Honduras - Third Agricultural Credit Project

Document of The World Bank FOR OFFICIAI, USE ONLY Report No. 8707 PROJECT COMPLETION REPORT HONDURAS THIRD AGRICULTURAL CREDIT PROJECT (LOAN 2284-HO) MAY 31, 1990 Agricultural Operations Division Country Department II Latin America and Caribbean Regional Office 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 8ank authorization.  BANADBS - National. Agricultural Development Bank PCR - Project Completion Report PU - Project Unit UNDP - United Nations Developmet Program  FOR GFICIAL USE ONLY THE WORtO BANK Washnton. D.C 20433 U.S.A. Oie cd Ovcuw-aGr.t Opr0w Evahat.w May 31, 1990 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on HONDURAS Third Agricultural Credit Project (Loan 2284-HO) Attached, for information, is a copy of a report entitled 'Project Completion Report on Honduras - Third Agricultural Credit Project (Loan 2284-HO)l prepared by the Latin America and Caribbean Regional Office with Part II of the report contributed by the Borrower. No audit of this project has been made by the Operations Evaluation Department at this time. Attachment ' This document has a restricted distrbution and may be used by recipients only in the performanee of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  FOR OMCLIL USE ONLY PROJECT COMPLETION REPORT HONDMUR THMD AGRICULTURAL CREDIT PROJECT (LOAN 2284-HO) TABLE OF CONTENTS Page No. PREFACE . .......................................................... i EVALUATION SUMMARY ................................................. ii PART I. PROJECT REVIEW FROM BANK PERSPECTIVE .................... 1 1. Project Identity .................................... 3 2. Background .......... ............................... . 3 3. Project Objectives and Description ............... ... 4 4. Project Preparation and Design ...................... 4 5. Project Implementation ..... .................... 5 6. Project Results ..................................... 13 7. Bank's Performance .................................. 15 8. Project Relationship ...... .......................... 16 PART II. PROJECT REVIEW FROM BORROWER PERSPECTIVE .... ............ 17 1. Introduction ..9...................................... 9 2. The Role of the World Bank ........... .. ............. 19 3. The Role of the Executing Agencies .. ................ 19 4. Overall Impact of the Project ....................... 20 PART III. SUMMARY OF STATISTICAL DATA . . ........................... 23 Table 1. Related Bank Loans and/or Credits ....................... 25 Table 2. Project Timetable ....................................... 26 Table 3. Cumulative Estimated and Actual Disbursements ........... 28 Table 4. Direct Benefits ........... .............................. 30 Table 5. Project Costs ........................................... 31 Table 6. Loan Portfolio by Year and Type of Activity ............. 33 Table 7. Subloans Granted by Banks .............. .. ............... 34 Table 8. Balance of Portfolio and Arrears of Participating Private Banks ..... ...................... 35 Table 9. Project Financing ....................................... 37 Table 10. Internal Rate of Return ............... .. ................ 38 Table 11. Status of Covenants ....... .............. * ............... 39 Table 12. Use of Staff Resources .................................. 41 Table 13. Interest Rates for Agricultural Lines of Credit ........ . 42 GRAPH 1. Shares of Agricultural Exports .................. ........ 43 MAP - IBRD 16747R1 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents mnay not otherwise be disclosed without World Bank authorization.  PROJECT COMPLETION REPORT HONDURAS THIRD AGRICULTURAL CREDIT PROJECT (LOAN 2284-HO) PREFACE This is the Project Completion Report (PCR) for the Third Agricul- tural Credit Project in Honduras, for which Loan 2284-HO in the amount of US$45 was approved on May 19, 1983. The loan was closed on December 31, 1987, on schedule. It was fully disbursed and the last disbursement was on October 3, 1988. The PCR was prepared jointly by Agricultural Operations Division II of the Latin America and the Caribbean Regional Office (Preface, Evalua- tion Summary, Parts I and III), and the Borrower (Part II). Borrower prepared and submitted a PCR in March 1989. On the basis of that report and, inter alia, on a review of the Staff Appraisal Report, the Legal Documents, supervision reports, correspondence between the Bank and the Borrower, and internal Bank memoranda, the Region prepared the PCR. Part II is a summary of the Borrower's original PCR prepared by the Region; the borrower has examined that summary (and the rest of the PCR) and has certified that the summary is accurate and reflects the spirit of its draft PCR.  - il . PROJET C0ONILMTOE REPORT HONDURAS THIRD AGRICULTURAL CREDIT PROJECT (LOAN 2284-E0) EVALUATION SUMMARY 1. This is the fifth agricultural credit project financed by the Bank in Honduras and administered by a qualified Project Unit (PU) established in the Central Bank of Honduras to carry out this type of operations. At the tine of project appraisal, agriculture was, and still is, a key sector In the economy. The world economic recession which existed at that time resulted in stagnant demand and depressed prices for Honduras' main exports and In international trading restrictions. These outcomes lead to relatively weak grokth prospects for its major foreign exchange-earning exports, particularly banana, coffee and sugar. 2. The Government's main agricultural policy objectives as stated in the President's Report of this Project, were to increase overall agricultural production and, in particular, to further increase and diversify agricultural exports. A second major government objective for the sector was to achieve self-sufficiency in food production in the face of rapidly expanding population growth and to improve the income earning ability of the rural poor, especially that of small-scale farmers. It was expected that the resources to be provided under the project for increasing farm investments would help to achieve such increases in agricultural output and to improve income, consumption, production and employment levels of the rural population. Obiectives 3. The main objectives of the project were to promote agricultural exports.so as to improve the country's balance of payments, and to increase rural production, incomes, consumption and employment. It included a line of credit to finance on- and off-farm investments, incremental staff and equipment for the PU, and training for staff of both the PU and the financial intermediaries and for small-scale farmers. Ium2n1e2ation E*perience 4. Successive Bank-financed livestock and agricultural credit operations in Honduras have enjoyed high priority since 1971 when the first livestock project was undertaken. Over time, the project concept has been successfully promoted and has reached a growing number of potential borrowers and financial intermediaries. The Third Agricultural Credit project tried to incorporate new elements, bringing it closer to a general credit line for agriculture than previous operations, which tended to target more specific activities, i.e., livestock activities, specific crops, areas, etc., and included some elements to encourage more commercial bank lending to small-scale farmers.  - ili - 5. During project implemntation, tbhre have been controversial aspects arising from the project's reliance on the Honduran banking system and from the d4iffulties of monitoring performance of participacing banks. In fact three commercial banks accounting for about 33% of the project funds are considered to have substantial financial difficulties with two of the being under special scrutiny by the Bank Superintendency. 6. ^Agenios in aonduras are not controlling the banking syr-em effectively. The Bank Superintendency is carrying out thorough and cempetent audits but it has not ben provided with adequate powers to disclose and enforce its findings. The PU was able to implement an efficient follow-up system for subloan repaymonts by an intermediary, but it lacked quantifiable criteria and power to take effective measures against nouperforuing intemediaries. Therefore, neither the Bank Superintendency for overall compliance with regulatory norms nor the PU for compliance wl.th conditions related to Loan 2284-HO, has been able to achieve improvemnts in the inred eiarl financial performance. 7. Other problems of implementation were an excessive bias towards livestock investments, partly due to tradition as the program was initiated with two consecutive livestock projects (1970-80), and partly to the fact that there is a natural tendency for intermediaries and PU to finance the same type of investments over time. PU should review its promotional policies which still place heavy emphasis on this subsector. S. Another issue during implementation was the prQject's relative failure to incorporate a larger number of small-scale farmers than estimated at appraisal. Its predecessor, the Second Agricultural Credit Project was the first to explicitly include a special loan category to facilitate the monitoring of loans to this type of beneficiaries and allowed higher rediscounting percentages to banks for small-scale farmer loans. In spite of the fact that rediscounting margining was further increased under the Third Agricultural Credit Project, banks have shied away from small-scale farmers as the costs and risks of lending to this sector are not commensurate with permitted lending rates under the Project. In fact the way in which rediscount and on-lending rates were determined under the Third Agricultural Credit Project proved to be insensitive to internal market conditions. Situations arose during project implementation where rates on project funds were above, but more often than not substantially below, going commercial ra.es. Results 9. In retrospect, one could state that there was not sufficient concordance between project objectives and design. The Third Agricultural Credit Project is essentially a financial sector operation while its main objectives, i.e., to increase sectoral output, exports, and employment, are sectoral variables subject to sector and macroeconomic constraints which were totally outside the scope of the project. 10. While the project had significant impact in increased agricultural output and income of directly benefitted farmers, it is not evident that the project had significant impact in increased overall agricultural output and exports. The effects of increased credit was possibly offset by  - iv - advOJe variatiovs tu other variables such an term of trade and on an export trade regime that discriminates against exports. Moreover, efforts to diversify project portfolios over time were weak to have a major impact in diversifying Honduras' agricultural exports. In fact, the composition of agricultural exports in 1987 was substantially the same as the one existing in 1982 (Part III, Graph 1). 11. The main accomplishment of the Third Agricultural Credit Project is to have promitted an insider view into the workings of the financial system in Honduras. The monitoring and auditing requirements accompanying this and previous Sank-financed agricultural credit projects have induced the Bank Superintnd4ency and the Central BSnk to pay more attention to control systems and to the analysis of bank performence than heretofore. Successive agricultural credit operations have in fact come accompanied by technical assistance components directed at improving commercial banks ability to assess and monitor agricultural projects, and its impact has been quite positive. This Project also included a technical assistance component for the BSank Superintendency, which has improved the ability of its staff to monitor the banking system. In addition, in the case of BANADESA, the technical assistance component under the Project has been more ambitious and has sought to redress longer-term problems of defective loan portfolio administration practices and collections. Sustainnbilt1 12. The project is supported by a local Project Unit (PU) which has been in operation for over 18 years and has enjoyed relative stability. It is staffed with well-qualified professionals and has enjoyed insulation from political interference. The PU has been quite successful in promoting the credit line amongst commercial banks and potential beneficiaries, and in providing technical advice to sub-borrowers to carry out the financed investments efficiently. It has been less effective in enforcing financial controls on participating intermediaries. The PU was able to implement an efficient follow up system for subloan repayments by an intermediary, but the project did not provide PU with criteria and power to take effective measures against nonperforming intermediaries. Quantifiable criteria could have been based on the information submitted by the intermediaries to the Bank Superintendency or on that contained in the audit reports which provide a thorough assessment of the situation of each intermediary vis-a- vis the project, as well as its overall performance. 13. If provided with enforceable criteria and the power to take effective measures against nonperforming intermediaries, the PU could serve as the specialized agricultural credit unit within the Central Bank for all agricultural credit lines financed by this institution and not to be confined only to Bank-financed credit operations. In fact, starting in November 1988, and as part of the Fourth Agricultural Credit Project, the PU was provided with a set of enforceable criteria and power to sanction nonperfo:ming intermediaries. The Government started also to implement a plan to consolidate all agricultural credit lines under this unit as a first step to rationalizing the rural financial sector in Honduras. The PU will need to be trained in financial matters, and, in particular, in banking analysis.  