ASIAN DEVELOPMENT BANK PCR:THA 31058

Transcription

1 ASIAN DEVELOPMENT BANK PCR:THA PROJECT COMPLETION REPORT ON THE RURAL ENTERPRISE CREDIT PROJECT (Loan 1540-THA) IN THAILAND December 2001

2 CURRENCY EQUIVALENTS Currency Unit Baht (B) At Appraisal (July 1997) At Project Completion Review (September 2001) B1.00 = $ $ $1.00 = B31.56 B44.17 ABBREVIATIONS ADB Asian Development Bank BAAC Bank for Agriculture and Agricultural Cooperatives BDO business development officer BOT Bank of Thailand DMP debt moratorium program MIS management information system MLR minimum lending rate MOF Ministry of Finance p.a. per annum PCR Project Completion Report RECP Rural Enterprise Credit Project SFCP Small Farmer Credit Project TA technical assistance NOTES (i) (ii) The fiscal year (FY) of the Bank for Agriculture and Agricultural Cooperatives ends on 31 March. FY before a calendar year denotes the year in which the fiscal year ends. In this report, "$" refers to US dollars.

3 CONTENTS Page BASIC DATA MAP ii v I. BACKGROUND 1 A. History 1 B. Scope of Operations 1 C. Relationship with the Asian Development Bank and Other Lenders 1 D. Rationale for the Asian Development Bank Loan 2 II. IMPLEMENTATION 2 A. Lending Policies 3 B. Characteristics of Subloans 4 C. Implementation and Internal Operations of Subprojects 4 D. Operational Performance of the Bank for Agriculture and Agricultural Cooperatives 6 E. Financial Performance of the Bank for Agriculture and Agricultural Cooperatives 9 F. Financial Statements and Ratios 10 G. Covenants 11 H. Performance of the Asian Development Bank 11 III. TECHNICAL ASSISTANCE 11 IV. EVALUATION 12 A. Loan Appraisal 12 B. Implementation 13 C. Project Impact 13 V. CONCLUSIONS AND RECOMMENDATIONS 14 A. Conclusions 14 B. Recommendations 14 C. Lessons Learned 15 APPENDIXES 16

4 BASIC DATA A. Loan Identification 1. Country Thailand 2. Loan Number 1540-THA 3. Project Title Rural Enterprise Credit Project 4. Borrower Bank for Agriculture and Agricultural Cooperatives 5. Executing Agency Bank for Agriculture and Agricultural Cooperatives 6. Amount of Loan Original: $200,000, Revised: $93,958, (net of cancellation) 7. PCR Number PCR: THA 671 B. Loan Data 1. Appraisal - Date Started 10 July Date Completed 20 July Loan Negotiations - Date Started 13 August Date Completed 15 August Date of Board Approval 18 September Date of Loan Agreement 29 September Date of Loan Effectiveness - In Loan Agreement 28 December Actual 26 November Number of Extensions None 6. Closing Date - In Loan Agreement 26 November Actual 11 October Number of Extensions None 7. Terms of Loan - Interest Rate Variable - Maturity 15 years - Grace Period 4 years - Repayment Terms Semi-annual 8. Terms of Relending - Subloan Amount Not to exceed B5,000,000 - Interest Rate for Subloans Not less than an amount adequate for the borrower to recover its costs of funds, intermediation costs, foreign exchange and interest variation risks, and operating costs

5 iii 9. Disbursements a. Dates - Initial Disbursement 28 November Final Disbursement 17 November Time Interval 11.8 months - Effective Date 26 November Closing Date 11 October Time Interval 22.8 months b. Amount Disbursed $93,958, C. Implementation Data 1. Number of Subloans 65, Distribution of Subloans by Purpose Purpose Number of Subloans Amount (Baht million) 1. Working Capital 43,160 3, Food Processing/Preservation Milling Machine/Plant Grinding/Crushing Machinery Marketing Transportation Assembly/Grading Machinery Selling Fingerlings/Seedlings Building/Warehouse 2, Mixing/Curing Machinery/Plant Woodwork Cottage Industry Weaving/Sewing Machinery Tractor/Land Preparation 2, Harvesting/Threshing Services Transportation Others 12,392 1,040.6 Total 65,193 5, Distribution of Subloans by Region Region Number of Subloans Amount (Baht million) 1. Northern Region 25,434 1, Northeastern Region 17,813 1, Central and Eastern Regions 8, Southern and Western Regions 13,152 1,555.3 Total 65,193 5,558.9

6 iv 4. Average Size of Subloans Range Number of Subloans Aggregate Amount (Baht million) Over Baht 1,000, Over Baht 500, Over Baht 400, Over Baht 300, Over Baht 200,000 1, Over Baht 100,000 17,100 2,290.9 Over Baht 50,000 29,600 2,031.7 Less than Baht 50,000 16, Total 65,193 5,558.9 D. Data on Bank Missions Name of Mission Date No. of Persons No. of Persondays Specialization of Members Loan Reconnaissance Jan a, b Loan Reconnaissance 2 10 Mar-3 Apr b, c, d Loan Fact-Finding May b, c, e, f, g Appraisal Jul b, h, i Review May j Special Project Administration Sep b, k Review Mar b Project Completion Review Sep g, c, k E. Related Loans Loan No. Date of Agreement Amount 664-THA: Agricultural Credit Project 7 May 1984 $50,000, THA: Brackishwater Shrimp Culture Development Project 30 Sep 1986 $11,110, THA: Small Farmer Credit Project 14 Mar 1996 $50,000,000 a-manager, b-senior financial analyst, c-staff consultant/agricultural economist, d-staff consultant/credit specialist, e- senior programs officer, f-environment specialist, g-project economist, h-senior control officer, i-programs officer, j- senior project analyst, k-assistant project analyst

