Tajikistan: Microfinance Systems Development Program

Validation Report Reference Number: PCV: TAJ 2010-55 Project Number: 33040 Loan Number: 2000; 2001 November 2010 Tajikistan: Microfinance Systems Development Program Independent Evaluation Department

ABBREVIATIONS ADB Asian Development Bank ADF Asian Development Fund LIBOR London interbank offered rate MDO microcredit deposit organization MFI microfinance institution MPSC Microfinance Policy Steering Committee MSDP Microfinance Systems Development Program NBT National Bank of Tajikistan NGO nongovernment organization PCR program completion report PMU project management unit RRP report and recommendation of the President SDR special drawing rights TA technical assistance TWG Technical Working Group NOTE In this report, $ refers to US dollars. Key Words adb, asian development bank, microfinance, program completion report, systems development, tajikistan, validation Director Team leader Team members R. B. Adhikari, Independent Evaluation Division 1 (IED1), Independent Evaluation Department (IED) H. Feig, Principal Evaluation Specialist, IED1, IED N. Gamo, Evaluation Officer, IED1, IED S. Labayen, Senior Operations Evaluation Assistant, IED1, IED In preparing any evaluation report, or by making any designation of or reference to a particular territory or geographic area in this document, the Independent Evaluation Department does not intend to make any judgments as to the legal or other status of any territory or area.

PROGRAM COMPLETION REPORT VALIDATION FORM A. Basic Program Data PCR Validation Date: November 2010 Project Number: 33040 Approved Actual Loan Numbers: TA Number: 2000-TAJ, 2001-TAJ 4132-TAJ Program Name: Microfinance Systems Total Project Costs ($): 9,951,282 10,296,614 Development Program Country: Tajikistan Loan/TA ($): 2000-TAJ 4,000,000 4,378,099 2001-TAJ TA equivalent 4,000,000 equivalent 640,000 3,474,667 404,520 Sector(s): Finance Total Cofinancing ($): ADB Financing ($): ADF: 8,000,000 Borrower ($): 389,173 262,606 equivalent OCR: 0 922,109 1,776,722 Other External Financing: MFIs/MFI clients ($): Approval Date: 26 June 2003 Effective Date: 2000-TAJ 2001-TAJ Signing Date: 10 November 2003 Closing Date: 2000-TAJ Program Officers: Validator: Quality Control Reviewer/Peer Reviewer: Name: R. T. Moyes B. Wilkinson A. Chi R. Narasimham N. P. Knoll Tetsu Ito, Consultant Henrike Feig, Principal Evaluation Specialist, IED1 2001-TAJ Location (HQ or RM): HQ HQ HQ HQ HQ Director: 8 February 2004 8 February 2004 7 February 2006 30 June 2008 From 2000 2003 2006 2007 2008 21 July 2004 21 July 2004 12 December 2006 17 October 2008 To 2003 2006 2007 2008 2009 Ramesh B. Adhikari, IED1 ADB = Asian Development Bank, ADF = Asian Development Fund, HQ = headquarters, IED 1 = Independent Evaluation Division 1, MFI = microfinance institution, OCR = ordinary capital resources, PCR = program completion report, RM = resident mission, TA = technical assistance, TAJ = Tajikistan.