Ffndinsl and Lesson (a) By utilizing the financLal system, the project became an instrument for testing the efficiency and needs of the financial Intermediaries involved. It became apparent that the manner in which credit lines are structured and priced is as important in the final balance as the availability of credit. Moreover, it was also apparent that ahievement of stated objectives depends heavily on the financial viability and managerial efficiency of the Institutions on which the project depends for on-lending projoet fuds. Therefore, it was recognized that any new credit operation had to Include an assesmant of th overall viability of the intermediaries to be involved and to set minimum enforceable standards to participate in the project. (b) The project provided a unique opportunity for understanding the manner in which the risks of lending to small-scale farmers are perceLved by formal lenders. Commercial banks in Honduras would aot lend to small farmers even with the incentives provided under the Third Agricultural Credit Project. Undertaking the study to facilitate agricultural lending on chattel mortgages, as proposed at the time of project preparation and appraisal, could have provided new possibilities for increased lendlng to this type of beneficiaries. This study is being carried out under the Fourth Agricultural Credit Project. Also, due to lack of interest of commercial banks, it has become urgent to explore and analyze innovative forms of channeling more credit to small-scale farmers such as credit unions and other types of nongovernment organizations. (c) In the case of BANADESA, association with the Bank through the project was the means for establishing a long-term working program for institutional strengthening, and it provided BANADESA with support vis-a-vis the Government in its quest for a higher degree of autonomy. (d) Duriug appraisal of the Third Agricultural Credit Project, the Bank requested the Government to parcel out the areas that would be serviced by its lines and by other donors. In particular this request was addressed to credit allocations for small-scale farmers in an effort to eliminate lower cost competition for Loan 2284-HO. Targeting credit operations to specific types of beneficiaries, at different rediscount and on-lending rates, has contributed to accentuating the existing rural financial market imperfections and segmentation in Honduras. The need for rationalizing the rural financial system is now fully recognized. To that end, in 1988, the Government began to readjust rediscount and on-lending rates on all agricultural credit lines so as to unify them and bring them closer to the marginal cost of funds to financial intermediaries and market on-lending rates respectively. (e) Pricing of Bank funds for financial Intermediaries and final users linked to the cost of Bank funds to the Honduran Government proved to be insensitlve to local market conditions. Situations arose  - Vi - 4iu project lpl a where rates on the project were above. but more often substantially below. going doestic rates. Therefore, it is recognized that rates to financial intermediaries nd final user should be set in accordance with the marginal cost of funds to financial intermediaries and market onlending rates respectively.  PROJECT COiMPLETION REPORT HONDURAS THfID AGRICULTURAL CREDIT PROJECT (LOAN 2284-HO) PART I PROJECT REVIEW FROM BANK PESRPECTIVE  - 3- PART . PROJECT REVIEWR FR BANK PESPECTIVE 1. tJC fI! Proj ct Naem: Third Agricultural Credit Project Loan No. 2284-HO RVP Unit: lA42A Comtry: Honduras Sector: Agriculture Subsector: Agricultural Credit 2. BACKGROUND 1.2.1 Sector DevelgMnt Obleoties: At t2e time of apiraisal, world economic recession resulted In stagnant demand and depressed prices for Honduras" main exports, and in international trading restrictions. These outcomes lead to relatively weak growth prospects for its major foreign exchange-earning exports, particularly banana, coffee and sugar. Consequently, the Government's main agricultural policy objectives, as stated in the President's Report of the Project, were to increase overall agricultural production and to further increase and diversify agricultural exports. A second major Government objective for the agricultural sector was to achieve self-sufficiency in food production in the face of rapidly expanding population growth and to improve the income earning ability of the rural poor, especially that of small-scale farmers. 1.2.2 Policy Context: Increased access to domestic credit for private investors was expected to help to achieve the above-mentioned production and export objectives. Special features to entice commercial bank lending to small-scale farmers was expected to improve the income earning ability vf this type of beneficiaries. 1.2.3 Linkaees between nroiect. sector. and macronolicv objectives: The Third Agricultural Credit Project was implemented immediately after a major setback of economic and per capita growth in Honduras. Between 1980 and 1983, unemployment reached 20% of the labor force and per capita incomes fell by 10%. Weak public finances during 1980-1982 brought about a larger than usual increase in the use of domestic credit by the public sector and a tightening of credit to the private sector. Investment as a percentage of GDP declined from 26% in 1980 to 15% in 1983 as a consequence of a 50% drop in private investment. In 1983, as part of a stand-by agreement with the IMF, the Government took significant revenue measures that were expected to loosen credit to the private sector. The project, by providing long-term financing together with technical advice to agricultural private sector investors, was expected to contribute to the sectoral objectives, in particular to increase agricultural production and expand exports.  - 4 - 3. PRQJECT OBJECTIVES ND DESCRIPTION 1.3.1 Project Objectives: The project's objectives were to (a) promote agricultural exports; (b) Improve the income, consumption and production levels in rural areas; (c) increase exployment; (d) improve subloan collection by BANADESA; (e) upgrade the coumercial banks' agricultural loan officers' project appraisal techniques; (f) improve small farmers' managerial and bookkeeping capacity; and (g) reduce paper work for subloan approval. 1.3.2 Project Components: Main project components were (a) a line of short- and medium-term credit to channel funds to farmers for rural investments through commercial banks and BANADESA (93.8% of Loan proceeds), (b) incremental PU administration costs, and (c) training for PU staff, financial intermediaries staff and small-scale farmers. 4. PROJECT PREPARATION AND DESIGN A. Identification. Prepsration and Appraisal 1.4.1 The conceptual foundation of the project was sound for its time. The Government assigned high priority to this type of project as they provided medium- and long-term financing to increase agricultural investment and hence agricultural production and exports. Loan 2284-Ho was similar to the four previous Bank-financed agricultural credit operations (two livestock projects and two agricultural credit projects). Project identification consisted of expressions of interest on the part of the Honduran authorities in late 1981 in renewing the then on-going Second Agricultural Credit Project which became effective in September 1980. The on-going project highlighted the need to restructure BANADESA's loan portfolio administration capabilities and find new formulas to involve commercial banks in lending to small-scale farmers. 