7 v

8 I. BACKGROUND A. History 1. The Bank for Agriculture and Agricultural Cooperatives (BAAC) was founded as a stateowned bank in Thailand in 1966 to provide financial services to farmers and farmers associations. It also supports activities and invests in ventures to develop agricultural knowledge to increase the income and improve the quality of living of farmers and their families. BAAC has emerged as the dominant source for rural financial services with an outreach of 5.12 million clients (91 percent of all farm families) and branches in 76 provinces and 877 districts. B. Scope of Operations 2. The clients, 80 percent of whom are borrowers, are individual farmers (3.55 million) and members of agricultural cooperatives and farmers associations (1.57 million). As a licensed specialized bank, BAAC provides a range of deposit products and can borrow in money markets and issue bonds. BAAC finances agricultural production and agriculture-related as well as nonfarm enterprises owned by farmers, i.e., those with more than half their income derived from agriculture. At the end of fiscal year (FY) 2001, BAAC had mobilized B215 billion of rural savings and had B244 billion of loans outstanding. 3. Under Government directives BAAC extends nonfinancial services to raise agricultural productivity, reduce production costs, and expand marketing opportunities. It is also developing electronic databases for main crops. BAAC has helped establish agricultural marketing cooperatives and supports them for bulk purchasing of inputs and produce marketing. Most of the BAAC clients are members. The Thai Agribusiness Company, part-owned by BAAC, provides collective purchasing and marketing power at the national level. BAAC pledges farm produce and provides loans against these pledges. C. Relationship with the Asian Development Bank and Other Lenders 4. External assistance to BAAC has declined in recent years with BAAC s ability to mobilize local resources. The nature of assistance has also shifted from agricultural financing to portfolio diversification, microfinance, and institutional strengthening. The Agricultural Credit Project, approved in 1983, was the first Asian Development Bank (ADB) loan to BAAC. 1 The ADB-funded Small Farmer Credit Project (SFCP) 2 and the Rural Enterprise Credit Project (RECP) 3 catalyzed enterprise financing. Related technical assistance (TA) developed enterprise financing 4 and risk management skills. 5 An ongoing ADB-funded TA 6 is supporting institutional restructuring. 5. The Japan Bank for International Cooperation and Kreditanstalt für Wiederaufbau supported agricultural financing for small farmers. BAAC received assistance from the United States Agency for International Development for agriculture development and silkworm projects Loan 664-THA: Agricultural Credit Project, for $50 million, approved on 6 December BAAC was the executing agency for Loan 787-THA: Brackishwater Shrimp Culture Development Project, for $11 million, approved on 11 September Loan 1423-THA: Small Farmer Credit Project, for $50 million, approved on 16 January Loan 1540-THA: Rural Enterprise Credit Project, for $200 million, approved on 18 September TA 2524-THA: Institutional Strengthening of BAAC, for $500,000, approved on 16 January TA 2953-THA: Strengthening Project Loan Appraisal and Risk Management of BAAC, for $600,000, approved on 19 December TA 3355-THA: Restructuring of Specialized Financial Institutions, for $3 million, approved on 21 December 1999.

9 2 The World Bank and the International Fund for Agricultural Development supported medium- and long-term agricultural lending. The World Bank s ongoing TA (para. 41) aims to assist BAAC in institutionalizing the expected revisions in prudential norms for specialized financial institutions. The German Agency for Technical Cooperation is piloting microfinance delivery at two branches. 6. At the end of FY2001, project-related external borrowings (Appendix 1) were only 7.1 percent of the total liabilities and equity. BAAC s operations are predominantly funded by deposits, which constitute 70 percent of the total liabilities and equity, and local borrowings. D. Rationale for the Asian Development Bank Loan 7. The Eighth National Economic and Social Development Plan ( ) recognized that low income in the rural sector arises from low land and labor productivity. The Government, therefore, sought to upgrade agricultural technology, encourage high-value crops and livestock, increase agricultural productivity, and help farmers to diversify into new enterprises. One of the key constraints in promoting rural enterprises was the availability of medium- and long-term loans. Accordingly, the BAAC Act was amended to enable the financing of rural enterprises owned by farm families. 7 The SFCP was the first ADB project in support of BAAC s rural enterprise lending, and its implementation was considered successful. 8 BAAC emerged as the most appropriate agency to promote rural enterprises in view of its (i) country-wide network, (ii) significant outreach, (iii) in-depth knowledge of farm families income patterns, and (iv) ability to adapt products and services to client demand. These provided the rationale to continue ADB support to BAAC for the expansion of its enterprise loan portfolio. 8. Similar to the SFCP in objective, scope, and design, the RECP was expected to build on the momentum by expanding BAAC s enterprise loan portfolio toward creating employment in line with the Government s efforts to improve the living standards of the rural people. A loan of $200 million to BAAC for the RECP was approved by ADB from its ordinary capital resources on 18 September 1997 for onlending to enterprises. Specific objectives of the RECP were to (i) provide term loans to 54,000 rural enterprises, including 20,000 new and 34,000 existing enterprises owned by farm families; and (ii) improve loan processing, supervision, and customer services by BAAC for its rural enterprise clients. An advisory TA (footnote 5) was approved to complement the second objective. The RECP became effective on 26 November 1997 and was due for closing on 26 November 2002, but was closed about three years ahead of schedule on 11 October 1999 after 47 percent utilization. BAAC requested the closure due to weak credit demand, high liquidity, and the higher interest cost of ADB funds. II. IMPLEMENTATION 9. Financial Crisis. The RECP was approved within a month of the International Monetary Fund Stabilization Program that aimed to reestablish fundamental conditions for growth in view of the regional financial crisis (the crisis) that erupted in The RECP implementation occurred when the economy, compared with its pre-crisis level (1996), experienced severe contraction, per capita income declined, and nearly two million urban workers lost their jobs. Public expenditure on social and rural development was slashed. The crisis increased rural 7 8 Rural enterprises include both agriculture-related and nonfarm enterprises. Agriculture-related enterprises include small-scale processing of agriculture produce and byproducts into semifinished and finished goods, production and supply of agricultural inputs, agricultural marketing, and provision of agricultural services. PCR THA: Loan 1423-THA: Small Farmer Credit Project. February 2001.

