Tunisia - Third and Fourth Agricultural Credit Projects

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 11977 PERFORMANCE AUDIT REPORT REPUBLIC OF TUNISIA THIRD AND FOURTH AGRICULTURAL CREDIT PROJECTS (LOANS 1885-TUN AND 2865-TUN) JUNE 11, 1993 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. CURRENCY EQUIVALENTS Currency Unit - Tunisian Dinar (TD) (year end) 1980: US$ 1.00 = TD 0.42 1981: = TD 0.52 1982: = TD 0.62 1983: = TD 0.73 1984: = TD 0.87 1985: = TD 0.76 1986: = TD 0.84 1987: = TD 0.78 1988: = TD 0.90 1989: = TD 0.90 1990: = TD 0.84 1991: = TD 0.86 1992: = TD 0.90 The US$ loan amounts in this report should be understood to mean US$ equivalent. GLOSSARY ACU - Agricultural Credit Unit APIA - Agricultural Investment Promotion Agency ASAL - Agricultural Sector Adjustment Loan BOT - Central Bank of Tunisia BNA - National Agricultural Bank (resulting from the merger of BNDA & BNT) BNDA - National Bank for Agricultural Development BNT - National Bank of Tunisia CLCM - Local Mutual Credit Bank (co-operative bank) CNCA - Caisse Nationale de Cr6dit Agricole (of Morocco) CNEA - National Center for Agricultural Studies CRDA - Regional Bureau for Agricultural Development (of the Ministry of Agriculture) EEC/EIB - European Economic Community / European Investment Bank FNG - National Guarantee Fund FODERI - Fund for Integrated Rural Development FOSDA - Special Fund for Agricultural Development FOSEP - Special Fund for Fisheries Development GDP - Gross Domestic Product GOT - Government of Tunisia M&E - Monitoring & Evaluation MIS - Management Information System MOA - Ministry of Agriculture MOF - Ministry of Finance OED - Operations Evaluation Department PAR - Performance Audit Report PCR - Project Completion Report PDR - Rural Development Program SAR - Staff Appraisal Report SCMA - Agricultural Mutual Guarantee Association WEIGHTS AND MEASURES Metric System FISCAL YEAR January 1 - December 31 FOR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Office of Director-General Operations Evaluation June 11, 1993 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Performance Audit Report on Tunisia Third and Fourth Agricultural Credit Projects (Loans 1885-TUN and 2865-TUN) Attached is the Performance Audit Report on TUNISIA - Third and Fourth Agricultural Credit Projects (Loans 1885-TUN and 2865-TUN) prepared by the Operations Evaluation Department. Both projects performed satisfactorily, and the investments financed by the subloans are judged to be worthwhile. Institutional developments at the National Agricultural Bank (BNA) were acceptable, notably expansion and computerization of the branch network, although loan collection performance deteriorated. The Fourth Project made more progress on key sectoral objectives than the Third; however, interest rates on agricultural loans remain below market rates and do not cover costs. The PAR gives more attention than the two PCRs to BNA's losses on its agricultural portfolio - resulting from arrears and from low on-lending rates. These have increased BNA's dependence on cross-subsidization from its commercial lending, as well as on borrowings from the money market to address liquidity problems. A follow-up project is presently being prepared. Project design aims to improve financial performance and BNA's operations for reaching credit-worthy small farmers. Attachment This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  FOR OFFICIAL USE ONLY PERFORMANCE AUDIT REPORT REPUBLIC OF TUNISIA THIRD AND FOURTH AGRICULTURAL CREDIT PROJECTS (Loans 1885-TUN and 2865-TUN) TABLE OF CONTENTS Page No. Preface .......................................................... Basic Data Sheet .....................................................iii Evaluation Summary .................................................. vii . BACKGROUND ................................................ 1 IL DESIGN AND IMPLEMENTATION ................................ 3 A. Project Design ............................................... 3 B. Project Implementation ......................................... 5 III. PROJECT OUTCOME ......................................... 9 A. Farm-Level Impact ............................................ 9 B. Institutional Development ....................................... 11 C. Impact on Sectoral Reform ..................................... 12 D. Impact on BNA's Financial Health ................................ 15 IV. FINDINGS AND ISSUES ......................................... 21 A. Overall Assessment of the Projects ................................ 21 B. The Future of Agricultural Credit in Tunisia ......................... 22 Figures 1. BNA Profits on All Lending vs. Losses on Agricultural Loans ............................................ 21 Tables 1. Appraisal vs. Actual Project Costs ..................................... 6 2. Appraisal vs. Actual Disbursements .................................. 6 3. Number of Subloans .............................................. 10 4. Preferential Agricultural Interest Rates, Selected Market Rates and Inflation ................................. 12 5. Recoveries on the First Three Bank Projects, as of End-1991 ............... 16 6. BNA's Loan Portfolio in 1990 .............. ....18 7. BNA's Agricultural Portfolio: Sources of Funds, Cash Flow and Profitability ................................................. 20 8. BNA Financial and Profitability Ratios ................................ 21 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. Page No. Annexes 1. BNA's Balance Sheet .............................................. 27 2. BNA's Income Statement and Financial Ratios .......................... 28 3. BNA's Income Statement and Financial Ratios Excluding Agriculture ............................................ 29 4. BNA's Income Statement and Financial Ratios for its Agricultural Portfolio ....................................... 30 5. BNA's Sources and Applications Table for its Agricultural Portfolio ....................................... 31 6. BNA's Cash Flow for Agricultural Loans ............................... 32 7. BNA's Overdues on Agricultural Credit, November 30, 1991 ................ 33 8. BNA's Recovery Rate by Line of Credit, 1987-1991 ....................... 34 9. Third Agricultural Credit Project (Loan 1885-TUN): Breakdown of Recoveries by Province ................................ 35 10. BNA's comments ................................................. 37 Map IBRD 24516 PERFORMANCE AUDIT REPORT REPUBLIC OF TUNISIA THIRD AND FOURTH AGRICULTURAL CREDIT PROJECTS (Loans 1885-TUN and 2865-TUN) PREFACE This is a Performance Audit Report (PAR) on the Third and Fourth Agricultural Credit Projects in Tunisia, involving IBRD loans in the amounts of US$30 million equivalent each. The objectives of the third project (Loan 1885-TUN) were to increase agricultural production, reduce Tunisia's dependence on imported food and raise the incomes of small and medium scale farmers, as well as to establish a credit extension service in which the Ministry of Agriculture (MOA) and the Banque Nationale de Tunisie (BNT), a Government-owned commercial bank, would share responsibilities. The loan was approved on June 26, 1980, became effective on June 24, 1981, and was fully disbursed on March 6, 1987, more than three years after the original Closing Date of December 31, 1983. The objective of the fourth project (Loan 2865-TUN) was to assist the Government of Tunisia (GOT) and BNT in promoting the financial viability of agricultural lending by increasing savings mobilization and boosting agricultural productivity and rural incomes through credit operations. The loan was approved on July 7, 1987, became effective on January 22, 1988, and was fully disbursed and closed on November 19, 1991, six weeks before the original Closing Date of December 31, 1991. The PAR is based on the Project Completion Report (PCR) for the third project, prepared by the Agricultural Operations Division in Country Department I of the Middle East and North Africa Regional Office (MNIAG)-' and submitted to the Board, the draft PCR for the third project prepared by BNT and submitted to the Bank in December 1986, the PCR for the fourth project, prepared jointly by MN1AG and by the National Agricultural Bank (BNA, which is the successor to BNT, following the 1989 merger of BNT and the National Agricultural Development Bank (BNDA)) and submitted to the Board, the Staff Appraisal Reports (SAR), the President's Reports, the loan documents, the transcripts of the Executive Directors' meetings at which the projects were considered, on a study of project files, and on discussions with Bank staff. An OED mission visited Tunisia in November/December 1991, and discussed the effectiveness of the Bank's assistance with BNA, the Ministries of Agriculture and of Finance, the Central Bank and other relevant Government agencies. Their kind cooperation and valuable assistance in the preparation of this report is gratefully acknowledged. Y Formerly known as the Agricultural Operations Division In Country Department II of the Europe, Middle East and North Africa Region. - 1i - The PCRs provide a comprehensive account and assessment of the project experience, and discuss the performance of the Bank and the project executing agencies. The PAR elaborates on particular aspects of the overall lending period (1981-1991), including the weak record of compliance on key covenants in the loan and guarantee agreements particularly under the third project, and the increasing impact of arrears and subsidized credit on institutional viability. Following standard OED procedures, copies of the draft were sent to Government and BNA for comments in January, 1993. BNA's response was received in April, 1993. It has been attached as an Annex, and where relevant the text of the audit has been adjusted. - li - PROJECT AUDIT REPORT REPUBLIC OF TUNISIA THIRD AGRICULTURAL CREDIT PROJECT (LOAN 1885-TUN) BASIC DATA SHEET Key Prolect Data Item Appraisal Actual or Actual as % of Expectation Current Estimate Appraisal Estimate Total Project Costs (US$ million) 23,740.0 1' 45,343.0 190 Loan Amount US$ million 30.0 30.0 100 Number of Subloans 3,365 9,147 272 Economic Rate of Return 18-92% 18-65% institutional Performance Partial Cumulative Estimated and Actual Disbursements FY81 FY82 FY83 FY84 FY85 FY86 FY87 Appraisal Estimate (US$ million) 17.1 13.3 24.5 30.0 Actual (US$ million) 1.7 5.6 10.0 19.9 26.1 30.0 Actual as % of Appraisal (%) 0% 13% 23% 33% Date of Final Disbursement: March 6, 1987 Prolect Dates Original Plan Revisions Actual Appraisal Report 04/80 05/27/90 Negotiations 04/80 05/12-16/80 Board Approval 06/26/80 06/26/80 Loan Agreement 08/06/80 Loan Effectiveness 11/06/80 06/24/81 Loan Amendment 10/06/86 Completion 06/30/83 09/30/86 Closing 12/31/83 12/31/84 12/31/86 06/30/85 06/30/86 I' Project costs exclude incremental personnel costs (estimated at appraisal at about D 580,000) to make them comparable to Project costs as estimated by BNT. Project costs in dollar terms were probably less than estimated at appraisal because of Dinar devaluation over the Project period. - iv - Staff Inputs (staff weeks) FY72 FY75 FY78 FY80 FY81 FY82 FY83 FY84 FY85 FY86 FY87 TOTAL Pro-appraIsal 0.3 47.3 0.2 6.8 54.6 Appraisal 50.4 50.4 Negotiation 5.9 5.9 Supervision 1.6 4.3 1.2 11.9 7.5 5.8 3.5 6.3 42.1 Other 15.9 15.9 Total 0.3 47.3 0.2 80.6 4.3 1.2 11.9 7.5 5.8 3.5 6.3 168.9 Mission Data Mission Dates No.of Man/Days Specializations Performance Trend Types of (mo.yr) Persons In field represented V Rating W Problem ! Identification/ Preparation 10/79 3 42 E A ACS Appraisal 11/79 4 84 (2) E A Supervision 1 03/81 3 7 EA ACS 2 1 FO Supervision 2 10/81 1 3 E 2 3 Supervision 3 07/82 1 7 E 3 2 MFO Supervision 4 11/82 2 10 E 3 2 MFO Supervision 5 06/83 1 3 E 3 2 FO Supervision 6 02/84 1 4 E 3 2 MFO Supervision 7 10/85 1 7 FA 2 1 FO TOTAL 17 167 Other Proeect Data Borrower: Banque Nationale de Tunisie (BNT) Executing Agencies: Banque Nationale de Tunisle (BNT) Follow-On Prolect Name: Fourth Agricultural Credit Project Loan Number 2865-TUN Amount (US$ million) US$30.0 Approval Date July 7, 1987 E = Economist; FA = financial analyst; A = agronomist; ACS = agricultural credit specialist. - 1 = Problem-free or minor problems; 2 = moderate problems; 3 = major problems. 1 = Improving; 2 = stationary; 3 = deteriorating. F = Financial; M = managerial; T = technical; P = political; 0 = other. PROJECT AUDIT REPORT REPUBLIC OF TUNISIA FOURTH AGRICULTURAL CREDIT PROJECT (LOAN 2865-TUN) Basic Data Sheet Key Proiect Data Item Appraisal Actual or Actual as % of Expectation Current Estimate Appraisal Estimate Total Project Costs (USS million) 59.0 57.2 97 Loan Amount US$ millIon 30.0 30.0 100 Number of Subloans 4,840 7,290 151 Economic Rate of Return 16-37% n.a. Institutional Performance Partial Cumulative Estimated and Actual Disbursements FY88 FY89 FY90 FY91 FY92 Appraisal Estimate (US$ million) 5.00 12.00 20.00 27.00 30.00 Actual (US$ million) 3.70 8.10 14.12 26.00 30.00 Actual as % of Appraisal (%) 74% 68% 70% 96% 100% Date of Final Disbursement: November 19, 1991 Project Dates Original Plan Actual Appraisal Report 10/86 11/86 Negotiations 04/87 4/13-17/87 Board Approval 06/87 07/07/87 Loan Agreement 07/22/87 Loan Effectiveness 10/87 01/22/88 Completion 12/31/90 09/30/91 Closing 12/31/91 11/19/91 -vi - Staff Inputs (staff weeks) FY84 FY85 FY86 FY87 FY88 FY89 FY90 FY91 FY92 TOTAL Pre-appraisal 1.1 32.0 31.7 64.7 Appraisal 23.4 23.4 Negotiation 4.9 0.9 5.8 Supervision 6.2 4.3 10.0 10.0 15.6 46.1 Other 0.1 1.4 5.3 1.1 7.9 Total 0.1 1.1 33.4 65.2 8.2 4.3 10.0 10.0 15.6 147.9 Mission Data Date No. of Days In Specializations Performance Mission LM& Persons Field'! Represented 8ating I Types of Problem identification/Preparation 9/85 2 19 FA.ACS Preparation 12/85 1 21 FA Preparation 4/86 1 14 FA Pre-Appraisal 6/86 7 21 FA,ACS,DS,RS,TMS,MIS,CB Appraisal 11/86 6 14 FA,E,CB,ACS,TMS Post-Appraisal 3/87 1 7 FA Sub-Total 18 96 Supervision 1 12/87 1 7 FA 1 Loan effectIveness Supervision 2 4/88 1 14 FA 2 Drought Special SPN on policy 1/89 3 7 FA,ACS 2 Policy reform reform8/d Supervision 3 9/89 1 7 FA 2 Drought Supervision 4 12/89 1 14 FA 2 Merger of BNT/BNDA Supervision 5 7/90 2 7 FA,ACS 2 Fungibility of funds Supervision 6 6/91 2 14 FA,ACS 2 BNA reorganization PCR preparation 2/92 2 8 FA,ACS 2 Sub-Total 13 78 Other Prolect Data Borrower: Banque Nationale de Tunisle (BNT) Executing Agencies: Banque Nationale do Tunisle (BNT) Follow-On Project Name: National Rural Finance Project (under preparation). i Man/days In field include time spent on supervising other projects. 'k E = Economist; FA = Financial Analyst; ACS = Agricultural Credit Specialist; DS = Discounting Specialist; TMS = Training and Monitoring Specialist; RS = Rural Savings; MIS = MIS and EDP Systems; CB = Commercial Banking. 'S Status Rating: I = problem free or minor problems; 2 = moderate problems; 3 = major problems - appropriate actions are being taken to moderate these problems; 4 = major problems - they are not being addressed adequately. 'd International donors meeting on the policy and institutional reforms. - vil - PERFORMANCE AUDIT REPORT REPUBLIC OF TUNISIA THIRD AND FOURTH AGRICULTURAL CREDIT PROJECTS (Loans 1885-TUN and 2865-TUN) EVALUATION SUMMARY Introduction BNA's technical agricultural staff at headquar- - ters. Assurances were also obtained that a new 1. The Bank's association with the National accounting system would be established at BNA Agricultural Bank (BNA) dates back to an to identify the cost of lending to agriculture. US$18 million cooperative farm project (Loan The Bank loan of US$30 million to BNA was to 484/Credit 99) approved in 1967. In 1971, the finance 49% of the total project cost of US$60.8 Bank began financing what became a series of million. four agricultural credit projects, totalling US$80 million equivalent. The credit program began by 3. The fourth project switched the emphasis lending exclusively to large farmers, but later from promoting agricultural production to broadened its scope to include small and medium promoting the development of a sustainable and farmers, cooperatives, agro-industries and fisher- financially sound rural credit system. It pro- ies. In addition, increasing emphasis was placed posed specific reforms such as increases in on- on institutional development and sectoral re- lending rates to make agricultural lending more forms. The third and fourth projects in the attractive for BNA and other commercial banks, series are the subject of this Audit. A follow-up harmonization of terms and conditions on van- project is currently being prepared. ous lines of credit, improved access to credit for small and medium scale farmers and greater Objectives savings mobilization. In order to secure the Government's support for these objectives, 2. The primary goal of the third project was to preparation of the fourth project was coordinat- increase agricultural production by financing a ed with the preparation of the first Agricultural part of BNA's medium and long-term agricultur- Sector Adjustment Loan (ASAL 1, Loan 2754- al lending over a three-year period (1981-1983) TUN), approved in September, 1986. Credit for the following categories of borrowers: small was provided under the project for small, medi- and medium scale farmers, production coopera- um and large scale farmers, production and tives, service cooperatives, commercial farmers service cooperatives, agro-industries and, for the and agro-industries. The project also sought to first time, fisheries. The institutional develop- simplify and decentralize appraisal, approval and ment component committed funds for BNA's monitoring of subloans by establishing agricultur- management information system (MIS), training al credit units (ACUs) in each regional bureau program and consultancy services, and included for agricultural development (CRDA) of the proposals to strengthen sectoral policy coordina- Ministry of Agriculture (MOA), and by strength- tion among relevant Government agencies. The ening MOA's and BNA's field staff and training Bank loan of US$30 million to BNT was to - vill - finance 51% of the total project cost of US$59 sectors. There is also considerable inertia in million. BNA and GOT to provide, respectively, the means and the incentive to improve recovery Implementation Experience rates. Finally, BNA appears to be reluctant to commit its own funds to agricultural production 4. Both projects became effective a few loans unless borrowers are demonstrably credit- months behind schedule, the third project in worthy. Thus access to BNA loans is con- June, 1981, and the fourth project in January, strained for small farmers who are potentially 1988. The third project experienced serious viable clients. delays in disbursements and the Closing Date was extended four times from December, 1983 Results to December, 1986. The fourth project was the first in the series to envisage a 4/2-year, rather 6. Both projects exceeded appraisal estimates than a 3-year, disbursement period and was also in terms of the number of subloans granted. the first in the series to be disbursed and closed The fact that the two projects provided more on time. Project costs under both projects were than twice as many subloans to small and medi- close to appraisal estimates in US Dollar terms, ur scale farmers as expected (13,957 vs. 6,650) but the longer implementation period of the is particularly impressive. Although no ex-post third project, coupled with an inflation rate of evaluation was attempted under the fourth around 10% (rather than the 7% expected at project, BNXs M&E unit conducted surveys appraisal) and a large depreciation of the Tuni- during the third project, evaluating investments sian Dinar, entailed a 91% cost overrun for the on farms similar to those envisaged in appraisal third project in local currency. farm models. Re-estimated financial rates of return ranged from 24% to 81%, compared to 5. Although the Bank appropriately placed 18% to 81% at appraisal, whereas re-estimated increasing emphasis on sectoral reforms, perfor- economic rates of return ranged from 18% to mance was weakest in this area for both projects 65%, compared to 18% to 92% at appraisal. (paras. 