Mexico - Agricultural Credit Projects

Document of The World Bank I' FOR OFFICIAL USE ONLY Report No. 8926 PROJECT COMPLETION REPORT MEXICO EIGHTH AGRIUCLTURAL CREDIT PROJECT (LOAN 2454-ME) AND AGRICULTURAL CREDIT PROJECT (LOAN 2610-ME) JULY 25, 1990 Agrlculture Operations Division Country Department II Latin America and Caribbean Regional Office This docmen$ has a reficted dsbtribo ad ay be used by nden y in de pearinam i of er o dudes. Id se be wboeWodd Ilank _,o  ABBREVTATIONS ACF/CPP Average Cost of Funds index published by Bank of Mexico BANXICO Mexican Central Bank BANRURAL National Rural Credit Bank CETES 28-day Treasury Bill Interest Rate CPI Consumer Price Index CONASUPO National Commission for Distribution of Basic Food FEFA Trust Fund for Investment Credits FEGA Trust Fund for Agricultural Insurance and Technical Assistance FIRA Agricultural Trust Fund comprised of FONDO, FEFA, FEGA and FOPESCA FONDO Trust Fund for Working Credits GIRA General Interest Rate Agreement GOM Government of Mexico IDB InterAmerican Development Bank IBRD International Bank for Reconstruction and Development (World Bank) LIPIPBI Low Income Producer MIP/PIH Medium Income Producer OTP Other Producer PCR Project Completion Report SAM Government-sponsored Mexican Food Program SAP Special Action Program SHCP Finance Ministry/Secretarin de Hacienda, de Credito Publico  FOR02 OMFCIALUS ONLY THE WORLD BANK Washmgton, D.C. 20433 USA.- Of*of ice t Dtvmnsra July 25, 1990 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECTs Project Completion Report on Mexico Eighth Ag2icultural Credit Project (Loan 2454-ME) and Agricultural Credit Project (Loan 2610-ME) Attached, for information, is a copy of a report entitled 'Project Completion Report on Mexico - Eighth Ag-!icultural Credit Project (Loan 2454-ME) and Agricaltural Credit Project -(Loan 2610-ME)' 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. Attacbment - Thi docunmsn a artrictedbdtbution n may be ud bW repents ony i te efonnune of their ofki duts. its contenuts may not oterwie be dbicled wihout Word Bank autwoItinL  FOR OMFFICAL USE ONLY PROJECT COMPLETION REPORT XMCO RIGHTE AND INlTRuM AGRICULTURAL CREDIT PROJECTS (LOANS 2454-HE AND 2610-ME) Table of Contents Page No. PREFACE . ......................................................... i EVALUATION SUMMARY ................................................. iii PART I: PROJECT REVIEW FROM BANKS'S PERSPECTIVE .....1.............. 1. Project Identity ................................... ..... 1 2. Background ............................................... 1 3. Project Objectives and Description ................... .... 3 4. Project Preparation and Design ..... ...................... 4 5. Project Implementation .... ............................... 5 6. Results .................................................. 11 7. Institutional Performance ...... ......................3.... i 8. Project Relationship ....... .............................. 15 PART IIs PROJECT REVIEW FROM THE BORROWER'S PERSPECTIVE .. ....... 17 1. Introduction ......... .................................... 21 2. The Role of the Bank ..................................... 21 3. The Role of the Executing Agency .......... .. ............. 22 4. Categories of Beneficiaries . . . 23 5. Project Cost and Financing ............. .. ................ 23 6. Characteristics of the Investments and Subloans .......... 24 7. Lessons Learned from the Project ......................... 26 PART IIIs STATISTICAL TABLES ...... * .............................. 29 1. Related Bank Loans and/or Credits ........................ 31 2. Project Timetable/Project Schedule ......... .. ............ 33 3. Distribution of Loans Resources by Category of Beneficiary ........ 4 ................................ 37 4. Project Costs by Source of Financing ..................... 39 5. Cumulative Estimated and Actual Disbursements ............ 41 6. Direct Benefits ........ .................................. 43 7. Rates of Return ........ .................................. 44 8. Support Services Component ...... ......................... 45 9. Status of Covenants ..................... I .. 47 10. Use of Staff Resources ....... ........................... 49 11. Total Number of Loans Granted by FIRA ............ ....... 51 12. Weighted Intermediation Margins to Participating Banks .... 52 MAP - IBRD 11789R2 * For the sake of clarity, Loan 2610-ME, the Agricultural Credit Project, is herlein called the Interim Agricultural Credit Project. This document has a festricted distribution and ma be used by repients only In the performance of their offici duties. Its contents may not otherwie be disclosed without World Dank authormtion.  - i.- PROJECT COmPLETION EPORT MEXICO BIGTHT AND INRTEWDI AGRICULTURAL CREDIT PROJECTS * (LOANS 2454-ME AND 2610-ME) PREFACE 1. This document is the Project Completion Report (PCR) for the Eighth and Interim Agricultural Credit Projects (Loans 2454-ME and 2610- ME). The Interim Loan did not involve a separate appraisal process; it only required the preparation of a President's Report as it provided additional resources to the Mexican Government's agricultural credit system based on the then recently fully appraised Eighth Agricultural Credit Project. The Guarantee, Loan and Project Agreements for Loan 2454-ME were signed on August 8, 1984, while those for Loan 2610-ME were signed on September 26, 1985. These two loans were fully disbursed on January 21, 1986, and on December 22, 1987, respectively. 