10 3 poverty and widened the income gap. Rural unemployment doubled. Remigration of family members and decline in remittances to villages stretched the resources of farm families. 10. Coping strategies to weather the crisis became the priority. Repayments to banks were affected, the banking industry experienced credit contraction, and the credit demand shifted to short-term loans whereas the RECP emphasized term loans. Cognizant that financial sector weaknesses aggravated the crisis, BAAC preferred to limit its term loan exposure and obtained ADB s consent to provide short-term and working capital loans under the RECP. In view of the rising rural poverty across the country, the initial project area of the 44 poorest provinces was extended to cover all provinces except for Bangkok and five surrounding provinces. The crisis and its response affected lending policies, subloan characteristics, and repayments. A. Lending Policies 11. BAAC provides short-, medium-, and long-term loans. Short-term loans have maturities up to 1 year and are provided mainly for agricultural production, farm produce pledges, and cash credit facilities. Medium-term loans are those with maturities between 1 and 5 years, and are primarily for purchasing machinery and livestock. Long-term loans have maturities up to 20 years, and are generally for tree crops, enterprises, and to refinance old debts. No minimum loan amount requirement applies. BAAC can lend secured loans up to B15 million to individuals. Loans above this limit must be approved by BAAC s board. BAAC provides loans to individual farmers (90 percent of the portfolio) and wholesale loans to agricultural cooperatives and farmers groups. Appraisal of individual loans considers collateral, cash flow, rate of return, and repayment record. Wholesale loans are based on the financial and institutional status of the borrowing entity. Full collateral is required for individual loans, but loans with joint liability are also available to farmers. Delinquency is managed through individual as well as group-based monitoring. 12. Loans are further classified into normal lending, special project lending, and governmentsponsored projects. Normal loans originate from internal procedures. Special project loans are either BAAC s own initiative, such as to develop a new product type, or in pursuance of government policies. For loans under government-sponsored projects (about 12 percent of the portfolio at the end of FY2001), BAAC receives partial or full refinance facility and a service fee. Most of these loans are meant to rehabilitate lost production. Specific borrower selection criteria and lending procedures apply for special projects and government-sponsored projects. 13. The lending interest rate is linked to client classification based on repayment record. Since February 2001, the prime (AAA) clients pay the minimum lending rate (MLR) of 8 percent per annum (p.a.). For other categories the rates are: MLR+1 percent p.a. for very good (AA); MLR+2 for good (A); MLR+3 for the general category i.e., those with overdue or new clients; and MLR+3+3 for defaults without reasonable cause. Defaults due to natural calamities do not affect the classification. For agricultural cooperatives the MLR is 6 percent p.a. BAAC lowered the MLR twice during the RECP to maintain parity with the interest rate in the banking industry. For subloans less than B60,000 a 1 percent service charge was applied under the RECP. 14. Debt Relief. The Government announced a three-year debt moratorium program (DMP), effective 1 April 2001, ostensibly to ameliorate rural indebtedness due to the crisis. The DMP covers nearly half the BAAC clients and one third of the loan portfolio. All farmers with less than B100,000 in loans outstanding are eligible for (i) Option 1: suspension of principal and interest repayment for three years, or (ii) Option 2: reduction of 3 percentage points in the lending interest rate for three years. Out of the 2.37 million eligible farmers, 2.25 million (95 percent) joined the DMP, and the rest will join soon. Farmers selecting Option 1 are ineligible for BAAC