3.10-3.19). The obstacles encountered Based on the findings, the Audit believes that under the third project included low on-lending most investments could have supported loans at rates, growing arrears and different loan terms market rates of interest, so that the overall rate for competing lines of credit, all of which en- of investment would probably not have been tailed breaches in loan and guarantee covenants. lower in the absence of administered rates. Indeed, disbursement delays under the third Some investments with low returns, such as project are attributable to routine direction of tractors for low-acreage farms, might not have subloan requests to a cheaper, Government- been made without the preferential interest funded, alternative source of credit (known as rates. FOSDA, the Special Fund for Agricultural Development). The fourth project received a 7. There have been a number of positive more favorable rating ("2", or moderate prob- institutional developments at BNA over the past lems) than the third project ("3", or major prob- decade. At the Bank's instigation, BNA devel- lems) from supervision missions. Progress was oped a new cost accounting system under the made in upgrading institutions (BNA and third project which allowed it to identify the CRDAs), e.g. through training and decentraliza- losses associated with its agricultural lending tion, Government funding for credit was reduced (paras. 3.27 ff). Between 1984 and 1991 it as scheduled, and agricultural interest rates were expanded its network from 75 local branches and increased, becoming positive in real terms. 9 regional offices to 126 branches and 13 offices. However, they remain well below rates to other These are being supplied with personal comput- - ix - ers that are being linked to an upgraded main- and was far more restrictive. Thus many farmers frame computer at headquarters. Progress was benetitted from GOT-funded loans and subsidies made on a computerized management informa- on more favorable terms than Bank-funded tion system (MIS) under the fourth project. subloans. The Bank worked hard to persuade Loan authorizations have also been decentral- GOT to revise its definition and harmonize ized to regional offices, so that headquarters lending terms, but: a 1988 redefinition still did now reviews only those loan applications that not rely on income as the principal criterion; a exceed TD120,000 (US$134,000). Other devel- rural development program (PDR) is providing opments that BNA tackled included the 1989 loans to some BNA borrowers that substitute for merger of BNT and BNDA and the 1990 trans- their own equity contributions under other loan fer to a new headquarters built by BNA. One of programs, thereby violating minimum equity the most important issues that remains to be requirements; supervised credits in kind, among resolved is coordination of BNA and CRDAs' others, are still offered at more beneficial rates credit-related activities (para. 3.8). For example, than budget-funded or Bank-funded loans; and neither CRDAs nor BNA have devoted much a new integrated rural development fund attention to the pursuit of defaulters on loans (FODERI) was established in 1985 that requires made with Government resources. lower equity contributions and lends at less than half the interest rate on Bank funds. Govern- 8. The two key sectoral reforms that the ment budgetary support for FODERI has been fourth project and, to a limited extent, the third increasing rapidly (para. 3.15 if). project sought to address were increases in agricultural interest rates and harmonization of 10. BNA's recovery rate is very good for its lending terms. Although interest rates to agri- own funds (over 90%), but tends to be signifi- culture were negative in real terms at appraisal cantly lower on Government or external donor of the third project in 1979, GOT insisted that funds, for which it bears partial or no credit risk they were not negotiable and the Bank post- (paras. 3.20 ff)Y The average recovery rate for poned specific action on the matter. In April, all funds has been stable at around 75% in spite 1985, GOT agreed to the first and only increase of variable climatic conditions, suggesting an in interest rates under the third project. The underlying structural problem. A comparison of fourth project established a timetable for annual its clientele with those benefitting from loans adjustments in agricultural interest rates, so that financed with Bank or Government resources they would equal market rates by December, suggests that BNA selects a safer pool of clients, 1991. The goal was not achieved as market rates especially commercial farmers seeking short term increased and as the Bank agreed to GOT's loans. Although irrecoverable loan losses are request to defer the adjustments scheduled for small for BNA-funded loans, they appear to be 1988 and 1989 in the light of consecutive around 15-20% for Bank-funded loans under the droughts. GOT has continued to raise agricul- first two projects in the series, and are likely to tural interest rates since December 1991; howev- be much larger for Government-funded loans. er, they remain below market rates. 11. The share of agriculture (excluding agro- 9. The harmonization of lending terms and industry) in BNA's portfolio has declined steadi- conditions was not achieved. In particular, in ly, because agricultural lending is unprofitable 1983 GOT defined small and medium scale (paras. 3.26-3.29). This was easily demonstrated farmers in such a way as to allow over 95% of once a separate cost accounting system for all farmers to benefit from subsidies. The Bank's agricultural loans was established at the Bank's definition of small and medium scale farmers was insistence under the third project (Annex Tables based on income rather than size of landholding 3-6). Annual losses increased from TD1.2 - x - million in 1980 to TD12 million in 1990. A Bank was optimistic that substantive progress major reason for this sharp increase in losses is could be made on ensuring the viability of that BNA borrowed funds in the money market BNA's agricultural credit operations. Unfortu- at rates of 9% to 11.8% in order to cover a nately, back-to-back droughts in 1987/88 and shortfall in Government funding for agricultural 1988/89 slowed progress on interest rates, which loans for which GOT had assumed responsibility. are still Goverment-controlled and below The funds were on-lent at preferential rates, market rates. Furthermore, although budgetary which were as low as 4% in the case of contributions for FOSDA subloans were cut, FODERI loans. Currently BNA and GOT are GOT expanded lending under a cheaper, alter- involved in a dispute over who should pay for native line of credit, namely FODERI Thus, the money market advances. The Bank is re- the Audit concludes that only partial progress quiring that the matter be resolved before it was made on key sectoral issues during the proceeds with a follow-up project. fourth project. Sustainability 15. The Audit identities the following issues and lessons for future operations: 12. That BNA's financial situation is unsustain- able is demonstrated by the temporary suspen- 0 The sequencing of reforms is important for sion of its agricultural lending in November, some (but not all) measures: in particular, 1991, until more Government funds were forth- agricultural interest rates must be raised to coming. BNA is unwilling to continue using market levels before BNA can bear 100% of proits in other sectors to cross-subsidize loss- the credit risk and full responsibility for loan making loans to agriculture. Such cross-subsidi- approvals (especially for small farmer sub- zation may in any case be unfeasible since profits loans). in other sectors have been overstated by main- taining inadequate provisions for bad debts. The M Future credit operations should continue to dismal performance on the agricultural portfolio co-ordinate with sectoral adjustment opera- has contributed to a two-thirds decline in BNA's tions, with a view to eliminating administered overall return on equity during the past decade - interest rates and to securing the right for to around 8% in 1990, equal to the rate of BNA to freely determine rates on all its agri- inflation, cultural loans. Findings and Lessons 1 The Bank should continue its dialogue with the Government with a view to harmonizing 13. Although the projects were successful in loan terms on al lines of credit, including promoting institutional development and avail- FODERI. However, at present, BNA and able data suggest that investments financed at GOT continue to treat FODERI and a few the farm level were generally viable, only partial other lines of credit as exceptions (see BNA's progress was made on key sectoral issues, on comments in Annex 10). balance, the overall performance of both pro- jects is assessed as satisfactory but this is no M As lending to agriculture becomes more remu- grounds for complacency given the poor sustain- nerative, BNA must assume substantial re- ability prospects mentioned above. sponsibility for loan approvals. However, as a first step, BNA representatives should chair 14. During negotiations for ASAL I, GOT and meetings held by regional credit committees the Bank forged a consensus for sectoral reforms and their veto should be definitive. that carried over to the fourth project, and the co ---ont wihsetrl-dusmn-oea - xl - " The low administrative cost of BNA's opera- agricultural lending based on whether BNA or tions confirms the Audit's observation that GOT bore the credit risk. This would help to little effort has been made to recover loans. distinguish losses made by BNA from those This will only change if BNA bears substantial made by GOT. credit risk for all loans. Until it does, its spread on special funds should be tied to the 0 Initiatives are required to ensure that access recovery rate on those funds. BNA's policy on to BNA credit is not restricted for viable small provisions should also be reviewed in order to farmers. Possible measures include dialogue establish adequate provisions for loan losses. with senior staff, training, additional resources Finally, a well-managed, private-sector operat- at the branch level, and greater emphasis on ed calamity insurance fund, supported by the savings mobilization to establish a relationship Government, should be set up to replace the with the clientele. Further measures are also calamity grants implicit in the practice of debt required to address the problems of insecure forgiveness in drought years. land tenure and fragmentation of land- holdings, so as to attract more domestic re- " During the third project, BNA introduced sources into investing in the agricultural sec- accounting changes that made it possible to tor. review the agricultural portfolio separately. These useful changes should be complemented by further revisions that would break down ' If funds are distributed by responsible party rather than by source, one finds that the recovery rate on loans for which BNA bears the full credN risk Is 83%, compared to 52% on loans for which the Government bears the full risk.  PERFORMANCE AUDIT REPORT REPUBLIC OF TUNISIA THIRD AND FOURTH AGRICULTURAL CREDIT PROJECTS (Loans 1885-TUN and 2865-TUN) I. BACKGROUND 1.1 THE TUNISIAN ECONOMY. The Tunisian economy grew at around 7% per annum during the 1970s as a result of terms of trade gains from higher prices on oil exports, a gradual shift towards a more market-oriented economy and a high rate of investment (around 30% of GDP). However the world recession of the early 1980s coupled with a decline in oil prices contributed to a slowdown in growth and macroeconomic imbalances became more pronounced: annual inflation increased to 10.5% during 1981-84; the budget deficit averaged 5.6% during the same period; open unemployment increased to 14% in 1983 and the current account deficit reached 11% of GDP in 1984. Although the VIth Plan (1982-86) supported several policy reforms aimed at employment generation and export promotion, macroeconomic performance deteriorated during the plan period. In the mid-1980s the Government developed an agenda for enhancing efficiency in the economy by promoting the private sector and increasing its outward orientation. Beginning in FY87, the Bank supported GOT's program with five adjustment operations promoting reforms in the financial sector, taxation, public enterprises, industrial and trade policy, and agriculture. 1.2 Agriculture accounted for about 13% of GDP, 35% of employment and 8% of export earnings in 1985. The number of farms in Tunisia is estimated at 354,000, of which 301,000 (85%) occupy less than 20 hectares. Most agricultural production is in rainfed areas (8,750,000 ha) that are subject to severe fluctuations in rainfall, with droughts typically occurring every two or three years. In the past decade, the most severe droughts were in 1982/1983, 1985/1986, 1987/1988 and 1988/1989. Tunisia's limited irrigation potential, equal to 3% of arable land (250,000 ha), has been fully tapped. The agricultural sector grew at an average rate of 3.3% during the late 1970s and early 1980s, which was not sufficient to keep up with increased domestic demand. Thus the ratio of sector exports (fresh fruits and vegetables, fish and processed products including olive oil and wine) to sector imports (cereals, dairy products, sugar and meat) fell from 81% in 1976 to 49% a decade later. 1.3 AGRICULTURAL CREDIT INSTITUTIONS. The National Agricultural Bank (BNA) accounts for most short term and almost all medium and long term agricultural credit. Established in 1959 as a Government-owned commercial bank specializing in agriculture, it became a full-service bank and was renamed the National Bank of Tunisia (BNT) in 1969. Its name was changed back to BNA following a merger with the National Agricultural Development Bank (BNDA) in October, 1989, and it is now the largest commercial bank in Tunisia, with a network of 126 branches and 13 regional offices throughout the country.y The share of agriculture in BNA's portfolio has declined steadily (from 44% in 1981 to 33% in 1991) as growth in industrial and commercial lending has outstripped growth in agricultural credit. BNA restricts the use of its own funds in agriculture to seasonal credits for the 1' The Audit will use the name BNA throughout the report to refer to BNT before the 1989 merger and to BNA after the merger. -2- most creditworthy clients. Medium and long-term loans (and short-term loans to less creditworthy clients) are financed primarily by Government budget funds and by external donors. 1.4 BNT's merger partner, BNDA, was set up in 1983 in an unsuccessful attempt to address problems in the agricultural credit sector by establishing a new institution. It provided a channel for external funds, mainly of the European Economic Community and European Investment Bank, to be on-lent to large scale farmers for integrated farm development projects. With only one office in Tunis and limited ability to mobilize domestic resources, it appeared not to be viable in the long run and the merger was a logical solution. However, it is feared that as the grace periods on long-term BNDA loans run out, significant arrears on these loans may emerge, further contaminating BNA's portfolio (paras. 