2. The PCR was prepared by the Latin America and Caribbean Regior, Country Department II, Agriculture Operations Division. Preparation of the PCR was based on project files and data provided by the Borrower. Project. papers reviewed include the Staff Appraisal Report for the Eighth Agricultural Credit Project dated June 4, 1984, the President's Reports for the Eighth and Interim Agricultural Credit Projects dated June 4, 1984 and June 28, 1985, respectively, the Guarantee, Loan and Project Agreements for the two loans, supervision reports, correspondence with the Borrower and internal Bank memoranda on the complete project cycle. This report was also based on the Borrower's Project Completion Report dated October 20, 1988. For the sake of clarity, Loan 2610-ME, the Agricultural Credit Project,; is herein called the Interim Agricult~ral Credit Project.  - iii - -PROJET COMPLETION REPORT MEXICO DIGQUT ARD INTERIM AGRICULTURAL CREDIT PROJECTS (LOANS 2454-ME AND 2610-ME) EVALUATION SUMMARY Introduction 1. Through the Eighth and Interim Agricultural Credit Projects, which disbursed from December 1984 to December 1987, the Bank provided US$480 million equivalent to the Government of Mexico (GOM) in support of the agricultural credit system. These loans were administered by a qualified second-tier agricultural trust fund established in the Central Bank of Mexico, FIRA. Prior to these two loans, the Bank made seven similar agri- cultural credit 'oans to Mexico for US$925 million equivalent administered by the same trust fund, which disbursed from 1966 through 1984. The Bank made also a similar subsequent loan for US$400 million equivalent, the Ninth Agricultural Credit Project, which started disbursing in 1987, and is still current. 2. The general economic conditions present in Mexico during the decade of 19809 greatly affected the design and implementation of the Eighth and Interim Agricultural Credit Projects. In the late 1970s, GOM became increasingly concerned with the impact of the agricultural trade deficit, and the inability to achieve self-sufficiency in the production of grains. To address these issues, the GOM embarked on an ambitious plan of expansionary spending in the agricultural sector. This expansionary spend- ing policy had a negative macroeconomic impact, one of which was to fuel inflation, increasing from 26Z in 1980 to 1022 in 1983. While after 1982 COM started to work on reducing these inflationary pressures and managed to bring down inflation to about 60X per annum during 1984 and 1985, the abrupt oil price decline in 1986 reversed the recovery process. 3. While inflation was very high, the agricultural sector was being heavily taxed through administered prices and other marketing controls in order to promote industrialization. To partly compensate the deteriorating terms of trade in agriculture, interest rates for agricultural activities were maintaired at levels substantially below the rate of inflation. 4. By the end of 1987, the crisis was so severe that GOM negotiated an agreement between labor, farmer and industry leaders (the Solidarity Pact) in a concerted attempt to lower inflation which continued in triple digits, to renew structural reform efforts in order to cut down GOM subsi- dies, to liberalize trade, and to further tighten fiscal and monetary poli- cies. Substantial progress has been made along these lines in 1988 and 1989; GOM subsidies have been heavily reduced while the inflation rate has been dropped to itg current level of about 20? per annum. These severe  - iv - economic conditions presented the frazewort in which the Eighth and Interim Loans were prepared and implemented. Objective 5. The Eighth Agricultural Credit Project moved forward during 1984, with the important instrument of the General Interest Rate Agreement (GIRA) which was expected to provide a protective armor against further decapita- lization of the agricultural lending institutions. GIRA provided for reduction of subsidies via negative interest rates by narrowing the gap between the interest rates for agriculture and the Average Cost of Funds (ACF) for commercial banks. In addition, the proceeds of the Eighth and Interim Agricultural Credit Loans were to be contributed to FIRA as capital to reinforce its financial position. 6. Therefore, the main objectives of the Eighth and Interim Agricul- tural Credit Projects weres a) to decrease rural financial market imperfections by reducing interest rate subsidies and restricting them to specific target groups; b) to continue to strengthen FIRA's institutional framework for efficient allocation of financial resources by: (i) increasing PIPA's capital, and (ii) supporting FIRA's productive support services to producers and commercial bank staff through techni- cal assistance and training-, and c) as a consequence, foster economically sound investments geared at increasing agricultural productivity and production. Imlementation Experience 7. GIRA was to be the main tool for substantially reducing interest, rate subsidies under the project. Therefore, a significant part of the supervision time was spent on discussing and negotiating compliance of the interest rates conditionality under GIRA. However, implementation of the GIRA for agricultural credit lines became politically difficult. During the implementation of the twvo projects, the inflation and ACF rate did not decrease as predicted at the time of appraisal; instead, they increased substantially, while production costs were increasing more rapidly than output prices. The options available to the Bank were either to (i) stop the flow of Bank funds to the agricultural credit system and leave interest rates for agricultural subloans to be even more negative than otherwise, or (ii) accede to waivers and postponements of interest rate adjustments as stipulated under GIRA with the view to closing the gap between interest rates for agriculture and the ACP, while at the same time maintaining the flow of funds to support the system. After protracted negotiations, the Bank and the GOM opted for the second option. 