2 B. Program Description (Summarized from the Report and Recommendation of the President 1 ) 1. Rationale Tajikistan has seen the emergence of a highly active microfinance sector. By the end of 2002, around 17 nongovernment organizations (NGOs) were carrying out lending activities, which were generally funded through international grant assistance. They developed a substantial level of trust with their clients, registered a high percentage of loan repayment, and were developing approaches to enhance their sustainability, but required support for the transformation to fully licensed and regulated microfinance institutions (MFIs). The policy, legal, and regulatory frameworks needed to be developed so that NGO programs and new MFIs could operate legally and under effective supervision that supported prudential financial intermediation, including deposit mobilization. The Microfinance Systems Development Program (MSDP) was formulated in line with the four pillars of the microfinance development strategy 2 of the Asian Development Bank (ADB) to: (i) create an enabling policy environment; (ii) improve the financial infrastructure, specifically the legal and regulatory environment; (iii) build the institutional strength of MFIs to enhance their sustainability and outreach; and (iv) encourage innovative programs that support the poor. 2. Impact The MSDP goal, or expected impact, was to create a strong, commercially viable microfinance sector. The program framework specified the following non-quantified impact performance targets: (i) increased volume of deposits in the financial sector, (ii) substantial linkages (deposits and loans) between MFIs and commercial banks, and (iii) increased number and delivery of financial institutions. 3. Objectives or Expected Outcomes The MSDP objective, or expected outcome, was to establish a regulated microfinance market where microand small enterprises and poor households have access to financial products provided through sustainable MFIs. The program framework specifies 60,000 households reached by MFIs as the overall outcome performance target, with specific targets for the policy loan (i.e., more NGO microfinance organizations registered; more MFIs licensed and regulated; and donor and NGO activity directed to form MFIs), and the investment loan (i.e., at least two MFIs in total compliance with legal and regulatory requirements; outreach of MFIs increased by 100%; about 75% of borrowers initially poor, 50% of clients female, and 10% of the loan portfolio of each MFI matched by savings deposits; and MFIs achieving or surpassing specific performance indicators developed under MSDP). 4. Components and Outputs The MSDP, which was designed in 2002 and approved in June 2003 as a sector development program, comprised (i) a policy loan (2000-TAJ) for a committed amount of SDR2,934,000 ($4 million equivalent) from the Asian Development Fund (ADF), (ii) an investment loan (2001-TAJ) for the same committed amount also from the ADF, and (iii) a technical assistance (TA) grant (4132-TAJ) of up to $640,000. The completion report of this TA was prepared in 2008. 3 The MSDP included four components: (i) the adoption of policy, regulatory, and institutional actions to develop a microfinance industry, (ii) capacity building of the National Bank of Tajikistan (NBT) to regulate and supervise MFIs, (iii) capacity building of MFIs, and (iv) transformation of microcredit programs operated by NGOs into formally licensed and regulated MFIs. 1 ADB. 2003. Report and Recommendation of the President on Proposed Loans and Technical Assistance Grant to Tajikistan for the Microfinance Systems Development Program. Manila. 2 ADB. 2000. Finance for the Poor: Microfinance Development Strategy. Manila. 3 ADB. 2008. Technical Assistance Completion Report on Institutional Strengthening of Microfinance Systems. Manila.

3 The policy loan aimed to create a legal, regulatory, and supervisory environment that supports the development of a strong, commercially viable microfinance sector. The policy loan was to be released in two tranches: the first, $1.5 million equivalent, was to be released upon loan effectiveness; and the second, $2.5 million equivalent, was to be released within 24 months of the effective date. The policy matrix comprised 14 policy measures grouped under the following five expected outputs: (i) formulation of a policy and strategy to develop microfinance services; (ii) establishment of appropriate legislation for MFIs; (iii) establishment of a tax environment supportive of the establishment and development of MFIs; (iv) development of effective regulations and enforcement mechanisms; and (v) development of an institutional structure for microfinance, and licensing of MFIs. Of the 14 policy measures, two were conditions for loan effectiveness: (i) establishment of a Microfinance Policy Steering Committee (MPSC), and (ii) establishment of a technical working group (TWG). Four were conditions for second tranche release: (i) ratification of a law on MFIs; (ii) amendment to the tax code to provide a fair and consistent tax regime appropriate to MFIs, approved by Parliament; (iii) development of prudential and nonprudential regulations appropriate for diverse categories of MFIs; and (iv) issuance of operating licenses to two deposit-taking MFIs. ADB was to consider the provision of an incentive tranche of $1 million equivalent to be released up to 6 months before the second tranche if two of the policy measures, i.e., ratification of the microfinance law and approval of regulations for MFIs, were taken. The investment loan aimed to expand the outreach and sustainability of MFIs that serve micro- and small enterprises and poor households by transforming unregulated credit programs to licensed, regulated financial institutions. The investment loan financed (i) consulting services and equipment to assist with the drafting of MFI regulations, the transformation of NGOs into MFIs through capacity development support, including the development of savings products, the establishment of microfinance services in rural areas by commercial banks, comprehensive capacity development support for the National Bank of Tajikistan (NBT), microfinance regulation and supervision, and the installation of an accounting and management information system for the project management unit (PMU) and other logistical support; and (ii) financial support for the establishment of a pilot liquidity guarantee fund for MFIs. The loan had 11 expected outputs with specified, although not always quantified, indicators: (i) a legal environment favorable to MFIs; (ii) MFI