3.20 ff). 1.5 Other commercial banks have limited their agricultural loans to large farmers with good collateral who are important depositors and are therefore well known to them. Few farmers can meet commercial banks' strict collateral requirements. Because Government policy has mandated below- market rates of interest for agriculture, many banks have preferred to accept the penalty of unremunerated deposits frozen in the Central Bank of Tunisia (BCT) rather than meet Government lending targets for the agricultural sector. Unless Government policy changes, commercial bank participation in agricultural credit is likely to remain limited. 1.6 In the first two decades following independence, Tunisia emphasized cooperative farming schemes rather than individual small farmers. However, by the early 1970s it became clear that collective farming was not delivering the results that had been expected. Lending through cooperative banks, known as Local Mutual Credit Banks (CLCMs), proved to be equally unsuccessful and most were converted into BNA branches during the 1980s. Mutual guarantee associations for agriculture (SCMAs), established at the Government's initiative in the early 1970s in order to improve individual small farmers' access to credit, have been saddled with arrears. GOT closed down several bankrupt SCMAs in 1983 and assumed responsibility for paying off their debts to BNA. 1.7 GOVERNMENT CREDff AND SUBSIDIES FOR AGRICULTURE. The Agricultural Investment Code of 1982 provides for several financial incentives for farms of various sizes. These consist of agricultural credit at rates below market rates, of (further) interest subsidies, of grants and of so- called reimbursable contributions to farmers to help meet equity requirements. The Agricultural Investment Promotion Agency (APIA) helps to direct these Government-funded benefits to farmers interested in integrated farm development projects and monitors these projects. For most farmers interested in individual farm-level investments, the Ministry of Agriculture's (MOA) regional bureaus for agricultural development (CRDAs) identify and prepare sub-projects for financing as part of their extension activities. CRDAs also monitor farm investments after the loans and subsidies are approved. Loans for these investments are approved by regional committees composed of CRDA and other Government officials, BNA and a farmer representative. Loan funds are provided by BNA if it approves the loans. BNA selects the best seasonal clients and, until recently, had ready access to rediscounts from the Central Bank of Tunisia (BCT). If the loans are medium or long term loans or if the regional loan committee overrides BNA's veto of a short term loan, BNA manages the loan for a commission, but the loan funds are provided from resources known as special funds. These special funds are provided by the Government, World Bank and other external sources. BNA's commission is 2% on outstanding FOSDA loans (see below) and 3% on Bank funds (as well as compensation if the Bank's lending rate to BNA is higher than BNA's lending rate to clients). BNA -3. bears partial or no credit risk for loans using special funds. For example, BNA bore the credit risk for only 40% of loans to small and medium farmers or to cooperatives financed by the third project. BNA bore full risk for all loans financed by the fourth project, but 70% of defaults on loans to small and medium farmers were covered by the Government's National Guarantee Fund (FNG) for agricultural credit risks. Currently, BNA classifies its subloans by source of funds - it manages 48 different lines of credit - although a classification of subloans by whether BNA or GOT bears the credit risk would also be useful. 1.8 The most important of the special funds are FOSDA, the Special Fund for Agricultural Development, and FOSEP, its counterpart for fisheries. Other important lines of credit are FODERI, a trust fund at BNA which is controlled by the Ministry of Planning and is used to finance integrated rural development in areas of the country requiring special assistance, and Rural Development Program (PDR) funds that are controlled at the provincial level. All four funds are financed by the Government. Overall, BNA merely acts as a bookkeeper for over half the resources that it lends to agriculture. 1.9 WORLD BANK SUPPORT FOR BNA. The Bank's association with BNA dates back to an US$18 million Cooperative Farm project (Loan 484/Credit 99) approved in 1967. In 1971 it began financing what became a series of four agricultural credit projects. Under the first one, (Loan 779/Credit 263), credit was devoted entirely to large farmers. Later projects supported small and medium farmers, agro-industries and both production and service cooperatives, and placed growing emphasis on institutional components. A total of US$80 million was committed under the four projects. The third and fourth agricultural credit projects are the subject of this audit. A follow-up project is currently being prepared. II. DESIGN AND IMPLEMENTATION A. Project Design 2.1 THE THiRD PRoJEcr. BNA identified and prepared the third project in 1979, with assistance from Bank supervision missions to the second project. The project's agro-industry component was prepared by the FAO/World Bank Cooperative Program. Unlike its precursor, which had concentrated on increasing agricultural production, the third project also addressed institutional development and BNA's financial performance. However, the primary objective of the third project was to increase agricultural production and thereby reduce dependence on imported foodstuffs. This was addressed by financing a part of BNA's medium and long term agricultural lending over a three- year period (1981-1983) for the following categories of borrowers: small and medium scale farms, production cooperatives, service cooperatives, commercial farmers and agro-industries. The project emphasized improving the incomes of small and medium farmers by allocating 45% of project funds to them, either directly or as members of production cooperatives. 2.2 The second objective of the project was to promote institutional development by simplifying and decentralizing appraisal, approval and monitoring of subloans. This was to be done -4- by establishing agricultural credit units (ACUs) in each CRDA, strengthening MOA's and BNA's field staff and training BNA's technical agricultural staff at headquarters. 2.3 Finally, an unstated objective of the project was to improve BNA's financial position. It was assumed during preparation that several pertinent issues had already been resolved, particularly the standardization of appraisal procedures and of interest rates on sub-loans funded by various lines of credit administered by BNA, as well as a determination to reach small and medium scale farmers. Unresolved questions raised during preparation and appraisal included the lack of data on arrears and on the cost of administering the agricultural portfolio, as well as the inadequacy of interest rates on subloans. During negotiations, assurances were obtained that a new accounting system would be established at BNA to identify the cost of lending to agriculture. The Bank's proposal that interest rates be reviewed annually with a view to making them both positive in real terms and sufficient to cover BNA's costs of lending was removed from the loan agreement and added to a supplemental letter at the Government's request. GOT also agreed in a supplemental letter to ensure the viability of BNA's agricultural portfolio, but did not commit to implementing any proposals that BNA might make on the matter. 2.4 The Bank loan of US$30 million to BNA was allocated entirely to the credit component. The loan covered 49% of the appraisal project cost of US$60.8 million, with BNA, GOT and subborrowers providing the remainder. 2.5 THE FOURTH PRoJEcr. The fourth project was also prepared by BNA with assistance from Bank supervision missions. In this project the emphasis was switched from promoting agricultural production to promoting the development of a sustainable and financially sound rural credit system. The fourth project addressed a number of the obstacles encountered during implementation of the third project, including on-lending rates below market rates of interest, growing arrears and different loan terms for competing lines of credit offered by BNA. The project sought to increase on-lending rates in order to make agricultural lending more attractive for BNA and other commercial banks, to harmonize credit terms and conditions, to improve small and medium scale farmers' access to credit and to promote savings mobilization. In order to secure the Government's support for these objectives, preparation of the fourth project was coordinated with the preparation of the first Agricultural Sector Adjustment Loan (ASAL, Loan 2754-TUN), approved in September, 1986. 2.6 In May, 1987, the Bank received a policy letter on agricultural credit from the Ministry of Planning and Finance, outlining GOT's intentions to increase savings mobilization and reduce the contribution from the budget, gradually harmonize lending terms and conditions for loans to agriculture and fisheries, transfer a greater share of the credit risk to the banking sector, tighten the definition of "small and medium farmers" by adding an income criterion (thereby reducing the number of farms eligible for subsidies), and raise interest rates to cover reasonable costs of lending to agriculture. During negotiations, GOT committed itself to three important financial reforms: i) annual increases in agricultural on-lending rates with a view to aligning them with the money market rate of interest by 1991; ii) lower budgetary contributions to FOSDA and iii) harmonization of loan terms on various lines of credit by 1988. 2.7 The project as approved included credit components for small, medium and large scale farmers, production and service cooperatives, agro-industries and, for the first time, fisheries. Although an earlier project brief refers to a component for agricultural loans by other commercial -5- banks, it appears to have been dropped after a meeting with their chief executive officers. The project's institutional development component committed funds for BNA's management information system (MIS), training program and consultancy services, and sought to strengthen policy coordination on agricultural credit between the Central Bank and the Ministries of Planning, of Finance and of Agriculture, and between BNT and BNDA. The Bank loan of US$30 million to BNT was to finance 51% of the total project cost of US$59 million. US$27.5 million of the loan were provided for the credit component and US$2.5 million were allocated to the institutional development component. B. Project Implementation 2.8 GEmRmu PERFloMANcE The third project became effective in June, 1981, after a delay of seven months. The fourth project became effective with a three-month delay in January, 1988. There were serious implementation problems throughout the life of the third project, and most supervision missions gave it a "3" rating (major problems). The main problem was a slow rate of disbursement of Bank funds, which arose because CRDA staff gave priority to disbursing (cheaper) FOSDA funds. As a result, the Closing Date had to be extended four times, from December, 1983 to December, 1986. The other major problems were Government resistance to increases in agricultural interest rates and BNA's unwillingness or inability to take the necessary action to improve recovery rates. 2.9 The fourth project generally received a "2" rating (moderate problems) from supervision missions. Unlike its predecessors, it was disbursed on time and closed on time. It was, however, also the first project to envisage a four-and-one-half-year rather than a three-year implementation period at appraisal, and it faced less competition from alternative sources of funds (para. 2.15). Performance in upgrading institutions (BNA and CRDAs), e.g. through training and decentralization, was better than performance in implementing financial reforms. In particular, GOT was reluctant to abide by its commitment to raise interest rates annually in the light of consecutive droughts in 1987/1988 and 1988/1989 (paras. 3.12-3.14). 2.10 Overall, the Audit mission found that i) even in the absence of droughts, some BNA and MOA staff continue to express reservations about harmonizing agricultural interest rates with market rates of interest; ii) there is considerable inertia in GOT and BNA to provide, respectively, the incentive and the means to improve recovery rates, and iii) BNA appears to be reluctant to commit its own funds to agricultural production loans unless borrowers are demonstrably creditworthy. Thus access to BNA loans is constrained for small farmers who are potentially viable clients. 2.11 PRoJECT CosTs. The third project saw a 91% cost overrun, measured in Tunisian Dinars, but a modest decrease in US Dollar terms. Costs were higher in local currency due to a six-year rather than a three-year implementation period and an average inflation rate of over 10% during the early 1980s, rather than 7% as assumed at appraisal. They decreased slightly in US Dollar terms due to the depreciation of the Dinar over the life of the project. Project costs were on target for the fourth project, both in local currency (Table 1) and in US Dollars. They include a US$2.2 million grant furnished by the EEC after BNT merged with BNDA. -6, Table 1: APPRAISAL VS. ACTUAL PROJECT COSTS (TD '000) -- Third Project -- -- Fourth Project - Appraisal Actual Appraisal Actual Small/Medium Scale Farmers 5,240 11,932 15,400 14,846 Production Cooperatives 5,690 14,561 4,800 5,496 Service Cooperatives 1,280 1,751 - - Commercial Farmers 8,120 8,800 15,400 11,954 Agro-industries 3,400 8,299 6,500 6,240 Fisheries - - 5,700 11,127 Institutional Development 590 - 3,500 1,498 Total Project Cost 24,320 45,343 51,300 51,161 2.12 During the third project, the demand for mechanization was lower in real terms than estimated at appraisal, because of sharp increases in unit prices. Small and medium scale farmers' demand for irrigation equipment and livestock, both of which had high rates of return, exceeded appraisal forecasts. 2.13 The fourth project saw a larger than expected contribution to the fisheries component, due in part to droughts in the agricultural sector and in part to the Bank's decision to approve the sale of existing trawlers to assist with the privatization of the National Fisheries Office (ONP). This element of the fisheries component was not envisaged at appraisal, and resulted in considerably larger fisheries loans for far fewer beneficiaries than was expected at appraisal (Tables 1, 2 and 3). While it helped to maintain the pace of disbursements, it essentially funded the transfer rather than the creation of real assets. 2.14 DisBUsEmEwrs. Overall, Bank disbursements were in line with appraisal estimates, as shown in Table 2 below, although the third project took twice as long as expected to disburse. At the original Closing Date of December, 1983, only 24% of Bank funds had been disbursed, primarily for commercial farmers. Sub-loan demand was especially weak for small and medium scale farmers, whose applications were routinely directed by the CRDAs to FOSDA funds, which offered more favorable loan terms. Indeed, the Audit mission learned that, when FOSDA funds ran low, farmers were sometimes prepared to wait until the next season rather than finance their projects with alternatives such as the Bank's funds. Similarly, the majority of commercial farm applications for Bank support were for mechanization loans, since these were not eligible for FOSDA subsidies. Table 2: APPRAISAL VS. ACTUAL DISBURSEMENTS (US$ 9000) -- Third Project -- -- Fourth Project - Appraisal Actual Appraisal Actual Small/Medium Scale Farmers 7,400 7,491 8,000 8,090 Production Cooperatives 9,660 9,580 3,500 3,327 Service Cooperatives 1,070 1,071 Commercial Farmers 6,520 6,550 8,000 6,931 Agro-industries 5,350 5,308 4,500 3,678 Fisheries 3,500 6,370 Institutional Development 2,500 1,604 Total Bank Disbursements 30,000 30,000 30,000 30,000 -7- 2.15 Because of slow progress on the resolution of sectoral issues (paras. 3.10-3.19), the Bank at first recommended that the balance of the third loan be cancelled in December, 1983. However, the Government agreed to raise interest rates for commercial farmers and to extend greater legal powers, known as privilkge d'Etat, to BNA to help it recover arrears. After a fourth extension of the Closing Date to December, 1986, the loan was fully committed. As a result of the hiatus between the third and fourth projects, disbursements started promptly under the fourth project with 10% of the loan being used for retroactive financing. Although disbursements slowed down in the drought years, they picked up thereafter as i) FOSDA and Bank lending terms became harmonized, and ii) reductions in budget contributions to FOSDA increased demand for alternative sources of medium and long term credit, including the Bank's funds. 2.16 MONrlORING AND EVALUATION (M&E). A sound methodology for evaluating subloans was developed during the second project, based on the collection of information from a sample of farms, both during and after the sub-loan investment period. The findings for subloans granted under the third project are described below (paras. 