8. - In spite of the fact that the levels of interest rates for agri- cultural activities were increased above the initial expectations with respect to the nominal ACF, the effective interest rate for agriculture  \~~~~~~~ continued to be highly negative during the implementation of the two projects. This was mainly due to high inflation and the way interest rates for agricultural activities are charged to borrowers.l/ Under these conditions FIRA's annual lending decreased by about 20Z in real terms during 1984-1987 as the GOM was not able to fully compensate FIRA for the losses of its portfolio due to negative interest rates. FIRA's lending decreased in spite of the fact that producers' credit demand remained high and institutional agricultural credit outside that of FIRA and the GOM's rural development bank, was quite small. 9. An important element of the project design which helped FIRA stabilize its financial position was the condition that the proceeds of the two loans under review would be contributed cost-free as equity to FIRA. This condition was introduced in response tc. the alarming decapitalization of FIRA during the period of 1981-1983, when the Seventh Agricultural Loan was implemented. Implementation of the Eighth Agricultural Credit Project was compressed into a third of its original time frame due mainly to the strong motivations that the GOM and FIRA had to substitute internal financ- ing from Banco de Mexico to FIRA with Bank resources. By speeding up the disbursements of Bank loans, FIRA received free resources, instead of receiving borrowed funds from Banco de Mexico, while the GOM received badly needed foreign exchange. Implementation of the Interim Loan lagged behind its original closing date due in part to a substitution of Bank funds for IDB funds resulting from an mRB 1oan whiea disbursed between 1986-87. The slowdown of the Bank-financed loan was also due to high onlending interest rates; while still negative in real terms, these were so high that farmer demand for investment loans dropped substantially. On the other hand, demand for short-term lending increased at a time when commercial banks did not have sufficient resources for counterpart funding. In those circum- stances, the Bank agreed to reallocate funds from long-term to short-term lending, and allow PIRA to increase its rediscounting margin to commercial banks. With these modifications, the Interim Loan was fully disbursed with a six-month delay from the appraisal estimate. Results 10. Productive Imnacts The results from the Monitoring and Evaluation unit of FIRA suggest that investments financed by FIRA were financially and economically sound. However, due to fungibility problems, it is difficult to evaluate to what extent investments rediscounted by FIRA and financed by the Bank are net incremental investments. 11. In spite of the above, FIRA's close supervision ensures that real investment in agriculture is rediscounted by commercial and development banks. Moreover, even though negative interest rates lessen borrower's incentives to invest carefully, they do not remove it altogether. During project implementation, sectoral distortions discouraged agricultural investment and could have cancelled out incentives via subsidized credit. 1/ Interests in agriculture are calculated and paid once a year based on the nominal ACF, while commercial lending, based on the same indicator, requires monthly payment of interest.  - vi - Thersfore, if some FIRA loans were accompanied by their borrowers' trans- fers of resources out of the sector, it was the sectoral environment and not FIRA, which was responsible for that. It is quite likely that there has been net agricultural disinvestment in Mexico during the implementation of the two projects despite FIRA's financed investments, but there could have been even more disinvestment without FIRA financing. 12. Institutional Impacts At the end of the project period, the gap between interest rates for agricultural activities and the AC! was smaller than anticipated at ape.raisal although interest rates remained highly nega- tive in real terms and substantially below commercial rates throughout. Consequently, FIRA's financial assets were further eroded. The GOM absorb- ed the majority of FIRA's losses by providing FIRA with relatively cheap credit through rediscounts from Banco de Mexico, and by granting the proceeds of external loans (mainly from IBRD and IDB), as capital contribu- tions. During 1984-1987, the implementation period of the two projects under review, the estimated subsidy via cheap credit to FIRA (in 1988 constant dollars) amounted to US$1.9 billion equivalent (Mex$4,392 bil- lions) while GOM capital contributions from proceeds of external loans amounted to US$1.0 billion (Mex$2,377,763 million) in 1988 currency. Howe- ver, FIRA's capital during 1984-1987 increased only by US$0.2 billion in 1988 currency (from Mex$874,627 million to Mex$1,288,490); the balance US$1.7 billion subsidy received by FIRA through cheap funding and capizal contributions was passed on to the final borrowers through negative lnte- rest rates. 13. The above-mentioned implicit and explicit transfers to the FIRA system were quite insignificant compared with the resources that the GOM transferred to the EANRURAL system (BANRURAL, FICART and ANAGSA) during the same period. It is estimated that during the period 1984- 1987 the BANRURAL system received US$7 billion (in 1988 currency value) in the form of transfers and as GOM assumption of internal and external debt, while the implicit subsidy via cheap credit is estimated at US$3 billion (in 1988 currency). Most of these subsidies were required to maintain a system with little credit discipline and hence poor repayment performance. In contrast, the commercial banking system which rediscounts over 90 of FIRA's financing has been able to maintain a good repayment performance during the project's implementation period and beyond. 