11 4 loans for three years. Those under Option 2 are eligible for new loans and credit upgrades that will further reduce their interest costs. Nearly one sixth can win free insurance cover up to B100,000. Due to these incentives, 49 percent of eligible farmers chose Option 2. An additional one percentage point interest up to B50,000 of savings has been offered under both the options. The Government will compensate BAAC for the loss of interest income due to the DMP. B. Characteristics of Subloans 15. Number and Size. It was expected that 54,000 subloans would be disbursed with an average of B257,140 per subloan. At the close of the RECP, B5.56 billion had been disbursed to 65,193 subloans with an average of B85,270 per subloan. Out of this, 17,777 subloans with an outstanding amount of B1.67 billion, or 30 percent of the disbursements, are current (Appendix 2: Tables A2.1 and A2.2). The number of subloans was 21 percent higher while the average amount was one third of the appraisal estimate and the average subloan size under the SFCP. 9 The reason for the smaller than expected average subloan was the preference for shortterm and smaller loans during the crisis. Short-term subloans constituted 66 percent of all subloans. As specified in the appraisal report, BAAC and subborrowers contributed 55 percent of the subproject cost, and nearly 61 percent, or 39,767, of subloans were to BAAC clients with existing enterprises and the rest to clients with new enterprises. 16. Geographical and Sector Distribution. While the RECP had no region based disbursement allocation, about 65 percent of the subloans were in the relatively poorer North and Northeast regions (Appendix 2), 10 as the initial project area covered the 44 poorest provinces. Subloans have been classified into 15 enterprise types and grouped into four broad categories: small manufacturing (29,140), services (15,780), marketing (13,952), and food processing (6,321). Small manufacturing subloans include furniture, weaving, and sewing. Most of the subloans classified as warehousing, building, assembly, grading, and others are related to the marketing of processed farm produce. Services subloans include tractor, transport, harvesting, and threshing. 17. Income Distribution. The average annual net income of the subborrowers was B140,000 p.a., out of which B79,000 or 56 percent was from farming. 11 This is considerably lower than B250,000 for the SFCP subborrowers. The average land holding size of the subborrowers is about 3.3 hectares which is below the national average of 4.0 hectares. The average age of subborrowers was 41 years. The lower income and land holding of the subborrowers and the smaller subloan size compared to those under the SFCP indicates outreach aimed more towards smaller farmers under the RECP. C. Implementation and Internal Operations of Subprojects 18. The Mission visited 50 subprojects, reviewed subproject files covered by previous BAAC studies (footnote 11), and discussed subproject implementation at BAAC branches, provincial Comparison with the SFCP and references to the SFCP Project Completion Report (footnote 8) have been made, as the RECP is essentially a follow-on project. Aspects that have not changed since the SFCP Project Completion Review Mission, have, in some instances, been maintained to present a self-contained document. Regional poverty incidence Northeastern (37.5 percent), Northern (32.2 percent), Southern (21.5 percent), Central Plains (16 percent), and Bangkok Metropolitan Region (3.4 percent). (i) BAAC Project Completion Report of ADB Loan 1540-THA: Rural Enterprise Credit Project. Bangkok. Sample size: 215 subprojects; (ii) Thailand Development Research Institute Foundation An Evaluation of BAAC Rural Enterprise Project. Bangkok. Sample size: 94; and (iii) BAAC. March Annual Evaluation Report of ADB Loan 1540-THA: Rural Enterprise Credit Project. Bangkok. Sample size: 94.

12 5 offices, and the head office. The effect of the crisis varied with subproject type. While subprojects relating to construction (e.g., brick making) were adversely affected, small food processing enterprises were relatively less affected. Those procuring raw materials from the immediate vicinity and selling in niche markets fared well. Most of the subprojects, though, operated below their rated capacity due to depressed demand. Nevertheless, the weighted financial internal rate of return and economic internal rate of return of subprojects 12 were 17 and 21 percent, respectively; and for successful subprojects, 26 percent and 33 percent. Representative subproject profiles are in Appendix Subprojects are essentially family-operated units with semiskilled and unskilled labor inputs. During the crisis the ability to adjust to price, cost, and demand was the key success variable rather than the quality of the machinery or technical competence of the owner manager. Experience and familiarity with the business was, therefore, crucial. Subprojects in general did not experience cost overruns. Delays due to slow completion of legal formalities, civil works, and procurement and installation of machinery were not observed. 20. Successful subprojects were also characterized by (i) strong complementarity with farm production; (ii) business process and technology consistent with subborrower capacity; (iii) subborrower motivation and commitment; (iv) little competition in the area for the produce or services; (v) quality consciousness, especially among food processors; (vi) lower debt service burden due to large equity contribution; and (vii) other income sources to meet transient cash flow problems. Irrespective of the subloan status, subborrowers generally had no plans for increasing output levels in the immediate term. Therefore, short-term subloans to maintain output level to meet debt service and fixed costs were the stated priority for the majority. Unlike during the SFCP, BAAC was more conscious of the working capital needs of the subprojects. 21. The main reasons for subproject failures were weak appraisal and indifference of the subborrowers. BAAC staff were unable to analyze subproject risks, ignored multiple borrowings that compounded the risk, and relied only on collateral coverage. Unsuccessful subprojects that targeted local markets found smaller than expected markets, and those that aimed to sell outside the provinces faced unfavorable product prices. Indifference of the subborrowers stemmed from the expectation of the DMP and the leniency of BAAC in enforcing credit discipline. Other factors include (i) overinvestment in plant and machinery based on unrealistic projections; (ii) inability to position subprojects for a low demand scenario; (iii) little, if any, experience of subproject activity; and (iv) unforeseen and unavoidable expenses (e.g., illness) that reduced family labor inputs. Despite the problems, most of these subprojects are potentially viable and can be restructured with greater subborrower commitment. Subprojects run a low risk of failure on technical or managerial grounds. 22. About 80 percent of the subprojects are considered successful. On average, the annual income of the subborrowers increased by 45 percent (compared with a without-subproject scenario) instead of the appraisal report estimate of 50 percent, and helped diversify sources of family income. 13 Field observations confirm that most subborrowers, being long-standing clients, received repeat loans from BAAC s own sources. The subprojects created about 75,000 jobs, at an average investment of B110,000 per job, that provide work for the unemployed urban workers who remigrated to rural areas during the crisis and for the partly employed family Based on the analysis during the Project Completion Review Mission and previous studies (footnote 11). The positive impact of the RECP on employment and changes in occupation of the clients were also noted by IES : Impact Evaluation Study on ADB s Rural Credit Assistance in Bangladesh, People s Republic of China, Indonesia, Nepal, Philippines, Sri Lanka and Thailand