regulations established and enforced; (iii) community awareness of all MFI laws, regulations, and activities; (iv) strong, well-managed MFIs; (v) diversified savings and loan products and other services that meet customer needs developed; (vi) pilot liquidity guarantee fund established; (vii) microfinance supervision capacity in the NBT established; (viii) supervision methods and procedures implemented by trained NBT staff; (ix) capable domestic staff operating MFIs; (x) development of management and human resources in MFIs; and (xi) effective project management. The investment loan of $4 million was to finance 75.3% of the total project cost of $5.31 million equivalent ($1.82 million in foreign-exchange costs and the remaining in local currency costs), of which $2.6 million would finance (i) a credit line, relent through commercial banks, which act as apex institutions to MFIs for onlending to microenterprises; and (ii) a liquidity guarantee fund (50% of the cost paid by ADB) to ensure that sufficient funds are available for unexpected customer deposit withdrawals. The remaining $1.4 million would finance (i) vehicles, equipment, consultants, and overseas and domestic training for the NBT; (ii) vehicles, equipment and furniture, consultants, and operational costs for the PMU; and (iii) the interest charge during implementation. A small-scale TA project 4 (3555-TAJ) helped develop policies and legal and regulatory measures, virtually all of which were later adopted under the program. The program s advisory TA (4132-TAJ) aimed to (i) formulate MFI financial models that fit microfinance needs and NGO capabilities; (ii) work closely with each MFI and candidate MFI in planning, evaluation of activities, and development of staff benefit packages linked to staff performance; (iii) work closely with the NBT MFI staff supervision section in refining off-site monitoring and appraisal of MFIs; (iv) work with NBT staff in developing efficient and effective procedures for on-site inspections; (v) explore possible linkages between commercial banks and MFIs in the joint development of products, placing of deposits, credit lines, and equity investment; (vi) prepare a baseline 4 ADB. 2000. Technical Assistance to Tajikistan for Support to Rural Financial Systems Development. Manila.

4 survey of MFI customers and a system for monitoring benefits received by households during the project; and (vii) obtain market information from households about their demands for a range of financial services. C. Evaluation of Design and Implementation (PCR Assessment and Validation) 1. Relevance of Design and Formulation The program completion report 5 (PCR) notes that the MSDP s design was consistent with the strategies of the government and ADB regarding poverty reduction and microfinance. This validation agrees with the above PCR observation, and notes that lessons from ADB s experience in rural finance in transitional economies 6 were incorporated into the design of the MSDP. All four of the MSDP components mutually reinforced each other and helped achieve MSDP outputs and outcomes. However, the report and recommendation of the President (RRP) did not specify details of the credit line or the proposed liquidity guarantee fund, which might indicate that these project components were not fully designed at appraisal. Implementation arrangements for the credit line were subsequently changed and the liquidity guarantee fund was never set up. 2. Program Outputs The PCR concludes that the MSDP achieved most of the expected outputs, noting that the government had enacted the microfinance law; amended tax legislation to support the establishment and development of MFIs; adopted prudential and nonprudential regulations for MFIs; significantly improved the NBT s capacity to regulate and supervise MFIs; strengthened the institutional capacity of MFIs in Tajikistan through 89 training sessions on nine topics, including best practices for MFI operations, the development of new products, financial reporting, gender issues, registration of pledge agreements, and taxation, which were attended by 414 participants from 39 organizations; and transformed more than 50 NGOs into formally licensed and regulated MFIs. All of these outputs resulted in a thriving microfinance industry in Tajikistan. 7 The output that was not met, according to the PCR, was the establishment of a pilot liquidity guarantee fund. 8 The PCR notes that, of the $3.47 million actually disbursed under the investment loan, $2.024 million was used to support a credit line to MFIs. The government relent this fund to the NBT, through a subsidiary loan agreement, at an annual interest rate of 1.5%. In turn, the NBT onlent this fund to two commercial banks at the 6-month London interbank offered rate (LIBOR), with interest payable semiannually. 9 The two banks then onlent the funds to 14 MFIs at 6-month LIBOR plus 4.5%, also with interest payable semiannually. The MFIs further onlent the funds received from the two banks to 5,419 MFI clients at rates of 2.5 3.0% per month, for periods of between 1 and 24 months. The loans, which were made between February 2007 5 ADB. 2009. Program Completion Report: Microfinance Systems Development Program (Tajikistan). Manila. 6 Such lessons highlight the importance of (i) following a financial systems development approach to provide an enabling environment for commercial microfinance, (ii) focusing on poor clients to reduce poverty, (iii) working within the capacity of participating financial institutions to minimize distraction from their main functions and to avoid raising their administrative costs, and (iv) maximizing the impact of capacity building by providing TA over a period of time and not as an isolated instance (RRP, para. 32). 7 The PCR further notes that at the end of 2008, the industry s aggregate gross loan portfolio totaled more than $79 million, with aggregate capital of almost $37 million. Profitability has been good, enabling MFIs to use their retained earnings to expand their outreach. For example, the members of the Association of Microfinance Organizations of Tajikistan had an average operational self-sufficiency ratio of 153.3% in 2007 and 154.5% in 2008. Performance has been sufficiently impressive to enable some MFIs to dramatically expand their loan portfolios using funds borrowed from foreign lenders. 8 The PCR does not explain why a pilot liquidity guarantee fund was not created. 9 Semi-annual amortization payments on the loans from the NBT to the two banks, and on the loans from the two banks to the MFIs, began on 31 December 2009 and will end on 30 June 2016.

