institution (percent) Sakhi Indicator PMMS Samiti PRADAN CGB PANI Average Return on assets without adjustment for loan loss provisions or SHPI costs 9 15 18 7 11 12 Return on assets after adjustment for loan loss provisions 9 12 16 1 7 9 Return on assets after adjustments for loan loss and SHPI costs 2 7 ­1 1 ­9 0 Note: For additional financial information, see Annex I, Table IA-12. Several SHG promoters believe that the higher based on solidarity groups or village banks, and cost, NGO-led model generates greater social these costs should not be excluded when looking benefits in the form of member empowerment and at SHGs. well being. But this study's results do not consis- Thus the third stage includes the SHPIs' costs tently find that higher promotional costs lead to and amortizes them over three years. Once these higher portfolio quality, deeper outreach, higher costs are included, the profitability picture is SHG ratings, or greater profitability after adjusting mixed. The return on assets at this final stage for loan provisions. The results are more nuanced ranges from very poor (a 9% loss for PANI) to very and depend on the promoting institution. strong (a 7% profit for Sakhi Samiti). By way of comparison, returns on assets in most commercial Financial sustainability (profitability) banks are 1­2%.12 The study measured profitability in three stages. The first-stage return on assets was calculated Conclusion from SHG records with no adjustment. By this SHGs in India reach almost 33 million households measure, the SHGs in all five programs were and provide loans, empowerment, and social serv- highly profitable, with a return averaging 13 per- ices in addition to limited, largely compulsory, sav- cent. But this result cannot be considered a reli- ings mechanisms. It is true that many SHGs and able indicator of financial sustainability because it SHG programs do not perform well. But the same does not take into account the fact that there will has been true historically of most other microfi- likely be losses when uncollectible loans are finally nance institutions. If the purpose is to determine written off. whether the Indian SHG model is a sustainable Thus, the second-stage return on assets includes one, it is more relevant to look at the programs provisions (expenses) for estimated loan losses, that are performing well than those that are not. spread over two years. This stage of measurement One can expect that over time, the weak SHG pro- reflects the SHGs' ability to continue operating into the future. At this level, all the SHG models were still profitable, averaging a healthy 9 percent 12Financial analyses of microcredit programs often include a "subsidized- return. But this stage of measurement ignores the cost-of-funds adjustment" if the program receives loans at rates signifi- cantly lower than market rates. There has been no adjustment for cost fact that some part of the costs of each SHG is of funds in this analysis because the rate SHGs pay banks (10­12%) is paid by external actors: their SHPIs. These are real higher than the 90-day deposit rate in the banking system (5.0­6.5% a costs, and SHGs cannot be formed without them. year during 2002­04) and close to the average bank lending rate (11%). Further, given India's long history of priority-sector lending require- The costs of the promoter would always be ments, SHGs will probably continue to have access to large quantities of included in an analysis of a microfinance program loans on similar terms in the future. 14 OP 12-I.qxd 8/6/07 6:55 PM Page 15 grams will stagnate or close, and most outreach and monitoring more difficult, but the SHGs pro- will be dominated by well-managed programs. moted by PMMS have managed this challenge. This is what has happened among conventional Members do not use SHGs as savings vehicles. microfinance institutions over the past three Another concern is that SHGs mobilize only mod- decades. est amounts of member savings, mainly through Thus, this study analyzed five better performing compulsory deposits that members make, not SHG programs. The findings paint a largely posi- because they want to save, but only because the tive picture: Based on results to date, the Indian deposits are required to get a loan. Only a limited SHG model can work sustainably in well-managed number of SHGs offer voluntary savings, possibly programs. Compared to many other microfinance because groups or promoting institutions do not approaches early in their development, the SHG want to address the complexities that voluntary model seems to be producing more rapid outreach savings entail, including liquidity management, and (as Part II will show) lower costs. But there more staff or volunteer time to meet member are also concerns. requests for access to their savings, and more record-keeping for SHG managers. Other reasons How effective are Indian SHGs in reaching for low voluntary savings may be that members vulnerable and marginalized groups? have other satisfactory options for savings or may SHGs reach poor and excluded groups. Most mem- not consider savings to be secure or accessible in an bers of the SHGs studied come from poor house- SHG. holds, including marginalized groups; most say that participation has improved their lives. How financially sustainable are Indian SHGs? SHG financial services are not fully matched to In well-managed programs, SHGs can be profitable, member need. Many group members pay their and many are. High income from their loan port- loans late, suggesting that SHG loan terms do not folios and low operating expenses enabled most of match members' needs and cash flows. SHGs the SHGs in the study to be profitable, even after under the PMMS model offer more individualized adjusting for loan loss provisions and the costs of loan terms and experience lower late payments and launching, supporting, and monitoring in their ini- default rates. Many SHPIs are concerned that tial three years. (Part II of this Occasional Paper adapting loan terms would make loan origination addresses the sustainability of another group of Table I-13. Average Operational and Outreach Indicators for SHGs, by Promoting Institution Sakhi Indicator PMMS Samiti PRADAN CGB PANI Average Portfolio at risk > 90 days (percent) 7 18 21 49 25 24 Return on assets after adjustment for loan loss provisions and promotional costs (percent) 2 7 ­1 1 ­9 0 Promotion and support cost per group member (US$) 32 18 20 3 16 18 Promotion and support cost per group (US$) 443 232 361 50 210 259 Group members from scheduled castes and tribes (percent) 22 58 97 35 68 56 Group distance from paved road/highway (miles) 3.3 1.7 10.6 2.4 8.2 5.2 15 OP 12-I.qxd 8/6/07 6:55 PM Page 16 SHG programs, with greater focus on the costs of costs, such as PMMS and PRADAN, have varying ongoing external support.) results. PMMS's SHGs scored higher on overall SHG ratings and have better loan repayment, but Do Indian SHGs provide value for money? PRADAN has deeper outreach and better prof- The sample SHGs reach poor and marginalized itability. SHGs promoted by CGB, the least expen- groups with loans and other services and appear to sive model studied, had weaker outreach, collec- produce social and economic benefits for mem- tion, and profitability net of potential loan losses. bers. It is clear that they can be financially sustain- This shows that "money saved" in the careful pro- able, and most were profitable. Although costs of motion of SHGs may be lost later through loan promotion and support have been subsidized, losses and other problems with the SHGs. SHGs compare favorably with many other microfi- The SHG movement is still relatively young. nance approaches in terms of the subsidy required Final assessment of its effectiveness will have to per client and financial sustainability. The study's wait until more experience is acquired. One impor- financial analysis indicates that most of the SHGs tant question for the future is whether the increas- would be financially sustainable even if they had to ing patronage by politicians and the resulting high pay the costs of external promotion and support. growth rate of SHGs and their bank linkage could How do differences in SHG models lead to high default rates. Another is the extent to affect performance? which SHGs can adapt their lending and savings Is more intensive group formation and support products to provide an appropriate fit with their worth the added cost? No simple answer emerges members' financial preferences. from the study. Programs with higher promotion 16 OP 12-II.qxd 8/17/07 3:29 PM Page 17 Part II. Designing SHG Programs for must fulfill two conditions: first, the program must the Long Term provide groups with an ongoing set of essential support services; second, the program must collect NGOs, government agencies, and banks in India enough revenue to cover the cost of providing choose to promote SHGs for a variety of reasons, these continuing services. not all of which are focused on long-term financial service access. Some of these SHPIs see SHGs pri- Judging Sustainability marily as a vehicle for grassroots social mobiliza- In mid-2005, a leading Indian commercial bank tion. Such promoters sometimes suggest that the asked CGAP to review its SHG lending operation social empowerment goals of their SHG programs and to suggest improvements. In response, CGAP are achieved even if the groups cease operations researchers tried to identify those elements that are after a few credit cycles. Other promoters see essential for long-term success. Our first step was SHGs as part of a permanent system providing vil- to review the literature. In addition, we studied lage-level financial services. four Indian SHG programs reported to be leading SHGs vary widely in approach and quality. Some examples and sought to learn from their experi- are little more than one-time events, often organ- ences. ized by government workers after hours, while oth- A desk review of the SHG literature revealed ers have been carefully built by NGOs as stable that few studies have tackled the problem of sus- providers of various development services. In tainability of SHG programs. The studies that do many programs, SHGs are organized into federa- treat financial performance (summarized in Table tions that provide services to individual groups and II-1) provide only a partial understanding of the build the leadership capacity of SHG members. issue. Some studies focus on the costs of running Some observers question whether many SHGs an SHG program, such as the costs of forming or will offer permanent access to finance. In confer- maintaining the group, but do not look at revenue. ence presentations, proponents of the SHG These studies reveal little about whether SHGs can approach tend to focus on the numbers of clients pay for the costs of forming and maintaining the served rather than on how to maintain the sys- group. Other studies look at whether lending to tem's financial viability. APMAS, a specialized sup- SHGs is profitable for banks, but ignore the costs port institution for SHGs, has trained thousands incurred by NGOs and others in forming and sup- of professionals and analyzed over 300 SHG fed- porting these groups. These studies create an arti- erations. It claims that only a minority of SHGs ficially optimistic impression of the viability of across India are of "good quality," or rated "A" SHG programs. using the APMAS grading scale. Nearly 40 percent Only the studies done by Nair (2005) and Reddy of the SHGs it examined in one district of Andhra and Prakash (2003) directly address the question of Pradesh state in 2002 had "grossly neglected" or sustainability. These studies analyze superstructures nonexistent book-keeping. Groups that lack such (SHG federations) designed to provide support a basic tool of financial service delivery are unlikely services to SHGs, including capacity building, per- to last long. 13 formance monitoring, and helping to access bank How can an SHPI ensure that SHGs continue credit. But neither study examines whether other to operate over the long term? Its SHG program external actors provide services to SHGs alongside the federations. Nair's analysis does not account for the setup costs of SHG federations in examin- 13See "The Study of SHG Movement in Visakhapatnam District." Avail- able on www.apmas.org. ing their sustainability (Christen 2005). 