13 6 members. The preponderance of short-term working capital subloans to ongoing enterprises led to fewer than the appraisal report estimate of 127,000 jobs being created. 23. Most of the subborrowers live in extended families where the working age members contribute in the household farm and nonfarm activities. The intensity of participation of the family members, particularly women, positively correlates with the quality of management of the subprojects in general. Women were more aware than their spouses of subproject cash flows and viability. As in the case of the SFCP, field visits indicate that women managed 40 percent of the subprojects and actively participated in another 40 percent. Collateral requirements favor loans to be taken in the name of the head of the family, therefore, only 27 percent of the subborrowers were women, compared to the appraisal report expectation of 33 percent. As farm families have more than one source of income, the women often took charge of the ongoing farm or nonfarm activity to enable other family members to take advantage of RECP subloans, and also helped in the management of the new enterprises. For relatively poorer farm families in the North and Northeast regions, this served as a significant risk mitigating factor. Accordingly, some BAAC branches consider the skills and potential contributions of family members while appraising subprojects. D. Operational Performance of the Bank for Agriculture and Agricultural Cooperatives 1. Organization, Management, and Staffing 24. The Government, through the Ministry of Finance (MOF), owns 99.9 percent of the common shares. BAAC is governed by the BAAC Act. The Board of Directors, headed by the finance minister, consists of 15 members. An executive committee headed by the president and a team of five senior executive vice presidents are responsible for day-to-day management. The senior-management team comprises 14 senior vice presidents who head 14 departments and offices (Appendix 4). Each department and office consists of functional divisions headed by a director. Functional responsibility is delegated to provincial offices, and 90 percent of the 12,754 staff are placed in 1,474 branches and field offices. BAAC is audited by the auditor general and supervised by the Bank of Thailand (BOT). BOT prudential norms though are not yet applicable. 25. BAAC has a five-year rolling plan cycle and annual plans linked to annual budgets. The objectives of the corporate plan include demand-based services; good governance; sustainability of BAAC, its clients, and the environment; and strengthening farmers institutions. While well intentioned, these objectives transcend normal banking functions. Institutional stress is apparent, as BAAC strives to balance the competing considerations with little control on product pricing. The systems and procedures that once served BAAC well 14 now struggle to meet multiple challenges due to adverse economic conditions and changing customer demand. 26. BAAC s institutional strengthening efforts are incremental and inadequate due to the insufficiency of internal resources. The key challenges are in the area of management autonomy, credit risk management, asset and liability management, and segregation of financial and nonfinancial services. Strengthening of the management information system (MIS) is the critical underpinning for institutional strengthening. BAAC has an extensive MIS with four systems that deal with different aspects of financial and portfolio management. However, these systems largely stand alone in the absence of a standardized platform and centralized data 14 The World Bank, Washington Rural Finance: Issues, Design, and Concepts. September 1997.

14 7 management core. This diminishes real time and relational database capabilities and restricts analytical reporting. Client information at branches remains underutilized. 27. The Government s influence on BAAC s operations is inevitable. While not a negative factor in itself, lately the distinction between government ownership and management has weakened. Government directives largely ignore BAAC s capacity to implement them without undermining its sustainability and image as a bank. Having set the mandate, the owners need to allow BAAC autonomy to set its strategy, achieve its objectives, and present the true costs. 2. Personnel Administration 28. The Human Resources Management Department and the Human Resources Development Department discharge personnel-related functions. BAAC has simplified job descriptions for all grades and a transparent performance appraisal system to reward staff. BAAC staff are committed and diligent, though their salaries (based on Government pay scales) are lower than in the banking industry. The human resource development plan aims to meet the skill gaps in financial management, diversified banking services, and information technology. 3. Lending Operations a. Processing of Subloans 29. Business development officers (BDOs), formerly credit officers, appraised subloans. Disbursements were made at the branches without any delays. The main criteria for granting subloans were (i) good repayment record, (ii) collateral coverage, and (iii) adequate net income from the subproject to meet repayment obligations. Cash flow analysis for short-term subloans and internal rate of return for term subloans were used in appraisal. Rigorous analysis was observed for term loans involving larger amounts. While BDOs are familiar with the financial aspects of the appraisal, they are unable to assess potential markets for products and services in the absence of any database or linkages with business advisory services. 30. The subloan appraisal was better than during the SFCP. This is because of the wider dissemination of appraisal skills provided under ADB TAs (footnotes 4 and 5) and most of the subborrowers being long-standing clients. BAAC branches were conscious about the working capital needs of the subprojects and made provisions for such loans from their own resources till short-term subloans were permitted under the RECP. However, field visits indicate that branches had difficulty in assessing the impact of subloans on revenue flows of existing enterprises. Branches and BDOs with weaker appraisal skills relied on collateral and past repayment record. This, however, did cause problems in some cases (para. 21). Consolidation and analysis of data on subproject implementation can significantly enhance appraisal quality. b. Maturity of Subloans 31. The appraisal report allowed the subloan maturity to be consistent with subproject implementation and debt service capacity. The maturity of short-term subloans, 66 percent of all subloans, were up to 1 year, and those of the term subloans vary from 3 to 12 years. More than 90 percent of the term subloans mature between 5 and 8 years. Term subloans do not have any grace period, and the first installment is due at the end of the first year. By the end of FY2001, 47,416 subloans, constituting 73 percent of all subloans and 70 percent of the amount disbursed, had been repaid. Of the remaining 17,777 subloans, 9,598, or 54 percent, are being