5 and June 2008, totaled somoni (TJS) 10.371 million. 10 The loans supported activities in the following sectors: trade, 42%; production, 32%; agriculture, 15%; and other, 11%. More than 80% of the loans were made in rural areas, and (although not all loan installments have become due yet) the performance of loans has been very good, with more than 97% of the aggregate loan principal already repaid. This validation finds that the PCR fairly discusses the (i) status of key MSDP outputs at the time of the PCR assessment, and (ii) reasons for deviations from planned outputs, although more analysis could have been provided of the quality of approved microfinance regulations and the relevance of proposed changes in the taxation of MFIs. 11 3. Program Cost, Disbursements, Borrower Contribution, and Conformance to Schedule The PCR fairly and completely discusses the program cost, borrower contribution, disbursements, and conformance to schedule as summarized in the ensuing two paragraphs. The actual total program cost was $10.30 million $8.62 million provided by ADB ($4.39 million through the policy loan, $3.47 million through the investment loan, and $405,000 through the TA grant), $263,000 by the government, $702,000 by MFIs, and $1,075,000 by MFI clients. Compared with appraisal projections, (i) the amount disbursed under the policy loan was higher because of the appreciation of the SDR against the dollar; (ii) the amount disbursed under the investment loan was lower, as expenses in almost all categories were lower than expected; (iii) government expenditures were lower, because less equipment was purchased than anticipated; and (iv) the amount attributable to MFI clients, which consisted of savings deposited in MFIs, was much higher because more-than-expected savings were collected. The PCR notes that expenses under the investment loan were lower than projected because, among other things, (i) MFIs refused to borrow from the Government to acquire equipment, software, and furniture; (ii) the staff of the PMU and the individuals providing training to MFIs were local rather than international hires contracted through a firm; (iii) no liquidity guarantee fund was established; (iv) delays in establishing the PMU resulted in PMU expenses being incurred during only part of the program period, and (v) more than 85% of the loan was disbursed in the last 2 years of program implementation, resulting in low capitalized interest charges. The initial tranche of the policy loan was disbursed in April 2005. The incentive tranche was disbursed in November 2005 after the two conditions for its release were fulfilled, and the final tranche was disbursed in December 2006, about 5 months behind schedule, after the fulfillment of the remaining two conditions. Disbursement under the investment loan began in 2005, one year behind schedule, and ended in 2008, as scheduled. The original disbursement schedule for the policy loan, which contemplated disbursements of the three tranches in 2003, 2004, and 2005, was overly optimistic because it did not anticipate (i) the delay in loan effectiveness, 12 (ii) the extended time required for Parliament to enact the microfinance law, and (iii) the 1-year interval between the passage of the microfinance law and the approval of regulations implementing the law. Disbursements of the investment loan were delayed because the establishment of the PMU was also delayed. 13 10 The PCR gives the following assessment of the foreign exchange rate risk associated with MFI loans: as required by law, the loans were denominated in somoni. Typically, the loans were not indexed to the US dollar, so the MFI retained the exchange rate risk that had been transferred to it. However, if the loans were indexed to the dollar, the exchange rate risk was transferred to the clients of the MFIs. 11 The progress report for second tranche release noted that all but one policy condition included in the policy matrix had been complied with. The one remaining condition, i.e., the provision of a letter of instruction to the NBT clarifying the tax level for each category of MFIs, was considered as substantially complied with. The progress report commented on this policy measure as follows: although the letter of instruction was not issued, following Parliament s approval of the tax code revision in December 2004, the tax level for each category of MFIs has been discussed and clarified in meetings with the Ministry of State Revenue and Duties, NBT, and MFIs. 12 The policy loan s effectiveness was delayed because it was conditional on the establishment of the MPSC, the TWG, and the PMU, which took longer than expected. 13 The establishment of the PMU was delayed mainly because (i) the government had initially taken steps to establish the PMU as a separate, autonomous legal entity rather than as a part of the NBT (the PMU was eventually set up