3.1-3.6). The Audit Mission was unable to review the calculations of ex-post economic rates of return for the project. These had been archived during the transfer to the new headquarters. Although BNA affirmed that the calculations had been sent to the Bank, the Audit was not able to locate them. 2.17 The appraisal report of the fourth project stressed the need for adequate monitoring of subloans but was virtually silent on the question of ex-post evaluation. During the fourth project, BNA decided to stop collecting information on a sample of subloans for ex-post evaluation. It redeployed a number of staff from the central M&E unit to other functions after the merger, and the unit now concentrates on loan appraisals that are beyond the purview of branch staff. Although BNA has considered resuming its farm surveys to verify whether appraisal expectations have been met for sub-projects, it believes that results will be imprecise unless it conducts regular visits throughout the sub-projects' implementation period. Furthermore, it is difficult to secure full cooperation from farmers. 2.18 It is not clear that BNA should devote its own resources to ex-post evaluation. APIA would like to develop ex-post evaluation capabilities, and specialized staff are already in place at MOA and other agencies, such as the National Center for Agricultural studies (CNEA). This Center carries out surveys and prepares agricultural studies and could arguably be contracted to provide an assessment of the returns on agricultural investments financed by BNA, or indeed by any Tunisian bank. Such an arrangement might ensure that funds are directed towards investments that are indeed remunerative, while avoiding waste through duplication of effort by various agencies. Although evaluation capacity could usefully be centralized, each bank should of course continue to monitor the use of its own loans. 2.19 REPORTING AND AUDITING. BNA regularly sent quarterly and annual reports to the Bank, documenting disbursements by category of borrower. Occasionally, the reports included information on arrears. In May, 1985, BNA sent a report to the Bank based on a monitoring survey of 286 loan files. The report presented the farmers' perspectives on problems with sub-loan appraisal and implementation. BNA also submitted annual audit reports that were generally on time and of good quality. -8- 2.20 COMPLIANCE WrH CovENA.S. BNA and GOT complied with most covenants in the loan and guarantee agreements, however the covenants and side-letter commitments that they failed to fulfil were almost without exception indispensable for reforming agricultural credit. In particular, during the third project, BNA did not increase its recovery rate to the levels established in a side letter (85% by May, 1982). Nor did the Government honor its commitments, spelled out in side letters, to transfer more of the credit risk on agricultural loans to the banking sector, to ensure the viability of BNA's agricultural portfolio, and to raise interest rates to levels that would be both positive in real terms and sufficient to cover reasonable costs associated with agricultural credit. Furthermore, there was a breach of the covenant in the third project's guarantee agreement that specified that credits for purposes similar to the project's be made on terms and conditions spelled out in the loan agreement. In particular, the 1982 Agricultural Investment Code that itemized benefits for small and medium farmers and the 1983 decree that defined "small and medium" to include 95% of Tunisian farmers entailed breaches in the covenant. On the other hand, the Bank did not object when a 1988 decree redefining small and medium farmers had a similar effect. Finally, ACUs were not established in the CRDAs as foreseen. Although BNA proceeded to decentralize and train its own branch level staff, CRDAs still appraise most small loans financed with special resources (including Bank resources), with little or no coordination with BNA. 2.21 Commitments under the third project that were honored by BNA included strengthening of its central technical unit, establishment of a separate cost accounting system for the agricultural portfolio, analysis of income from and costs of agricultural lending, and decentralization of authority over loan approvals. In addition, GOT compensated BNA for losses on loans for which GOT bore the credit risk. 2.22 Under the fourth project, BNA established an MIS and a computer managing committee and prepared an annual training program as agreed. GOT also established an Agricultural Credit Coordinating Committee and reduced its budgetary contributions to agricultural credit programs as agreed. However the most important covenant in the guarantee agreement was not fulfilled, namely annual adjustments in agricultural interest rates to reach rates equal to money market rates by 1991. GOT did increase rates in 1987 as a condition of release of the second tranche of ASAL I (Loan 2754-TUN), and again in 1990 and 1991, but two years of drought (1988 and 1989), coupled with a temporary decline in the money market rate, led GOT to persuade the Bank to agree to deferment of the planned increases for those years. The PCR (para. 6.03) nevertheless characterizes the progress made on interest rates as "commendable", particularly in the light of the political sensitivity attached to them. 2.23 BANK PERFORMANCE. Both projects adequately foresaw and addressed needs identified under the credit components and under the hardware and training elements of the institutional components. The Bank also appropriately placed increasing emphasis on sectoral reform. However, performance was mixed with regard to sectoral issues and some aspects of institutional development. In particular, covenants proved not to be sufficient to ensure success on policy reforms, in spite of coordination with agricultural sector adjustment operations. 2.24 The Bank worked hard to promote essential reforms and at times adopted quite a firm stance. For example, project staff initially recommended cancellation of the third project in December, 1983, due in large measure to lack of progress on sectoral issues. After 1983, supervision was reduced in favor of intensive sector work and agricultural credit sectoral issues were mainly -9- discussed in the framework of the ASALs. In 1987, there appeared to be an important change in perspective on the merits of subsidized credit among senior Government officials, and GOT committed itself to sectoral reforms under the fourth project. However, the Bank felt obliged to garner support from other donors at a 1990 meeting held in Tunis, in order to ensure continued progress on policy reforms. While project ratings (generally a "3") were severe but accurate during the third project, the Audit believes that the average "2" rating in the course of the fourth project was rather optimistic. 2.25 While the Audit does not criticize time spent addressing sectoral issues, it believes that certain project matters deserved more attention, particularly helping BNA staff to find viable ways of lending to smallholders. After the Bank's expectation that BNA and CRDAs could cooperate on loan appraisals proved wrong, no other technical proposals were made to establish closer links between small farmers and BNA. BNA branches still lack adequate staff and transport to develop a working relationship with or good knowledge of their small farmer clientele. Current proposals for a follow-up project rightly consider ways to improve rural BNA branch outreach to small farmers. Finally, the Bank should have been struck by the relatively low operating costs for credit operations in Tunisia, which clearly contribute to the low recovery rates observed. III. PROJECT OUTCOME A. Farm-level Impact 3.1 NUMBER OF BENEFICIRIES. Both projects exceeded appraisal estimates in terms of the number of subloans granted (Table 3). The fact that the projects provided more than twice as many subloans to small and medium scale farmers as expected (13,957 versus 6,650) is particularly impressive.? At first, the Audit was struck by the modesty of the appraisal targets when compared to a population of about 350,000 farm families. It discussed the issue of credit penetration with staff at BNA and MOA. BNA argued that it maintains about 150,000 active accounts, including accounts for cooperatives that would indirectly benefit 100 or more members on average; therefore a large share of the rural population is reached.? GOT also claimed that most farmers who are not considered eligible for commercial bank credit because of the small size of their holdings are reached through special lines of credit, notably FOSDA, FODERI, supervised credits in kind, and funds managed by provincial authorities (PDR). Yet a survey conducted in one of the poorest areas of the province of Sfax (the Sidi M'hadeb plateau) in June, 1991, showed that only 10% of all farms had received credit and/or investment subsidies in one year. 91 It is not clear how many of the sub-loans were made to repeat borrowers. Since these were investment loans, the frequency of repeat borrowing within the span of one project was probably very low. Y This is particularly true if one accepts CNEA's estimate that only about 110,000 of the farmers can be considered creditworthy. - 10 - Table 3: NUMBER OF SUB-LOANS -- Third Project -- -- Fourth Project - Appraisal Actual Appraisal Actual Small/Medium Scale Farmers 2,250 7,947 3,400 6,010 Production Cooperatives 170 172 160 209 Service Cooperatives 25 21 ComDercial Farmera 900 985 615 843 Agro-industries 20 22 25 36 Fisheries 640 192 Total Number of Subloana 3,365 9,147 4,840 7,290 3.2 RETURNS ON INVESTMENTS. BNA's ex-post evaluation results for the third project are based on surveys of farms that undertook investments similar to those identified in appraisal farm models. Re-estimated financial rates of return ranged from 24% to 81%, compared to 18% to 81% at appraisal, whereas re-estimated economic rates of return ranged from 18% to 65%, compared to 18% to 92% at appraisal. Increases in farm incomes were reported to be much larger than expected at appraisal. No re-estimates were attempted under the fourth project, but BNA staff suggest that financial rates of return are relatively low (under 20%) for farm mechanization and high (around 50%) for orchards and for investment in irrigation. It would be worthwhile to confirm these findings, since low returns on farm investments have often been used as an argument for lending to agriculture at rates below market rates. 3.3 EFFICIENCY OF RESOURCE ALLOCATION. In credit schemes involving the distribution of investment loans at subsidized rates of interest, there is in theory an important risk of misallocation of resources. The Audit mission was not able to assess the extent of misallocation of resources in practice. Investment in tractors would probably have been lower if market rates of interest had prevailed, since some tractor purchases were being financed for use on acreages that were too small to make the investments worthwhile in the absence of additional income from renting out the tractors. However, rates of return on farm-level investments calculated by BNA after the third project suggest that most of them were viable at market rates. Thus the Audit believes that in the absence of interest subsidies, the pattern of agricultural investments might have differed slightly but the overall rate of investment in agriculture during the past decade would probably not have been lower. Based on available information, the economic justification for subsidized interest rates appears to be weak. 3.4 On the more serious risk of diversion of loan funds, one official told the Audit mission that diversion was estimated at no more than 4-5% of loans. For BNA disbursements of sub-loans, the requirement that three proforma invoices be provided to document investments has recently been reduced to one invoice, in order to simplify procedures. This may contribute to an increase in abuses, but BNA staff are cognizant of unit prices for the inputs and equipment that are commonly financed.y Y In other cases, credit Is delivered by reimbursing borrowers for investments that they have agreed to undertake. CRDAs, BNA and occasionally an Independent MOA unit monitor the loanees to verify that the scheduled Investments have Indeed been made. - 11 - 3.5 There is too little information on rural incomes and savings patterns to ascertain the degree to which substitution of sub-loan funds for other funds occurs. It is more likely to arise in the case of commercial farmers and some cooperatives that have access to short term credit at market rates. However, the Audit visited one service cooperative in the northeastern province of Nabeul which financed capital investments with medium to long term loans from BNA, but had enough working capital to satisfy seasonal requirements without recourse to credit and indeed did not take out seasonal loans. 3.6 Although the fourth project envisaged a savings mobilization program that would provide BNA with an opportunity to establish banking relations with a rural clientele, the Audit has no information about the outcome of the program and presumes that it was not successfully implemented. B. Institutional Development 3.7 There have been a number of positive institutional developments at BNA over the past decade. At the Bank's instigation, BNA developed a new cost accounting system under the third project which allowed it to identify the losses associated with its agricultural lending (paras. 3.27-3.29). Between 1984 and 1991 it expanded its network from 75 local branches and 9 regional offices to 126 branches and 13 offices. These are being supplied with personal computers that are being linked to an upgraded mainframe computer at headquarters. Progress was made on a computerized management information system (MIS) under the fourth project. Loan authorizations have also been decentralized to regional offices, so that headquarters now reviews only those loan applications that exceed TD120,000 (US$134,000). Other developments that BNA tackled included the 1989 merger of BNT and BNDA and the 1990 transfer to a new headquarters built by BNA. 3.8 Coordination of BNA and CRDAs' credit-related activities proved less successful. The third project envisaged the establishment of agricultural credit units (ACUs) at each CRDA to identify and appraise sub-projects to be presented to BNA for financing. Appraisal procedures were to be simplified and the ACUs were to be staffed with competent agents trained by BNA who would cooperate with BNA on improving recovery rates. A study group was set up as required under the third project to recommend job descriptions and training programs, but after three years of disagreements between BNA and MOA on their respective responsibilities, the Bank concurred with BNA's decision to focus on training its own branch level staff. CRDAs were reorganized satisfactorily with Bank support, but procedures for processing loan applications could still be improved. The role of CRDA credit units is limited to desk reviews of sub-project proposals that are then presented to regional credit committees. Most of the field work (identification and appraisal) is conducted by CRDA extension agents, whose activities are still not closely coordinated with those of BNA's branches. For example, neither CRDAs nor BNA actively pursue defaulters on loans made with special funds: BNA bears only partial credit risk on these funds and has little incentive to deploy more staff to recover these loans, whereas CRDAs, which have more field staff and could perform this task more easily, limit themselves to keeping loan files and recording disbursements. 3.9 Finally, an existing interministerial task force on agricultural credit was converted into a standing committee for operational coordination as required under the fourth project. The Audit mission could not assess the extent to which the Agricultural Credit Reforms Committee contributed to the process of harmonizing loan terms and reducing subsidies. - 12 - C. Impact on Sectoral Reform 3.10 INrERES RATES. Since 1988, BNA has been allowed to set interest rates on short term agricultural production loans to large farmers and on marketing loans, subject only to Tunisia's general ceiling on interest rates, equal to the money market rate plus 3%Y On the other hand, rates on supervised credits (short term loans in kind), on loans to small and medium scale farmers, and on medium and long term loans are still controlled by the Government. These controls apply on loans made with Bank resources (and other special funds) and have resulted in losses for BNA. Although the Bank tried hard to encourage GOT to lift the controls, the Government for various reasons resisted pressure to allow agricultural interest rates to rise to market levels (Table 4). Table 4: PREFERENTIAL AGRICULTURAL INTEREST RATES, SELECTED MARKET RATES AND INFLATION (Percent) 1980 1985 1986 1987 1988 1989 1990 1991 Inflation Rate 10.00 7.55 6.15 8.20 7.15 7.70 6.80 8.20 Money Market Rate 8.50* 10.28 9.95 10.00 8.63 11.31 11.81 11.81 Deposit Rate 2.50 5.35 6.75 7.22 6.63 9.25 9.63 9.63 Preferential Axricultural Rates Short Term 6.00 6.75 6.75 7.00 7.00 7.00 8.00 9.00 Medium Term 6.00 7.00 7.00 7.50 7.50 7.50 8.00 9.00 Long Term 6.00 7.00 7.00 8.00 8.00 8.00 8.50 9.50 * Figure for 1981. 