15 8 repaid on schedule (Table A2.3, Appendix 2). However, 8,179 subloans (46 percent) had arrears, most of which represent delayed repayment in anticipation of the DMP. c. Administration of Subloans 32. Subloans were classified as agriculture-related loans and were appraised, approved, disbursed, and managed at branches. Disbursements were usually made within two weeks of filing the loan application. All the RECP subloans were disbursed over a 16-month period or at a rate of 4,074 per month. Monitoring and administration of subloans were mainstreamed through the respective line departments (Appendix 4). The MIS recorded the subloans with an identification tag that is deactivated after full repayment. Branches were informed on the procurement procedures to be adopted by the subborrowers. Procurements were made by subborrowers either from registered contractors or BAAC-approved outlets. The Fund Management Division of the Financial Management Department ensured that TA inputs were administered to coincide with the disbursements, conduct impact assessment, and inform ADB of implementation issues. d. Environmental Aspects 33. As with the SFCP, support for small-scale enterprises was consistent with local resource endowment. Adverse impact on the physical environment was negligible. The Rural and Environmental Loan Division reporting to a senior vice president mainstreams the environmental concerns. Lists of ineligible enterprises are continually updated and the review of compliance of environmental guidelines is an integral part of the annual audit. The BDOs are environmentally conscious. The Mission noted advice to a subborrower to regularly de-slug the waste ponds to stop the wastewater from overflowing into the public waterway. Subborrowers, though, consider compliance with environmental standards as a burden. A revised environmental guideline developed with ADB assistance were awaiting BAAC approval at the time of the Project Completion Review Mission. e. Project Monitoring 34. Branches maintain computerized subborrower and subproject data. BDOs visit subprojects and update subloan files. Problem subloans are also analyzed at branches in the presence of subborrowers to amicably address issues. BAAC was advised to adequately classify subloans to avoid the difficulties encountered in monitoring the SFCP subloans. The subloans were accordingly classified into four broad categories and 15 enterprise types. As the current MIS operating platform cannot provide real time relational portfolio information, the corporate office and the provincial offices could assess portfolio quality only after a time-lag. Dissemination of such analysis to branches took additional time, as only selected branches are online. BAAC hopes that upgrading the MIS will enhance monitoring and risk management. RECP impact assessment was also constrained by limited sample size and inadequate analysis. 35. The data fields in subloan files do not adequately capture pre-subloan production, sale, and profitability. In addition, many subborrowers do not maintain financial records despite having the ability to do so. This is the key constraint in subloan monitoring. The workload of BDOs, who are expected to monitor clients each, also affected subloan monitoring, which requires more frequent interactions than agricultural production loans. However, the workload factor has been offset, to an extent, because of subloans to ongoing enterprises and improvements in appraisal quality.