6 4. Implementation Arrangements, Conditions and Covenants, Related Technical Assistance, and Procurement and Consultant Performance The PCR fairly and completely discusses the subject matters as summarized below. The NBT, as the executing agency for the MSDP, consulted with the Ministry of Finance, Ministry of Justice, Office of the President, and Ministry of State Revenue and Duties on measures to be implemented. The MSDP also provided for the creation of the MPSC, the TWG, and the PMU. The MPSC and the TWG each of which included representatives from the Ministry of Finance, Ministry of Justice, Ministry of State Revenue and Duties, NBT, NGOs, and the Office of the President were designed to facilitate cooperation between the various represented entities; they were merged, with ADB s approval, in 2006. The PMU, formed in the NBT, was responsible for the day-to-day implementation of the investment loan. The PCR does not highlight any particular issue in these MSDP implementation arrangements. As regards the policy loan, all policy conditions (both second tranche and monitoring conditions) had been complied with before the second tranche release in December 2006. The compliance with reporting and meeting requirements was only partly satisfactory. For example, quarterly reports on the policy matrix implementation were not filed as required under the loan agreement, and the MPSC did not meet regularly. However, these did not negatively impact the work of the NBT and other stakeholders. As regards the investment loan, the PMU submitted regular and timely reports on its activities. However, seven covenants did not meet full compliance. 14 The associated TA successfully delivered six of the seven planned outputs, and the TA completion report rated this TA highly successful. All consultants under both the TA grant and the investment loan performed satisfactorily. Procurement under the investment loan was uneventful, with no problem regarding suppliers. 5. Performance of the Borrower and Executing Agency The PCR rates the performance of the borrower and the NBT satisfactory, citing the NBT s critical role in developing and implementing laws and regulations governing MFIs. Such satisfactory efforts outweighed the negative performance of the borrower and the NBT at the program s onset, achieving loan effectiveness and fulfilling the other prerequisites to disbursement of the policy loan s initial tranche. This validation agrees with this rating and assessment, with a general observation that (i) political support for reforms was secured throughout the program period, and (ii) the executing agency effectively managed the MSDP implementation. 6. Performance of the Asian Development Bank The PCR rates the performance of ADB satisfactory, considering that (i) ADB worked actively with the NBT to address the delay in loan effectiveness, fielding four missions over 4 months; (ii) ADB s disagreements with the NBT over the selection of the project manager 15 and on whether the 2004 changes to the tax code were sufficient (both of which were ultimately resolved in the NBT s favor) delayed the disbursements of the as part of the NBT as originally envisaged), and (ii) it took a while for ADB and the NBT to reach consensus over the selection of a project manager at the PMU (reference to footnote 12 on further details). 14 For example, participating banks and MFIs submitted audited financial statements chronically late; two MFIs did not meet the 35% minimum thresholds for women borrowers; three MFIs each violated one financial covenant; one of the two participating commercial banks did not receive a satisfactory rating from the NBT in its on-site examination reports because of insufficient loan loss reserves, thereby violating one of the loan covenants; and in one year, the same bank violated a covenant requiring it to have a return on assets of at least 2% (the ratio was 0.7%). The investment loan agreement provided that any participating commercial bank or MFI that violated an eligibility requirement (e.g., a financial covenant) for more than 3 months would be automatically excluded from the program, and that it would be required to immediately repay any amounts borrowed under the credit line component. However, ADB and the NBT did not enforce these provisions. 15 The PCR further notes that in several previous projects, ADB had not objected when the government directly appointed the project manager [neither funded under the investment loan nor the TA] instead of utilizing an open recruitment process that excluded existing government staff and included public solicitation and ranking of qualified