17 OP 12-II.qxd 8/21/07 2:09 AM Page 18 Table II-1 Studies on the Costs and Sustainability of the SHG-Bank Linkage Model Author Objective of analysis Actors examined Findings Srinivasan (1999) Profitability of SHG 1 RRB Cost/SHG/year is US$18a model for banks Gross margin (interest revenue minus costs) is (0.07%) to 2.05% Harper (2002a) SHG promotion costs 20 promoters, 34 SHGs, Cost/SHG unit linkage: under various models 16 banks NGO-promoted (US$25­US$424) Bank-promoted (US$25­US$182) Agent-promoted (US$8­US$87) Government-promoted (US$4­US$145) Individual-promoted (US$69)b Seibel and Profitability of SHG 1 commercial bank, 1 Revenues/average costs is Harishkumar (2002) model for banks RRB, 1 district central 101­165% cooperative bank Return on assets is 1.4­7.5% Tankha (2002) SHG promotion costs 7 SHG programs Cost of promoting SHG over and model various periods from 15 months sustainability to 7 years is US$93­US$517c (no sustainability indicators provided) Sinha (2003) Profitability of SHG 5 RRBs SHG lending is unprofitable model for banks Portfolio yield of 12.5­13.0%, but operating cost of 19.0% Reddy and Prakash Efficiency and 26 federations of SHGs All 3-tier SHG federations (2003) sustainability of SHG unprofitable: federations Operating costs/average portfolio is 10­25% Revenues/costs is 24­98% RRB = rural regional bank aConverted from rupees using rate at December 31, 1998. bConverted from rupees using rate at November 30, 2002. cConverted from rupees using rate at August 31, 2002. To remain viable over the long term, SHGs are necessary for the long-term stability of community like other community-level savings and loan finance models.14 They include the following: groups in that they require an adequate level of external support (Christen 2005). Given that most I Promotion--Groups need help forming and Indian SHGs borrow from commercial banks, one maintaining their structure, especially in man- might argue that they need even more support aging member exits and entrances. than other forms of community-level finance that 14 do not assume external liabilities. Christen (2005) Rotating savings and credit associations (ROSCAs) are informal groups where all members contribute a fixed quota at each meeting, and mem- discusses the types of support that have proven bers take turns receiving the entire amount collected at each meeting. Extremely simple mechanisms like ROSCAs can operate without contin- uing external support, but more complex arrangements seldom can. 18 OP 12-II.qxd 8/17/07 3:29 PM Page 19 I Training--SHG members must be trained in that provides these services must generate enough basic operations if they are to maintain quality income from the SHGs to cover its costs. service, especially in light of limited human We studied four well-regarded SHG programs resources available at the community level. in India to see whether they meet this two-part test Clients also need to be trained to understand and to uncover lessons they might have in helping the products being offered and the proce- SHGs achieve long-term viability. dures they need to follow to access these products and services. Four SHG Promoters I Standardized Products and Norms--Local The four SHG promoters we selected were reputed entities, such as SHGs, are usually better off to be building sustainable SHG programs, based when they can offer standardized products on conversations with government officials, micro- that have been developed by a higher level finance practitioners/consultants, and others. organization that is in a better position to They are the Oriental Bank of Commerce (OBC), develop supporting management information headquartered in New Delhi; Sarvodaya systems, rules, and risk mitigation strategies. Nanofinance Ltd. (Sarvodaya), in Tamil Nadu; I Administration--Sometimes regular operat- Dhan Foundation (Dhan), in Tamil Nadu; and the ing functions, such as book-keeping, process- Microcredit Foundation of India (MFI), in Tamil ing transactions, and serving clients must be Nadu.15 performed by nonmembers. These organizations are not representative of all SHG promoters; three of the four are located in I Oversight/Intervention in Operations-- When corrupt or unhealthy practices, includ- just one of India's 28 states, Tamil Nadu in south- ing capture by SHG leaders, occur, someone ern India. They are not the largest, the most prof- outside the group should be able to intervene itable, or the best-known practitioners of the SHG and help correct the problems. approach. Rather, these organizations were chosen because their programs appeared to provide ade- I Liquidity--SHGs can provide better service if quate support services--as confirmed by strong group members can draw from an external loan collection--and because their data and expe- fund when cash flow needs are higher than rience were available to us. The four included pro- usual, and deposit into that fund when excess grams that were reported to serve large numbers of cash is available. An external fund also may be poor people and/or to operate particularly effi- used to invest excess cash in liquid instru- cient programs. ments. Dhan Foundation has several SHG programs in Tamil Nadu; only the one run by the Kurinji We do not argue that all of the above support serv- Vattara Kalanjiam (KVK) federation of SHGs, in ices are essential for every community finance Madurai, was studied for this paper. OBC is an model. But some robust combination of most of Indian commercial bank with 1148 branches16; the these services seems to be important for long-term stability. When community-level financial service 15 programs fail, the problem usually can be traced to Data from these organizations presented in this paper were assembled through interviews and correspondence with their management and/or inadequate external support structures. review of public sources, including annual reports, the Microfinance In- As noted earlier, long-term SHG viability formation eXchange (MIX), and case studies published in Small Cus- depends on two conditions. First, the individual tomers, Big Market by Sukhwinder Singh Arora and Malcolm Harper (ITDG 2005). SHGs must receive an adequate package of exter- 16As of April 29, 2006, according to the OBC Web site (www.obcindia. nal support services. Second, the SHG program com/knowus/knowus_ourachivements.html). 19 OP 12-II.qxd 8/17/07 3:29 PM Page 20 Table II-2. Key Features of the SHG Programs Studied Unweighted average of 58 MFI SNFL OBC Dhan SHG promoters Features (at 3/31/06) (at 3/31/05) (at 3/31/03) (at 3/31/03) (M-CRIL)a Legal status Section 25 NBFI Scheduled Trust -- (not-for-profit) commercial company bank Region South South North South -- (Tamil Nadu) (Tamil Nadu) (Uttaranchal) (Tamil Nadu) SHG members 517,784 47,282 4,949 164,552 21,057 Total assets of 4.5b 6.9 7,132.8 2.0 0.5 promoter organization (US$ millions) Promoter's return NA 0.32% 1.0% NA ­27.6% on average assets Portfolio at risk 0.33% 0.7% 0.0% 2.94% 27.7% of loans to SHGsc (90 days) (60 days) (90 days)d (30 days)e (60 days) a Calculated from "M-CRIL Microfinance Review 2003 (revised Feb. 2004)," Micro-Credit Ratings International Ltd., Gurgaon India. The numbers in this table do not always match numbers in similar tables from the original report, because some of those tables show weighted calculations and include results from non-Indian SHG programs. b Authors' estimate. c Outstanding balance of all loans that are late by more than a given number of days, divided by total outstanding balance of the whole loan portfolio. This statistic reflects repayment of external lending, not the loans within the individual SHGs. d For Rudrapur program only. e For KVK program only. SHG program studied in this paper operates in their own performance. Thus, they probably are one branch, at Rudrapur in the northern state of more focused on sustainability than the average Uttaranchal. Indian SHG program. Table II-2 compares the four SHG promoters with an average of 58 Indian SHG promoters in Sustainability Test Part 1: Are Essential India--including NGOs, cooperative bodies, and Support Services Provided? nonbank financial intermediaries (NBFIs)--that The first part of our sustainability test was applied were rated by M-CRIL, an Indian rating agency to the four SHG programs in the study by examin- for microfinance institutions, between September ing whether an adequate package of support serv- 1998 and June 2003.17 According to M-CRIL, the ices is provided. In each of the four programs, sup- SHG programs it rated include some of the largest port services are performed by more than one in India. These programs probably sought ratings actor, including the following: either to access external funding or to improve I A support organization, such as an NGO, non- profit corporation, or NBFI, that is usually the 17Sarvodaya Nanofinance Ltd, one of the four promoters analyzed in this study, is also included in the M-CRIL sample of 58 SHG promoters. SHG promoter and that supervises the overall 20 OP 12-II.qxd 8/17/07 3:29 PM Page 21 functioning of the program (used in the MFI, MFI gets one-third of the interest, or 6 percent, as Dhan, and Sarvodaya programs) a "service fee," and ICICI Bank keeps the remain- I An SHG federation, or collective of SHGs, ing 12 percent. usually with its own management and balance Sarvodaya Nanofinance Ltd. is a licensed non- sheet, that forms, trains, or otherwise sup- bank financial institution that acts as a support ports individual SHGs or smaller groups of organization. It establishes system-wide policies SHGs, such as local-level associations (used in and products, borrows at commercial rates from Dhan and Sarvodaya) banks, and on-lends to about 50 SHG federations that then on-lend to SHGs. Sarvodaya's field exec- I An individual agent or facilitator contracted utives support and monitor these SHG federations, by the support organization, SHG promoter, and the federations' field officers handle all group or the SHG itself to handle bookkeeping, cash formation, monitoring, and cash transactions. transactions, and other support services with Saravodaya and the SHG federations cover their or without pay (used in OBC) costs through interest spreads. Saravodaya borrows I One or more bank staff, such as branch man- from