1.4.2 * There were two short missions in the spring of 1982 staffed with one person to finalize detaLls of the Government's request. Most of the discussions with the Government at the time centered around: (a) new audit procedures to be used by the Bank Superintendency in monitoring the finanuial intermediarier, (b) the dismal results of BANADESA in 1981 and the eligibility of this institution to participate in the new loan, and (c) alternatives to entice commercial banks to lend to small-scale farmers. Among the alternatives considered at that stage were the creation of a guarantee fund and the preparation of a study and the enactment of a chattel mortgage law. The project was prepared in July 1982 by the Central Ba.ak of Honduras, assisted by the UNDP Regional Unit for Technical Assistance. 1.4.3 Appraisal by the Bank took place in September and October of that same year. To avoid competition of available lower cost funds with the Third Agricultural Credit Project, the Bank requested the Government to parcel out the areas that were to be serviced by this line and those of  -5- other donors. Also, Inclusion of BANADLSA was then recommended in spite of Its poor performance, as it was recognized to be the only agency lending to small-scale farmers. The appraisal mission, however, requested the creation of a special unit within BANADESA to handle Bank funds. The only item included in the new operation to entice commercial banks to increase lending to small-scale farmers was an increase in rediscount margining for this type of sub ,ns. B. Lgan =iltk . Negotiations and Board A&roval 1.4.4 The only substantial changes after appraisal and during the Loan Coimittee were a tightening of conditionality regarding the performance of BANADESA. Performance targets were established for (i) loan recovery rates on the Bank-financed section of the portfolio, (ii) submission of a loan recovery plan for its overall portfolio, and (iii) execution of this recovery plan. Failure to comply with such targets would signify suspension of disbursements to BANADESA. 1.4.5 Negotiations took place in Washington in April 1983 and the project was approved by the Board on May 19, 1983. During negotiations, it was agreed that submission of BANADESA's recovery plan would be a condition of loan effectiveness and that special requirements to be included in BANADESA's Project Administration Contract were to be included in a supplementary letter to the Loan Agreement. The loan documents were signed on SeptemLer 29 of that same year, over four months after Board approval, due to delays in congressional approval of the Project's legal documents. Total project costs were estimated at US$74.1 million with the Bank financing US$45.0 million. 5. PROJECT IMPLEMENTATION 1.5.1 Loan 2284-HO became effective in January 1984, four months after the legal documents were signed and only one month after it was originally scheduled. Project was implemented very close to appraisal estimates, although disbursements did lag about one year behind schedule: Actual project costs were about 11% higher (US$82.1 million) than estimates at appraisal, mainly due to a reallocation of funds from small- to large-scale farmers for US$7.3 million (Part III, Tables 3 and 4). A. Implementation of Project Comnonents 1.5.2 Agricultural Credit Components: Livestock activities and large and medium producers were the major beneficiaries of the Third Agricultural Credit Project (Part III, Tables 5 and 6). A total of 3,200 subloans were granted to 6,074 beneficiaries. About 60% of project funds went to finance livestock projects. Loan commitments for small-scale farmers lagged behind original estimates from early stages of project implementation, and an amendment had to be made in December 1986 to reallocate funds from the categories of small-scale to those of large-scale farmers. The actual value of subloans to medium and large producers exceeded original  -6- prdaetiors by 63%, duh to Miaher credit demand by this typ of b.noficiarles and to the existence of alternative lines of credit for small-scale farmers financed by other donors at lower interest rates than those stipulated under the Third Agricultural Credit Project. Table 1! Project Subloans by Tve of, Benef4icia No. of Value No. of Type of beneficiary subloans (L '000) beneficiaries Small-scale farmers 1,615 36,835.9 4,622 Other Producers 1.585 123.010.8 1X452 Total 3,200 159,846.7 6,074 1.5.3 The method used for setting interest rates for final beneficiaries under the project proved impractical and produced a cost of credit on this line that in some periods bore no relation to local market conditions. Rediscount and on-lending rates to final borrowers under the Third Agrieultural Credit Project were linked to the cost of Bank funds to the Honduran Government with no direct link to internal market conditions. Situations arose during project implementation where rates were either too high, but most often too low, compared with the going commercial lending rate (Part III, Table 13). 1.5.4 Train comugWnt: The project allocated US$700,000 for training of Central Bank staff in a number of areas including management, computer, auditing, crop and animal husbandry, etc. It also included training for commereial bank staff in the appraisal of agricultural related projects and training for small farmers. Implementation of this component in particular the one for small-scale beneficiaries lagged substantially behind appraisal estimates and two modifications in the Loan Agreement were necessary to reallocate funds from this component to those on credit (Part III, Tables 3 and 4). 1.5.5 A training coordinator was hired by the PU to upgrade the training activities of both the PU and the other entities associated with the project, ministries, banks, etc. In retrospect, it is not very clear that thls position was justified, as this activity could have been performed by the deputy manager of the PU. The technical advice provided by PU to sub- borrowers on an intensive individual basis through contract with the ?inistry of Natural Resources and other research and extension agencies, although effective, became too expensive and limited for too small a number of farmers. Intents were made to reformulate this training program from direct individual sessions to group demonstrations to farmers but the bulk of the training was provided on an individual basis and oriented mainly to larger cattle ranchers. Also, two professionals from the Bank Superintendency benefited from this component. Results were seemingly satisfactory.  -7.. B. PUs P.rforannea 1.5.6 PU's staff is considered to be professionally competent and stable. Staffing has beeo insulated from political influences and there is little rotation of personnel. There are 80 staff members at the PU, of which some 30 have professional degrees. However, only five of them have training tn finance. In adoition, there are another 65 staff members who work with participating banking institutions; they are engaged in agricultural leoding financed under the Project and by the commercial banks' own resources. A special unit at BNADESA was created under the Project to be in charge only of the funds rediscounted on account of the Project. 