14. FIRA's operating costs have remained nearly constant in real terms; therefore, as percentage of its declining average loan portfolio, these costs show a rising trend during the life of the projects. Overall operating costs averaged over 42 per annum. Although these costs might be high for a second-tier financial institution, they are mostly associated with services provided by FIRA to ensure sound investments through the provision of technical assistance, and to increase the access of small- scale farmers to commercial lending through the provision of a guarantee fund. 15. FIRA's operating deficit between its average operating costs and the 3.5Z spread provided by Banco de Mexico for its rediscount operations has been compensated by the GOM through transfer of funds from externally financed loans as capital contributions to FIRA, and through fiscal trans-  - vii - fers to cover FIRA's technical assistance costs. However, fiscal transfers have been substantially reduced and, since 1989, foreign financed loans are transferred to FIRA at CETES rate. Sustainabilit, 16. As mentioned above the FIRA system, and hence the project, could not be self-sustained. In spite of GOM's increasing transfers via Banco d't Mexico and resources lent to the GOM by the Bank and IDB, FIRA could not even maintain its level of lending in real terms. FIRA's contracting effect on agricultural lending was even more pronounced if one considers that, during the project period, rediscounting proportions for lending to other producers were revised upward; therefore, lesser and lesser resources from the commercial banking system were mobilized to the agricultural sector. By maintaining interest rates to agricultural lending substan- tially below commercial interest rates and the level of inflation, the GOM displaced savings that would have otherwise been mobilized for agricultural investments either by commercial banks and/or borrowers. Currently, inte- rest rates for agricultural activities are positive in real terms but are still substantially below commercial rates and are insufficient to cover the marginal cost ot funding and transaction costs of rediscounting agen- cies and development banks. An adjustment of interest rates would be necessary for the system to become self-sustained in the medium _2-rm. Findings 17. The project was reasonably successful in maintaining some level of investment in the sector in spite of the adverse conditions under which these investments were made. The evaluation of VIRA's activities, and of the projects in particular, suggests that investments financed by FIRA were financially and economically sound. However, due to fungibility problems, it is difficult to determine to what extent investment rediscounted j VIRA and financed by the Bank are net incremental investments. 18. Disbursements of the Eighth Agricultural Credit Loan were accele- rated to a third of its original time frame mainly due to the strong moti- vations that GOM and FIRA had to substitute internal financing from Banco de Mexico to FIRA with Bank resources. This accelerated disbursement was done at the expense of a stronger Bank position to a decrease in interest rate subsidy. However, the Bank loan helped the GOM in its difficult balance of payment situation during the implementation of the two projects. 19. The GIRA was successful in closing the gap between interest rates for agriculture and ACF but was marginally successful in closing the gap of negative interest rates in periods of high inflation. It was also less successful in ensuring financial self-sufficiency of the financial institu- tions. However, the GIRA mechanism for consultation and negotiation between GOM and the Bank was mostly effective in raising the level of consciousness of the GOM as to the devastating effect that negative inte- rest rates have on its financial institutions, and the associated high fiscal burden placed on the Central Gov'vrnment to replenish thls capital through fiscal transfers.  _ viii - 20. Because fiscal transfers-to IRA have been substantially elimninta- ed while foreigL-financed loans are now transferred to FIRA at CETES rate, FIRA would have to accelerate the transfer of costs of supporting services to producers and to commercial banks. In addition, it would need to increase its rediscount rate to a level sufficient to cover its marg%inal cost of borrowed funds plus its financial intermediation costs to pt:vent further decapitalization. Not to do so would limit the catalytic role of this institution to induce comercial lending to the agricultural sector, in particular, in light of the recent liberalization of the financial system. 21. The cost to the GOM of maintaining the FIRA system has been unsus- tainable. However, such cost has been only about one tenth of the cost the G0 has had to bear to malu ain the BANRURAL system where misallocation of resources has been widely pread and credit discipline has deteriorated. Therefore, to address the kisues of the rural financial sector in the future it would be necessary to include reforms not only to the FIRA system. but also to the BANRURAL system.  