commercial banks at 7.5 to 8 percent per agers and loan officers who handle local SHG annum and lends to the federations at 12 percent, business; these usually follow policies leaving it with a spread of 4 to 4.5 percent. The designed by, and reported to, a higher author- federations on-lend to the SHGs at about 22 per- ity in the bank, such as a microfinance project cent, leaving a spread of 10 percent. unit (used in all programs) In the OBC program, the only role of the bank's Rudrapur branch is to service SHGs. The Based on short field visits, interviews with the branch's two officers oversee about 1,000 five- SHG promoters, and the financial results of their member SHGs and perform most support func- programs, it appears that each program is provid- tions. OBC charges SHGs 11 percent per annum ing an acceptable level of all the support services on loans to cover the costs of funds, support serv- identified earlier. This is not to suggest that these ices, and headquarters overhead. Day-to-day trans- programs could not be improved, but rather that action and book-keeping services are provided the level of support services they offer is adequate directly to groups by individual "facilitators" who to keep most of their SHGs healthy. Annex II-2 are identified by branch staff. Each facilitator is provides details of the support services offered in contracted by about 200 SHGs and is paid 1 per- each program. Organizational structure and cost cent of each group's outstanding loans for his or coverage are described below. her support. In the MFI program, MFI acts as a support In the Dhan program, the Dhan Foundation organization whose staff form, support, and mon- NGO is a support organization that forms SHG itor all SHGs. ICICI Bank, India's largest private federations to run SHG programs. In this case, it bank, is the only lender to SHGs and also handles formed the KVK federation in 1997. The federa- cash deposits and withdrawals for SHGs at its tion serves 350 SHGs at two levels: 16 local or branches. The costs of these and other support cluster-level associations of 10­15 SHGs each, and services are borne by SHG members, who pay 18 the block-level federation of all the cluster-level percent per annum on loans from ICICI Bank.18 associations. Cluster-level associations train and monitor SHGs and help them conduct bank trans- actions. The block-level KVK federation borrows 18All interest rates in Part II of this Occasional Paper are stated on an effective (declining balance) basis. from banks, on-lends to SHGs and cluster-level 21 OP 12-II.qxd 8/17/07 3:29 PM Page 22 associations, and trains and monitors the associa- this analysis. The income may be paid by SHGs tions. Federation staff at both levels estimate their directly--through fees to an individual facilitator costs at the beginning of each year and collect this or interest payments to a lender--or indirectly, by from SHGs in proportion to their loans outstand- paying interest to an SHG federation that, in turn, ing. At the end of the year, any surplus contribu- pays interest to a local bank. Still, the revenue tion is returned to the SHGs. earned by the bank is, at its origin, paid by an SHG. Banks participating in the Dhan program lend External actors also earn nonoperating income to the KVK federation; they also lend directly to from supporting SHGs. For example, a support SHGs and handle their cash transactions at organization that collects fees from SHGs may branches. They cover their costs through the inter- deposit the sum in a bank and earn interest. This est rate they charge on loans to SHGs and SHG type of nonoperating income is not predictable and federations (typically 11­12% per annum). was not counted in this analysis. Quantifying costs is less straightforward. The Sustainability Test Part 2: Are SHG Programs most accurate approach is to measure the precise Able to Cover the Costs of the External cost to each actor of performing each support serv- Support They Provide? ice, separating out the cost of any activity unrelated After ensuring the SHG programs in our study to serving SHGs. Because doing this is not practi- offer adequate support services, we looked at cal, several simple assumptions and estimates were whether the total costs of providing support serv- made instead. ices are paid for out of operating income earned First, in some cases, it was assumed that all of an from SHGs. A three-step process was applied to actor's operating costs relate to providing support each program. services to SHGs. For example, in MFI's SHG pro- We did not investigate whether SHG internal gram, MFI plays the role of a support organization revenues exceed internal costs, in part because we whose main purpose is to perform administration, did not have access to internal group results. supervision, liquidity, and other functions for However, the findings of Part I of this Occasional SHGs. But MFI also performs nonfinancial serv- Paper have shown that, in general, group financial ices, such as teaching SHG members about health performance is generally positive. Also, in a study and environmental issues. To be conservative, all of cited earlier, Srinivasan found that SHG adminis- MFI's operating costs are accounted for in this trative costs--such as stationery and travel-- analysis, even though some costs were probably amount to roughly US$17 per SHG per year. We incurred to deliver nonfinancial services. assumed that even SHGs that save only a few dol- Second, assumptions were made for banks' cost lars each month19 would be able to meet these of funds and their cost of handling SHG transac- annual expenses out of savings or the interest tions at branches. The Reserve Bank of India's rate earned on savings. for reverse repurchase agreements was used as a proxy for cost of funds when the actual cost was Step 1. Account for income earned and unavailable. For processing SHG loans and group the costs of support services transactions at branches, a cost to banks of 3 per- External actors earn two types of operating cent of loans outstanding was assumed. This was income from SHGs: fees and interest on loans. reduced to 1 percent of loans outstanding in the Both types of operating income are included in case of the Sarvodaya program, because group transactions are not handled by the bank. These 19The bank-organized groups we visited usually saved more--about US$20 per month per group. operating cost assumptions are conservative (that 22 OP 12-II.qxd 8/21/07 2:10 AM Page 23 is, probably on the high side), based on conversa- earlier SHG program, and these costs would have tions with bank officials and on the 2002 Seibel been fully amortized by 2005, the year for which and Harishkumar study. That study found that a we analyzed the organization. commercial bank, a regional rural bank, and a dis- Third, we ensured that, for each SHG program, trict central cooperative bank incurred only a small subsidized funds received from international and marginal cost for processing SHG transactions, local donor agencies were not included in operat- because of underutilized branch capacity and ing income. By accounting for these subsidies sep- group-based transactions.20 arately, we are able to arrive at an understanding of Third, some costs we judged to be relatively the system's inherent sustainability--the compari- insignificant were excluded. In most cases, the son of its internally generated income to its total apportioned cost of the bank's head office over- operating costs. This gives a better picture of the head was not considered, because SHG lending is program's ability to expand when subsidies are no usually only a tiny portion of a bank branch's longer available. operations in rural areas, and head office overhead One adjustment we did not make is to increase is allocated among a large number of bank branches. the interest rate banks charge SHGs to a commer- cial level. Critics of the SHG approach argue that Step 2. Adjust Income and Costs rates of 8­12 percent per annum on loans to SHGs After tallying the income and costs, three impor- are below market rates and that the true cost to tant adjustments were made to get a full picture of banks of making and servicing these loans is much each SHG program's sustainability. higher. They argue that treating the bank loans as First, loan loss provisioning was standardized at if they carried a market interest rate presents a 2 percent of average portfolio outstanding in each truer picture of SHG sustainability. We did not case. This is because each SHG program maintains make this adjustment because it would presuppose a different policy on how much to provision the conclusion to our main research question: how against loan losses, and we wanted to be able to are external actors (including banks) in four SHG compare them without giving any one an unfair programs servicing SHGs, and are they covering advantage. their costs? Second, the cost of forming SHGs and SHG federations was amortized over five years when Step 3. Estimate loan portfolio and total assets this activity was not included in ongoing costs of Finally, we estimated the total value of external an SHG program. For instance, Dhan Foundation loans to SHGs and the total assets managed in each incurred a cost of US$200 per group in forming SHG program to help us understand how effi- its SHGs, its local-level associations (CLAs), and ciently the SHG program operates: the greater the the KVK federation over three years, beginning in loan portfolio and assets compared with the cost of 1997. We counted the amortized cost (one-fifth of providing support services, the more efficient the the total) when we analyzed the program's 2003 program. performance. On the other hand, most of the External loans to SHGs include loans from com- SHGs and federations managed by Sarvodaya mercial banks, SHG federations, and support Nanofinance Ltd. were created before 2000 by an organizations. In the Dhan program, for example, total external loans to SHGs include loans from 20In most cases, marginal costs were minimal because of excess capac- the KVK federation and from Canara Bank. ity in the branch and no additional personnel requirement. The study The total assets managed by an SHG program also found that the SHG lending business resulted in returns on assets of 4.6 percent to 11.8 percent versus ­1.7 percent to 2.3 percent for include external loans, the total deposits SHGs the banks as a whole. maintain in bank accounts, and all the assets that 23 OP 12-II.qxd 8/17/07 3:29 PM Page 24 Table II-3 Dhan's KVK SHG Program: Estimated Adjusted Income, Costs, and Assets (in US$) Dhan KVK Foundation Federation -- Canara Bank TOTAL Support SHG Type of actor organization federation Individual(s) Bank All actors Interest income 0 41,800 -- 60,811 102,611 Fee income 0 2,800 -- 0 2,800 TOTAL INCOME 0 44,600 -- 60,811 105,411 Staff costs 0 8,600 -- 0 8,600 Administrative expenses 0 6,900 -- 15,864 22,764 Training costs 0 2,700 -- 0 2,700 SHG promotion costs (amortized) 0 0 -- 0 0 SHG federation promotion costs (amortized) 13,920 0 -- 0 13,920 Total Operating Costs 13,920 18,200 -- 15,864 47,984 Estimated cost of funds 0 26,400 -- 37,016 63,416 Adjusted loan loss provisions 0 4,880 -- 5,696 10,576 TOTAL COSTS 13,920 49,480 -- 58,576 121,976 Startup subsidies (amortized) 13,920 0 -- 0 