1.5.7 PU has been successful at establishing and following procedures demanded by the Loan and Project Agreements. However, as PU staff is less familiar with comercial sspects of credit, it was less successful in implementing the financial aspects contained in Loan 2284-HO; in particular, it was unable to resolve or improve the arrears situation of participating intermediaries. In addition, some intermediaries feel that PU staff is not quite flexible in such aspects as frequency of repayment, minimu size of subloan, monitoring requirements and approval procedures. 1.5.8 The project did not provide PU with a solid set of criteria to evaluate the performance of participating intermediaries. Therefore, given its deficiency of commercial banking knowledge, PU was often more concerned with collecting information on arrears than with analyzing it so as to facilitate subsequent efforts to collect late loans. Persistently high arrears have been a concern over the life of this project (see para 1.5.19). These shortcomings indicated the need to review and revise the job descriptions of PU personnel and the responsibilities of PU itself. The Fourth Agricultural Credit Project accordingly includes more restrictive rules for qualifying intermediaries based on both leverage and arrears records. Punitive actions for intermediaries not complying with loan covenants are to be enforced by PU with assistance from the Executive Committee (see para 1.5.12). 1.5.9 PU appears to allocate substantial resources to the promotional aspects of the project. Thus its staff has been heavily involved in the preparation of manuals, calendars and other publications which both promote use of the credit line and provide didactic instructions to beneficiaries. There are several aspects here that deserve review. 1.5.10 First, the level of resources allocated to these activities seems rather large if the time spent by high level PU staff is imputed. In contrast with the initial years, most farmers in Honduras (particularly larger farmers as are most of PU's beneficiaries) know of the PU credit lines, and, therefore, promotion should be heavily concentrated on commercial banks and other intermediaries on the latest adjustments to lending terms on conditions. Second, PU's marketing effort tends to be too concentrated in livestock investments. This could explain why a large proportion of subloans continue to be destined to livestock activities. PU should prepare a plan each year showing goals in terms of activities to be promoted, profiles of customers to be targeted, intermediation channels to be used, and loan funds to be placed. Supporting activities such as marketing and promotion should be tailored to achieve this annual plan.  -8.- C. Cgordinating Mechanisms 1.5.11 PU has a unique situation vis-a-vis other institutions related to the project. It is subject to general policy guidelines and needs occasional assistance from other government entities such as the M4inistry of Natural Resources, the Central Bank, the Bank Superintendency, etc. At the same time, PU needs to preserve its autonomy in credit operations. A balance has been maintained so far by the general manager of PU who has a reputation for integrity and professional capability, but there is a perceived need to institutionalize coordinating mechanisms by means of an executive coinittee. Such a committee has been a feature in the Third Agricultural Credit Project and in previous operations but not much has been achieved. The Third Agricultural Credit Project reduced the number of members in the Committee, but even with these simplifications, it failed to play any relevant role. Because the Committee was given no specific function at appraisal or thereafter by the Project Director, either representation of executing agencies was vested in individuals of relatively low rank or the meetings were not attended at all. 1.5.12 With more power being provided to PU to take specific actions against nonperforming intermediaries, the Executive Committee is in a good position to support PU's enacting suspension measures and other important aspects of loan monitoring and collection. D. The Bankin System 1.5.13 The Bank has utilized PU and the rediscounting mechanism since 1971 to channel funds into the Honduran agricultural sector. Initially, in the first operation only seven banks participated. In Loan 2284-Ho some 12 private banks plus BANADESA were active in the transfer of project funds. BANADESA has been the main channel for lending to smaller-scale farmers, while commercial banks tended to operate with larger-scale clients. BANADESA plays an important role in reaching this segment of the farmer population and, in spite of its operational problems, efforts have been made to keep it in the program. A total of 6,137 individual projects have been financed since 1971, of which 3,200 correspond to the Third Agricultural Credit Project Loan 2284-HO. The table below illustrates the growing number of projects financed in this manner.  -~~9 Table 2. iculurl Credit Proeram First Second First Second Third Livestock Livestock Agricul. Agricul. Agricul. Program Program Program Program Program Period 1971-73 1973-79 1977-83 1980-85 1984-88 Number 79 310 535 1,602 3,200 Value (L million) 6.6 19.5 37.2 76.5 159.8 1.5.14 Commercia BankPric2inatmi : The credit lines financed through the PU represent the major source of long-term funds to agriculture in the country, and for some banks it represents the only portion of the portfolio in agriculture. However, the relative importance of PU's portfolio on the whole banking system has been declining since 1982 with only one bank as an exception. On average, loans rediscounted through PU represented about 6% of total loans of each bank in 1988, down from 8% in 1982. For some banks, however, the percentage is as high as 15% (Part III, Table 8). It is a matter of concern that major users of project funds present systematically poorer recovery rates. In 1988, the two banks with a PU-financed slhare of their portfolio of over 10% have recovery rates of about 11 points below the average for the banking system as a whole (58% compared to 69%). 1.5.15 The table below illustrates the relative importance of participating intermediaries in Loan 2284-HO. Table 3: Relative Imnortance of Participating Intermediarieg -Aloans & =Aount Avera&2 amount Banco No. % L '000 % L '000 ATLANTICA 48 2.0 3,743.4 2 77.9 OCCIDENTE 262 8.0 12,556.8 8 47.9 AHORRO 36 1.0 3,188.9 2 88.6 BANCAHSA 81 3.0 8,218.7 5 101.5 BANADESA 2,143 67.0 61,704.2 39 28.8 SOGERIN 162 5.0 20,241.8 13 124.9 COMERCIO 25 0.5 2,153.5 1 86.1 CONTINENTAL 169 5.0 18,504.9 12 109.5 MERCANTIL 69 2.0 10,484.0 7 151.9 FICENSA 25 0.5 2,616.8 2 104.7 BANHCAFE 35 1.0 2,080.9 1 59.5 BANFFAA 50 2.0 5,455.5 3 109.1 TRABAJADORES 95 3.0 8.887.3 6 93.6 Total 3,200 100.0 159,836.7 101  - 10 - 1.5.16 There have been some important developments in the banking sector. The most significant perhaps is the agreement under the Fourth Agricultural Credit Project to go towards a unification of interest rates (the Central Bank administers a number of credit lines for rediscount. each with different conditions and pricing terms). Another interesting development is the imcreas. in the financing of governoent by the commercial banks, both In direct loans and as investments in government securities accepted an part of reserve requirements. Reserve requirements in Honduras are quite high at 34% of the deposits, but the funding of commercial banks by the Central Bank via the rediscount mechanism is also very substantial. 