PRCM=T COMPLETION REPORT MEXICO 31GM! AGRICULTRAL AND INTERiM AGRICULTURAL CREDIT PROJECTS' (LOANS 2454-HE AND 2610-ME) PART It PROJECT REVIEW FROM BANK'S PERSPECTIVE 1. Project Identity Project Names: Eighth Agriculturail Credit Project * Agricultural Credit Project Loans No.: Loan 2454-ME Loan 2610-ME RVP Unit: LA2AG Country: Mexico Sector: Agriculture Subsector: Agnicultural Credit 2. Background 1.2.1 The general economic conditions present in Mexico during the decade of the 1980's greatly affected the design and implementation of-the- Eighth and Interim Project. In the late 1970's, the Government of Mexico (GOM) became increasingly concerned at the impact of the agricultural trade deficit (about US$ 1.1 billion, or 2 percent of total merchandise exports) and the inability to acbh ve self sufficiency in the production of baslc grains. With the resources provided by increased oil revenues, the GOH embarked on an ambitious subsidy plan which included increases in pro4i%cer prices through the CORASUPO enterprise, maintenance of consumer prices at very low levels, increasing GOM expenditure on agticulture and th& expan- sion of the bureaucratic machinery to implement these strategies. Total GOM expenditures increased trom 28.0 percent of GDP in 1979 to 35.1 percent in 1981 and to 41.4 percent in 1982. This expansionary spending policy had negative macroeconomic repercussions, one of which was to fuel inflation, which increased from 26.3 percent in 1980 to 58.8 percent in 1982 and 101.9 percent in 1983. While the Adminiistration which= took office in 1982 was working on containing and reducing these inflationary policies, the abrupt ' For the sBke of clarity, Loan 2610-ME, the Agricultural Credit Project, Is herein cqlled-the Interim Agricultural Credit Project. I.~~~~~~~~~, -3 I / -  -2- oil price decline in 1986, which partially forced a sharp devaluation of the Mexican peso, further delayed the recovery process. The subsequent targeting of the exchange rate, together with an increase in the frequency of wage and cost adjustments, added to the instability of the economic system. By the end of 1987, the crisis continued to be so spvere that the Government negotiated an agreement between labor, farmer and industry leaders (the Solidarity Pact) in a concerted attempt to lower inflation which continued in triple digits, to renew structural reform efforts in order to cut down GOM subsidies and liberalize trade, and to further tighten fiscal and monetary policies. Substantial progress has been made along these lines in 1988 and 1989 with GOM subsidies being reduced or eliminated and the inflation rate dropping to its current level of about 20 percent annually. This severe economic crisis, which is briefly outlined above, presented the environment in which the Eighth and Interim loans were disbursed. These loans, as weil as the many other Bank Loans approved and disbursed during the last decade, have assisted Mexico to evolve from this crisis. 1.2.2 Mexico's agricultural sector contributes about 10 percent of the gross domestic product, employs about one-third of the country's active labor force and contributes about 6 percent of the country's exports. During the period 1965 to 1980, agricultural production growth rates have been progressively declining to a point at which, in 1979, the country experienced its first agricultural trade deficit. In 1981, the deficit exceeded US$1 billion. As a result of government intervention in 1981 and 1982 in the form of substantially increased farm gate prices and the intro- duction of a range of input subsidies in the context of the Government- sponsored Mexican Food Program (SAM), agricultural output rebounded at an annual growth rate of 8.5 percent. Although successful in increasing food production, the program proved too costly and was abandoned in 1982. 1.2.3 While agricultural zredit grew continuously up to 1981, growth came to an abrupt end in 1982 when institutional agricultural credit fell nationwide by about 28 percent in real terms, and again in 1983 it fell by 25 percent. It recuperated in 1984-85 and fell again during the next three years. In 1982 and 1983, the Government found itself unable to finance the large gap between the available funds generated through the recoveries on the existing loan portfolios at devalued amounts, and the demand for new agricultural loans. It is estimated that, in 1983, the banking system's impact on agricultural financing, measured as a proptrtion of GNP, fell from 25 percent in 1980 to about 20 percent, and therefore it satisfied a significantly smaller proportion of the agricultural credit demand. Exter- nal sources of funding such as that provided by the multilateral develop- ment banks became especially important at this time. In late 1983, the Bank agreed to a general Special Action Program (SAP) for Mex4co which had the effect of assisting in filling this funding gap via accelerated disbursements of Bank loans allocated to FIRA. 1.2.4 As noted in paragraph 1.2.1 above, the economic conditions present during the period 1980 to 1984 were not favorable. The combination of foreign debt demands, fiscal deficits and persistently high inflation rates. diminished the options available to the Government to address the credit demand in the agricultural sector. FIRA had suffered the alarmingly  -3- high decapitalization rate of an average of about 30 percent per year during 1981, 1982 and 1983. And, interest rates to agricultural producers remained highly negative as compared to the annual inflation rate until the second half of 1988. Currently, interest rates to agricultural producers are highly positive although substantially below market rates. 1.2.5 Mexican agriculture remained basically stagnated during the period 1984-1988. Combined output for the main crops of cereal, oilseeds and beans increased substantially in 1985 (14.5 percent) to 29.6 million tons, but then dropped again in the period 1986-88 to about the 1983 level of production (25.2 million tons). Yields per hectare for these same crops remained very stable, fluctuating between 1.9 and 2.1 tons per hectare during the period 1983-1988. Imports of basic grains peaked at 9.4 million tons in 1983 and were approximately halved to 4.6 million tons by 1987. As indicated in paragraph 1.6.1 below, the targeted beneficiaries of FIRA subsidized credits, generally showed improvement in farm income and genera- tion of employment for the period 1984-1987. 