13,920 Ongoing subsidies 0 0 -- 0 0 MARGIN (Income ­ costs) (13,920) (4,880) -- 2,235 (16,565) Estimated average external loans to SHGs 0 244,000 -- 284,793 528,793 Average SHG deposits 0 0 -- 227,471a 227,471 Estimated average assets used for SHG program 0 77,872 -- 0 77,872 EST TOTAL ASSETS MANAGED 0 321,872 -- 512,264 834,136 aThese deposits are liabilities on the bank's books, but are financial assets of the SHGs. external actors use to provide support services to serving SHGs, we made very rough estimates of SHGs. For instance, the total estimated assets of allocations. the Sarvodaya SHG program include external Data that emerged from this three-step process loans to SHGs from the Sarvodaya support organ- were used to construct a simple table that lists ization, the aggregate savings held by the SHGs in income, costs, and assets for each program. Table their own accounts, and all the assets owned by II-3 is a sample of this for the Dhan program. the support organization and the 50 SHG federa- Table II-3 suggests that Dhan's KVK program is tions that support SHGs. These include office running a deficit in providing support services to equipment, working capital, and other fixed and SHGs and could lose its ability to provide these liquid assets. In cases where an external actor uses services over time if adjustments are not made. these fixed and liquid assets for more than just External actors involved in the program spent a 24 OP 12-II.qxd 8/17/07 3:29 PM Page 25 Table II-4. Four SHG Programs: Estimated Basic Financial Results, Adjusted (in US$) OBC Dhan Mean MFI Sarvodaya (Rudrapur) (KVK Federation) (unweighted) Operations SHG members 517,784 47,282 4,949 6,264 144,070 SHG borrowers 304,380 67,061 4,848 5,356 95,411 Staff of external actors supporting SHGs 1,846 613 10 35 626 Income Operating income earned by external actors 7,315,521 1,718,635 85,775 105,411 2,306,336 Costs Estimated operating costs incurred by external actors (adjusted) 2,701,918 742,701 20,525 47,984 878,282 Estimated total costs incurred by external actors (adjusted) 5,562,373 1,818,753 82,175 121,975 1,896,319 Margin (Income ­ Costs) 1,753,148 (100,118) 3,600 (16,564) 410,017 Loans and assets Average external loans to SHGs 40,863,636 4,306,169 602,500 528,793 11,575,275 Estimated average total assets managed 45,358,636 9,500,218 987,291 834,136 14,170,070 total of about US$114,000 to support SHGs but I What is the cost to support one borrower? collected only US$105,000 from them directly or Indicator: Total cost/SHG borrowers indirectly. To correct this problem, Dhan I How productive is the collective staff of these Foundation or KVK federation needs to reduce SHG programs? Indicator: SHG borrowers/ the program's cost of providing services or staff members increase the interest rates or fees charged to SHGs. I How big are the external loans provided to Basic operational and financial results for each borrowers? Indicator: Period-average external of the four SHG programs are summarized in loans/SHG borrowers Table II-4. (See Annex II-B for detailed tables.) I What effective interest rate do borrowers pay on external loans? Indicator: Income/Period- Assessing the Financial Performance of SHG average external loans Programs I What is the administrative cost to keep a given Data from Table II-4 can be used to calculate sim- loan amount outstanding? Indicator: Operating ple indicators that can help explain several aspects cost/Period-average external loans of the financial performance of the SHG pro- grams. The most important question in this study The indicators calculated in Table II-5 are simple is whether these programs are sustainable: proxies for the industry standard indicators most analysts use in assessing a single microfinance insti- I Do SHG programs cover the costs of provid- tution. The precise data required for those latter ing support services? Indicator: Income/costs 25 OP 12-II.qxd 8/17/07 3:29 PM Page 26 Table II-5. Key Financial Results for SHG Programs OBC Dhan Mean MFI Sarvodaya (Rudrapur) (KVK Federation) (unweighted) Portfolio at risk on 0.33% 0.7% 0.0% 2.94% -- loans to SHGs (>90 days) (>30 days) (>90 days) (>30 days) Income/costs 131.5% 94.5% 104.4% 86.4% 104.2% Estimated operating costs/ borrower per year $8.9 $11.1 $4.2 $9.0 $8.3 Borrowers per staff person 165 109 485 153 228 Average external loan balance outstanding per borrower $134.3 $64.2 $124.3 $98.7 $105.4 Income/average external loans to SHGs 17.9% 22.2%a 14.2% 19.9% 18.6% Estimated operating costs/ average external loans to SHGs 6.6% 17.2% 3.4% 9.1% 9.1% aThis is interest income earned by the federations for lending to SHGs. The federations, in turn, pay interest on loans to the Sarvodaya support organization. That income is not included here. indicators are not available because the SHG pro- and loan loss provisions. This administrative cost grams analyzed involve several external actors. turns out to be less than 10 percent. This is The indicators reveal that, on average, the four approximately the same as the full interest rate that SHG programs cover all (104 percent) of the costs many banks charge on SHG loans. Where a bank is of providing support services to SHGs. Two of the the only external actor supporting SHGs, it is dif- programs cover all of their costs. The other two ficult to see how the bank can charge this rate and must find ways to increase their revenues from cover all of its costs--but still expect to achieve SHGs or to become more efficient. On average, long-term group viability. the four programs spend US$8 per SHG member Table II-6 compares the performance of the to provide support services. four SHG programs with benchmarks from other The most efficient program in the group stud- groups of microfinance providers. All of those ied is OBC, at Rudrapur, which uses a small num- comparator groups consist of single-institution, ber of bank branch staff and external facilitators to stand-alone providers. It is not completely precise manage all SHG operations. Remarkably, nearly to compare single-institution performance indica- 500 SHG borrowers are served by each staff per- tors against proxy indicators that consolidate the son (three branch officers and seven individual performance of multiple actors. Even so, we think facilitators). Yet the program's low outstanding that the comparison sheds some meaningful light. balance per borrower and its location in a rural The first comparator set is the group of 58 mountainous region suggest that it reaches among Indian SHG programs rated by M-CRIL between the poorest clients of the programs studied. September 1998 and June 2003 and shown in How much do SHGs pay for credit? The aver- Table II-2. The second comparator set is the group age yield of the four programs is 18.6 percent. In of nine Grameen-style microfinance institutions other words, SHGs pay about 19 percent interest that were also rated by M-CRIL during this for credit and a full set of support services. period. The third comparator set is a group of 37 The last indicator in the table shows the admin- leading microfinance institutions in India, of vari- istrative cost of keeping US$1 of loans outstand- ous types, that had voluntarily provided unad- ing to SHGs. It does not include the cost of funds justed financial information to the Microfinance 26 OP 12-II.qxd 8/17/07 3:29 PM Page 27 Table II-6. Key Benchmarks 9 Grameen- 58 SHG style 37 leading 302 4 SHG microfinance microfinance microfinance microfinance programs institutions institutions institutions institutions (proxy in India in India in India (global) indicators) MixMarket/ M-CRIL M-CRIL MicroBanking Source of data (2003)a (2003) MixMarketb Bulletin c This study Average borrowers (unweighted) 5,912 27,847 44,031 62,246 95,411 Portfolio at risk 19.3% <3.0% (>60 days, 4.2% 4.4% 3.9% (>30 or 90 days, loans to SHGs) (>60 days) (>30 days) (>30 days) loans to SHGs) Operational self-sufficiency (income/costs) 48% 85% 98.5% 123.6% 104.2% Average outstanding loan balance per borrower US$37 US$54 US$134 US$814 US$105 (as % of GNI p.c.) (6.0%) (8.7%) (21.6%) (62.3%) (17.0%) Operating costs per borrower per year US$24 US$18 US$14 US$153 US$8 (as % of GNI p.c.) (3.8%) (2.9%) (2.3%) (NA) (1.3%) Operating costs/average loan portfolio 63.8% 33.4% 15.5% 28.9% 9.1% Borrowers per staff person 118 142 439 140 228 Yield on gross loan portfolio 12.6% 28.8% 20.7%d 38.1% 18.6% aCalculations based on M-CRIL (2004). The numbers in this table do not always match numbers in similar tables from the original report, because some of those tables show weighted calculations and include results from non-Indian SHG programs. bMIX Market and World Bank (2005). cMIX (2006). dFinancial revenue ratio (financial revenue as a percentage of total assets). Yield on gross portfolio was not available. Information eXchange (MIX) up to September The single clearest measure of success or failure 2005.21 Microfinance institutions that report to in a microfinance program is the ability to recover the MIX are generally more commercially oriented money that is loaned. By this standard, the 58 programs, so the sample of 37 microfinance insti- SHG promoters rated by M-CRIL are not success- tutions is called "leading Indian MFIs" here. The ful, given their average portfolio-at-risk (>60 days) last comparator set is based on adjusted data for a of 19.3 percent. In other words, almost a fifth of group of 302 microfinance institutions worldwide the loans outstanding to SHGs by banks and other that voluntarily provided data for the 2004 fiscal lenders are two months or more late and thus at year on a confidential basis to the MIX; five-sixths substantial risk of default. Experience in other of these institutions have external audits, ratings, microfinance models suggests that when portfolio or assessments to support their performance infor- at risk measured at one repayment period rises mation. above 10 percent, it usually becomes unsustain- In general, the four SHG programs in our study able--that is, it must be reduced quickly or it spins compare highly favorably to the benchmarks. In out of control.22 particular, they seem to outperform the 58 stand- 22It is important to note that the portfolio-at-risk figures reported here alone SHG-promoting microfinance institutions are for external loans to the SHGs. It appears that SHGs can sometimes analyzed by M-CRIL. sustain high PAR on internal lending by groups to their members, while still ultimately collecting a very high percentage of these loans (see Table 21Some, but not all, of the MFIs had external audits, ratings, or assess- II-6). But this does not mean that high PAR on external loans to the ments to back up their financial information. SHGs is sustainable. 