1.5.17 lsses of Perf2Lbne Control: At present, it becomes difficult to monitor the performance of the banking system in Honduras due to deficiencies in the banking regulatory and legal norms, in reporting and analysis systems, and to the lack of disclosure and homogeneity in the classification of information supplied by the banks. Financial indicators prepared on the basis of financial accounting data are usually too late to be an effective screening procedure. This is particularly important and partially explains PMs inability to have had better control of loan management practices of participating intermediaries. Compliance with auditing and reporting requirements was generally good, but some violations of normal financial practices were reported. There have been improvements in compliance with the auditing and reporting requirements spelled in the various agreements. Repeated violations such as those shown in Part III, Table 7, evidence (a) lack of effective authority over banks on the part of the Bank Superintendency, (b) lack of resolve on the part of PU in exerting its powers over the banks,1 and (c) the difficulties in monitoring the use of funds. Some improvement can be observed in terms of a smaller number of violations by banks between 1985 and 1987. 1.5.18 Through its participation in various bank sponsored operations, the Bank Superintendency has been receiving technical assistance to upgrade its personnel and systems. In addition, a number of specialists in bank analysis are expecte.& to arrive shortly in Honduras to provide the Central Bank, and in particular the Bank Superintendency, with technical assistance on bank analysis, auditing, and follow up controls. 1.5.19 Arrears: Arrears have been as persistent a problem on the Third Agricultural Credit Project as it was on previous credit projects. Available information on the quality of overall loan portfolios of banks in Honduras indicates a significant deterioration since 1980. The share of loans on time has gone down from 89% in 1982 to 77% in 1987. Given the high level of lending to related parties by banks, as reported ih many of the banks' audit reports, these figures could underestimate the arrears problem since many loans "on time" could be refinanced subloans. Furthermore, reserves for loan losses are reportedly low and in two of the largest participating banks, underprovisioning could be masking a situation of potential insolvency. PU's portfolio reflects some of the problems affecting the overall commercial bank loan portfolio. I/ Under the Fourth Agricultural Credit Project, PU is given unambiguous instructions about actions to take in the event of noncompliance by the intermediaries.  - 11 - Table 4, State of Arrears of Partciltating Intermediaries PU: Loan 2284-HO inem e erO Ital* WTogking capital* Total % Total % Bank. Bal nce Dora rec. Balance mora rec. AZLARTIDA 3,150.1 201.5 63 1,176.7 80.7 84 OCCIDS T 8,350.5 702.3 49 2,720.8 1,309.2 59 AHOlRO 2,063.4 260.7 60 264.9 67.4 84 BANCAHSA 5,151.3 195.8 73 612.4 194.0 66 BAJADESA 48,057.1 7,049.8 42 10,307.8 4,243.9 55 SOGU m 15,028.3 1,205.9 54 2,958.6 393.2 83 COx mCI0 1,279.8 89.7 62 24.0 3.1 90 5CWITXNUTAL 12,546.7 615.8 71 2,037.5 1,162.0 60 MC&NTIL 8,083.9 541.6 61 818.4 79.7 90 TICENSA 2.114.8 142.5 74 310.4 72.4 82 BANHCAPE 904.1 26.2 83 100.0 5.5 8 BANFFM 4,164.1 300.8 67 913.8 313.2 74 TBABAJADORES 6,631.9 262.9 75 2,566.0 198.3 91 Total 117,526.0 11,595.5 - 24.811.3 8,122.6 - Source: Superint. Auditing/BCH as of 12/31/87; * Capital + Interests. BANADESA 1.5.20 As in previous credit projects, BANADESA intermediated the largest amount of ftnds under the Third Agricultural Credit Project. As of December 1988, BANADESA had utilized 40% of the funds and placed about 67% of the number of subloans attributed to the project (see Table 3 above). 1.5.21 The participation of BANADESA in agricultural credit projects has been important in reaching smaller-scale farmers, but the Bank and the PU have continuously expressed concern over its ability to manage project funds. Therefore, and because of its less than satisfactory financial performance, the Third Agricultural Credit Project included conditions related to the management of BANADESA's loan portfolio for its participation in the project. The project contemplated the adoption of a loan recovery plan, a new system for measuring arrears, the creation of a special unit to oversee progress on arrears reduction on the portion of the portfolio financed by PU, and a major review of its organizational and financial situation. BANADESA has been afflicted with managerial and financial problems since it began operations in 1980. The most significant are lack of autonomy, overextension of functions, and unrealistic and conflicting objectives, all of which cont-ibute to persistently high levels of arrears. While it still experiences operational difficulties, it has made substantial improvements in crucial areas of loan portfolio management. The effects of such adjustments, however, would not materialize in improved financial results for some time.  - 12 - 1.5.22 Devlouments: In 1985/86, after several years of implementation failure of several piecemeal plans, including the Recovery Plan presented to the Bank as a condition of Loan Effectiveness, BANADESA embarked in a major restructuring exercise. Several actions were identified: (a) improvement in lendAing and recoveries, (b) divestiture of BANADESA's nonfinancial intermediaries, (c) better savins mobilization effort, and (d) lovering of transaction costs and better design of its reporting systems for management. The main benefits of these actions have been (a) to gain an understanding of the main causes for arrears and the means to reducing them, (b) to provide middle and higher management with intensive on the job training in the areas of loan administration, and (c) to place BSADESA in a position to negotiate for more autonomy and ndependence in respect to the government. 1.5.23 Retrueturing of the Lgan Portfolio: BANADESA is putting new methods in place to control the quality of its lending. Most of 1987 was spent in the preparation of a strategy for identifying the problems and training line personnel in the implementation of a loan recovery plan. During 1988, the plan was adopted in 15 of the 32 agencies, recovery targets were agreed upon for each area with new coordinating mechanisms including active participation from top management. Much of this effort was accelerated with the negotiations of the Fourth Agricultural Credit Project. 