3. Prolect Oblectives and Description 1.3.1 Proiect Obiectivess Within the difficult financial context described above, which was well documented by the Bank, the Eighth Agricul- tural Credit Project moved forward with the important element of the Gene- ral Interest Rote Agreement vhich was expected to provide the protective armor against further decapitalization of the agricultural lending institu- tions. GIRA provided for the reduction of subsidies via negative interest rates. The stated objectives of the project were much in line with the previous seven agricultural credit projects channelled through FIRA. They were as followst a. continue to strengthen PIRA's institutional framework for effi- cient allocation of financial resources; and b. improve agricultural sector policy by reducing interest rate subsidies and restrict them to special target groups. c. increase agricultural productivity and production with a view to (i) increasing exports, (ii) improving food self- suffi- ciency, and (iii) increasing farm income; 1.3.2 Project Description: The specific functional project activi- ties were as follows: a. investment credit and short-term loans to farmers for crops, livestock and agroindustries (US$477 million proceeds from Bank Loan targeted to this component); b. technical assistance to farmers to enable appropriate transfex and adoption of improved technology (Part of US$3 million proceeds from Bank Loan earmarked for Production Support was targeted to this component); and  -4- c. training to FIRA's technical staff and that of participating banks, and to strengthen FIRA's medium and long term develop- ment planning (Part of US$3 million from Bank Loan earmarked for Production Support was targeted to this component.) 4. Proiect Preparation and Design 1.4.1 Project Identification. Preparation and Appraisal. The Eighth and Interim Agricultural Credit Project design was to a large degree developed in direct response to the implementation difficulties experienced in the then ongoing Seventh Agricultural Credit Project with FIRA. During the latter stages of the Seventh Loan, the Government was experiencing fiscal shortages which limited its counterpart fund availability; and the higher than expected devaluation rate over the inflation rate for 1982 and 1983 had the effect of slowing down Bank disbursements on the loan. In 1983, the Bank implemented a Special Action Program (SAP) to Mexico to increase the disbursement percentages for Bank financed projects in Mexico, which had the effect of rapidly disbursing the remaining 50 percent balance of the Seventh loan and setting the stage for an accelerated disbursement of the Eighth Loan. 1.4.2 In early 1983 when the Initial Project Brief was circulated, Programs commented that the proposed project could not be a continuation of previous projects, but would have to be targeted to specific groups, regions and subsectors. However, it was then considered that to follow such recommendations would have increased the segmentation of the rural financial market, and therefore, it was decided that the Eighth Project would become a continuation of the preceding projects, with two exceptions. The first was that 30 percent of the loan was allocated to short-term work- ing capital credits; all previous Bank loans financed only medium- and long-term subloans. The second was that the overall disbursement percent- age was set at 63 percent to facilitate fast disbursement within the frame- work of the Special Action Program for Mexico. 1.4.3 The Eighth Agricultural Credit Project (Loan 2454-ME for US$300 million) was appraised in November 1983, and approved in mid- 1984, coin-' ciding with the finalization of the General Interest Rate Agreement (GIRA), the umbrella interest rate agreement for all credit operations financed by the Bank in Mexico (see paras 1.5.6 to 1.5.8). GIRA formed the basis for approval of the loan by the Loan Committee and the Board. 1.4.4 The project appraisal report for the Eighth Loan identified a cumulative projected cash flow shortage in FIRA's operations of US$490 millions which would be met by the rescheduling of existing loans or through the negotiation of new loans. This apparently was not possible in the short term and the rapid disbursement of the Eighth Loan filled the gap. 1.4.5 The Eighth Agricultural Credit Project (Loan 2454-ME) was planned to be carried out over a four year period with the Completion Date set for March 1988 and the Closing Date for September 1988. The Loan was fully committed by FIRA as of August 1985 and fully disbursed by the Bank as of  -5- January 1986. This acceleration of the execution of the Eighth Project led to a Government request for the Interim Loan (2610-ME). No appraisal report was prepared for this loan which was approved by the Board in Decem- ber 1985. 1.4.6 The organization of the projects called for loan proceeds to be channeled through the FIRA crganization, with which the Bank has had a long and good relationship. It was decided to exclude the Government's specia- lized Rural Development Bank, BANRURAL, as a direct executor of loan proceeds. BANRURAL was heavily subsidized (due not only to negative on- lending interest rates but also to lack of credit discipline and hence poor recovery performance) and not ready to accept institutional reforms of the kind FIRA had been gradually introducing over the years. The amounts included in the loan to assist FIRA in improving its productive support activities to producers and commercial bank staff through technical assis- tance and training were merely a continuation and refinement of this type of support included in the previous seven projects with FIRA. 1.4.7 An important element of the design of these two projects which helped FIRA stabilize its financial position (even though at the expense of the Government as a whole) was the condition that the proceeds of these loans would be contributed cost-free as equity to FIRA. This was done in response to the alarming decapitalization of FIRA during the Seventh Agri- cultural Credit Loan in the highly inflationary period of 1981-1983. This action