3.11 When the third project was appraised in 1979, it was recognized that interest rates to agriculture were negative in real terms. Nevertheless, GOT was adamant that interest rates were not negotiable and the Bank limited itself to a covenant requiring that a new cost accounting system be established for agricultural loans at BNA. It was hoped that, by identifying the losses made on agricultural loans, BNA would be able to argue for higher rates. The issue of interest rate adjustments was relegated to a side letter. A decision on interest rates was postponed in December, 1983, when the first request for extension of the project was considered, as the Bank argued that a cross-sectoral analysis would be required before it could insist on adjustments affecting agriculture alone. In April, 1985, GOT agreed to the first (and last) adjustment in agricultural interest rates during the life of the third project, raising them by 0.75-1%. 3.12 The fourth project, in conjunction with the first ASAL, attempted to resolve the interest rate issue once and for all. It established a program of annual interest rate adjustments, beginning Y The Bank's Economic and Financial Reforms Support Loan (3424-TUN), currently under implementation, calls for the removal of this interest ceiling by the end of 1993. In a transition phase prior to removal of the cap, the 3% limit is to be applied to the 'average' spreads of each bank's lending rates over the money market rate. - 13 - in 1987, that was to end with rates equal to market rates by 1991. The "market rate" was defined in the Loan Agreement as "an agricultural interest rate, covering average cost of funds, reasonable operating costs and reasonable lending risk of the Borrower". The second tranche of ASAL I was conditional, inter alia, on an adjustment of interest rates in 1987. With droughts in 1988 and 1989, the Bank accommodated GOT"s request that scheduled increases be deferred. Further interest rate adjustments were made in August, 1990 and in July, 1991. At project closing preferential rates were 8% for short and medium term loans and 8.5% for long term loans. The increase to 9% on short and medium term loans and 9.5% on long term loans that was approved in July, 1991, went into effect in January, 1992. After the project was closed, a further increase in preferential agricultural rates was approved in July, 1992, raising short and medium term rates to 10% and long term rates to 10.5%. The new rates became effective in November, 1992, after most seasonal credits for the year 1992-93 had been approved. 3.13 In the light of GOT's earlier reluctance to countenance increases in preferential rates for agriculture, the interest rate adjustments approved during the project were significant, particularly since agricultural interest rates became positive in real terms. Nevertheless, project objectives were not attained as agricultural rates remain well below the (market) rate that would have permitted BNA to cover the cost of funds, reasonable operating costs and, importantly, reasonable lending risk. Supervision documents suggest that in 1991, the Bank regarded the market rate to be in the order of 11.5% to 14.8%, which reflects what BNA might have earned by placing its funds on the money market or by providing loans to industry and commerce. In comparing preferential agricultural rates with the money market rate of interest, the fourth project's PCR (para. 6.03 and Table 2) reveals that less than half the original gap of 2.5-3.5% between agricultural and market rates was closed over the past five years (see also Table 4 above). Clearly supervision and the PCR apply the term "market rate" in a sense closer to normal usage than did the Loan Agreement. The 1991 Economic and Financial Reforms Support Loan 3242-TUN included conditionality pertaining to preferential interest rates in all sectors, and has made a release of the second tranche conditional on a 2% increase. It does not, however, call for the removal of administered rates, which was one of the ultimate objectives of the fourth project (SAR para. 6.03). 3.14 Thus the Bank has adopted a gradualist approach in liberalizing agricultural rates. It has de facto concurred with GOT's view that increasing them towards market levels may at times (e.g. during drought years) have negative social effects that outweigh any economic advantages. The Audit believes that, unless Bank policy on interest rates is revised, it is unlikely that the objective of market- determined agricultural interest rates will be achieved in the next several years, due to the possibility of future droughts, coupled with GOT reluctance to hand over control of agricultural interest rates to commercial banks (because of the political sensitivity attached to these rates). This has important implications for BNA's profitability. 3.15 HARmONIzATION OF LENDING TERms. The existence of multiple lines of credit (more than 30 are listed in Annexes 7 and 8), each with its own terms and conditions, has been a major weakness V Furthermore, the effective rate of Interest on long term agricultural loans Is lower than the nominal rate since interest charged during the grace period Is not compounded and is only payable during the last five years of the loan repayment period. For example, a nominal interest rate of 10% on a loan for 8 years with 3 years grace would result In an effective rate of 8%. - 14 - and source of inefficiency in the agricultural credit sector in Tunisia. This view is, however, not shared by everyone involved in extending loans to agriculture. The Audit mission met with BNA branch staff who approved of the principle of selecting the most favorable line of credit from a menu of options and offering it to the borrower. Neither the third nor the fourth project has resolved the issue entirely. 3.16 Under the third project, Government agreed that terms and conditions for subloans financed with FOSDA resources would not differ substantially from those funded with Bank resources. Schedule 4 of the loan agreement limited access to Government subsidies to small and medium farmers and to production cooperatives, and stipulated the minimum equity contribution required from borrowers as well as the maximum share of sub-project costs that could be covered by Government subsidies. However the statement in Schedule 4 that defined small and medium farmers failed to reflect the tighter, income-based definition given in the appraisal report, namely those farmers whose annual value of farm production does not exceed US$5,000, equivalent to a net income of US$2,000 in 1980 prices. 3.17 On GOT's side, the 1983 decree that defined small and medium farmers eligible for Government subsidies ignored income-based criteria and ended up including over 95% of all farms. At the Bank's insistence, GOT agreed to revise the definition under ASAL I, but the new decree issued in June, 1988, again failed to rely on income as the main criterion. BNA estimates that farmers with revenues of US$15,000 or more can still benefit from subsidies. An MOF decree also issued in 1988 limits FOSDA subsidies to sub-projects of less than TD120,000 (US$134,000), which is a substantial figure for a "small or medium farmer". 3.18 A further departure from loan covenants is the violation of minimum equity contributions. These were set at 5% to 20% under the third project and 10% under the fourth project. It turns out that the minimum equity contribution is in many cases financed by Government credit distributed by regional authorities under the PDR programme. Although GOT claims that these subsidies are only distributed under stricter income criteria, this practice nonetheless reduces certain borrowers' vested interest in their sub-projects and goes against basic credit principles. This problem was never raised by supervision missions. 3.19 Although there has been some progress on harmonization of loan terms and conditions, much remains to be done on this front. In particular, i) although GOT eliminated budget contributions to FOSDA for credit, with a view to turning FOSDA into a fund for subsidy grants only, recoveries on former FOSDA loans are still being used for on-lending; ii) supervised credits in kind are still offered on more beneficial terms than FOSDA or Bank funds; iii) FODERI loans, which require a minimum equity contribution of only 5% and offer credit at 4%, have grown in importance since their inception in 1985. Indeed, by 1990 FODERI commitments had reached TD14.5 million annually, almost as much as FOSDA. Annual budgetary contributions to FODERI are projected to increase to around TD20 million by 1996. The emergence of this rival program at a time when the - 15- Bank has succeeded in convincing GOT to phase out FOSDA loans is disconcerting.' At present, BNA and GOT continue to treat FODERI and a few other lines of credit as exceptions (see BNA's comments in Annex 10). D. Impact on BNA's Financial Health 3.20 ARREARs. The most notable feature of BNA's performance on loan recoveries is the clear positive correlation between the degree of credit risk borne by BNA and the degree of success in recovering loans. The Bank has repeatedly pointed this out to the Tunisian authorities. Figures in Part III of the fourth project's PCR indicate that BNA has recovered over 90% of the principal due on loans made with its own resources, between 58% and 84% on Bank resources (for which it bears risks of 25%-100%), between 10% and 79% on supervised credits (for which it bears risks of 0% to 40%), and 52% on FOSDA/FOSEP (for which it bears no credit risk at all). Dividing the agricultural portfolio into those loans for which BNA bears 100% of the credit risk and those for which Government bears 100% of the credit risk, one finds that the recovery rate was 83% on the former and 52% on the latter loans, as of September, 1992. The average recovery rate was 72%. This represents a decline relative to December, 1988, when the respective recovery rates were 90% and 60%, for an average of 79%. 3.21 The other notable features of BNA's performance on loan recoveries, as evinced by detailed data for the third project (Table 5 below and Annex Table 9), are: i) recovery rates are highest in the North (74%), which has the highest average rainfall, and lowest in the arid southern (38%) and central-southern regions (10%). Within regions, there are large differences at the provincial level that may reflect the effectiveness of BNA branch staff; ii) recoveries are highest for commercial farmers (86% at end-1991 for the third project) and lowest for small farmers (11%). Some of the patterns exhibited by the third project are also reflected in the data for the fourth project: in particular, recoveries are presently about 90% for all categories except small and medium farmers, for which the recovery rate is only 16%. Although the weak repayment performance of small and medium farmers relative to commercial farmers in Tunisia belies the conventional wisdom that small farmers are generally more likely to repay loans than large farmers, there are additional factors in the Tunisian case that may explain the pattern of repayments: first, BNA bore only 40% of the credit risk on small farmer loans compared to 100% of the risk for all other categories under the third project, and second, two-thirds of all loans to small farmers were made in the arid southern part of the country, even though the North has far greater potential for agriculture and received three-quarters of the project's resources overall. Z Annual budgetary contributions to the FOSDA fund for new loans and grants declined from TD18.9 million in 1987 to TD13 million in 1991. Since 1990, the bulk of the new funds have been used for grants rather than loans. However, most of the recoveries on Government-funded subloans (TD7-10 million annually during 1986-90) are on FOSDA/FOSEP loans and are being re-lent. Furthermore, increased lending through FODERI and by external donors has resulted in continued growth in agricultural credit at preferential rates. - 16 - Table 5: RECOVERIES ON THE FIRST THREE BANK PROJECTS, AS OF END-1991 TD'000 Total Not Yet Recovery Loans Due Due Repaid Overdue Rate First Project 5,353 17 5,336 4,492 844 84% Second Project 5,324 875 4,449 3,346 1,103 75% Third Project 24,621 3,632 20,988 12,229 8,760 58% Smallholdera 6,290 857 5,433 5,989 4,834 11% Producer Coops 7,852 1,605 6,247 5,370 877 86% Service Coops 471 37 434 269 165 62% Commercial Farms 4,711 303 4,409 3,305 1,104 75% Agro-industries 5,296 830 4,465 2,686 1,779 60% 3.22 BNA's average recovery rate for all resources of about 72% applies after taking into account rescheduling due to climatic conditions and after a 1989 political decision by GOT to write off TD45 million of loans to some 130,000 small farmers. If these actions had not been taken, BNA's recovery rate would have been below 70%. This still compares favorably with Bank-supported agricultural credit operations in some other countries, but is far below the top performers. 3.23 According to BNA, loan losses on subloans that it finances are very small. The figures for Bank-funded subloans are less encouraging. Judging from the recovery rates on the first two Bank projects, irrecoverable losses appear to be around 15-20% of Bank-funded subloans. For the first project, even though subloans were provided only to commercial farmers and BNA bore 100% of the credit risk, over 80% of arrears are more than five years overdue and current recoveries on those subloans are negligible. Although recoveries are still sizable for the second project, overdues of five years or more account for 56% of all arrears. Irrecoverable losses are probably much larger for Government-funded subloans. 3.24 The fact that BNA's recovery ratio has remained stable over time suggests that its performance on recoveries is not attributable to year-to-year variations in climatic conditions, but rather to an underlying structural problem. If one considers the difference in performance between subloans financed with BNA's own funds and subloans funded with Government resources, or between those for which BNA bears the full credit risk and those for which GOT bears the full credit risk, it becomes clear that the problem is primarily one of client selection and credit risk. By agreement with GOT, BNA selects the safest loan applicants and bears full risk for its pool of clients. These tend to be commercial farmers seeking short term loans, secured by a crop lien, whose crops are usually marketed through official channels (the Office National des C6r6ales). Since BNA does not have the power to veto loans approved by regional credit committees chaired by CRDA V The Impact of the rescheduling undertaken in the mid-1980s on subsequent recovery performance is unclear. However, the Audit believes that the political decision to write off all credit contracted by farmers prior to 1987 with a principal of less than TD1,000 leads borrowers to expect future write-offs, thereby discouraging timely repayment. - 17 - representatives, even though in some cases the applicant is already in default, it bears no risk for loans that it objects to but that are approved. Many applicants fail BNA's financial and technical criteria (e.g. minimum acreage for tractors with a given horsepower) but meet less stringent criteria set by CRDAs.- Thus the Government funds a more risky pool of clients and bears the credit risk. With no credit risk for these clients, and BNA commissions based on loans outstanding rather than on recoveries, BNA has little incentive to invest the staff time required to ensure a solid recovery rate on special funds such as FOSDA, or subloans funded by the Bank or other external donors for which it bears no credit risk.ML' BNA does not even make provisions against bad debt for loans made with FOSDA funds and other funds for which it bears no credit risk. A final and important reason for BNA's relative success with its own clients is that it allocates any partial payments by borrowers first to the settlement of any interest or principal due on BNA-funded subloans, then to interest or principal on loans financed by external donors such as the Bank, and finally to loans made with Government resources. This sound financial rule improves BNA's recovery performance at the expense of other lenders'. 