7 initial and final tranche of the policy loan; (iii) although ADB provided 3 days of training on loan administration to six NBT employees in December 2003, as well as additional training in 2004 and 2005, it failed to provide any training to the replacement project manager who was hired in July 2005; and (iv) by comparison, ADB worked diligently to negotiate a memorandum of understanding requested by the NBT in April 2008 to facilitate disbursement of the remainder of the credit-line component of the investment loan. This validation agrees with this rating and assessment, with an observation that (i) ADB s experience in rural finance in other transitional economies added value to the MSDP design, and (ii) ADB s supervision of MSDP implementation generally contributed to achieving the results. D. Evaluation of Performance (PCR Assessment and Validation) 1. Relevance The PCR rates the MSDP highly relevant, as (i) it was consistent with the government s development priorities and ADB s country and microfinance sector strategies; (ii) a sector development program was the appropriate modality, as an integrated approach was required to meet the needs of the microfinance sector in Tajikistan; and (iii) stakeholders including the NBT and MFIs were highly committed to achieving program results. This validation supports this rating and the assessment, although the need for the initially contemplated pilot liquidity fund is debatable in hindsight. 2. Effectiveness in Achieving Outcome The PCR rates the MSDP highly effective, as it successfully resulted in the creation of a sound legal and regulatory environment for Tajikistan s microfinance industry in a relatively short period of time, which in turn contributed to a rapidly expanding, profitable microfinance sector that attracted additional investment in the form of wholesale loans and contributions to the equity of newly formed MFIs. All unlicensed entities providing microfinance services ceased their operations or obtained proper approvals to operate. To support this assessment, the PCR provides sets of quantitative information to show the (i) achievement of all policy loan outcome targets, 16 (ii) achievement of most investment loan outcome targets 17 except one, 18 and (iii) good performance of 13 of the 14 MFIs participating in the credit-line component of the investment loan. This validation confirms the PCR rating, while noting the following: The outcome target for domestic savings mobilization was not met, which reflects to some extent the fact that industry growth has been driven by foreign resources, which might have reduced incentives for MFI savings mobilization efforts. A number of MFI performance indicators, including outreach (i.e., the number of active clients), likely declined in 2009, as a result of macroeconomic conditions, which probably also affected some of the 14 MFIs directly supported under the MSDP, particularly those whose operational and financial sustainability ratios were below industry averages in 2008. applicants. ADB initially insisted on an open recruitment process, but eventually acceded to the government s request for direct appointment. 16 Such achievements included the following: (i) between 2005 and 2008, more than 50 NGOs engaged in microlending registered under the law on microfinance organizations, and (ii) the number of licenses and regulated MFIs has successively increased from 0 in 2003 and 2004 to 102 in April 2009. 17 Such achievements included the following: (i) eleven MFIs were taking deposits as of 31 May 2009; (ii) outreach increased by more than 350% from the end of 2004 to the end of 2008 (more than 93,000 borrowers as of the end 2008), with an average loan of $765 (i.e., less than per-capita gross domestic product [$795]) and about 45% of borrowers being women; (iii) loans outstanding totaled about TJS326 million as of 31 May 2009; and (iv) write-offs for 2008 were about 0.34% of gross loan portfolio. 18 The PCR explains the unfulfilled project loan outcome target on the savings-loans ratio of minimum 10% as follows: since the target for loans outstanding was TJS9.0 million, TJS0.9 million in savings deposits would need to be collected for the 10% savings-loans target to be achieved. In fact, TJS11.0 million was deposited in MFIs by the end of May 2009 (thereby exceeding the savings target by twelve times). However, loan growth exceeded the program target by an even greater multiple (36 times), so the saving-loans target was not achieved.