did nothing to ameliorate the distortion in the credit system caused by negative interest rates, but did provide a temporary extension of the financial viability of FIRA. Moreover, the granting by GOM of "cost-freea funds as contributions to FIRA equity could have been a disincentive to a more cost effective provision of its support services. Overall FIRA tech- nical assistance and administrative costs, expressed as a percentage of average annual portfolio, increased 2.5 times during the period 1980- 1988. A large portion of this cost is, however, due to technical assistance to farmers and commercial banks; these institution-building activities have been very important and their impact is reflected in PIRA's impressive reputation throughout the region. However, this increasing trend in FIRA's account should be reversed by transferring such costs to its beneficiaries on an accelerated basis. 5. Project Implementation 1.5.1 Loan Effectiveness and Project Start-up: Although Loan 2454-HE was approved on June 27, 1984 and signed on August 8, 1984, it did not become effective until December 3, 1984. A key consideration by the Board for loan approval was the recently negotiated GIRA for Mexico (see para. 1.4.3). As stated in the SAR, interest rates stipulated in the GIRA were to apply to FIRA total lending program (para 2.04 of SAR). The Mexican Government, however, did not adjust interest rates in accordance with GIRA. The loan, therefore, did not become effective until the Government took this initial step in December 1984. 1.5.2 In October 1984, the consent of retroactive financing under loan 2454-HE was increased from US$20 million to US$30 million. In the mean-  -6- time, the Bank had agreed to allow FIRA to commit and disburse subloans at the old interest rates until end December 1984, at which time the revised rate was to be introduced. FIRA issued instructions on the new interest rate structure in March 1985, and these new rates cam into effect in April 1985. FIRA disbursements between January and March 1985 did not conform to GIRA; a continuous exchange of reimbursement applications was necessary to ensure that the Bank would disburse against subloans which complied with GIRA. This difficulty in complying with GIRA at the outset was to be indi- cative of Government sensitivity towards decreasing interest rate subsidies throughout the life of the Project and beyond. Interest rate subsidies have been justified to compensate producers for below-market output prices due to price controls and other market regulations, and to provide neces- sary assistance to agricultural producers whose enterprises are not econo- mically viable. The latter type of beneficiaries were mainly financed by BANRURAL. 1.5.3 Implementation Schedule: The execution of the Eighth Agricultu- ral Credit Project was compressed into a third of the original time frame. Contrary to the four year estimated disbursement schedule (December 1984 to September 1988), the Eighth Loan was fully committed by August 1985 and disbursed in 14 months, by January of 1986. However, during the period of the Interim Loan (Loan 2610-ME), which was approved by the board in Decem- ber 1985 and become effective immediately thereafter, the rate of commit- ments and disbursements slowed down. This was due in part to a substantial substitution for IBRD funds with IDB funds resulting from an IDB loan to FIRA which was disbursed in 1986-1987. Additionally, Bank disbursements slowed as on-lending interest rates rose to a point that, while still negative in real terms, farmer demand for investment loans dropped substan- tially due to the shortening of repayment periods in real terms under high inflation and interest rates combined with traditional forms of amortizing long-term loans. On the other hand, demand for short-term working-capital loans increased. In these circumstances, the Bank agreed to the Borrower's request to reallocate funds from investment to working-capital loans and to allow FIRA to increase its rediscounting margin from 50 percent to 80 percent of subloans to the category of Other Producers beneficiaries involved in production of basic grains. Pending these reallocations, the Government requested and obtained a six-month extension of the Closing Date for Loan 2610-ME, which proved to be sufficient time to fully disburse the balance of the loan. 1.5.4 Loan Allocations: The original and revised allocations and actual costs for Loan 2454-ME and 2610-ME are shown on Tables 3 and 4 in Part III. The most significant outstanding feature of the loan allocation was that of the actual application of the two Bank loans. Of the US$480 million, US$156.0 million, or about 33 percent of the loans, was disbursed for working capital loans, a significant departure from previous loans to FIRA which had been almost exclusively earmarked for investment credits. (The first Bank loan used for working-capital subloans was lIRA VII which was amended under the SAP to allocate about 9 percent of the total to subloans of this type). This change vas in direct response to the lack of liquidity in the commercial banking system and the Federal Government it- self, and the shifting demand from investment to working capital credits on the part of producers.  -7- 1.S.S External Funding Relative to Total Funding: As might be expected, during the life of the project, external sources of funding, principally from the' World Bank and 1DB, became more critical to the total financial capacity of FIRA. Expressed as a percentage of total inflow of funds to FIRA on an annual basis, external sources increased four fold as the following table demonstratess Pesos Billions 1988 Constant Total Inflow Internal* 2 External 2 1982 1,444 li238 86 206 14 1983 1,214 988 81 226 19 1984 1,294 529 41 765 59 1985 1,513 893 59 620 41 1986 1,303 512 39 791 61 1987 1,300 557 43 743 57 * Principally rediscounts from BANXICO It is apparent that external sources of funds, which, since December 1, 1983 have been granted to FIRA as equity, were progressively substituting for domestic sources of funs. The 'role of external funding as a catalyst for increased credit availability has been eroded 'in the process. This fact takes on added weight in light of the parallel decapitalization of the institution. 