3.25 The Bank has repeatedly asked BNA and GOT to improve recovery rates. In 1983, GOT granted BNA the same legal powers that it uses to enforce tax collection, the privilWge d'Etat. The Bank had asked for this since Morocco's counterpart to BNA, the Caisse Nationale de Cr6dit Agricole, had the privilge d'Etat and its recovery rate was higher than BNA's. In practice, these legal powers are rarely used against small farmers, whose repayment performance is weakest. Although the fourth project's PCR (para. 6.08) reports that in January, 1992, BNA launched a loan recovery program through collection teams with decentralized authority in order to improve the recovery rate, the program appears not to have gotten off the ground. Nevertheless, the Audit believes that the problem of arrears is likely to persist as long as BNA does not have full responsibility for approving clients and does not bear virtually all the credit risk.A 3.26 BNA LENDING FOR AGRICULTURE. Although progress was made on decentralization and on expansion of the branch network with a view to reaching out to the agricultural sector, the Audit is concerned that BNA may not take full advantage of these developments. Notwithstanding BNA's V The PCR for the fourth project (Evaluation Summary, para. 4) reports that there has been a gradual shift in the declslon-making process from MOA to BNA as a result of a 1990 agreement between GOT, BNA and International donors, and some staff at BNA claim that it does have the right to veto applications. The Audit Is certain that this is not the case, at least not in every region. BNA's *no* does carry more weight when it reports that the applicant has defaulted on a previous loan. L' All loans, including the smallest ones, are granted on sound technical grounds, Irrespective of whether they are appraised by CRDAs or by BNA, although in some Instances appraisal forms do not include all the required Information. Based on an earlier mission to Tunisia by a member of the Audit team, the Audit can attest that punitive measures are taken against CRDA staff who provide false information to regional credit commissions. "V A significant volume of Irrecoverable subloans granted through lines of credit that have been inactive for years are still on BNA's balance sheet because it earns a 2% commission on them from GOT. These should be written off. -2 For example, under the third project, BNA bore 40% of the risk on loans to small and medium farmers. This means that it bore the full risk on (up to) 40% of the loans, with GOT bearing 100% of the risk on the remaining 60%. -1 BNA has commented that it intends to Improve the recovery rate on loans made with its own funds (see Annex 10). - 18 - stated purpose to lend for agricultural development, its agricultural portfolio as a share of total lending has shrunk steadily over the past decade. Table 6 below shows that barely half (51.4%) of its outstanding loans in 1990 were devoted to agriculture. The proportion falls to one-third (34%) if one considers only agricultural production and excludes marketing. Furthermore, if we consider BNA's own resources, as opposed to special funds (including Bank funds), only 35.4% were committed to agriculture and a mere 12.9% to agricultural production. Thus BNA is really a commercial bank that lends first and foremost for industry and commerce and provides support to agriculture mainly as a service to GOT, using resources that are to a great extent not its own and that are largely free of credit risk.-' Table 6: BNA'S LOAN PORTFOLIO IN 1990 TD '000 Overall Portfolio BNA's Own Funds Special Funds Agricultural Production Short term 92,352 6.1% 69,184 6.2% 23,168 5.95 Medium/long term 419,494 27.91 75,189 6.7% 344,305 88.2% Agricultural Marketinx Short term 261,629 17.4% 261,629 23.52 -- -- Total for Agriculture 773,475 51.4% 406,002 36.4! 367,473 94.1% Total for Industry, Trade & Tourism 732,239 48.6% 709,522 63.61 22,717 5.91 Total Outstanding Loan. 1,505,715 100.0% 1,115,524 100.01 390,191 100.02 3.27 PROFrABIUTY OF BNA's AGiCULTuRa PoRTouo. The reason for BNA's gradual withdrawal from agriculture (in relative terms) is straightforward: GOT regulation of agricultural lending makes it unprofitable, or at best less profitable than lending to other sectors. This is true for any financial institution in Tunisia. The Bank has tried hard to address this problem. Under the third project, it insisted that BNA develop a cost accounting system that could isolate costs and incomes associated with agricultural credit. GOT was reluctant to release the data to the Bank, but did so at the Bank's insistence just before the first extension of the project's Closing Date in December, 1983. A complete set of figures is now available for 1980 to 1990 (Annex Table 4). 3.28 Obviously, the Bank had little difficulty in demonstrating that BNA's spread on agricultural lending was insufficient to cover its operating costs and repeatedly used this argument to prompt GOT to increase interest rates. In 1983, BNA's analysis suggested that the spread be increased by at least 1% (from around 2%) for BNA to break even, or by 2% for a satisfactory return on equity. ly Although agriculture's share in the total portfolio declined, Annex Table 6 shows that annual disbursements of agricultural loans rose by 226% during the 1980s. Growth In disbursements funded by the Bank and other external donors outpaced growth In BNA-funded or GOT-funded disbursements. Thus the share funded by the Bank and other external donors rose from 5% to 18% during the period 1981-1990. GOT's share fell from 47% to 36%, while BNA's fell from 48% to 46%. - 19 - Now, an increase of at least 3% would be required to break even, as the situation has deteriorated. Although administrative costs have declined, provisions for bad debts have risen and the cost of funds has risen sharply, from 2.1% in 1983 to over 4% since 1989. Even though gross income from interest rose from 2.4% to 3.8% of outstanding agricultural loans over the same period, BNA's net interest income in 1990 was negative for the first time.2 Annual losses on the agricultural portfolio increased from TD1.2 million in 1980 to TD12 million (or 64% of profits on non-agricultural lending) in 1990, for a cumulative total of TD44.5 million for the decade. 3.29 A major reason for this sharp increase in losses is that BNA borrowed funds in the money market at high rates of interest in order to cover a shortfall in Government funding for agricultural loans for which GOT had assumed responsibility. The funds were then on-lent at preferential rates of interest to agriculture. Annex Tables 4-6 and Table 7 below show that the difference between loan commitments to be funded with Government resources and available Government funds rose from less than TD5 million in 1987 to TD15 million in 1988, TD40 million in 1989 and TD47 million in 1990. This problem was compounded by monetary reforms which reduced BNA's access to central bank rediscounts on agricultural loans (which declined from TD69 million in 1988 to TD38 million by 1990). Together with the poor recovery rate, these factors contributed to a major cash flow problem. In the late 1980s, BNA became a net borrower on the money market, raising TD288 million in 1989 and TD327 million in 1990 at rates of 9.5% to 11.8%. Beginning in 1988, it also issued long term bonds, totalling TD45.3 million in 1990 at an average interest rate of 9%. It is therefore not surprising that BNA's interest expense has exceeded its interest income and that losses have grown rapidly. It is not clear why the losses were allowed to increase rapidly for four to five years (1987-1991) before the problem came to the fore. The Audit mission was informed that GOT, while prevailing upon BNA to meet Government lending objectives, restricted its own funding for the lending program. Indeed, one of the aims of the fourth project was to reduce budgetary allocations to FOSDA. On the other hand, shortsighted management at BNA and a lack of communication between BNA and GOT may have left the Government unaware of the extent of the shortfall and of the cost of financing the gap. Furthermore, GOT has argued that the shortfall might not have arisen if BNA's performance on recoveries had been better. Currently BNA and GOT are involved in a dispute over who should pay for the money market advances. The Bank is rightly requiring that the matter be resolved before it proceeds with a follow-up project.L" ly The figures for Interest income are well below the nominal rates charged (about 6-8% during 1983-1990) as BNA only records the commission, not the Interest paid, on FOSDA, Bank and other special funds. In addition, unpaid Interest on loans in arrears Is not counted as income. The arrears are proving to be costly: interest lost on cumulative arrears of TD295 million In 1990 (for all BNA lending) amounted to about TD54 million. Arrears on agricultural loans accounted for 56% of all arrears. M8 BNA has commented that initiatives to obtain reimbursement from the Government are going well (see Annex 10). - 20 - Table 7: BNA'S AGRICULTURAL PORTFOLIO: SOURCES OF FUNDS, CASH FLOW AND PROFITABILITY, (TD'000) 1981 1983 1985 1987 1988 1989 1990 Net GOT Funding for Subloans (152) (5,430) (3,589) (4,743) (15,060) (40,627) (46,738) Central Bank Rediscounta 5,746 12,190 (6,426) 8,464 12,548 (3,609) (27,056) New Long-Term Bonds Issued 25,000 20,291 Money Market Activity * Loans 9,229 18,977 11,395 * Borrowings 13,784 7,993 71,930 55,547 BNA's Gross Cash Flow (13,145) (32,817) (32,202) (18,081) (49,716) (50,686) (54,383) BNA's Net Profit (Loss) (2,013) (1,226) (2,711) (3,459) (3,953) (6,394) (11,991) 3.30 SUSTAINABILrrY OF BNA's AGRICULTURAL OPERATIONS. It is clear that the objective of ensuring the financial viability of credit operations has not been met. BNA's unprofitable agricultural lending is being sustained by cross-subsidization from non-agricultural lending. GOT is also incurring losses on its resources that are not reflected in BNA's balance sheets and income statements. Unfortunately the current presentation of BNA's financial position does not allow for a calculation of the degree to which BNA's agricultural lending is dependent on subsidies. Such a calculation would reflect not only the net losses on operations, but also the need to impute an opportunity cost to the equity used in agricultural lending, the subsidy inherent in the use of resources borrowed on concessionaire terms, as well as other direct subsidies such as the CRDA loan appraisal and approval services that are provided free of charge to BNA. The Bank has developed an index that could be used in future projects to identify such subsidies and thereby help the Tunisian authorities evaluate the true economic costs of subsidized agricultural lending.2' 3.31 Perhaps the clearest indication of the unsustainability of BNA's agricultural lending was the temporary suspension of agricultural credit by BNA in November, 1991. Lending was resumed when GOT provided more funds to BNA. 3.32 BNA's OVERALL FINANCIAL PERFORMANCE. Because of the dismal performance of the agricultural portfolio, BNA's overall profitability is barely growing in real terms, in spite of its apparently sound performance in other sectors (Figure 1). Financial ratios are well below those projected at appraisal of the fourth project (Table 8). Although annual profits on the non- agricultural portfolio increased by nearly 20% in real terms between 1980 and 1990, BNA's overall return on equity has declined from about 24% in the early 1980s to around 8% in 1990, equal to the rate of inflation; growth in equity was below the rate of inflation in 1990. Therefore, the possibility of decapitalization of BNA in real terms is a serious concern. Indeed, at the end of 1991, measures were taken to strengthen BNA's equity base through a new capital increase and to increase profits by charging more user fees to clients (fourth project's PCR, para. 6.08). -n For further details on the Subsidy Dependence Index (SDI), see J.Yaron: "Assessing Development Finance Institutions: A Public interest Analysis", World Bank Discussion Paper No.174, 1992. - 21 - Figure 1: BNA: PROFITS ON ALL LENDING VS. LOSSES ON AGRICULTURAL LOANS (CONSTANT 1985 TD MILLIONS) a. 8 8 MAL PWITS - (12) 1951 1882 1983 1944 19S HiS 1987 1688 1 ea 1190 Table 8: BNA FINANCIAL & PROFITABILITY RATIOS ------ Actual ------ Appraisal 1981 1985 1990 1990 Financial Ratios Risk Assets/Deposits 1.11 1.12 2.05 0.98 Equity/Risk Assets (2) 9.65 8.70 5.55 9.36 Provisions/Risk Assets (2) 1.97 3.17 2.94 12.01 Return on Equity (2) 20.34 24.72 8.26 22.08 Profitability Ratios Income/Risk Assets (2) 6.15 7.94 7.02 8.11 Cost of Resources (2) 2.91 4.53 5.34 4.96 Financial Margin/Risk Assets (2) 6.10 6.74 3.04 6.65 Operating Costs/Total Assets (%) 2.35 2.73 1.73 2.26 IV. FINDINGS AND ISSUES A. Overall Assessment of the Projects 4.1 Although the projects were successful in promoting institutional development and available data suggest that investments at the farm level were generally viable and successful, only partial progress was made on key sectoral reform issues. Overall, however, both projects are assessed as satisfactory but their sustainability is rated as uncertain. 4.2 Though the impact of agricultural credit is hard to measure, the projects contributed to growth in agricultural production and productivity over the past decade by financing inputs and equipment used in proven production processes. The use of inputs increased by 3% annually over - 22 - the past five years in spite of adverse climatic conditions due in large measure to the dynamism of producers encouraged by reforms such as liberalized agricultural prices. 4.3 The third project was the first in the series to begin to tackle some of the sectoral constraints that were hampering efficient extension of credit to agriculture. Little progress was made on such issues as below-market rates of interest, multiple lines of credit with different terms and conditions, or high arrearages, but, for the first time, the costs of agricultural lending were identified, thus providing the basis for negotiations with GOT. 4.4 During negotiations for ASAL I, GOT and the Bank forged a consensus for sectoral reforms that carried over to the fourth project. In May, 1987, the Bank received a policy letter from GOT outlining GOT's intentions to harmonize lending terms on agricultural loans, reduce budgetary contributions to FOSDA and raise interest rates to cover reasonable costs of lending to agriculture. Thus the Bank was optimistic that substantive progress could be made under the fourth project on ensuring the viability of BNA's agricultural credit operations. Unfortunately, back-to-back droughts in 1987/88 and 1988/89 slowed progress on interest rates, which are still Government-controlled and below market rates. Furthermore, although budgetary contributions to FOSDA were duly reduced, GOT introduced a cheaper, alternative source of Government funding for agricultural credit in 1985, namely FODERI. Annual funding for FODERI reached TD14.5 million by 1990 and is projected to increase over the next few years. Overall, the Audit mission found that there was not universal enthusiasm in GOT for liberalizing agricultural interest rates, for making BNA responsible for lending to smallholders, or for harmonizing terms on all lines of credit, including FODERI. No mention is made in the PCR for the fourth project of the savings mobilization programme envisaged at appraisal, and it is presumed that the program was not successfully implemented. Finally, progress on improving the recovery rate has been hampered by insufficient BNA resources devoted to the task. The Government's decision to write off the debts of some 130,000 farmers in 1989 may have compounded the problem by creating expectations of future write-offs, thereby discouraging repayment. Thus, the Audit concludes that, although GOT did cut budgetary funding for FOSDA loans, raise interest rates by 2-3% and harmonize lending terms for some lines of credit, only partial progress was made on key sectoral issues during the fourth project. 4.5 On the institutional side there were several important developments. The new cost accounting system was introduced for agricultural loans, the branch network was computerized and expanded by 65%, lending appraisal and authority for approvals were decentralized, a computerized MIS and a new mainframe were established at headquarters, and a training program was formulated and implemented under the fourth project. The most important institutional issue that remains to be tackled is the need for increased staffing and transport at the branch level so that BNA officers can become more familiar with their small farmer clientele. B. The Future of Agricultural Credit In Tunisia 4.6 Although agricultural production and productivity have increased over the past decade, it is likely that progress might have been more rapid with an undistorted credit system. The overall goals of future operations should be to i) continue to restructure the agricultural credit sector by removing distortions that reduce the profitability of agricultural loans, in order to provide an incentive for the banking sector as a whole to participate in lending to agriculture; ii) ensure that access to -23- credit is not constrained for creditworthy small farmers. The Audit concludes by identifying some important issues and lessons, and by making recommendations for future operations in the sector. 4.7 INTEREsT RATES AND THE SEQUENCING OF REFoRMs. Although a reform program will be more robust if it can be completed irrespective of the sequence of program items, the sequencing of certain reforms deserves attention. In particular, interest rates must be raised to market levels before 100% of the credit risk and responsibility for loan approvals (especially for small farmers) can be transferred to BNA. BNA can earn 11.8% with relatively little credit risk by placing its funds in the money market - it can earn as much as 14.8% on industrial and commercial sector loans. If BNA assumes full responsibility for all loan approvals, it is unlikely to maintain agricultural lending at current levels, particularly to small farmers, unless it is able to earn at least the money market rate of interest on agricultural loans. Thus interest rate reforms are a prerequisite for transferring responsibility for agricultural lending from GOT to the banking sector. This still leaves a number of reforms discussed below, including harmonization of lending terms and certain measures to improve recovery rates, that can be introduced at any time in the reform program. One important prerequisite for any future Bank operations is a solution to the problem of whether BNA or GOT is to bear the cost of the money market advances obtained by BNA over the past few years on GOT's behalf. 4.8 COORDINATION wrrT ADJUSTMENT OPERATIONS. The decision to orchestrate sectoral reforms under the fourth project together with ASAL I was a sound one that yielded a promising consensus. Coordination between the fourth project and adjustment operations was maintained through ASAL II (Loan 3078-TUN), in which agricultural credit sector reforms were included as key actions, rather than as conditions for tranche release, and through the on-going Economic and Financial Sector Reforms Support Loan (3242-TUN), in which the release of the second tranche is conditional on a 2% increase in preferential rates to all sectors. One important measure that calls for coordination between future project and non-project operations is the removal of administered rates of interest, so that BNA can freely determine rates on all its agricultural loans. 