27 OP 12-II.qxd 8/17/07 3:29 PM Page 28 This result leads us to suspect that, on average, But banks charge SHGs too little to cover the these 58 SHG promoters do not adequately per- costs of supporting the groups over the long run. form the support services necessary to ensure Banks normally fail to consider the sunk costs of group stability and loan repayment. In contrast, the NGOs that have organized and/or maintain the SHG programs studied here, and the other com- groups to which the bank is lending. If banks had parators, do a better job of collections. Their port- to provide all the support services needed to pro- folio-at-risk ranges from 0.0 percent to 2.94 percent tect the quality of their SHG loans, or pay NGOs (> 30 days), compared to 4.4 percent (>30 days) for or other actors to handle this function, they would the leading Indian microfinance institutions. have to charge SHGs higher interest rates or fees. Data from the four leading SHG programs International experience with grassroots-level studied indicate that, when it includes the neces- financial intermediaries has shown time and again sary support functions underpinning long-term that unless core external support functions are pro- sustainability, the SHG approach can compare vided in a sustainable manner, and are paid by rev- favorably with other models of financial service enue generated within the system itself, commu- delivery--despite the poor results of many Indian nity-level units will degrade over time and SHG programs, as represented by the M-CRIL eventually unravel. The history of savings and sample. Although three of the four programs do credit cooperatives, financial service associations, not yet cover all of their costs, they demonstrate community banks, and others have demonstrated the potential for doing so with relatively minor that the reputation and ultimate success of a sys- adjustments to cost and income. Moreover, they tem depends on the strength of centrally provided appear to serve people who are as poor as or support. In a few instances, this support has been poorer than clients of microfinance institutions in built over a period of several decades, allowing the India when measured by average loan outstanding. grassroots networks of membership-based financial The four SHG programs studied also compare intermediaries to compete into the modern era favorably to international benchmarks. Like most (Christen 2005). In most instances, this has not microfinance programs in South Asia, they make been the case. loans that are a far smaller percentage of average In those few instances where community-based income than in other regions. Low labor costs and financial systems have been successful, the support high population density in South Asia make tiny infrastructure was usually built during a second loans somewhat less expensive to deliver. phase, after the grassroots-level units were put in place. In India, therefore, it may not be too late, Strategic Implications for Proponents of the even though at present commercial banks and SHG Approach other promoters do not necessarily appreciate the The 2.2 million SHGs that currently exist provide necessity and the full cost of long-term support to a large and growing market for Indian banks. The maintain their SHG portfolio. results of this study show that well-run SHG pro- Making sure that such support is in place often grams compare favorably on outreach and opera- will require increases--usually modest--in the tional efficiency with alternatives. Thus there is a interest or fees charged to SHGs. But sustainable case to expand these types of programs. However, funding of such support is very much in the inter- a large number of SHG programs are not well run, ests of the members--and essential to the perma- and those need to be improved by adding essential nence of the SHG system. support services. 28 OP 12-III.qxd 8/17/07 3:27 PM Page 29 I I I Part I ANNEX Detailed Data on 150 Indian Self-Help Groups Table IA-1. SHG Location as Indicator of Outreach, by Promoting Institution (miles) Sakhi Average SHG distance from PMMS Samiti PRADAN CGB PANI Average Bank 19.8 5.0 8.6 2.7 3.5 7.9 Health center 9.4 1.6 4.8 2.7 5.8 4.9 Paved road/highway 3.3 1.7 10.6 2.4 8.2 5.2 Table IA-2. SHG Location: Type of Settlement, by Promoting Institution (percent) Sakhi Settlement PMMS Samiti PRADAN CGB PANI Average Few scattered houses 0.0 0.0 6.7 0.0 3.3 2.0 Small village 60.0 6.7 66.7 43.3 73.3 50.0 Large village 40.0 66.7 26.7 56.7 23.3 42.7 Town 0.0 26.7 0.0 0.0 0.0 5.3 Total 100.0 100.0 100.0 100.0 100.0 100.0 Table IA-3. SHG Members Belonging to Scheduled Tribes and Castes, by Promoting Institution (percent) Sakhi Caste category PMMS Samiti PRADAN CGB PANI Average Scheduled tribesa 0.9 0.0 84.6 25.9 19.3 26.1 Scheduled castesa 21.0 58.4 12.0 9.3 48.7 29.9 Subtotal: Scheduled tribes and castes 21.9 58.4 96.6 35.2 68.0 56.0 Minorities 1.6 11.6 0.0 2.7 7.4 4.7 Backward castesb 74.8 22.4 3.4 23.8 18.3 28.5 Other castes 1.6 7.7 0.0 38.3 6.4 10.8 Total 100.0 100.0 100.0 100.0 100.0 100.0 a Scheduled castes and scheduled tribes are communities accorded special status by the Constitution of India. These communities were considered "outcastes" and excluded from the four-caste system that was the social superstructure of Hindu society in the Indian subcon- tinent for thousands of years. These castes and tribes were relegated to the most menial labor, with no chance of upward mobility, and degenerated into the most economically and socially backward communities in the region. b "Backward castes" comprise the last of the four major castes in the Hindu social hierarchy. Though their status is above that of the scheduled tribes and castes, they are economically backward and have poor living conditions. 29 OP 12-III.qxd 8/17/07 3:27 PM Page 30 Table IA-4. Landholdings of SHG Members, by Promoting Institution (percent) Sakhi Landholdings PMMS Samiti PRADAN CGB PANI Average Landless (no homestead land) 11.0 44.2 0.8 21.9 8.5 17.3 Marginal: Owns no agricultural land 47.6 39.1 84.8 38.7 88.1 59.7 Subtotal: Landless and marginal 58.5 83.3 85.6 60.6 96.6 76.9 Small landholding 35.2 16.2 12.0 18.5 3.4 17.1 Large landholding 6.3 0.5 2.4 20.8 0.0 6.0 Total 100.0 100.0 100.0 100.0 100.0 100.0 Table IA-5. Primary Source of Income of SHG Members, by Promoting Institution (percent) Sakhi Source PMMS Samiti PRADAN CGB PANI Average Work own land 42 26 80 28 39 43 Agricultural labor 40 34 14 16 55 32 Subtotal: Agriculture-dependent 82 60 94 44 94 75 Nonagriculture wage labor 10 22 2 44 1 16 Microenterprise 5 17 3 11 2 8 Other 2 1 1 1 3 2 Total 100 100 100 100 100 100 Table IA-6. Education Level of SHG Members, by Promoting Institution (percent) Sakhi Level PMMS Samiti PRADAN CGB PANI Average Illiterate 17 43 61 14 74 42 Able to sign 71 46 25 0 19 32 Literate: Read and write 4 1 7 33 0 9 Primary school 4 10 1 23 2 8 Secondary school 3 1 5 9 4 4 Matriculate (passed 10th grade) and above 1 0 1 22 1 5 Subtotal: literate 12 12 14 87 7 26 Total 100 100 100 100 100 100 30 OP 12-III.qxd 8/17/07 3:27 PM Page 31 Table IA-7. SHG Organizational Details (categories used in rating the SHGs) Number of records Details Marginal A few basic records exist and are being maintained. Below average All basic records exist, but only a few are maintained. Quality of minutes book is poor, containing only details of meeting date, number of members, financial transactions, and signatures. Average All basic records exist and are maintained, but are not up to date. Quality of minutes book is average, with details of meeting date, number of members, financial transactions and discussions related to loan sanctions, monitoring, and default, along with signatures. Above average All basic records exist and are maintained and up to date, but not updated regularly each month. Quality of minutes book is average, with details of meeting date, number of members, financial transactions, discussions related to loan sanctions, monitoring, default, federation functioning (if any), and social aspects, along with signatures. Complete and up to date All basic records exist and are maintained and up to date, but contain errors and do not tally with financial statements. Virtually no errors All basic records exist and are maintained and up to date and have virtually no errors. Table IA-8. SHG Organizational Details, by Promoting Institution (percent) Sakhi Details PMMS Samiti PRADAN CGB PANI Average SHG meetings held in the morning 0 57 70 10 30 33 Records maintained by paid bookkeeper 97 100 80 10 0 57 Member's family sometimes 10 60 33 13 0 23 makes payment 31 OP 12-III.qxd 8/17/07 3:27 PM Page 32 Table IA-9. Field Staff Rating Criteria for SHGs Group cohesion rating Criteria Exceptional All members have similar backgrounds; the SHG makes decisions by consensus; all members attend meetings regularly and participate in discussions and decision-making. Above average Most members have similar backgrounds; SHG decisions are made by leaders and 2­3 members; all, except 1­2 members, attend meetings regularly and participate in discussions. Average At least half of members have similar backgrounds; SHG decisions are made by leaders; all, except 1­2 members, attend meetings regularly and are aware of group transactions. Below average At least half of members have similar backgrounds; SHG decisions are made by one leader; more than half of members attend meetings regularly and are aware of group transactions. Poor Members have different backgrounds; SHG decisions are made by leader; attendance is variable--meets only for financial transactions. Overall rating of group functioning Criteria Exceptional Group meetings and savings and loan plus interest repayments are regular. Group norms exist and those related to attendance and savings are implemented. SHG has accessed loan from bank or federation. Above average Group meetings and savings and interest repayments are regular. Group norms exist, and those related to attendance and savings are implemented. SHG has accessed loan from bank or federation. Average Group meetings and savings are regular. Group norms exist but some or all are not implemented. Below average Group meetings are regular but not scheduled. Savings are regular, but group norms are not articulated. Poor Group exists but does not meet regularly. Savings and loan repayments are highly irregular. Table IA-10. SHG Members' Use of Loans, by Promoting Institution (percent) Sakhi Use PMMS Samiti PRADAN CGB PANI Average Agricultural purposes 51.7 9.6 51.0 34.0 12.0 31.7 Animal husbandry 17.3 33.9 9.4 25.0 30.0 23.1 Consumption 11.0 23.5 16.0 7.0 28.0 17.1 House construction and repair 11.0 13.6 0.0 7.0 2.6 6.8 Microenterprise 6.2 14.8 17.0 9.0 21.0 13.6 Loans for repayment of other loans 0.0 0.0 0.0 5.0 2.0 1.4 Other 2.8 4.7 6.6 13.0 4.5 6.3 32 OP 12-III.qxd 8/17/07 3:27 PM Page 33 Table IA-11. Average Cost of Promotion of SHGs over Three Years, by Promoting Institution (amounts in US$, exchange rate of 47.65 rupees to US$1 for year ending 31 March 2003 and 44.125 rupees to US$1 for year ending 31 March 2004) Sakhi As % of Item PMMS Samiti PRADAN CGB PANI Average average cost Number of SHGs promoted 300 135 123 360 785 341 -- Cost of social mobilization Salaries, allowances, and honorariums 34 101 47 14 68 53 20.2 Cost of books and materials 5 10 -- -- -- 3 1.2 Training costs 39 15 38 13 33 28 10.5 Capital for entry-point activity of SHGs 105 -- -- -- -- 21 8.2 Subtotal 183 126 85 27 101 104 39.9 Overhead--project management @ 10% 18 13 8 3 10 10 4.0 Average social mobilization cost per SHG (A) 201 139 93 30 111 115 44.1 Support costs Staff costs (excluding field workers) 13 21 116 -- 29 36 13.8 Office administration costs, including meetings 12 47 152 20 65 59 22.6 Training of executive committee members and staff 7 26 -- -- 6 8 3.0 Endowment fund for federation (Mandal Samakhya) per SHG 210 -- -- -- -- 42 16.4 Average support cost per SHG (B) 242 94 268 20 100 145 55.9 Total cost of mobilization and support per SHG (A + B) 443 233 361 50 211 260 100.0 33 OP 12-III.qxd 8/17/07 3:27 PM Page 34 Figure I-A1. PMMS--Statement of Income and Expenditure 2001­02 2002­03 Operating Income Service charge (interest from loans) 368 494 Fines 1 2 Interest on account with bank -- -- Interest refund of SHG from Federation 1 6 Other: -- -- 1. Fees/membership 2 -- 2. Resource Fee -- -- Total operating income 372 502 Operating Expenses Salaries/honorarium 11 14 Stationery 1 1 Interest on borrowings from Bank/VO/Fed. 215 274 Interest paid on group savings -- -- Consumables -- -- Travel 2 3 Equipment -- -- Social mobilization costs 66 67 Support costs 79 80 Loan loss provision 11 12 Others 3 4 Total operating expenses 388 455 Net operating profit/(loss) (16) 47 Non-operational income (grants from NGO, etc.) -- -- Non-operational expenses 1 5 Total consolidated profit/(loss) (17) 42 Note: For all five SHG programs, financial information for the periods ending March 2001, March 2002, March 2003, and March 2004 was converted at Rs. 46.577 to US$1, Rs. 48.733 to US$1, Rs. 47.65 to US$1, and Rs. 44.125 to US$1, respectively. Given the variable annual exchange rates, balances carried forward from one year to the next will have different values. 