1.5.24 As a result of BANADESA's own efforts, the role of the PU unit at BANADESA's headquarters has become unclear and perhaps redundant. This unit is not well integrated with BANADESA's overall loan collection exercise. On the contrary, its existence has contributed to complicate the process of reassignment of loan portfolio responsibilities, since different accounting systems, follow up, and separate personnel is required for subloans financed under the Loan 2284-HO. 1.5.25 Effects on Banadesa's Lendini Activities: Over the past three years, BANADESA retrenched its operations, partly out of lack of liquidity and partly as a result of emphasizing the collection of outstanding obligations rather than new lending. Between 1984 and 1988, annual growth rates for loans went from 38% to less than 1%. At the same time, BANADESA carried out a thorough examination of its loan portfolio and was able to establish a more focused collection strategy. Loans in arrears were classified in accordance with a series of criteria suggested by line staff, and responsibilities for the administration of each subcategory was distributed accordingly. 1.5.26 BANADESA's financial situation continues to be precarious due to narrowing financial margins axd the low quality of its earning assets. However, BANADESA now has more accurate information on the value of its loan portfolio and is planning its activities more efficiently. The manner in which loan loss provisioning is calculated was carefully reviewed and agreed upon once a full assessment of the arrears situation was completed. For these efforts to culminate in permanent results, they would need to be supplemented with reforms in the legal norms governing BANADESA and a series of actions being pursued as part of a possible Agricultural Sector Operation.  - 13 - 6. PROJECT RESULTS 1.6.1 In retrospet, one could state that there was not sufficient cocordance between project objectives and desip. The Third Agricultural Credit Project is essentially a financial sector operation while its main objectives, i.e., to increase sectoral output, eports, and employment, are sectoral variables subject to sector and macroeconomic constraints which were totally outside the scope of the project. 1.6.2 The main accomplishment of the Third Agricultural Credit Project is to have permitted an inside view into the vorkings of the financial system In Honduras. The monitoring and auditing requirements accompanying this and previous Bank-financed agricultural credit projects have induced the Bank Superintendency and the Central Bank to pay more attention to control systems and the analysis of bank performance than heretofore. Successive agricultural credit operations, in fact, have come accompanied by technical assistance components directed at improving commercial banks' ability to assess and monitor agricultural projects, and its impact has been quite positive. This Project also included a technical assistance component for the lank Superintendency which has improved the ability of its staff to monitor the banking system. In addition, in the case of BANADESA, the technical assistance component under the project has been more ambitious and has sought to redress longer-term problems of defective loan portfolio adm',nistration practices and collections. In particular, the project had a clearer impact on strengthening BANADESA's portfolio management and on promoting the adoption of better audit and accounting systems. BANADESA's interest in carrying out reforms on its loan administration systems has been consistently supported by the Bank during the implementation of this Loan and now with the Fourth Agricultural Credit Project. 1.6.3 Project design reflected much higher expectations about the ability of comercial banks and BANADESA to reach small-scale farmers than was justified by previous experience. The project included US$17.6 million or about 40% of loan proceeds for small-scale farmer lending. Insufficient interest of commercial banks in lending to high-cost, high-risk, noncorporate borrowers, was part of the reason for not being able to achieve appraisal targets for small-scale farmer lending. But this was also partly due to the proliferation of lower interest credit lines financed by other donors at the time the Project was being implemented. Therefore, US$7.3 million had to be reassigned to service larger-scale farmers (Part III, Table 5). 1.6.4 The project has contributed to increased output and incomes of directly benefited farmers. However, it is recognized that it is extremely difficult to carry out an evaluation that could demonstrate an unequivocal link between increased credit to increased output. Therefore, it is difficult to evaluate the impact of this project on overall production and exports. The project represented a significant amount of new funds to agriculture (project funds represented, on average, 30% to 35% of new lending to the sector from banking sources during implementation years), but the effect of increased credit might have been offset by adverse vatlations in other variables such as deteriorating terms of trade for  _ 14 - agriculture, deteriorating price tmme for major expports, and a trad. regime that cr e aist epors. 1.6.5 The exception being the dearth of long-term finance in Honduras, which La an obstacle for Increased investmet and hence agricultural production and export growth. Therefore, a major contribution of this project was to provide long-term funds for investment finance which might have provented stronger shortags of term-financing and, hence, dbelines in agricuLtural private invostmnts and growth. The eistence of this administrative intervention is justified to the extent that long-term markets would be slow to develop and their absence is a serious constraint to sectoral growth. 1.6.6 The table below presents a sumasry of the various degrees of accomplishment of these obkectives.  - 15 - Degree of (a) Promote Unclar Deteriorating terms of trade for agricultural agriculture and existence of trade eports regims that discriminate against exports. (b) Improve Good for It is extremly difficult to carry icom _ direct out en evaluation that could consumption beneficiaries domonstrate unequivocal link between & production increased credit to increased output. (c) Increase Poor Many investments were capital mlayment intensive and financed mainly livestock investments. (d) Improve Satisfactory Constraints for further improvements collection are of a political nature. by BANADESA (e) Upgrade PU and Satisfactory Further training for PU staff is banksk' project required in financial and appraisal banking matters. techniques (f) Improve small- Poor Banks not very interested in scale farmer financing small-scale farmers. managerial Project did not explore alternative capacity vehicles to reach small-scale farmers. (g) Less paper- Good Further simplification and work loan decentralization of PU operations approval is still needed. 7. BANK'S PERFORMANCE 1.7.1 The Bank's overall performance was good. The main strengths were a good technical knowledge of the agricultural sector, and support to Government during project preparation and supervision. Weaknesses concerned (a) insufficient knowledge of the rural financial market, and hence insufficient emphasis on financial aspects of intermediaries, and (b) over optimistic assessment of the capability and willingness of the commercial banking system to lend to small-scale farmers. 