4.9 TRANSFERS FOR THE RURAL SECTOR. Although Tunisia's GNP per person was US$1,200 in 1990, over 30% of the population lives in rural areas, where the average GNP per person is only about US$200. Therefore political and equity arguments can be made in favor of the use of the budget process to transfer funds to the rural sector. In order to make these transfers without distorting incentives in the financial sector, the Bank should continue to encourage the replacement of subsidies granted through the credit system by transfers through other channels, (e.g. expenditures on rural infrastructure). This issue transcends the scope of rural credit projects per se and provides further justification for continued coordination between credit projects and adjustment operations. 4.10 HARMONIZATION OF LOAN TERms. The FODERI program, which provides credit at 4% rather than 9%, has grown in importance over the past five years to become a "second FOSDA" mechanism for providing cheap credit. Continued dialogue with GOT should lead to the harmonization of loan terms for all credit programs, including FODERI and the credit line of the Office de Diveloppement du Sud, both of which involve lending at 4%. 4.11 THE LOAN APPROVAL PRoCESS. As lending to agriculture becomes more remunerative, loan- approval decisions should be transferred entirely from regional credit committees to BNA branches. However, as a first step, BNA representatives should chair the meetings of the regional credit - 24 - committees and their veto should be definitive. Rather than override BNA's veto by providing loans to clients that do not meet BNA's technical criteria, GOT could negotiate standard criteria with BNA. With the increase in the branch network, increases in staff and transport are imperative, so as to enable BNA officers to develop the level of contact with farmers that is currently enjoyed by the extension services of the CRDAs. This would also permit the continued expansion of BNA's role in loan identification and appraisal. CRDAs could complement BNA branches by appraising loans for the smallest borrowers or by preparing loans in areas that are not adequately served by BNA branches. 4.12 LoAN RECoVEIES. Neither the PCRs nor the appraisal reports for the two projects have commented on the remarkably low ratio of operating costs to outstanding agricultural loans that BNA has (about 2.7%, compared e.g. to 4.8% for CNCA in Morocco). This no doubt contributes to an explanation for the weak recovery performance. The recovery rate will only improve significantly once BNA has an incentive to invest staff-time in improving collections. Ultimately, this entails transferring virtually all the credit risk for agricultural loans to BNA. In the meantime, BNA could be given an incentive to improve recoveries on special funds by relating the commission paid to BNA on these funds to its recovery performance. Currently BNA earns a 2-3% commission on most lines of credit, irrespective of the recovery rate. This not only discourages active attempts to improve collections but also discourages write-offs of irrecoverable loan losses. BNA's current policy on provisions should also be reviewed with a view to establishing adequate provisions for loan losses. 4.13 CALAmmTY FUND. The recovery rate might also be improved by establishing a well-managed, private sector-operated calamity fund, with crop insurance paid by subborrowers, complemented by budget funds. This idea was highlighted in the PCR of the fourth project (Evaluation Summary: Lessons Learned, and para. 6.09). Tunisia presently has both a calamity fund and national guarantee fund, but neither of them is working. A soundly designed and managed calamity fund could supplant the current calamity grants implicit in debt forgiveness in drought years, which undermines borrowers' incentive to repay in good years. 4.14 ACCOUNTING. Accounting changes introduced during the third project made it possible to review the agricultural portfolio independently of the industrial and commercial sector portfolio. These useful changes should be complemented by further revisions which would permit BNA to draw up separate balance sheets and income statements for agricultural lending based on whether BNA or GOT bears the credit risk. This would enable observers to distinguish the losses made by BNA as well as the losses made by GOT on agricultural credit. Such revisions would also help to identify the subsidies currently received and the subsidies needed by BNA and GOT to maintain their respective portfolios. Finally, such accounting revisions, by making more transparent the responsibilities of BNA and GOT for subloans, may serve to reduce the likelihood of a recurrence of the GOT funding shortfall that occurred during 1987-1991. 4.15 BNA's APPROACH TOWARDS AGRICULTURAL CRErr. BNA believes that it should not be involved in lending to non-bankable farmers. While this is not unreasonable, the Audit is concerned that many viable smallholders may incorrectly be regarded as non-bankable, and therefore experience difficulties in obtaining credit once GOT funding for agricultural credit is removed. The Audit believes that the Bank should devote more attention to the matter in future operations. This will involve dialogue with senior staff and a review of the training program, as well as encouraging BNA to provide additional resources such as staff and transport at the branch level. Savings mobilization -25 - tied to eligibility for credit was proposed under the fourth project but has received insufficient attention. Further study of the experience in other member countries in which local financial institutions enthusiastically address the challenge of lending to small farmers is also warranted. In addition, the problems of insecure land tenure and fragmentation of landholdings constitute obstacles to smallholder credit that merit attention in future agricultural projects in Tunisia. 4.16 CURRENT PROPOSALS FOR A FoLLow-up PRoJEcr. The Bank is currently preparing a National Rural Finance Project to follow up on the agricultural credit series. Current proposals for the project incorporate some of the lessons described above, notably a reorientation of BNA's training program towards small farmer lending, as well as introducing commendable new features such as diversification into non-farm rural finance and banking services for women. Renewed attempts are also planned to involve banks other than BNA in rural credit operations. Proposals for maintaining the impetus for reform provided by the fourth project still need to be elaborated, as they will be the key to attracting domestic resources into fruitful investments in the agricultural sector.  .ft6て01ユ○下』d:．○甲．開卼てlq叫ネょP r叩』0。昭粘－．加団日。．下rユ司叩田tフ」1馴り（.) 日国巨園国園国目国園園日国目国国 加5.16L GCf.566 nL-21田99!.9r9 ！初．ofr Llじ．rLりLヤ冨．乙ot LりL．に械ILり‘LヤどDgS．館寒でnKは ぶ園園価武湖階n冒園図』』』口 日国国園国口日日国国園国国園国国国国日国円国自国園国国国園園目国国国月国日目国国自国国日国国国国日国国円国国国日国■国国日国日国国国国日国■国園国国国国田国日目目国■園■■国国日■■国■園園国■□国国国■国■国日■■国り■■■国■□■■■国■□■■■■■■■■■■国■国園明■昭■ 6ヤり‘械冨．冨ヤ乙0‘乙今6．て100.0りC.1 9関‘900.t C貿．696 石的．gn 9f0.09L 926.frg orて．LgC 的9.ILヤUl国知の庁ロUl気田n1w割結 0L6.20 L的．日乙て10‘冨ぢ656．公f りLg&9ヤ，公6.gC t06．冨f 6て！.of L．乙．Lどり09．ロUzd知1で召協 国■園日日自月田価日日日目日りb園曲田柳月日園国倒陶目D価日園曲日自昭■園国陶昭■園■日国日国昭■昭■日国q■畑■り山咽■月■園■日田り国園■日国目日月日園国園国月日園■園■U陶瑠日価日日国国り月■月日■日日国国自園国日陶昭日国■園国日山昭■園■日国昭国国■日国昭園昭日園国日国国■昭■国国国■昭■日国日国園昭■日■日国昭■園国国自昭り国■園国国自■口昭■ 0L6.6令乙的．じヤ2籵0.6と6石6.OC t乙9．『t 令56．じ冨的6.0罵6にt‘日t L価L&gt ヤ0価．fl 飼U罵園口m 00り‘rt 000．『【000.t’ロり心‘&：口00&Ct 000.’て関0.’て関0&r[000.’て000.’て7は1』n 工エld湖 6Lヤ．[5!．『くtl&t鮪．て6L6.L日ヤ．1乙に［．冨に0.！館9.2て6 てにて．ort 0奮て．LりL 乙6L．に29 に鉱．6に石Tot.5りり口工孟1七Eワ11 11召冨 f館．91t gCヤ．りLと1てに．Lヤ零：tt．ヤ6t 【r日．L61 ど加．L乙1 1冨1．冨tl ●9ヤ‘9な！ヤ9ヤ．ヤC［的に‘乙t 網「ほ111田nlコ園区ル0 f巽‘Lg 56t.09 ヤの1.6ヤ今関．竹009．蛇OLと．ot 日LO‘じ2 gC石．LI 06L.21 令ど6.6 貼口町ロ7電1ロ』飼め1奪UロU [.&．釣0り0．館00［馴四田国月ロ孟ロ国『1 91て．95『LLじ．01に120．ヤ『冨61に．561 0冨0．【t[tL6.tりt LLt‘石ヤ1 tLT.6It てto‘冨0［てgt.6f 7rM昭価血日 如9&Lり［どり．どり［に今‘館EEt．今ど令99．と寒11図山ロ 10T ．曲2 に00．に2相．1どL曲9．今T りts.2て1ロ11ロ1園1 にて9&6[Lり令‘Lて6ど0‘ヤ［6【冨．ヤ［LX．ど［dl馴M 以ぢ‘tg 900‘て9 配t&LS Cてヤ．tヤL館‘どヤリロ馴M 馴四国』1マ1コ園d. ヤfC.9館．t 6的．L6T.1【W.L96 2ヤt&L6g 冨Cて．50石9て自．95ヤヤ90.96に6!9.0今f L10.06ど699.tfて7nり引町日 りLg.tC 0ちや‘Lに209.0て寒Lg．とt ZL6.0！令69.0!L96&L ZLL．乙666&9 ヤ0[&9 9観絡工Gnロ11加山口 2寒C.0：籵りL‘日1 Lじで．!1 6姉．LてLじt&0に0ヤt.9［ヤぢ！.01 石ヤ6.21 自55&0！に69．ど［B召国町園Mロ加臓D国口日1い月 2以‘械ヤ069.Z6f E館．けでにtに．9て1 6に田．て2 EL6&IL 日開‘曲5 fじと．けヤ閥5．田ヤ蛇り．ヤC 粘1馴湖園ロ月口昌 ぢ9ぢ‘C6で●tL．て！じZtL.9鍵0L0‘てCヤ切5．て6t 6［じ‘6館5L0.6てr L99&622 令56．【てて蛇令．石tl 粘19NI口ロ園切園ロ 1り［．り996どL．館ヤ村今．092 1.）』ロ馴臓1加に図り1国利lm』団ロ 園田ほ111月マ瓦1 国国園口国国D国国国国国国国国目月園国日日日国国国日園国国国国国国園国園国国国日園園園国国国開日園国日国国国園国国国園国口日園国国国国国園国国国日国田自国国国国国国日園国国日日国国園国国国国国国国国国国国国園国国馴日園H国国国園園目園自国国国国国園国国園国田国国国 雷ヤり．ヤ【て．2ヤ20．乙ヤ6.[100.0ヤら．[9関．9り0.19蛇．696 990&9鮪Lじ0.otL Lど6．【59 09て．L95 509.IL令引口田価71り割絡 関6．ヤ2 06自．0と9関．【［りtど．01 籵20.6 曽C6.L 16て‘乙て館‘石L6ぢ．ヤ寒5ヤ．ヤ9ユ1州N国ロ1』孟園国 【gt&CIC t65．りL七ヤ11.2ヤで69ぢ．奮fr tと0‘と6とらEC.9ぢとぢり9．けり寒ESg.L61 L99．とtl 切5&1と1 引ほ9留711図』0 tZg.t冨01!．に01[.q 2今［.rてて加‘りて的て．に1 ヤ91.It 6C9.0てて9石．6 W罵．田日Elひ欄1つ1ぷ「Mu1N湖 に0!&OCT 奮ら令．1ヤ［01籵．とに1 L06．ヤて［0t9.tol !ro‘ヤ6 Cじて．ヤL 【51．ど9 OLヤ．じ9 9ヤ9.9ヤ園1』11nつ園価り関国D田召園ロ町口起功 とot.C6り．[!0!‘ヤヤ1.［ヤ約‘LC6 9ら9.Cヤら［IL．蛇石t6石．2雷り00ヤ‘6［今に雷て．IK 916．雷OC 6！ヤ．乙9て別」997おm引嘱武旧て利K隔園ロ田 曲今9.6tて田20.6ヤ1 ！ら［．けCt 9り6&9 L【ヤ．r gtL．りりIC．り10L‘じて69.2 ott．今『mLに園71力罵図U凶 161.0館石6ヤ‘6Cf 51じ．Lヤどに約‘902 L乙2．ヤ6t 5て9‘ヤL［【！9.5令てどてて．LI[ ヤ【2．柏［石Ct‘乙9 （如訓四D馴nlり1コ園d価）飼nび】月ロふ Cヤて‘6雷t tヤ乙‘ヤ［ど鮪6‘ヤ冨ど的と‘2ど2 269&C!:9ヤ0.Lt[C6り．tr![［ヤ．乙11 60f.Lot ど6C&69 田n団陥ロでロ国7飼nび】用口孟－』罵田9 0どL-9にL LEt&0ケt 2知‘Lとf 20今．に11 90に．C!!TOZ.glt 0t6&0[[6石6．で11 てtg．ヤ6 2［に．ぢ日口1胡囲加514fnlt 今rg&0どて冨26&0ヤC L60.06[tC石．公てrn&9 966.！て初公．6てLK.9[6r6.L 69『．rど日lm月のN図飼n 曲－------------&------------------------－名口nり7 0661 6t6t ft6てL柳6Y n6［の6t 姉6！にけ61 9061 lt6[ りの口‘貼 1的昭5刃皿図ps 9心yNH 1 X3NNy －乙乙” -28- ANNEX 2 BNA's lumme Statement and Flinendel Itatios TD-000 1941 1962 1943 1964 1983 19" 1967 19" 1969 1990 MØDOM ---- ---- ---- E~T Om Laøs 20.746 24.733 29,943 60.712 63.184 64.315 72,873 90. "4 14.423 19.606 24.726 5,454 6,704 9,960 39,744 31.964 1,366 1.543 1,6*5 903 1,299 6,998 11.413 30,196 ------- ------- ------- ----- ------- ------- ------- ------- ------- SUITOTAL I~ T u~ 22.351 29,339 36.779 43,*42 56,356 47.04# 71.191 41.273 124.032 13 3. ON OTM 1~ 6.657 6,907 a,003 9,379 12,391 3,479 6,019 804 4.173 3,639 ------- ------- ------- ------- --- ---- ------- ------- ------- ------- ------- TOTAL 1~ 29,004 36.464 ",761 33,261 68,746 72.348 77,110 62,"l 118.203 136,647 Z~" U~T 9~u om DZMITS 9.124 11.521 14.913 17.635 20,817 16.931 42.706 46,033 en 90&~ 6.711 13.936 14,690 16.807 18,094 Z5,315 43,343 31,3" OTM 1.373 1.746 1,615 1,61% 4,316 4.114 4,763 11,663 ------- ------- ------- ------- -- - - -- ------- ------- ------- ------- ------- &V3TOTAL Z~ T XX~ 12,699 17,024 19.210 27.2Z5 34,218 39,291 43.237 44, 3" 90.614 111,316 791~ 3,671 7.200 8.764 9,447 10,148 11,255 12.223 13,629 16,124 16.103 DE= IATION 450 344 591 716 1,024 996 1,211 1,321 1.301 1,770 MM 07ZZATING x~ u 2,006 2,395 3,166 3,644 3.937 4,839 5.Z21 9,140 6.909 7,Z53 aulmvu lo& UD DUT& 2.733 2.666 4,744 6,3468 6,314 4.741 8,355 5,100 6.684 11.353 ------- ------- ------- ------- ------- ------- ------- ------- ------- ------- TOTAL Z~ ts 23,759 30,033 36.479 47,404 59,461 63.113 70, 2" 73.570 112.034 149,997 nZTAX D~ 5,249 6,433 6.302 7.t54 6,887 9.433 6.942 6,311 6,169 6,850 D~ TÅ= 2,877 3.659 41*00 4,047 4,694 4.S13 1,490 904 ------- ------- ------- ----- - ----- ------- ------- ------- ------- ------- MLT D~ 2.371 2.774 3,402 3.411 4.193 4,612 5,072 5,607 6,169 6,850 ALLMATION Or MMITS --------------------- UMAL U312v3 119 139 170 26 343 ZXT~ ~ LUXRTZ 300 1,270 750 750 1,430 1,360 2,000 1.300 1.700 9~ AUISTAffi= r~ Z50 271 275 320 350 400 700 "o 650 1.wo DU~ - r= Z3 Il 23 25 23 13 13 is 23 33 DITIDMD6 540 700 1,035 1,060 1,040 1,900 1,900 2.300 3.300 3,630 TAX-n= ingkvr 943 373 1,474 1,636 1,946 841 9z1 443 324 110 5~ raly» F~ 33 46 67 42 63 99 46 34 67 ------- ------- ------ . ------- - ---- ------- ------- ------- ------- ------- 2,412 2.810 3.444 3,679 4,234 4,695 5,172 5,634 6,233 6,903 Ml$ F13W9CIAL UTIOS 1961 1982 1#63 1984 1963 19" 1967 19*6 1969 1990 ~ UL &åTIOS ---- ---- ---- ---- ---- ---- --------------- DUTIZQ111TY 17.28 19.40 20.70 22.70 23.09 19.77 19.13 28.60 13.64 23.93 &M AssvTli~ lTs 1.11 1.13 1.10 1.11 1.12 1.16 0.94 1.37 1.63 2.03 ~ LisZ A"gn 9.65% Ø.WZ 4.34Z 7.83% 7.45% 6. 70% 9.M 3.33% 6.§OZ 3.55z ~ UOMIALL AM Augn(Z). 1.97% 2.301 2.49x J.161 3.17Z 3.3n 3.óóz 2.64 1. "z 2.94x IX ZQUITT (X) 7.722 8.39% 9.22% 9.26% 19.83% 15.60% -3.39% 31.649 3.in RISK ASIM (X) 13.52% 14.951 16.092 15.on 11.21% 2.13Z 70.8n 22.049 30.741 ~TABILITT AATIO6 ------------------- MO" MOM/TOTAL å~ 7.02X 7.33% 7.71Z 6.35X 7.tig 7.311 6.25% 7.33% 7.30Z 1~ CO6T 3.34x 3.39% 4.0*Z 4.73% 4.óóz 4.612 3.82% 3.60% 3. 72z ~ IAL, Mulm/Xlu 6.73% 7.439 7.18% 7.21X 6.33% 6.269 4.801 3.39Z 3.431 OFM~ C064 1~ å~ 2.30% 2.83z 1.siz I.M 1.601 2.63% 2.22Z 1.79% I.UZ M-TAX ~ m xrpirf 24.en 28.661 24.93% 25.61x 22.84% 13.44% 12.29z 9.42x "IL *0 A-U riAk 468*tø i=l&dø 9~ Z'£^ &***t@ pl» ~**ol.md ~p~s r«ørd»d øf£ -29- ANNEX 3 BNA's Income Statement and Financial Ratios Excluding Agriculture 1981 1962 1983 1984 1985 18 1987 198 1969 1990 IWIEREST Io 20,232 26,793 32,913 41,429 49,594 59,324 63,566 70,735 99,724 134,415 INTEREST REPgS 10,629 14,204 15,462 23,553 30,941 32,258 35,795 37,476 70,633 90,227 MET INTEREST ukxw 9,603 12,589 17,451 17,676 18,653 27,066 27,791 33,259 29,091 44,18 OTHER ICE 5,095 5.161 4,832 7,095 9,478 2.682 2,966 (2,566) 736 (220) FINANCIAL MARGI 14,696 17,750 22,283 24,971 28,131 29,748 30,757 30,693 29,627 43,968 PERSCM COSTS 4,512 5,606 6,740 7,094 7,565 8,601 9,392 10,513 11,466 13.538 OTHER DIRECT COSTS 301 487 748 886 671 1.022 1,009 3.478 1,293 1,435 MEPRACIATION 450 548 591 718 1,024 996 1,212 1,321 1,301 1,770 TOTAL OPERATING COST 5,263 6,641 8.079 8,696 9,460 10,619 11,613 15,312 14,062 16,743 1ET OPERATING roC 9,435 11,109 14,204 16,273 18,651 19,129 19,144 15,381 15,765 27,225 PROVISIOS 2.173 2,296 4,635 4,980 7,053 5,002 6.723 4,917 3,202 6,384 MET POFIT (LOSS) 7,262 8,813 9,569 11,293 11,598 14,127 10,421 10,464 12,563 18,641 SMAPS FINANCIAL RATIOS EXCLUDIM AGRICULTURE 1961 1962 1983 1984 1985 1986 1987 1966 1969 1990 EMT SALADCR SMET DATA ---- ---- ---- ---- RISE ASSETS 145,436 164,091 163,949 228,67 273,243 266,066 284,440 607,932 683,282 965,429 TOTAL ASSETS 331,917 400,259 443,469 533,236 589,918 650,061 729,602 1,135,566 1,339,724 1,564,351 PROV1SIOS 7,905 10.201 14,740 16,892 24,623 29,214 37,076 41.993 49,569 52.60 TMAL LIABILITIES 315,955 362,270 429,667 516,683 577,795 631,940 703,974 1,095,344 1,277,339 1,517,84 QUITT 15,962 17,989 13,582 16,553 12,123 18,121 25,626 40,524 62,385 66,465 FINANCIAL RATIOS GOSS INOM /TOTAL ASSETS 6.732 8.952 9.942 10.522 10.002 9.652 7.312 6.12 9.16 ITEREST EEFEMSE/LIUILITIES 4.072 3.612 4.96 5.652 5.332 5.342 4.172 5.952 6.462 OPERATING COSTS/RIS ASSETS 4.292 4.64Z 4.212 3.762 3.78 4.062 3.432 2.182 2.012 PMVSIOS/RISK ASSETS 5.44Z 6.222 6.01? 