34 OP 12-III.qxd 8/17/07 3:27 PM Page 35 Figure I-A2. PMMS--Balance Sheet 2001­02 2002­03 ASSETS Cash in hand balance 6 7 Bank balance 11 16 Total loan portfolio (with members) 2,069 2,324 Loan loss reserve (11) (23) Deposits (investments by the SHG) 2 2 Share capital (equity) in village organization 17 20 Net fixed assets (after depreciation) -- -- Total assets 2,094 2,346 LIABILITIES AND EQUITY Liabilities Savings: compulsory 450 583 Savings: voluntary -- -- Loans: financial institution* 145 179 Loans: village organization/federation 760 889 Other short-term liabilities 291 102 Total liabilities 1,646 1,753 Equity Paid-in-equity (membership share) -- -- Funds distribution (11) (80) Grant: revolving fund** 75 94 Grant from promoting institution 144 148 Previous years' retained earnings/losses 257 389 Current year retained earnings/loss (17) 42 Total equity 448 593 Total liabilities and equity 2,094 2,346 Note: For all five SHG programs, financial information for the periods ending March 2001, March 2002, March 2003, and March 2004 was converted at Rs. 46.577 to US$1, Rs. 48.733 to US$1, Rs. 47.65 to US$1, and Rs. 44.125 to US$1, respectively. Given the variable annual exchange rates, balances carried forward from one year to the next will have different values. *Borrowings from financial institutions were through the federation or village organization. **The District Rural Development Agency (a state government institution that promotes SHGs) gives SHGs older than one year 5,000­20,000 rupees to meet members' credit demands. 35 OP 12-III.qxd 8/21/07 2:11 AM Page 36 Figure IA-3. Sakhi Samiti--Statement of Income and Expenditure 2000­01 2001­02 2002­03 Operating income Service charge (interest from loans) 193 270 347 Fines 8 12 25 Interest on account with bank -- -- -- Interest refund to SHG from federation -- -- -- Other 1. Fee/Commission 9 16 17 2. Stationery and others 1 -- 1 Total operating income 211 298 390 Operating expenses Salaries/honorarium 39 44 54 Stationery 2 5 5 Interest on borrowings from bank/village organization/federation 32 42 62 Interest paid on group savings -- -- 1 Consumables -- -- -- Travel 1 1 1 Social mobilization costs 47 45 46 Support costs 32 31 31 Loan loss provision -- 66 67 Others 1 1 -- Total operating expenses 154 235 267 Net operating profit/(loss) 57 63 123 Nonoperational income (grants from NGO, etc.) -- -- -- Nonoperational expenses -- -- -- Total consolidated profit/(loss) 57 63 123 Note: For all five SHG programs, financial information for the periods ending March 2001, March 2002, March 2003, and March 2004 was converted at Rs. 46.577 to US$1, Rs. 48.733 to US$1, Rs. 47.65 to US$1, and Rs. 44.125 to US$1, respectively. Given the variable annual exchange rates, balances carried forward from one year to the next will have different values. 36 OP 12-III.qxd 8/17/07 3:27 PM Page 37 Figure IA-4. Sakhi Samiti--Balance Sheet* 2000­01 2001­02 2002­03 ASSETS Cash in hand balance 74 56 68 Bank balance 13 14 18 Total loan portfolio 1,041 1,373 1,918 Loan loss reserve -- (66) (72) Deposits (investments by the SHG) -- -- -- Equity in Federation (Sakhi Suvidha shares) 14 14 14 Net fixed assets (after depreciation) 3 -- -- Total assets 1,145 1,391 1,946 LIABILITIES AND EQUITY Liabilities Savings: compulsory 591 714 881 Savings: voluntary -- -- -- Loans: bank 431 527 811 Loans: Federation (Sakhi Suvidha) 7 18 6 Other short-term liabilities -- -- -- Total liabilities 1,029 1,259 1,698 Equity Distribution of funds/savings (54) (73) (31) Funds with federation (Sakhi Suvidha) 28 64 75 Promotional grant by promoting institution 79 76 77 Previous years' retained earnings/losses 6 2 4 Current year retained earnings/loss 57 63 123 Total equity 116 132 248 Total liabilities and equity 1,145 1,391 1,946 Note: For all five SHG programs, financial information for the periods ending March 2001, March 2002, March 2003, and March 2004 was converted at Rs. 46.577 to US$1, Rs. 48.733 to US$1, Rs. 47.65 to US$1, and Rs. 44.125 to US$1, respectively. Given the variable annual exchange rates, balances carried forward from one year to the next will have different values. *In the case of Sakhi Samiti, some SHGs were not carrying forward profits and choosing to distribute a portion of retained earnings after the close of the fiscal year. This affects calculations for retained earnings for the subsequent year. 37 OP 12-III.qxd 8/21/07 2:15 AM Page 38 Figure IA-5. PRADAN--Statement of Income and Expenditure 2000­01 2001­02 2002­03 Operating income Service charge (interest from loans) 64 88 131 Fines 14 14 16 Interest on account with bank -- -- -- Interest refund of SHG from federation -- -- -- Other 1. Fees/commissions -- -- -- 2. Stationery and other -- -- 1 Total operating income 78 102 148 Operating expenses Salaries/honorariums -- -- -- Stationery -- -- -- Interest on borrowings from bank/federation 1 1 8 Interest paid on group savings -- -- -- Consumables -- -- -- Travel -- -- -- Social mobilization costs 32 30 31 Support costs 91 87 89 Loan loss provision -- 17 18 Other 4 2 6 Total operating expenses 128 137 152 Net operating profit/(loss) (50) (35) (4) Nonoperational income (grants from NGO, etc.) -- -- -- Nonoperational expenses 16 14 7 Total consolidated profit/(loss) (66) (49) (11) Note: For all five SHG programs, financial information for the periods ending March 2001, March 2002, March 2003, and March 2004 was converted at Rs. 46.577 to US$1, Rs. 48.733 to US$1, Rs. 47.65 to US$1, and Rs. 44.125 to US$1, respectively. Given the variable annual exchange rates, balances carried forward from one year to the next will have different values. Figure IA-6. PRADAN--Balance Sheet* 2000­01 2001­02 2002­03 ASSETS Cash in hand balance 46 87 112 Bank balance 42 72 129 Total loan portfolio 284 366 509 Loan loss reserve -- (17) (18) Deposits (investments by the SHG) -- -- -- Net fixed assets (after depreciation) -- -- -- Total assets 372 508 732 LIABILITIES AND EQUITY Liabilities Savings: voluntary 295 422 573 Savings: compulsory -- -- -- Loans: bank 20 21 47 Loans: village organization/federation -- -- -- Other short-term liabilities -- 4 10 Total liabilities 315 447 630 Equity Distribution of funds/savings -- -- -- Promotional grant by promoting institution 123 118 120 Previous years' retained earnings/losses -- (8) (7) Current year retained earnings/loss (66) (49) (11) Total equity 57 61 102 Total liabilities and equity 372 508 732 Note: For all five SHG programs, financial information for the periods ending March 2001, March 2002, March 2003, and March 2004 was converted at Rs. 46.577 to US$1, Rs. 48.733 to US$1, Rs. 47.65 to US$1, and Rs. 44.125 to US$1, respectively. Given the variable annual exchange rates, balances carried forward from one year to the next will have different values. *In the case of PRADAN, profitable SHGs were carrying forward only 20% of the previous year's retained earnings and distributing 80%. However, loss-making SHGs would carry forward the losses. This affects calculations for retained earnings in subsequent years. 38 OP 12-III.qxd 8/21/07 2:16 AM Page 39 Figure IA-7. CGB--Statement of Income and Expenditure 2000­01 2001­02 2002­03 2003­04 Operating income Service charge (interest from loans) 90 129 241 219 Fines 2 2 1 1 Interest on account with bank 3 3 3 3 Interest refund of SHG from federation -- -- -- -- Other -- -- -- -- 1. Fees/commissions -- -- -- -- 2. Stationery and others -- -- -- -- Total operating income 95 134 245 223 Operating expenses Salaries/honorariums -- 2 2 2 Stationery 1 -- 1 -- Interest on borrowings from bank/federation 15 37 54 64 Interest paid on group savings -- -- -- -- Consumables -- -- -- -- Travel -- -- 1 1 Board -- -- -- -- Social mobilization costs 9 9 9 10 Support costs 6 6 6 7 Loan loss provision -- -- 118 128 Others 1 -- 3 2 Total operating expenses 32 54 194 214 Net operating profit/(loss) 63 80 51 9 Nonoperational income (grants from NGO, etc.) 4 -- -- -- Nonoperational expenses 12 1 14 6 Total consolidated profit/(loss) 55 79 37 3 Note: For all five SHG programs, financial information for the periods ending March 2001, March 2002, March 2003, and March 2004 was converted at Rs. 46.577 to US$1, Rs. 48.733 to US$1, Rs. 47.65 to US$1, and Rs. 44.125 to US$1, respectively. Given the variable annual exchange rates, balances carried forward from one year to the next will have different values. 39 OP 12-III.qxd 8/21/07 2:18 AM Page 40 Figure IA-8. CGB--Balance Sheet* 2000­01 2001­02 2002­03 2003­04 ASSETS Cash in hand balance -- 4 3 -- Bank balance 133 158 189 205 Total loan portfolio 508 1,000 1,716 2,088 Loan loss reserve -- -- (118) (243) Deposits (investments by the SHG) -- -- -- -- Net fixed assets (after depreciation) -- -- -- -- Other loans 9 -- -- 46 Total assets 650 1,162 1,790 2,096 LIABILITIES AND EQUITY Liabilities Savings: compulsory 376 645 987 1,102 Savings: voluntary -- -- -- -- Loans: bank 172 385 614 850 Loans: village organization/federation -- -- -- -- Other short-term liabilities 26 5 7 50 Total liabilities 574 1,035 1,608 2,002 Equity Distribution of funds/savings -- (39) -- (124) Sakhi Suvidha Fund 4 3 3 4 Promotional grant by promoting institution 16 15 15 17 Previous years' retained earnings/losses 1 69 127 194 Current year retained earnings/loss 55 79 37 3 Total equity 76 127 182 94 Total liabilities and equity 650 1,162 1,790 2,096 Note: For all five SHG programs, financial information for the periods ending March 2001, March 2002, March 2003, and March 2004 was converted at Rs. 46.577 to US$1, Rs. 48.733 to US$1, Rs. 47.65 to US$1, and Rs. 44.125 to US$1, respectively. Given the variable annual exchange rates, balances carried forward from one year to the next will have different values. *In the case of CGB, there was no distribution of earnings. Therefore, retained earnings in a given year will be the sum of Current Retained Earnings + Promotion Grant + Previous Retained Earnings - Distribution of Funds/Savings. Figure IA-9. PANI--Statement of Income and Expenditure 2000­01 2001­02 2002­03 2003­04 Operating income Service charge (interest from loans) 17 35 55 71 Fines -- -- -- -- Interest on account with bank 1 1 2 2 Interest refund to SHG from federation -- -- -- -- Other 1. Fee/commission -- -- -- -- 2. Stationery and other -- -- -- -- Total operating income 18 36 57 73 Operating expenses Salaries/honorarium -- -- -- 1 Stationery 1 2 2 3 Interest on borrowings from bank/village organization/federation 3 10 15 18 Interest paid on group savings -- -- -- -- Consumables -- -- -- -- Travel -- 1 -- -- Social mobilization costs 35 33 34 37 Support costs 32 30 31 33 Loan loss provision -- -- 19 21 Other -- -- 1 1 Total operating expenses 71 76 102 114 Net operating profit/(loss) (53) (40) (45) (41) Nonoperational income (grants from NGO, etc.) -- -- -- -- Nonoperational expenses -- -- 1 2 Total consolidated profit/(loss) (53) (40) (46) (43) Note: For all five SHG programs, financial information for the periods ending March 2001, March 2002, March 2003, and March 2004 was converted at Rs. 46.577 to US$1, Rs. 48.733 to US$1, Rs. 47.65 to US$1, and Rs. 44.125 to US$1, respectively. Given the variable annual exchange rates, balances carried forward from one year to the next will have different values. 