16 9 4. Subloan Repayment 36. By March 2001, 73 percent of the subborrowers had fully repaid their loans. The repayment rate for the ongoing subloans at 72 percent of the amount due as of March 2001 is marginally higher than the repayment rate for all BAAC enterprise loans at 70 percent, but lower than the overall repayment rate (farm and nonfarm) at 78 percent (Appendix 5). The delayed repayments by ongoing subloans (paras. 21 and 31) was primarily to take advantage of the impending DMP. Subloans are adequately covered by collateral. E. Financial Performance of the Bank for Agriculture and Agricultural Cooperatives 1. Size and Quality of Loan Portfolio 37. At 31 March 2001 the net loan portfolio of B244 billion (Table A6.1, Appendix 6) was about 3 percent of the aggregate bank credit and half of the rural credit. The portfolio grew at an average annual rate of 9 percent over the last five years. The Northeast region accounts for 31 percent of the portfolio, followed by North (28), South and West (25) and the Central and East (16). Of the portfolio, short-term loans account for 40 percent, medium-term loans 23 percent, and long-term loans 37 percent. The average outstanding loan size is about B60,000. The disbursements during FY2001 at B154 billion was 5.5 percent more than FY BAAC s pre-crisis repayment record has been good at percent of the demand. The arrears have been small and BAAC has been successful in gradually collecting the arrears, leading to very small eventual losses. However, the crisis led to a nearly 10 percent point reduction in the repayment rate. The repayment rate for FY2001 at 78 percent is a marginal increase over FY2000 at 75 percent. The aggregate arrears of B34.4 billion at 31 March 2001, a 1.2 percent decline over FY2000, is 14 percent of the outstanding loans. Of the total arrears, 35 percent was due for up to 1 year, 60 percent for more than 1 year but less than 5 years, and 5 percent for more than 5 years (Appendix 5). Cumulative loan loss provision at B28.8 billion, appropriated from earnings as well as received from MOF, is 84 percent of the arrears. 39. Measures to improve repayment performance initiated prior to the DMP include restructuring of 86,850 loans with genuine problems and waiving of penalty interest on prompt repayment of arrears. Delayed repayments due to the DMP was the single most important factor for the less than satisfactory improvement in repayment rate; other causes include (i) unfavorable economic conditions, (ii) an increase in rural household expenditure due to the return of unemployed urban labor, (iii) volatile commodity prices, and (iv) weak appraisal and monitoring at some branches. While no firm figures are available, it is estimated that 10,000 RECP subborrowers with an aggregate outstanding of B500 million have also joined the DMP. 40. BAAC will receive B6 billion p.a. from MOF to compensate for the loss of interest income due to the DMP. While this covers the immediate financial impact, the burden will be unbearable if the repayment of the principal amount is delayed after the DMP. Expectation of periodic debt relief is most likely to influence repayment behavior, as the DMP has erased the distinction between borrowers with good record (the majority) from willful defaulters and those that have benefited from rescheduling. Intensive monitoring of DMP loans, need-based rehabilitation support, and savings mobilization may mitigate some of the risks for BAAC.

17 10 2. Provisioning for Bad Debts 41. The income recognition and provisioning for bad debts by BAAC remain at variance with the BOT criteria. BAAC provides 10 percent p.a. of the arrears and accrual of interest is recognized up to 360 days of arrears, while the BOT requires commercial banks to stop interest accrual at 90 days of arrears and provision: 20 percent for substandard; 50 percent for loan losses; and 100 percent for write-offs. Notwithstanding BAAC s small eventual loan losses, the need for BAAC to adopt prudential norms that reflect the true portfolio quality is gaining acceptance. MOF and BOT are reviewing the World Bank TA report on Strengthening Supervision of Specialized Financial Institutions (para. 5) that recommends (i) full on-site inspection by 31 December 2002 based on upgraded prudential standards, (ii) training of BAAC and other institutions to facilitate the adoption of the standards, and (iii) the eventual adoption by September Concurrently, portfolio and financial audits of BAAC based on BOT prudential standards for commercial banks and international accounting standards are scheduled under the ongoing ADB TA (footnote 6). Provisioning guidelines for DMP loans are awaited. F. Financial Statements and Ratios BAAC s total assets increased from B207 billion in FY1997 to B309 billion in FY2001 (Table A6.1, Appendix 6). During the same period equity and reserves increased from B12 billion to B22 billion. The deposits at B215 billion constitute 70 percent of the total liabilities and equities, followed by borrowings at 17 percent. Borrowings have successively declined with a corresponding rise in lower cost deposits. 43. While the profits of BAAC nearly doubled in FY2001 (B536 million) compared with FY2000 (Table A6.2, Appendix 6), its profitability is very thin with return on equity of 2.39 percent and return on assets of 0.17 percent. BAAC operates on a very narrow spread (Appendix 7) due to the downward revision of the MLR and recurring foreign exchange rate losses. The MLR affected profitability by not differentiating between loan terms even though the costs of raising long-term resources are higher. The capital adequacy ratio at 31 March 2001, based on BAAC s income recognition and provisioning norms, was 8.45 percent. 44. Nearly 70 percent of the B39 billion of foreign exchange borrowings are unhedged, two thirds of which are in US dollars and the rest in Japanese yen. Large unrealized foreign exchange losses have been incurred since the floating of baht in BAAC has alternated between realizing the losses in the year they are incurred in line with international accounting standards or understating the year s losses by amortizing them. During FY2001, the six-month floating interest on the ADB loan was 6.7 percent and the foreign exchange loss charged to the income account was B1.9 billion. Unless MOF were to pay out the foreign exchange exposure or refinance them into baht, the best BAAC could do is to strengthen foreign exchange monitoring capacity. No foreign exchange borrowings have been made since FY On the one hand BAAC benefits from government guarantees and relaxed reserve, taxation, and prudential norms. On the other hand, it is expected to provide nonfinancial services, such as for marketing and farmer training, and is required to provide loans at interest rates that equate to the MLR in the banking industry, notwithstanding the higher transaction Based on the audited financial statements of BAAC. Application of internationally accepted prudential norms and accounting conventions could alter the financial results. For the RECP, a swap arrangement at the market rate was applied by BOT. However, due to increased risk perception in a floating rate regime, BOT stopped providing such hedging services.