8.252 9.012 10.142 13.032 6.912 7.252 5.36! - 30 - ANNEX 4 BNA's Income Statement and Financial Ratios for its Agricultural Portfollo 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 900CE FIUMCING "AP 6,944 6,777 7,993 71,929 55,547 CE~TEAL A 11,574 19,929 32,119 37.402 30,976 47,540 56,004 68,552 64,943 37,887 SIGW DEOITS 21,516 28,284 38,280 52,073 64,346 66,816 77,872 100,731 110,301 133.942 T D~POIT 191 495 350 245 1,153 2,275 2,700 3,470 4,522 4,022 CMEIT0M8 AC T 1,615 1,673 2,135 2,272 4,756 4,838 7,683 4,583 9,646 5,275 LuG TE 9~ 25.000 45.291 zT AL FMDB 5,989 7,164 8.250 11,991 15,907 19,313 21,172 24,121 122,417 137,821 svmyETAET FWD 79,106 89,126 100,764 120,206 127,010 139,555 152,137 180,155 166,121 193.00 DECIAMTM 892 1,046 1,221 2,072 2,541 3,022 3,577 3,822 4,643 5.513 P OmIIM 2,019 2,589 2,798 4,186 5,647 7,386 7,018 7,201 11,326 14,495 QUTT FD8 6,840 6,840 6,840 6,840 6,840 11,500 11,500 11,500 16,500 16,500 TTAL 136,686 163,923 200,750 237,287 259,176 302.245 339,663 404,135 607,348 650,093 AMLICTIM NECE8 wm 15,023 9,229 3,416 18,977 11,395 cTAL lA= 64,943 37,887 AICULTUAL LAMB 8EI1 TE ADfAfCE 3,187 3,872 4,813 6,112 2,633 9,209 6,461 7.122 9,205 7,744 DIS00TD LOM 11,574 19,929 32,119 37,402 30,975 47,540 56,004 68,552 64,943 37,887 MT-FDED L^AM1 21,240 21,781 22,518 14,139 21,991 24,537 16,102 25,150 67,626 100,129 EETf&LLy r1u D LTOM 6,584 7,*09 11,546 17,272 23,157 23,148 28,768 33,494 112,297 124,078 DUDET-1 LOIUS 79,259 91,300 106,194 115,595 130,599 144,191 156,881 195,214 206,748 240,538 0T ER 18 139 134 134 13 SUBTOTAL AGMCULTUAL L0818 121,983 144,825 177.334 190,533 209,355 248,625 264,216 329.532 460,819 510,376 CW UCIS OTEDUE 36,892 47,770 66,764 79.459 89,523 113,567 82,592 94,305 96,824 151,053 DE LITIGATM 4,485 4,939 5,390 5,896 6,766 11,613 11,841 11,822 13,489 13,980 TOTAL I Aå»EA~B 41,377 52,709 72,154 85,355 96,289 125,180 94,433 106,127 110,313 165.033 PUBLIC lOTES 4,168 5,626 7,904 10,019 14,036 15,856 17,404 20,361 20,978 25,280 BmD AM BU~8RE 3,799 4,149 4,309 4,624 5.962 8,042 8,114 6,337 12,944 14.017 rim A~ETS 2,450 2,655 3,269 5,718 6,510 7,533 8,719 10,324 15,088 17,968 AEUL L~08E1 2,014 2,380 1,266 3,436 2,711 4,691 3,460 3,952 6,394 11,991 PRETImUS C~L?ITE L.08E 2,273 4,287 6,667 7,933 11,369 14,081 18,772 22,232 26,184 32,578 TOTAL 136,687 163,922 200,749 237,284 259,172 302,244 339,662 404,133 607,350 650,097 Dote; Differeces betw *sources and application« totale ar* due to roundi% error. - 31 - ANNEX 5 BNA's Sources and Applications Table for its Agricultural Portfollo 1981 1982 1983 1984 1985 1986 1987 198 199 1990 80C FACIw WAP 6,944 6,777 7.993 71,929 55,547 CNTRAL Am 11,574 19,929 32,119 37.402 30,976 47,540 56.004 68,552 64,943 37,887 SIGR D0ITS 21,516 28,284 38,280 52,073 64,346 66,816 77,872 100,731 110,301 133,942 TEE DM ITs 191 495 350 245 1,153 2,275 2,700 3,470 4,522 4,022 CMITOm ACC ~rs 1,615 1,673 2,135 2,272 4,756 4.838 7,683 4,583 9,646 5,275 10 TEM 8D 25,000 45,291 EgEML ~UDS 5,989 7.164 8,250 11,991 15,907 19.313 21,172 24,121 122,417 137,621 9~DGETAY ~UDI 79,106 89,126 100,764 120,206 127,010 139,555 152,137 180,155 166,121 193,800 DEPEIATOM 892 1,046 1,221 2,072 2,541 3,022 3,577 3,822 4,643 5,513 1R~1SI~ 2,019 2,589 2,798 4,186 5,647 7,386 7,018 7,201 11,326 14,495 EQUITT FMuD 6,840 6,840 6,840 6,840 6,840 11,500 11,500 11,500 16,500 16,500 ÅTAL 136,686 163,923 200,750 237,287 259,176 302,245 339,663 404,135 607,348 650,093 APPLICATIM ECEs imnu 15,023 9,229 3,416 18,977 11,395 C ~TEAL ~AM 64,943 37,887 AGICULTUAL LOAM ET TE ADVACE 3,187 3,872 4,823 6,112 2,633 9,209 6,461 7,122 9,205 7,744 DISCOvTE LiAS 11,574 19,929 32,119 37,402 30,975 47,540 56,004 68,552 64,943 37,887 ufT-FoDE LOAM9 21,240 21,781 22,518 14,139 21,991 24,537 16,102 25,150 67,626 100,129 ErTET LLY roDD [AN 6,584 7,809 11,546 17,272 23,157 23,148 28,768 33,494 112,297 124,078 BUDET-Fe~D IA 79,259 91,300 106,194 115,595 130,599 144,191 156,881 195,214 206,748 240,538 orm 1R 85 139 134 134 13 SUBTOTAL &WC.TUAL lA~95 121,983 144,825 177,334 190,533 209,355 248,625 264,216 329,532 460,819 510,376 Or MCEI OVEDUE 36,892 47,770 66,764 79,459 89,523 113,567 82,592 94,305 96,824 151,053 DE LITIGATIN 4,485 4,939 5,390 5,896 6,766 11,613 11,841 11,822 13,489 13,980 TOTAL In aå~58an 41,377 52,709 72,154 85,355 96,289 125,180 94,433 106,127 110,313 165,033 PULIC NOTES 4,168 5,626 7,904 10,019 14,036 15,856 17,404 20,361 20,978 25,280 9DS AD 6A85E 3,799 4,149 4.309 4,624 5,962 8,042 8,114 6,337 12,944 14,017 FTZED AS=ETS 2,450 2,655 3,269 5,718 6,510 7,533 8,719 10,324 15.08 17,968 AMMAL L.08=S 2,014 2,380 1,266 3,436 2,711 4,691 3,460 3,952 6,394 11,991 REIOUS Cu~ÄTITE ös8u 2,273 4,287 6,667 7,933 11,369 14,081 18,772 22,232 26,184 32,578 TOTAL 136,687 163,922 200,749 237,286 259,172 302,244 339,662 404,133 607,350 650,097 ot Differees betmeen sourcas änd applicationc totala are dum to roundig error. - 32 - ANNEX 6 BNA's Cash Flow for Agrcultural an 1981 1982 1983 1984 1985 198 1987 198 1989 199o A) LAM DUSOUEM~n BET TE 1AAM Bzu FMD 15,608 20,163 24,649 29,161 34,912 35,632 43,183 46,846 47,036 32,994 GOE T MD 2,216 3,829 11,347 6,501 9,529 7,111 3,473 4,832 5,788 10,503 EEE&L6OTE D8 0 0 0 0 0 0 147 0 0 0 SUTOTAL 17,824 13,992 38,194 35,442 44,441 42,743 44,805 51,716 52,824 63,499 )MILM TEm Lom BUk FD 1,619 1,817 2,451 907 139 804 1,096 608 906 394 GOV ~lUT PUDI 14,514 17.341 13,316 12,054 22,548 17,588 17.105 26,103 30,130 31,84 ErEAL & OME yD8 1,738 3,318 4,672 8,736 8,53 9,847 4,041 6.381 3,649 20,583 SuiTOrAL 17,871 22,696 10,439 21,697 31,322 18,239 22,262 33,092 36,705 33,041 TOTAL DISMUmmT~ Buk ~UDI 17,227 21,980 29,100 30,068 37,151 34,456 44,281 47,494 47,942 53,390 OOV ~U=T FUDS 16,730 21,190 24,843 18,355 32,077 24,699 20,578 30,935 35,918 42,367 EEMAL 6 OTE FD 1,738 3,518 4,672 8,736 8,535 9,847 4,204 6,341 5,649 20,583 TOTAL 35,695 46,&88 5435 57,339 77,763 71,002 89,067 84,810 89,529 114,540 B) LAM EPAT~T Buk gUDI 15,500 15,18 15,519 25,125 30,246 25,4463 38,945 25,810 28,791 48,8U8 0OT ~T FUDS 6,082 5.934 8,937 10,195 12,924 9,774 9,279 7,277 6,924 9,725 E åAL & Tr VtDS 968 1,003 1,362 1,342 2,391 1,860 2,742 2,007 3,128 3,564 TOTAL 22,550 22,125 25,818 36,6482 45,561 37,101 50,9U 35,094 38,843 62.157 C) 088 Ca g 1a (13,145) (24,563) (32,817) (20,677) (32,202) (33,901) (18,081) (49,716) (50,684) (54,383) BU& FUDS (1,727) (6,794) (13,581) (4,943) (6,905) (10,991) (5,334) (21,684) (19,151) (4,722) GOm ~r rDS (10,64) (15,254) <15,924> (8,360) (19,153) (14,923) (11,299) (23,658) (28.994) <32,642) Em AL &T a r FDS (770) (2,515) (3,310) (7,374) (6,144) (7,987) (1,444) (4,374) (2,541) (17,019) (13,145) (24,563) (32,817) (20,677) (32,202) (33,901) (18,081) (49,716) (50,864) (54,343) D) ME CAM nVFJ ITABf cErAL Am 2EDIgC~Mrs 5,74" 8,355 12,190 3,283 (6,424) 16,564 8,464 12,54 (3,609) (27,054) CAM ~18 AFTEK REISCtMt (7,399) (16,208) (20,627) (15,394) (34,628) (17,337) (9,617) (37,168) (54,295) (81,439) 0om ~T FUDIMN Ir.AUrZ (152) (2,174) (3,430) 4,611 (3,589) (4,435) (4,743) (15,040) (40,627) (44,738) E~T M~ET ACTITITT LI~S 15,023 9,229 3,416 18,977 11,395 b0EROUI~8 13,784 2,490 7.993 71,930 53,547 - 33 - ANNEX 7 BNA's Overdues on Agricultural Credit, November 30, 1991 (TD'000) FUNDS X RISK UTILIZED NOT YET DUE ARREARS IN DISPUTE PAST DUE REPAYMENT % REPAID BNT 100 444,916 51,552 30,618 3,586 393,364 359,160 91% SCH 0-5 28,829 1,230 11,762 27 27,599 15,810 571 IBRD/IDA 100 5,397 10 773 120 5,387 4,494 83% IBRD 1340 TUN 10-25 5,324 703 1,034 212 4,621 3,375 73% IBRD 1885 TUN 40-100 29,729 3,797 8,896 901 25,932 16,135 621 IBRD 2865 TUN 26,978 19,961 3,768 33 7,017 3,216 461 FARM ROADS 25-100 401 8 91 2 393 300 761 FISHERIES 270 TUM 50 877 19 383 183 858 292 34% SECOND FISHERIES 1746 TUN 25 9,539 2,562 6,406 - 6,977 571 8% ONVVM/0NIVAN 10-25 1,917 111 297 6 1,806 1,503 831 FSA 25 7,118 756 655 296 6,362 5,411 85% BEI NEBHANA 25 808 - - 808 808 - - FMO - 2,450 1,343 442 - 1,107 665 601 BUDA GRANT AND LOAN 25-100 173 78 16 9 95 70 741 FCM GRANT AND LOANS - 44 44 - - - - * FODSA - 162,618 62,313 45,988 6,203 100,305 48,114 481 FOSEP - 37,520 9,378 10,969 3,514 28,142 13,659 491 PAN - 3,108 146 302 367 2,962 2,293 771 FADES - 609 238 158 - 371 213 571 FSPA 3 2 1 - 1 - - FSDR NANDIA 2,904 2,252 471 21 652 160 25% PANO 3,162 2,127 463 - 1,035 572 55% CENTRAL TUNISIA 0-5 3,330 956 2,090 - 2,374 284 121 SYLVO-PASTORAL 0-40 2,533 254 1,262 - 2,279 1,017 45% SIDA 0-5 1,097 34 194 25 1,063 844 79% APMANE (USAID) 0-40 52,549 2,104 15,363 120 50,445 34,962 691 FIDA 0-40 14,567 1,966 6,599 16 12,601 5,986 481 FIDA SIDI BOUZID 1 0-40 640 242 375 - 398 23 61 FIDA SIDI BOUZID II 985 [?] 984 54 - 61 7 [?] 11% FODERI 54,161 47,251 5,058 22 6,910 1,830 261 PDA KEF 3,102 2,973 117 - 129 12 9% PDRCF 56 40 13 - 16 3 191 ODS 2,946 2,839 93 - 107 14 131 INTEGRATED PROJECTS (DF) 126,410 106,356 14,191 - 20,054 5,863 291 MISCELLANEOUS 3,335 422 1,169 96 2,913 1,648 571 TOTAL 1,040,135 325,051 170,071 16,567 715.144 528,506 74% -34- ANNEX 8 BNA's Recovery Rate by Line of Credit, 1987-1991 (PERCENT) FUNDS 1967 1988 1989 1990 1991 AVERAGE BM 94 92 93 90 91 92 SCM 44 57 70 64 57 58 IBRD/IDA 779/263 TUN 86 84 83 83 83 84 IBRD 1340 TUN 73 76 78 75 73 75 IBRD 1885 TUN 62 65 64 62 62 63 IBRD 2865 TUN - 68 54 47 46 54 FARM ROADS 72 78 78 75 76 76 FISHERIES 270 TUN 31 33 33 33 34 33 FISHERIES 1746 TUN 4 4 5 7 8 6 OMVVM/OMIVAN 77 78 86 85 83 82 FSA 84 85 85 85 85 85 BEI NEBHANA 0 0 0 0 0 0 FMO - 34 64 59 60 54 BNDA GRANT AND LOAN - 0 11 29 74 29 FODSA 54 53 57 50 48 52 FOSEP 56 59 56 52 49 54 PAN 59 64 60 81 77 68 FADES 67 63 59 57 57 61 FSDR MANDIA - 54 28 28 25 34 PANO - - 92 43 55 63 CENTRAL TUNISIA 21 16 13 12 12 15 SYLVO-PASTORAL 62 66 85 47 45 61 SIDA 59 66 82 79 79 73 APANE (USAID) 76 82 82 71 70 76 FIDA 51 56 63 47 48 53 FIDA SIDI BOUZID I & II - 8 8 7 7 8 FODERI 26 28 27 30 26 27 PDA KEF - - 18 9 14 PDRCF - - - 17 19 18 DDS - - 25 12 13 17 BHAIER & BECHRI F.D. ZONES - - - - 29 29 MISCELLANEOUS 74 78 77 57 57 69 OVERALL RECOVERY RATE 78 79 81 76 74 78 WHWE RWWWWWWWWWWWWUWHEWWEUUUEE UUWWWHma UamEEmaWWaggamaWUammaWasassmagassmanagggaga må - TEI» AffICu.TA . CLDIT n ~OJT (Low Aus-TM) - 8z »OU l UD aCO lE Bn ~IOTINCE TD'000 C&TMO0EY As ALL w E& Fa~ CTOY t PRODUCTI0O COOPUATITEN CATM~T C* BRTICE C UERTITE~ TOTAL MOT YXT MC00ERY TOTAL MT yWT ~ET01ERY TOTAL MOT yET 810C1TEE Pa~VMCE M8 Du Dm RPAID OTEDU 1ATE IMB DUE DU EPAID OTEaDUE MA Ll DUE DU RAID OTDU lATm mEJA 58.8 9.3 49.3 13.4 36.1 27.1% 2866.9 391.7 2475.2 2312.6 162.6 93.41 km ARMUB 23.0 2.8 20.2 11.8 8.4 58.4z 145.3 145.5 113.3 32.2 77.9% ERTE 94.2 3.6 88.6 33.0 55.6 37.2z 990.1 112.2 877.9 672.0 205.9 76.52 31.2 31.2 31.2 100.0% JEx~WA 90.1 42.6 47.3 14.0 29.5 37.9 213.2 15.9 197.3 183.3 13.8 93.0% Lä~tam 92.0 3.9 8.1 42.8 45.3 48.62 212.1 8.7 203.4 166.2 37.2 81.7% 218.0 218.0 124.6 93.4 37.2Z ma . 193.3 10.9 184.4 68.9 115.5 37.42 255.8 0.9 254.9 250.8 4.1 98.4z 43.5 3.0 40.3 14.4 26.1 35.6x ZMMUAM 176.8 18.1 158.7 37.1 121.6 23.42 1328.0 233.6 1294.4 1164.2 130.2 89.92 AUTRE 113.0 16.1 96.9 14.2 82.7 14.72 ------- ---- --------------------------------------------------------------------------------------------------------------------------- MT6 843.2 109.3 733.9 239.2 494.7 32.62 6211.6 763.0 3449.6 4862.6 386.0 89.22 292.7 3.0 289.7 170.2 119.3 38.81 EAIRO081 609.6 89.9 319.7 36.3 483.4 7.02 10.2 10.2 10.2 100.02 135.7 6.3 129.2 9.9 119.3 7.1 661.1 77.3 383.8 394.6 189.2 67.6% 44.6 9.0 35.6 9.1 27.3 22.82 173.4 23.6 151.8 32.4 119.4 21.32 29.9 11.9 18.0 17.7 0.3 98.3 mATI 108.3 7.7 100.6 20.1 80.3 20.0x 5.0 3.0 3.8 1.2 76.02 8II.2ANA 160.3 13.6 146.9 27.0 119.9 18.42 121.3 8.7 112.9 112.8 100.0% 80098E 46.2 7.7 38.5 9.1 29.4 23.6z 42.8 42.8 62.0 3.2 56.8 42.2 14.6 74.32 ------- --- ------ -------------------------------------------------------------------------------------------------------------------- EETE-CUTRAL 1233.7 149.0 10~.7 134.8 951.9 12.42 825.4 128.8 696.6 307.4 189.2 72.8% 151.7 26.1 125.6 82.0 43.6 63.32 Ua~Sete 378.3 104.4 473.9 8.4 465.3 1.82 26.3 7.9 18.6 16.3 2.1 88.7% MPAX 1501.3 104.9 1396.6 82.3 1314.3 3.92 81DI 800 Z1D 1496.0 243.0 1251.0 37.3 1213.5 3.02 80~TM-CWRAL 3573.8 454.3 3121.3 128.2 2993.3 4.1% 0.0 0.0 0.0 0.0 0.0 n... 26.3 7.9 18.6 16.3 2.1 88.72 G~EB 134.7 26.0 128.7 33.4 95.3 26.02 GANA 25.9 15.7 10.2 1.0 9.2 9.8 00MBILI 79.5 79.5 113.9 99.7 14.2 14.2 0.02 mim 298.7 20.0 279.7 49.6 229.1 17.8% TATACm 76.0 3.4 72.6 12.0 60.6 16.3 TOE 0.9 0.2 0.7 0.7 100.0z 701.4 613.7 87.7 87.7 0.02 -------- ------ ------- ------- ------- ---- ------------- ------ ------ ------ ------ ------ ------ 80U11 635.7 144.8 490.9 96.7 394.2 19.7n 813.3 713.4 101.9 0.0 101.9 0.02 0.0 0.0 0.0 0.0 0.0 na. TUTAL& AL EMI(ou 6290.4 837.4 3433.0 398.9 4834.1 11.02 7832.3 1603.2 6247.1 3370.0 877.1 86.02 470.9 37.0 433.9 268.7 165.2 61.92 MMT 13.42 12.72 13.5 39.9N 10.22 79.12 47.5% 87.22 90.62 66.82 62.22 8.12 66.82 63.32 72.32 MM- TAL 19.62 17.41 20.0X 22.32 19.7% 10.5 8.02 11.2% 9.42 21.62 32.22 70.32 28.9% 30.32 26.4% 80UTM-ETRAL 36.82 33.02 57.3 21.42 41.9 0.0% 0.02 0.0 0.01 0.02 $.6Z 21.4% 4.32 6.12 1.32 80023 10.12 16.9x 9.02 16.12 8.29 10.42 44.42 1.6x 0.02 11.62 0.0% 0.02 0.01 0.02 0.02 m& - TmIED AIC~.TUAL CMDIT OJET (LO~ 1885-TO) - &U ~r10 OV RWGOVEI &T MOTIEK TD'000 CAT~RY D* CO~IAL la~ CA Y g& AMBOIMN T9IE TOTAL IM ALL CATEU TOTAL W7 O ET mEcoER TOTAL wrT C SVERT TOTAL MOT TXT aEC0TEK PRO~CE E~Na m Du AID oTEED mAT WaIS DU~ MO EPLIT DO&M O DM T aa DUK nog luPAI OTEDU &ATE MNJA 498.6 24.6 474.0 428.3 45.5 90.41 3424.3 425.6 2996.7 2734.3 244.2 91.91 aM A0M 472.9 27.7 445.2 376.3 68.§ 84.5z 212.9 212.9 212.9 100.01 854.3 30.5 623.8 714.3 109.3 8.7Z BIZEM 647.2 29.6 617.6 512.6 105.0 83.02 56.0 56.0 48.6 7.4 8.8U 1818.7 147.4 1671.3 1297.4 373.9 77.6% J1&n~ 343.8 7.1 356.7 301.5 35.2 90.1I 847.1 65.6 801.3 703.0 98.5 87.7K LAKva& 668.1 61.8 606.3 419.1 187.2 69.1K 2974.3 614.1 2360.2 1946.2 414.0 82.S2 4164.5 688.5 3476.0 2698.9 777.1 77.6K MAEE. 496.1 38.9 437.2 244.4 212.8 53.SZ 1582.6 211.6 1371.0 85.3 1283.7 6.22 2573.3 265.3 2308.0 663.8 1644.2 28.8 ZAå~AN 219.1 19.2 199.9 171.2 28.7 85.62 1923.9 270.9 1633.0 1372.3 280.3 83.0 AUTEE 108.6 7.3 101.1 81.7 19.4 80.8Z 221.6 23.6 198.0 95.9 102.1 48.4U ------- ---------- ----------------------------------------------------------------------------------------------------------------------- 3674.4 216.4 3458.0 2735.3 722.7 79.1 4825.8 825.7 4000.1 2293.0 1707.1 57.32 15847.7 1917.4 13930.3 10300.3 3630.0 73.*K z ~ms8U 20.* 20.8 6.9 13.9 33.22 31.8 31.8 31.8 100.02 672.4 89.9 382.5 85.2 497.3 14.6K om 347.0 33.2 313.8 208.0 103.8 66.3% 1188.4 126.0 1062.4 620.6 441.8 58.4 UANI 27.3 1.2 26.1 16.7 9.4 64.02 232.6 36.7 195.9 66.8 129.1 34.1 le T I 0.3 0.3 0.5 0.02 113.8 7.7 106.1 23.9 82.2 22.3 SIaL 339.0 37.9 301.1 171.1 130.0 56.8 621.0 60.2 560.8 310.9 249.9 53.4% 50088E 65.7 65.7 32.9 12.8 80.32 216.7 35.7 161.0 104.2 36.8 64.7 ---- - ------ --- --- - ------ --------- ------ ----- ----------- ------ ---- - - -- MORT~-1AL 800.3 72.3 728.0 455.6 272.4 62.6% 31.8 0.0 31.8 31.8 0.0 100.0% 3044.9 376.2 2668.7 1211.6 1457.1 45.42 &å~8s9258 33.9 1.3 52.6 24.7 27.9 47.02 57.0 37.0 57.0 100.02 715.7 113.6 602.1 106.6 495.5 17.7Z SrAK 133.2 11.5 123.7 70.4 53.3 56.9 58.0 38.0 46.4 11.6 80.02 1694.7 116.4 1579.3 199.1 1379.2 12.6K SIDI ~M ZID 16.0 16.0 16.0 100.02 1512.0 243.0 1267.0 53.3 1213.5 4.2% ------ --- ------ ---------------------------------------------------------- ------------------------------------------------------ 0TE_-~ZåL 189.1 12.8 176.3 95.1 81.2 53.9K 131.0 0.0 131.0 119.4 11.6 91.1% 3922.4 475.0 3447.4 359.2 3086.2 10.4% 06ABE 10.2 0.8 9.4 3.5 3.9 37.22 164.9 26.8 136.1 36.9 101.2 26.7 06arA 6.2 6.2 4.0 2.2 64.5 20.0 4.5 15.3 12.3 3.2 79.4 32.1 20.2 31.9 17.3 14.6 54.2% ~KILu 193.4 179.2 14.2 0.0 14.2 0.0 ~DI 31.3 0.4 30.9 11.5 19.4 37.2Z 287.0 287.0 229.6 57.4 80.02 617.0 20.4 396.6 290.7 303.8 48.7K TATA 76.0 3.4 72.6 12.0 60.6 16.5K toME 702.3 613.9 "8.4 0.7 87.7 0.8% 808TE 47.7 1.2 46.5 19.0 27.5 40.9% 307.0 4.3 302.5 241.9 60.6 80.02 1803.7 63.9 941.6 357.6 384.2 38.02 TOTALt ALL MIMM 4711.5 302.7 4408.8 3305.0 1103.8 75.02 5295.6 830.2 4465.4 2686.1 1779.3 60.2% 24620.7 3632.5 20988.2 12228.7 8759.5 38.3Z MCTE 78.0 71.SZ 78.4Z 82.8% 65.32 9.1K 99.5% 89.6Z 85.4K 95.92 64.42 52.92 66.42 84.22 41.4I MTR-CMTRAL 17.02 13.9% 16.52 13.8 24.7n 0.61 0.0% 0.72 1.22 0.02 12.42 10.4% 12.7K n .92 16.61 sT-C ~AL 4.0 4.22 4.02 2.9% 7.42 2.52 0.0% 2.9% 4.4% 0.7 13.9% 13.12 16.42 2.9% 35.32 S08!! 1.02 0.4K 1.1K 0.6 2.52 3.82 0.5% 6.8 9.02 3.42 7.32 23.82 4.5 2.9z 6.7 - 37 - ANNEX 10 Banque Nationale Agricole Operations Evaluation Department April 1, 1993 International Bank for Reconstruction and Development 1818 H Street N.W. Washington, D.C. 20433 SUBJECT: Performance Audit Report on the Third and Fourth Agricultural Credit Projects (Loans 1885 TUN and 2865 TUN) REF. : Your letter of February 2, 1993 Gentlemen: In response to your above-referenced letter, we have the honor to provide you with our comments, prepared after reading the report in question and referring to the following points, in particular: 1. Interest rates (a) Small and medium farmers The interest rate for this category of farmer is uniform. It went from 6.75% in 1987 to 10% in 1992. (b) Large farmers (and others) The full interest rate is applied to this category of farmer, i.e. the money market rate of interest plus a 3% or 3.5% margin, with an interest subsidy of up to 3.5% in the case of projects approved by the Agricultural Investment Promotion Agency (APIA). 2. Recoveries (a) Status of overdues The Bank has prepared a program for the recovery of overdues, on the basis of which it will be possible to set recovery objectives by branch and by regional department. This program will be implemented in several successive phases geared to the different categories of farmers. An analysis of the structure of arrears by length of time overdue, amount and source of funds, has shown that it was in 1987 that the total number of overdues (88%), primarily on subloans financed with BNA's own funds, began to show a significant increase. The recovery campaign to be launched in 1993 will therefore, in its first phase, primarily target recent debts relating to this period. Rue de la Monnaie - 1001 Tunis - Republic of Tunisia - 38 - (b) Future recoveries There are plans to improve the recovery rate of overdues, particularly where subloans financed with BNA's own funds are concerned, using all kinds of recovery procedure, depending on the different client categories: sending a demand for payment, making direct contact in the case of clients prepared to settle their accounts, serving notice by a process server in the case of clients who are recalcitrant or do not act in good faith, attachment of bank accounts or property, or other measures. 3. Fungibility of budgetary funds managed by BNA The public authorities and BNA are in agreement on the subject of making the funds fungible. (a) Action taken: - Harmonization of lending terms: * Uniform interest rate * Same loan-granting procedure and distribution network * Single financing plan (b) Actions to be taken by phase: - Reorganizing of 48 or 50 different lines of credit into fewer than a dozen lines at most. - In future, with the exception of a few funds such as FODERI, ODS, the olive fund and the PAAF fund, which are ongoing projects limited in time and intended for funding a specific category of farmer (special rural development projects, etc.), all other projects will be grouped under a single source of funds. FOSDA will also be used primarily for granting subsidies to supplement loans and for financial incentives (reimbursable contributions, subsidies, interest subsidies, land loans). 5. Relations between the Government and BNA (page XVII and 47) With respect to the cumulative budgetary overruns, initiatives to obtain reimbursement from the Government are going well. 7. Government encouragement for establishing a supervised credit system A department for supervised agricultural credits has been set up within BNA for the further development of agricultural loans to small and medium farmers. We thank you for your collaboration and remain, Very truly yours, /s/ Mokhtar Atallah Assistant General Manager MAP SECTION  스：&