40 OP 12-III.qxd 8/17/07 3:27 PM Page 41 Figure IA-10. PANI--Balance Sheet* 2000­01 2001­02 2002­03 2003­04 ASSETS Cash in hand balance 1 5 1 4 Bank balance 35 39 75 98 Total loan portfolio (with members) 133 287 327 441 Loan loss reserve -- -- (19) (41) Deposits (investments by the SHG) -- -- -- -- Other -- -- -- 4 Net fixed assets (after depreciation) -- -- -- -- Total assets 169 331 384 506 LIABILITIES AND EQUITY Liabilities Savings: voluntary 77 105 131 166 Savings: compulsory -- -- -- -- Loans: bank 73 176 192 211 Loans: village organization/federation -- 9 2 42 Other short-term liabilities -- -- 1 1 Loan loss reserve -- -- -- -- Total liabilities 150 290 326 420 Equity Distribution of Funds/Savings -- -- -- -- Distribution of retained earnings -- (1) (4) (4) Promotional Grant by Promoting Institution 67 64 65 70 Previous years' retained earnings/losses 5 18 43 63 Current year retained earnings/loss (53) (40) (46) (43) Total equity 19 41 58 86 Total liabilities and equity 169 331 384 506 Note: For all five SHG programs, financial information for the periods ending March 2001, March 2002, March 2003, and March 2004 was converted at Rs. 46.577 to US$1, Rs. 48.733 to US$1, Rs. 47.65 to US$1, and Rs. 44.125 to US$1, respectively. Given the variable annual exchange rates, balances carried forward from one year to the next will have different values. *In the case of PANI, there was no distribution of earnings. Therefore, retained earnings in a given year will be the sum of Current Retained Earnings + Promotion Grant + Previous Retained Earnings ­ Distribution of Funds/Savings. 41 OP 12-III.qxd 8/17/07 3:27 PM Page 42 I I I Part II ANNEX A How System Functions Are Covered OBC--Rudrapur Dhan--KVK ORGANIZATION MFI SNFL Branch Federation Support organization, Bank branch staff STRUCTURE Support organization SHG federations Individual agents SHG federations FUNCTIONS Product design Support organization Support organization OBC staff design Block-level SHG (MFI) designs basic (SNFL) designs basic products that are federation (BLF) (loan terms and SHG products along SHG products. Financial uniform across all designs basic SHG membership with ICICI Bank team. products include branches doing products. Products requirements/fees) Products include mandatory and microfinance. Products include term loans from mandatory and voluntary savings, bank include mandatory and banks, consumption recurring savings, and internal group voluntary individual loans from BLF, emergency fund, and lending, and savings accounts, term revolving line of credit to bank and internal group investments in the deposits, bank and mature groups from lending. capital of the federation. internal group lending, bank, mandatory and and others. voluntary savings. Administration Support organization Support organization Every SHG is organized Group meetings and has tiered staffing has 4­5 person in a center of 6­8 bookkeeping are (bookkeeping, travel, structure, including operational staff, 3- groups, and about 12 handled by CLA management, promoter (manages 25 person finance team, centers are managed by manager and transactions) SHGs), assistant project CEO, and support staff. a facilitator who is paid accountant. Bank officer (APO) (manages Each federation has a for by the SHGs. The transactions are 8 promoters), project CEO, 2­10 field officers, facilitator is a local conducted at bank officer (PO) (manages 4 and an accountant. secondary school branches, although BLF APOs), area manager Group meetings, graduate who handles has set up an office (manages 2­3 POs), bookkeeping, and bookkeeping, across from the bank and zonal manager transactions are transactions, and client branch to help SHG (manages 2 area handled by field officers. meetings. Bank branch leaders complete managers). Promoters Support organization has 2­3 staff dedicated transaction forms. Bank and APOs and POs appoints federation to SHG business. staff occasionally handle bookkeeping and CEO and 2 members of Branch staff members attends SHG and group meetings. federation Board of supervise these federation meetings. Transactions are Directors. Computerized facilitators. OBC also processed at ICICI Bank MIS in place at head has a development branches. Computerized office. manager who forms and MIS now being installed. nurtures groups in the Rudrapur area. Support Promoters form and Federation field officers Bank branch manager CLAs, the second-tier nurture groups with are responsible for promotes groups for the organization at the (customer support of APOs and forming and nurturing first 2­3 months of SHG panchayat level, acquisition/group POs. Strict mechanisms groups and for operations, and then promote and strengthen formation, group in place for group exits managing exits and relies on facilitators who groups at the village nurturing, management and entrances, such as entrances. SHGs are emerge from groups or level, and also perform of exits/entrances) requiring new members members of the friends/relatives to community mobilization to invest 1/20th of group federation, with 1 vote handle group meetings and other social profit account to join. per group. New and supervision. OBC services. BLF, the third- Promoter organizes two members pay a one- requires members to be tier block-level meetings each month, time admissions fee-- from similar economic federation, supervises one for business and Rs. 1,000--per group to backgrounds. Elected these functions and one for discussion of join an MBT, but are not group and center supports CLAs with social development required to contribute leaders are also training, etc. topics picked by group. share capital to the responsible for group federation. nurturing and inter- group discipline. (Continued on next page) 42 OP 12-III.qxd 8/17/07 3:27 PM Page 43 Part II ANNEX A (continued) OBC--Rudrapur Dhan--KVK ORGANIZATION MFI SNFL Branch Federation Support organization, Bank branch staff STRUCTURE Support organization SHG federations Individual agents SHG federations FUNCTIONS Liquidity Treasury functions are Support organization Bank branch staff BLFs handle treasury performed by the three- handles all fundraising handle all treasury management, including (treasury and person head office. All from banks. Federation functions, including use borrowing from banks, fundraising) funding is sourced from staff handle treasury of deposits for lending, and lending to SHGs ICICI Bank. management for lending and requests from OBC through CLAs. to SHGs, with support headquarters for from support additional funds. organization's head office. Training Staff training conducted Contract with BASIX Bank branch staff trains CLAs conduct training by zonal and area (http://www.basixindia. facilitators and clients. for SHG members. BLFs (staff and client managers. Client com/), paid for by grants Clients receive skills conduct trainings and training) training conducted by and earnings, to pay for training from support CLA staff. CLAs project officers and area staff training, systems, government or master and BLFs are set up managers. and planning. craftsmen, and pay for and trained by Dhan this directly. Foundation over 2­3 years, funded by subsidies. Direct supervision/ APO, PO, area All SHGs have one SHG members monitor Cluster-level and intervention in manager, and zonal external audit annually. each others' loan use. federation-level operations manager staff perform Internal audit team OBC microfinance managers and all supervision supervises trusts (SHG officer (development accountants monitor (verification of functions, including federations), and manager) monitors credit use and procedures, action in monitoring bookkeeping, support staff supervise branch operations, repayments. Bank case of default) completing registers, and intervene in verifies bookkeeping, branch managers and and checking group federation operations, etc. Facilitators federation staff visit meetings. Each level is including monitoring supervise group groups to motivate responsible for repayments. operations and, in turn, members and solve intervention at lower are supervised by bank problems. levels. branch staff, who have power to intervene. 43 OP 12-III.qxd 8/17/07 3:27 PM Page 44 I I I Part II ANNEX B Detailed Financial Analyses Table IIB-1. Microcredit Foundation of India Program: Estimated Adjusted Income, Costs, and Assets (in US$, as of March 31, 2006) MFI ICICI Bank TOTAL Type of actor (Support organization) (Bank) (All actors) Interest income 0 4,903,636 4,903,636 Fee income 2,411,885 0 2,411,885 TOTAL INCOME 2,411,885 4,903,636 7,315,521 Staff costs 1,016,563 0 1,016,563 Administrative expenses 327,805 1,225,909 1,553,714 Training costs 115,033 0 115,033 SHG promotion costs (amortized) 16,608 0 16,608 SHG federation promotion costs (amortized) 0 0 0 Total Operating Costs 1,476,009 1,225,909 2,701,918 Estimated cost of funds 0 2,043,182a 2,043,182 Adjusted loan loss provisions 0 817,273 817,273 TOTAL COSTS 0 4,086,364 5,562,373 Startup subsidies (amortized) 0 0 0 Ongoing subsidies 0 0 0 MARGIN (Income ­ costs) 935,876 817,272 1,753,148 Estimated average external loans to SHGs 0 40,863,636 40,863,636 Average SHG deposits 0 NA NA Estimated average assets used for SHG program 4,495,000a 0 4,495,000 EST TOTAL ASSETS MANAGED 4,495,000 40,863,636 45,358,636 Note: For all five SHG programs, financial information for the periods ending March 2001, March 2002, March 2003, and March 2004 was converted at Rs. 46.577 to US$1, Rs. 48.733 to US$1, Rs. 47.65 to US$1, and Rs. 44.125 to US$1, respectively. Given the variable annual exchange rates, balances carried forward from one year to the next will have different values. aCGAP estimate. Table IIB-2. Sarvodaya Nanofinance Ltd.'s SHG Program: Estimated Adjusted Income, Costs, and Assets (in US$, as of March 31, 2005) Sarvodaya (Support SHG Federation Bank TOTAL Type of actor organization) (Trusts) (Banks) (All actors) Interest income 490,637 864,766 294,748 1,650,151 Fee income 4,567 63,917 0 68,484 TOTAL INCOME 495,204 928,683 294,748 1,718,635 Staff costs 77,666 271,728 0 349,394 Administrative expenses 89,016 143,829 43,062a 275,906 Training costs 117,401 0 0 117,401 SHG promotion costs (amortized) 0 0 0 0 SHG federation promotion costs (amortized) 0 0 0 0 Total Operating Costs 284,083 415,557 43,062 742,702 Estimated cost of funds 294,748 490,637 204,543 989,928 Adjusted loan loss provisions 0 86,123 0 86,123 TOTAL COSTS 578,831 992,317 247,605 1,818,753 Startup subsidies (amortized) 0 0 0 0 Ongoing subsidies 117,401 0 0 117,401 MARGIN (Income ­ costs) (83,627) (63,634) 47,143 (100,118) Estimated average external loans to SHGs 0 4,306,169 0 4,306,169 Average SHG deposits 0 0 NA NA Estimated average assets used for SHG program 1,180,829 4,013,220 NA 5,194,049 EST TOTAL ASSETS MANAGED 1,180,829 8,319,389 0 9,500,218 Note: For all five SHG programs, financial information for the periods ending March 2001, March 2002, March 2003, and March 2004 was converted at Rs. 46.577 to US$1, Rs. 48.733 to US$1, Rs. 47.65 to US$1, and Rs. 44.125 to US$1, respectively. Given the variable annual exchange rates, balances carried forward from one year to the next will have different values. aCGAP estimate. 