1.7.2 The Bank responded promptly to Govertment's request to start project preparation. Project preparation proceeded quickly, encountering few problems as this was to be the fifth operation of this kind in  _ 16 - HondUras. Ilosuo at that stage fousoed aSanly on the dismal result of BANADESA and its eligibility to participate in the proposed loan, and in alternatives to entice commercial banks to lend to small-scale farmers. During appraisal, and to prevent competition of available lower-cost funds with the proposed project funds, the Bank requested the Government to parcel out the areas that were to be serviced by the Bank and by other donors. It vould have been more efficient to request the Government to rationalize its rural financial system as part of this operatioa. 1.7.3 Supervision activlties carried out by the Bank were quite meticulous throughout the project implementation. Different missions visited all PU regional offices and a substantial number of final borrowers. The Bank acted in a supportive and flexible manner in responding to Government's request for reallocation of the proceeds of the Loan. Supervision missions were well staffed to adequately assess the ability of PU to provide technical assistance to farmers and in general have focused on understanding the nature of investments funded by the project. A new method to monitor arrears was introduced at the beginning of the project cycle. Thereafter, it seems that supervisions overlooked financial and institutional aspects, in particular, the need to review the financial viability of participating internediaries, which, in a somewhat loose way, was a condition of project fund disbursements. Although supervision staff weeks during FY86 and FY87 were unusually high, very little time was spent in the field. A total of 71 staff weeks were assigned to the supervision of this project (Part III, Table 12). 1.7.4 A more strict review of the financial viability of intermediaries by the Bank, and a subsequent request to PU to take effective actions against nonperforming intermediaries might have helped PU to achieve some improvements in intermediaries' financial performance, in particular with respect to arrears. 1.7.5 Lessons for future Bank-financed projects are: (a) to include a credit specialist, preferably an agricultural credit specialist, during the complete project cycle, and (b) to examine more carefully the relevance of project components with objectives during project review. 8. PROJECT RELATIONSHIP 1.8.1 Bank-Borrower relationship was very good during project preparation and appraisal, and negotiations were quite smooth. Relationship during project implementation was quite supportive. Borrower relationship with project executing agency was normal and the relationship of executing agency with financial intermediaries was generally good.  - 17 - PROJECT COMPLETION REPORT UOIDDURS THIRD AGRICULTURAL CREDIT PROJECT (LOAN 2284-0) PART II PROJECT REVI FROM BORROVEP E  - 19 - PART II. lROJW"T RIvEW FPME BOPOUER PERSPECTIVE 1._ INTRODUCTION 2.1.1 A project completion report was prepared in March 1989 by the Central Bank Project Unit (PU). The Government's PCI contains a detailed description of the project s implemaetation and the Government's perspective on various aspects of the project. It presents detailed data on the credit component by financial intermediary, type of benficiary and type of livestment financed. It also included information on Iupact of livestook investments financed by the project on productivity in different areso. TIis section presents summary information from the Government PCR, foesing on the Government s views and conclusions concerning its own and the Bank's performanc*, and the impact of the project. This part has been approved by the Borrower as stated in the letter dated March 21, 1990, also attached to this section. 2. THE ROLE OF THE WORLD RANI 2.2.1 In the view of the Project Unit, the role of the World Bank in the execution of this Project has been satisfactory. The Bank was prompt in providing support and in solving operational problems. The timing of supervision visits has been very adequate, especially with respect to the frequency and quality of the visits. The Government also appreciated the continuity of the supervisory staff who always worked in conformity with their terms of reference. 3. THE RGLE OF THE EXECUTING AGENCIES 2.3.1 With the exception of the Project Executive Comaittee, the role of the executing agencies has been mostly satisfactory. The aforementioned committee did not perform as expected; the Government feels it should be done away with, and its coordination tasks be reassigned to the Project Unit. 2.3.2 BANADESA has played a very important role as financial intermediary in channeling subloans to small-scale farmers. It processed a large number of subloans (about 67% of the total) although, on the average, the amounts involved were small. 2.3.3 In addition to BANADESA, twelve commercial banks partic'1pated as financial intermediaries in the execution of this Project, with four of them channeling more than 40% of the loan funds. 2.3.4 By the end of 1987, loan recuperation rates in most of the intermediaries ranged from 60% to 83% of capital investment loans, with the exception of BANADESA which showed a low 42%. This bank is still suffering from low loan recuperation rates mainly due to administrative deficiencies.  - 20- 2.3.5 As for the Project Unlt, Its financial performence was considerably affected by the financial burden caused by foreLgn exchange risks and increased administrative expenses due to the Unit's involvement ia activities outside the credit component. 4. VERLLIMPACT O H RJC 2.4.1 Overall, the Project was sucCessful in achieving a greater diversification of food crops, increasing livestock and milk production, and Improving the trade balance by contributing to increased exports of agicltural products and to Import substitution. The financial and economic impact of the credit component is considered acceptable It is estimated that the Project will continue to have a strong impact on the increasing agricultural production in the future.  _ 21 - TraslatiLon iuto Enlish of letter addressed to Mr. V. Nerc$sstanta, Portfollo Ia_ar, dated Mwch 21, 1990 from Mt. Raon Narvaz. Deputy Director, UPCA/BCH. Dear Sir, We would appreciate your lending us the Project Completion Report for the Third-Agricultural Credit Project (lan 2284-lID) prepard by yu, bae on the draft submitted to you by our Project Unit. At the same tim, I would like to exted my saicere apologies for the delay in his rply, caued by the nomination of Ing. Mario Nuflo Gmero, Director of the Project Unit, as Minister of Natural Resources of this county. I have revmeved the above-mentioned document and confirm my coWplete agreement with the contents of Part I1 of this document. Also, we are making note of the new guidelines of the World Bank in the form of the prosentation of the Project Completion Report, which will undoubtedly facilitate this task in the future. Sincerely, Signed by Ramon Narvaez Deputy Director UPGA/BCH P.S. The Spanish translation of the Report suffers from some minor deficiencies which could make its analysis and interpretation difficult for persons who may or may not be connected with the Project.  - 22 - Teuci