8 3. Efficiency in Achieving Outcome and Outputs The PCR rates the MSDP highly efficient because all four of its components were achieved as a result of loans totaling only $8 million: as a result of a $4-million policy loan, the legal, regulatory, and supervisory framework on microfinance was created; as a result of the $4-million investment loan, the NBT and more than three dozen MFIs received capacity building assistance and more than $3 million was loaned to 14 MFIs, which used the funds to provide micro loans to more than 5,500 clients. The capacity building and arrangements for disbursal of loans to the MFIs were accomplished at a cost of less than $154,000. This validation downgrades the rating to efficient, considering that (i) the implementation of the policy loan experienced some delays i.e., the final tranche was disbursed 5 months later than the original target date; and (ii) other donors contributed to the achievement of project outcomes in the drafting of microfinance legislation and capacity development. 4. Preliminary Assessment of Sustainability The PCR rates the MSDP s sustainability most likely, because MSDP was designed to provide financial services to the poor on a sustainable basis through several types of NBT-licensed MFIs, which generally have good profitability and adequate capital to withstand moderate increases in delinquency. However, the PCR also notes the following risks to the microfinance sector: (i) the potential impact of macroeconomic shocks, partly because some MFIs have significant open foreign currency positions as a result of borrowings denominated in foreign currency, including the loans supported under the MSDP investment loan; (ii) the substantial contribution of foreign remittances to the Tajik economy, which has accounted for 40% 50% of Tajikistan s gross domestic product in recent years, means that some households may find it more difficult to repay their microfinance loans when remittances decline and there is reduced economic activity; (iii) MFIs face additional competition from local banks that are downscaling or that specialize in microfinance loans; and (iv) the insolvency of a microcredit deposit organization (MDO), like the insolvency of any other deposit organization that resulted in losses to depositors, would make it extremely difficult for the remaining MDOs to attract deposits. This validation confirms the PCR rating of most likely sustainable. Despite recent drops in remittance transfer levels, inflation, and a 30% depreciation of the local currency that increased MFIs portfolios at risk to more than 5% on average during 2009, most MFIs appear to have weathered the economic crisis reasonably well, based on available, albeit largely self-reported, financial information. The fact that many of the larger MFIs have had access to longer-term foreign funding probably has had a stabilizing effect on the system, although effective borrowing costs have increased as a result of depreciation. NBT is increasing minimum capital requirements for deposit-taking MFIs to improve industry capitalization. The fact that the industry has grown substantially in recent years and continues to attract capital and other financial resources including from reputable international microfinance organizations is a testament to its perceived institutional capacity, growth potential and likely sustainability. However, NBT capacity for supervising deposit-taking MFIs likely requires further strengthening considering that a 2008 financial sector assessment by the International Monetary Fund identified general weaknesses in NBT capacity for banking supervision. Political commitment to market-based microfinance development in Tajikistan remains high. 5. Impact The PCR does not give a rating for this criterion, but notes that (i) many low-income households now have institutional sources of credit to support the creation and expansion of entrepreneurial activities, 19 (ii) microfinance lending has enabled households to increase their entrepreneurial activity by enabling them to increase their income as well as to diversify its sources; (iii) the emergence of MDOs means that 19 The average size of the MFI loans made under the MSDP credit-line component was $558, compared to Tajikistan s 2008 per-capita gross domestic product of $795. The PMU estimated that approximately 80% of the households receiving a loan under the program were below the poverty line or had no income-generating activity before receiving the loan.

9 households have more institutions in which to deposit their savings; (iv) the growth of the microfinance sector during the program period resulted in the creation of hundreds of reasonably well-paying jobs offered by MFIs, many of which are held by women; 20 (v) the credit-line component of the MSDP required the participating MFIs to submit annual audited financial statements (to facilitate monitoring of MFIs), contributing to the development of the Tajik audit industry, which previously had only limited experience in auditing financial institutions; (vi) the MSDP credit-line component provided local commercial banks with valuable experience in underwriting and administering loans to Tajik MFIs, an activity previously dominated by foreign institutions; and (vii) the MSDP credit-line component provided 11 MFIs with their first experience in borrowing on commercial terms, which is considered an important step in an MFI s institutional development. In addition, the updated policy matrix provides the following information on the MSDP impact performance targets between December 2004 and April 2009: (i) deposits increased from TJS292 million to TJS1,913 million; (ii) the number of financial institutions increased from 17 to 122, and the different types increased from two to five; (iii) the number of credit societies increased from five to seven; (iv) the number of MDOs increased from 0 to 18; (v) the number of microlending organizations increased from 0 to 42; and (vi) the number of microlending funds increased from 0 to 42. This validation fully supports the PCR assessment, and on this basis rates the MSDP impact significant. E. Overall Assessment, Lessons, and Recommendations (Validation of PCR Assessment) 1. Overall Assessment The PCR rated the MSDP highly successful, reflecting the weighted average of the individual ratings for four core criteria; relevance (20%), effectiveness (30%), efficiency (30%), and sustainability (20%). Despite a downgraded efficiency rating, this validation arrives at a highly successful rating using the same weight used to aggregate the core criteria. The MSDP was highly relevant, highly effective, efficient, most likely sustainable, and has had significant impact. 