44 OP 12-III.qxd 8/17/07 3:27 PM Page 45 Table IIB-3. Oriental Bank of Commerce's SHG Program: Estimated Adjusted Income, Costs, and Assets (in US$, as of March 31, 2003) Facilitators Oriental Bank TOTAL Type of actor (Individuals) of Commerce (Bank) (All actors) Interest income 0 79,550 79,550 Fee income 6,025 200 6,225 TOTAL INCOME 6,025 79,750 85,775 Staff costs 6,025a 9,950 15,975 Administrative expenses 0 4,550 4,550 Training costs 0 0 0 SHG promotion costs (amortized) 0 0 0 SHG federation promotion costs (amortized) 0 0 0 Total Operating Costs 6,025 14,500 20,525 Estimated cost of funds 0 49,600 49,600 Adjusted loan loss provisions 0 12,050 12,050 TOTAL COSTS 6,025 76,150 82,175 Startup subsidies (amortized) 0 0 0 Ongoing subsidies 0 0 0 MARGIN (Income ­ costs) 0 3,600 3,600 Estimated average external loans to SHGs 0 602,500 602,500 Average SHG deposits 0 364,344 364,344 Estimated average assets used for SHG program 0 20,447 20,447 EST TOTAL ASSETS MANAGED 0 987,291 987,291 Note: For all five SHG programs, financial information for the periods ending March 2001, March 2002, March 2003, and March 2004 was converted at Rs. 46.577 to US$1, Rs. 48.733 to US$1, Rs. 47.65 to US$1, and Rs. 44.125 to US$1, respectively. Given the variable annual exchange rates, balances carried forward from one year to the next will have different values. aCGAP estimate. Table IIB-4. Dhan's KVK SHG Program: Estimated Adjusted Income, Costs, and Assets (in US$, as of March 31, 2003) Dhan Foundation KVK Federation Canara (Support (SHG Bank TOTAL Type of actor organization) federation) (Bank) (All actors) Interest income 0 41,800 60,811a 102,611 Fee income 0 2,800 0 2,800 TOTAL INCOME 0 44,600 60,811 105,411 Staff costs 0 8,600 0 8,600 Administrative expenses 0 6,900 15,864a 22,764 Training costs 0 2,700 0 2,700 SHG promotion costs (amortized) 0 0 0 0 SHG federation promotion costs (amortized) 13,920 0 0 13,920 Total Operating Costs 13,920 18,200 15,864 47,984 Estimated cost of funds 0 26,400 37,016 63,416 Adjusted loan loss provisions 0 4,880 5,696 10,576 TOTAL COSTS 13,920 49,480 58,576 121,976 Startup subsidies (amortized) 13,920 0 0 13,920 Ongoing subsidies 0 0 0 0 MARGIN (Income ­ costs) (13,920) (4,880) 2,235 (16,565) Estimated average external loans to SHGs 0 244,000 284,793 528,793 Average SHG deposits 0 0 227,471 227,471 Estimated average assets used for SHG program 0 77,872 0 77,872 EST TOTAL ASSETS MANAGED 0 321,872 512,264 834,136 Note: For all five SHG programs, financial information for the periods ending March 2001, March 2002, March 2003, and March 2004 was converted at Rs. 46.577 to US$1, Rs. 48.733 to US$1, Rs. 47.65 to US$1, and Rs. 44.125 to US$1, respectively. Given the variable annual exchange rates, balances carried forward from one year to the next will have different values. aCGAP estimate. 45 OP 12-III.qxd 8/17/07 3:27 PM Page 46 I I I Acknowledgments grams shared valuable perspectives. Prakash Bakshi, Vijayalakshmi Das, Deep Joshi, Ajit The authors of Part I are grateful to several indi- Kanitkar, Narendranath, Vipin Sharma, Matthew viduals and institutions for valuable contributions Titus, V. Satyamurthi, Sanjay Sinha, and Jayshree to this paper. Elizabeth Littlefield of CGAP iden- Vyas provided intellectual support and guidance. tified the need for it and provided essential guid- Any errors or omissions are the responsibility of ance and support. P. Kotaiah and C. S. Reddy of the authors. APMAS provided extensive comments on the The authors of Part II would like to thank the paper's research and analysis of self-help groups. management of the organizations studied for their Consultant Ajay Tankha researched one of the cooperation, and in particular Mr. Manoharan and SHG programs and, with Jeffrey Ashe, helped Mr. Narayan of Microcredit Foundation of India develop the study's conceptual framework. (now Madura Micro Finance Ltd.), Mr. Ravinder Headquarters and local staff of UNDP SAPAP's Yadav of Oriental Bank of Commerce, Mr. Panagal Mandal Mahila Samakhya, Sakhi Samiti, Sowmithri of Sarvodaya Nanofinance Ltd., and Professional Assistance for Development Action, Mr. K. Narender of Dhan Foundation. Vijay Chitradurga Gramin Bank, and People's Action Mahajan of BASIX and Richard Rosenberg of for National Integration gave generously of their CGAP provided invaluable comments. time. Self-help group members from all the pro- Bibliography Kanitkar, Ajit. 2002. "Exploring Empowerment and Lead- ership at the Grassroots: Social Entrepreneurship in the SHG Movement in India." In Beyond Micro-Credit, Arora, Sukhwinder Singh, and Malcolm Harper. 2005. Thomas Fisher and M. S. Sriram, eds. New Delhi, In- Small Customers, Big Market. Warwickshire, England: dia: Vistaar Publications. ITDG Publishing. Karduck, Stefan, and Hans Dieter Seibel. 2005. "Transac- Bansal, Hema. 2002. "SHG-Bank Linkage Program in In- tion Costs of Self-Help Groups: A Study of dia: An Overview." Journal of Microfinance, 5 (1). NABARD's SHG Banking Program in India." Univer- Christen, Robert. 2005. "Microfinance and Sustainability: sity of Cologne Development Research Center. International Experiences and Lessons for India." Con- Kropp, Erhard W., and B. S. Suran. 2002. "Linking Banks ference paper for NABARD workshop, "Microfinance: and (financial) Self-Help Groups in India--An Assess- Challenges for the Future," New Delhi, 3­6 May. ment." Paper presented at the seminar on SHG-Bank Dunn, Elizabeth, and J. Gordon Arbuckle Jr. 2001. "The Linkage Programme at New Delhi , 25­26 November. Impacts of Microcredit: A Case Study from Peru." US- Mumbai: Micro Credit Innovations Department, Na- AID Office of Microenterprise Development. Washing- tional Bank for Agriculture and Rural Development, ton, D.C.: USAID AIMS Project. Thompson Press. EDA Rural Systems. 2004. "Impact Assessment of Micro- Micro-Credit Ratings International Ltd. 2004. M-CRIL finance: Interim Findings from a National Study of Microfinance Review 2003 revised Feb. 2004)." Gur- MFIs in India." New Delhi, India: EDA Rural Systems. gaon India: Micro-Credit Ratings International Ltd`. Hannover, Wolfgang. 2005. "The Impact of Microfinance MIX Market and the World Bank. 2005. "Performance and Linkage Banking in India on the Millennium Develop- Transparency: A Survey of Microfinance in South ment Goals (MDG), Summary of major reports from Asia." Washington, D.C.: Microfinance Information existing studies. NABARD/GTZ, May. eXchange and the World Bank, December. Harper, Malcolm. 2002a. "Promotion of Self-help Groups Murray, Jessica, and Richard Rosenberg. 2006. "Commu- under the SHG-Bank Linkage Programme in India." nity-Managed Loan Funds: Which Ones Work?" Focus Paper presented at the Seminar on SHG-Bank Linkage Note 36. Washington, D.C.: CGAP. Programme, New Delhi, 25­26 November. MYRADA. 2002. "Impact of Self-help Groups (Group ------. 2002b. "Grameen Bank Groups and Self-help Processes) on the Social/Empowerment Status of Groups; What Are the Differences?" Women Members in Southern India." Paper presented 46 OP 12-III.qxd 8/17/07 3:27 PM Page 47 at the seminar on SHG-Bank Linkage Programme, gramme in India, New Delhi, 25­26 November. Mum- New Delhi, 25­26 November. bai: Micro Credit Innovations Department, National Bank for Agriculture and Rural Development, Thomp- Nair, Ajai. 2005. "Sustainability of Microfinance Self-help son Press. Groups in India: Would Federating Help?" World Bank Policy Working Research Paper 3516. Washington, Sinha, Sanjay. 2003. "Financial Services for Low Income D.C.: South Asia Region, Finance and Private Sector Families: An Appraisal." IIMB Management Review, Development Unit, World Bank. June. Nanda, Y. S. "Significance of Establishing Linkages with Srinivasan, Girija. 1999. "Financial to social capital--Role Self-help Groups and Banks." Mumbai, India: National of Banks." Bank for Agriculture and Rural Development. Srinivasan, R. 2000. "Self-Help Groups as Financial Insti- Nandhini, Usha S., P. David Jawahar, and M. Shivachan- tutions: Policy Implications Using a Financial Model." dran. 2004. "Analysis of monthly cash flows of Self Journal of Microfinance 5 (1). Help Groups." Tankha, Ajay. 2002. "Self-help Groups as Financial Inter- Reddy, C. S., and L. B. Prakash. 2003. "Status of SHG Fed- mediaries in India: Cost of Promotion, Sustainability, erations in Andhra Pradesh." Paper presented at SHG and Impact." The Netherlands: Interchurch Organiza- Federation Workshop organized by NIPCCD and co- tion for Development Co-operation and ordinated by Sa-Dhan, 24­25 March. Memisa­Mensen in Nood-Vastenaktie (Cordaid).. Satish, P. 2001. "Institutional Alternatives for the Promo- Thorat, Y. S. P., and Nitin Bhatt. 2000. "India's Regional tion of Microfinance: Self-help Groups in India." Jour- Rural Banks: The Institutional Dimension of Re- nal of Microfinance 3(2): 49­74. forms." Journal of Microfinance 3 (1). Siebel, Hans Dieter. 2001. "SHG Banking: A Financial Titus, Mathew. 2002. "Costs in Micro-finance: What Do Technology for Reaching Marginal Areas and the Very Urban Self-Help Groups Tell Us?" In Beyond Micro- Poor; NABARD's Program of Promoting Local Finan- Credit. Thomas Fisher and M. S. Sriram, eds. New cial Intermediaries Owned and Managed by the Rural Delhi, India: Vistaar Publications. Poor in India." IFAD (International Fund for Agricul- Wilson, Kim. 2002. "The Role of the Self-help Group-Bank tural Development), Rural Finance Working Paper No. Linkage Programme in Preventing Rural Emergencies 167910. Rome, Italy: IFAD. in India." Paper presented at the seminar on SHG- Siebel, Hans Dieter, and R. Dave Harishkumar. 2002. bank Linkage Programme, New Delhi, 25­26 Novem- "Commercial Aspects of Self-Help Group Banking in ber. Mumbai: Micro Credit Innovations Department, India: A Study of Bank Transaction Costs." Paper pre- National Bank for Agriculture and Rural Development, sented at a seminar on the SHG Bank Linkage Pro- Thompson Press. 47 OP 12-III.qxd 8/17/07 3:27 PM Page 48 OccasionalPaper No. 12 Please share this Occasional Paper with your colleagues or request extra copies of this paper or others in this series. CGAP welcomes your comments on this paper. All CGAP publications are available on the CGAP Web site at www.cgap.org. CGAP 1818 H Street, NW MSN P3 - 300 Washington, DC 20433 USA Tel: 202-473-9594 Fax: 202-522-3744 Email: cgap@worldbank.org 48