18 11 costs and covariant risks in rural finance. During FY2002 FY2004, BAAC disbursements are expected to be one third less than in FY2001 due to the ineligibility of clients under the DMP (para. 14) for new BAAC loans. The resulting significant income decline during the period will only be partially offset by the cost control measures such as reduction in the rate of salary increase, merger of provincial offices, and recruitment only for replacement vacancies. G. Covenants 46. BAAC did not comply, or did not fully comply, with covenants relating to repayment rate, interest rate on loans, benefit monitoring and evaluation, revision of loan loss provisioning norms, and design and implementation of a financial MIS (Appendix 8). The deliberate withholding of amounts due in anticipation of the DMP resulted in a lower than covenanted repayment rate of 80 percent. BAAC continued to apply an average cost-based pricing structure that did not fully factor in the foreign exchange risks of the ADB loan. As a result, depending upon the client classification (para.13), the RECP subborrowers paid lending interest rates similar to other BAAC clients. Subloan monitoring was affected by the MIS inflexibility, insufficient data analysis, and inadequate small sample size as well as by the analytical rigor of the impact assessment study. Application of stricter prudential standards for BAAC, including for income recognition and loan loss provisioning, is being considered by MOF and BOT under the World Bank TA (paras. 5 and 41). The proposed plan for adopting the revised standards, however, exceeds the RECP duration. The covenant relating to the design and implementation of a financial MIS has been partially complied with. Although the design has been completed and installation commenced in 2001, bankwide implementation is not yet complete, in part due to resource constraints. Time available for compliance was considerably reduced by the loan cancellation (para. 53). H. Performance of the Asian Development Bank 47. No inception mission was fielded because the RECP was similar to SFCP in design (paras. 7 and 8) and BAAC continued as the borrower and the executing agency. ADB undertook three review missions in 1998 and 1999, including one special loan administration mission, during which the impact of the crisis on the RECP, including the slow pace of disbursements, was reviewed. The recommendations included expansion of BAAC s lending to viable entrepreneurs in other provinces and the financing of working capital subloans (para. 10). ADB maintained dialogue with BAAC and concurred with the loan cancellation request in view of factors external to the RECP, primarily the lower cost of local resources compared with the ADB loan, and the recurring foreign exchange rate losses (para. 53). BAAC s liquidity was considered adequate for the much lower than expected enterprise financing needs. ADB did not accede to BAAC s request for a waiver of commitment charges. 48. Missions focused on issues encountered during SFCP implementation such as those relating to monitoring and appraisal. Missions met a cross-section of staff and visited selected subprojects to obtain first-hand assessments of subborrowers implementation experience. The importance of environmental awareness and systematic subloan monitoring were continually emphasized. Nevertheless, benefit monitoring evaluation remained weak. ADB s performance in administering the RECP is considered satisfactory. III. TECHNICAL ASSISTANCE 49. A stand-alone TA (footnote 5) was provided to enhance subloan appraisal and supervision, risk management capacity, training capacity, and provision of diversified banking

19 12 services. The TA enabled BAAC to (i) test and adopt a medium-term development proposal for client advisory services and environment due diligence procedures, (ii) enhance risk management procedures, (iii) implement skills-focused participatory training techniques as well as a long-term business plan for a BAAC training center, and (iv) develop a long-term business development strategy. The TA-financed short courses, workshops, and seminars were attended by 1,071 people from the top management, senior staff, and branches (Appendix 9). Although the TA led to better appreciation of international best practices in risk management and appraisal, the potential impact can only be fully realized by upgrading the MIS. The full benefits of the TA could not accrue to BAAC due to the earlier than envisaged loan closing. 50. BAAC also built on the foundation of the TA (footnote 4) piggy-backed to SFCP which supported training courses on investment analysis, environmental awareness, and foreign exchange risk management. Dissemination of appraisal skills continued, in some cases in conjunction with the RECP. However, skills imparted were only partially put in practice. For instance, the repayment schedule and cash flows were often inadequately coordinated. IV. EVALUATION A. Loan Appraisal 51. Distribution of Loans. The number of subloans were more than the appraisal estimate, even though the RECP was canceled mid-way. This deviation is attributed to the large number of small short-term subloans due to the reluctance of the farmers to make capital investments during the crisis. Learning from the restructuring of banks and finance companies affected by the crisis, BAAC restrained portfolio expansion particularly for term loans to new enterprises and preferred relatively less risky short-term subloans. 52. Quality of Appraisal. The appraisal report provided a strong justification to expand and diversify farm family income through rural enterprises and to enhance BAAC s enterprise financing capacity. The financial and institutional analysis of BAAC undertaken during the SFCP was adequately updated. SFCP implementation was critically assessed and monitoring and appraisal of subloans were noted for specific attention during the RECP. Emphasis on capacity building for enterprise financing was reiterated. To strengthen BAAC s viability, the appraisal report appropriately recommended higher loan loss provisioning, a reserve fund to meet part of the foreign exchange risk, and 1 percent service charge for subloans. BAAC was also advised to hedge foreign exchange exposure. The appraisal report continued with the covenants that were not complied with, or only partially complied with, during the SFCP such as those relating to environment, lending interest rates, and repayment rate. The following are areas in which the appraisal report could have been better: (i) (ii) (iii) (iv) It could have strengthened the demand side through support for training of farmers to equip them as owners and managers of the enterprises. While the inflexibility of the MIS vis-à-vis subloan monitoring and evaluation were recognized, no support was built-in to adequately address the issue. As with the SFCP, additional measures required to facilitate and enhance the access of women to financial services were not explored. Likely risk factors due to the crisis could have been noted, particularly when a stabilization program was concurrently being negotiated by the Government.