2. Lessons The PCR identifies the following lessons from MSDP: (i) interagency committees may be unnecessary based on the successful achievements of outputs even though committees were not created on time and met only twice; (ii) enacting legislation and adopting regulations frequently takes more time than anticipated; and (iii) inconsistent policies result in confusion and program delays as exemplified by the experience in recruitment of the project manager (footnote 13). This validation agrees with all of the lessons highlighted by the PCR. The MSDP experience also underpins the need for adequate advisory services to help prepare required legal and regulatory changes to speed up program implementation. The program also shows the benefits of integrated approaches to microfinance development support that promote capacity development of MFIs in conjunction with necessary reforms in the underlying policy environment. 3. Recommendations This validation agrees with the PCR recommendations that ADB should (i) consider providing targeted assistance in the form of TA and funding for specific classes of MFIs (e.g., mid-size organizations that have not reached sufficient scale to attract commercial lenders), and for specific initiatives of MFIs (such as new products) to further promote MSDP performance and sustainability; and (ii) continue to monitor the legal and regulatory environment for microfinance in Tajikistan. 20 The PCR notes that, in 2007, about 65% of the employees of the 39 largest MFIs were women, and women served as heads of three of the five largest MFIs.

10 F. Monitoring and Evaluation Design, Implementation, and Utilization (PCR Assessment and Validation) The MSDP program framework specified the performance indicators and targets, which accurately reflected its impact, outcome, and outputs, and thus enabled a robust assessment of the program results. The PCR collected most performance indicators except the number of households reached by MFIs (instead, the PCR collected the number of MFI clients), which was the MSDP outcome indicator. Most banks and commercially oriented financial institutions typically lend money to (and take deposits from) individuals and/or firms, not households. Most MFIs probably did not collect the number of households that they reached and thus the aggregate of such data might not be readily available. A data constraint of this kind should have been considered when the program framework was prepared. The credit-line component of the MSDP investment loan set the (i) eligibility criteria for participating commercial banks and MFIs, and (ii) performance monitoring indicators for the participating MFIs. However, the eligibility criteria were not strictly enforced by NBT and ADB (footnote 14), which suggests that some of them may have been unduly rigorous. G. Other (e.g. Safeguards, including Governance and Anticorruption Provisions; Fiduciary Aspects; Government Assessment of the Program, as applicable) (PCR Assessment and Validation) The PCR does not discuss any major safeguard and fiduciary issues. H. Ratings PCR IED Review Reason for Disagreement/Comments Relevance Highly relevant Highly relevant Effectiveness in Highly effective Highly effective Achieving Outcome Efficiency in Achieving Outcome and Outputs Highly efficient Efficient There were (minor) implementation delays. Some program achievements were also partly due to efforts by other development partners. Preliminary Most likely Most likely Assessment of Sustainability Borrower and Satisfactory Satisfactory Executing Agency Performance of ADB Satisfactory Satisfactory Impact Not indicated Significant Overall Assessment Highly successful Highly successful Quality of PCR Satisfactory I. Comments on PCR Quality This validation finds the PCR to be fair and comprehensive, with generally adequate information to support its findings and conclusions. It would have benefited further from the inclusion of an updated program matrix containing a qualitative assessment of the status of approved policy and regulatory changes, as well as capacity improvements in MFIs and the NBT. Also, an explanation of why a pilot liquidity fund was not created as envisaged would have been informative.

11 J. Recommendation for IED Follow-up In view of the PCR quality and findings, it is not necessary to conduct a program performance evaluation in the future. K. Data Sources for Validation RRP, program processing and review documents, Progress Report Release of Incentive Tranche, Progress Report Release of Second Tranche, and the PCR.

REGIONAL DEPARTMENT'S RESPONSE TO THE PROGRAM COMPLETION REPORT VALIDATION REPORT On 22 November 2010, Director, Independent Evaluation Department, Division 1 (IED1), received the following comments from the Financial Sector, Public Management and Trade Division (CWFM), Central and West Asia Department (CWRD). On 4 November and 21 November 2010, the Financial Sector, Public Management and Trade Division of the Central and West Asia Department (CWRD) submitted comments to the Independent Evaluation Department on drafts of the program completion report validation report. CWRD is appreciative that almost all of comments were incorporated in the final draft.