1.5.6 GIRA Effectiveness, The GIRA mechanism for closing the gap on neg4tive interest rates did not produce the desired results during the life of the Project (1984-1987) due to higher than expected inflation rates and the lack of a mechanism to ensure that the nominal ACF and CPI would move close together. Interest rates became positive or near equilibrium only recently in mid-1988 when the average consumer price index fell signifi- cantly. During the project period, the CPI was moving in the opposite direction contrary to expectations while ACF was lagging behind. The trend in the various relevant rates since 1982 was as follows: Annual Inc- reases in Average Relationship to CPI ACF Lending CPI ACF (1)' (2) (3) 3/1 3/2 1982 58.9 40.4 31.4 53? 78? 1983 101.5 56.7 37.6 37 66 1984 65.5 51.1 36.9 56 72 1985 57.7 56.1 41.9 72 75 1986 86.2 80.9 63.4 74 78 1987 131.8 94.6 83.8 64 68 1988 114.1 67.6 63.3 55 94 April 1989 48.4 48.1 99 As the above table indicates, movement of interest rates towards ACF was delayed but not foregone. Furthermore, through protracted negotiations  with respect to compliance with GIRA. the average interest rate on agricul- tural lending was moved progressively closer to ACF. However, movements of the ACF lagged substantially behind increases in the CPI. Therefore, the agricultural interest rate structure and commercial interest rates, as compared to the average annual CPI, continued to be significantly negative, which suggests that GIRAh itself was a weak tool for decreasing interest subsidies to borrowers in the agricultural sector. 1.5.7 This environment of accelerating inflation, fiscal deficits, and decreasing farm productivity made compliance with GIRA a very difficult task for the Government to implement on a timely basis. Alignment of inte- rest rates to the much higher CPI rate was even more remote. The options available to the Bank were tc accede to waivers and postponements to GIRA with the view to closing the gap between interest for agricultural loans and ACF, while at the same time providing required funds to maintain the system, or stop the flow of funds and leave interest rates for loans to this sector to be ever more negative than otherwise. In retrospect, howe- ver, the negative interest rates put into effect had the unfortunate result of the further decapitalization of FIRA. 1.5.8 As noted above, the GIRA accord was successful in closing the gap between interest rates for agriculture and ACF but was not successful in closing the gap of negative interest rates on agricultural credits during the Project period. However, the GIRA mechanism for consultation and nego- tiation between the GOM and the Bank did serve to raise the level of consciousness of the GOM of the devastating effect that negative interest rates have on their financial institutions' capital and the associated high fiscal burden placed on Central Government to replenish this capital through transfers. Even though the GIRA accord proved to be only partially successful during the Project, it did lead to eventual discussions with the Bank for a Subsidy Control and Budgetary Agreement which was approved to limit subsidies as a percent of GDP. The proposed agreement was to be a condition of the Ninth Agricultural Credit Project (appraised and approved in 1987). However, it was considered not necessary as the Mexican Govern- ment itself implemented a subsidy reduction program which exceeded the proposed targets. In this light, then, the GIRA accord can be viewed posi- tively in the sense that it contributed to the process leading to the reduction of financial subsidies to the agricultural sector, during a period of severe economic crisis in Mexico. 1.5.9 Compliance with Loan Conditions3 The project successfully channeled in excess of 50 percent of the total available funds to the targeted LIP group (see Table 3, Part III). At the same time, a substan- tial portion of funds originally earmarked for investment loans was trans- ferred to the working capital category, duly authorized in a formal ptoject modification. 1.5.10 With regards to project financing, the producers and participating banks contributed 35 percent instead of the 28 percent estimated at apprai- sal, while FIRA contributed 25 percent, instead of the 33 percent estimated at appraisal (US$ 299 million as compared to US$ 411 million). More impor- tantly, the Bank Loan took on added significance in light of the overall decline in real terms in the FIRA lending program during the petiod. The  -9- total FIRA lending program (discounts) during the project life was as follows 1988 Constant Pesos Millions 1984 3,253,449 1985 3,780,802 1986 2,691,766 1987 2,652,122 This 18 percent drop in FIRA's new lending was due to a 50 percent reduc- tion in GOM funding (principally Banxico rediscounts) as shown in the table in paragraph 1.5.5 above. This funding gap was partially filled by the IBRDIIDB loans during the period. Although the counterpart funding level vas not a condition of the Loans, they are reflected in the appraisal esti- mates and were expected to be forthcoming. 1.5.11 The original Project Agreement for Loan 2454-HE called for percentage points as margins to participating banks ranging from 2 percent (subloans to Other Producer category of beneficiaries) to 6.5 percent (subloans to Low Income Producer category of beneficiaries). In consulta- tion with the Bank, these margins were modified over the life of the 2454- ME loan and the interim 2610-ME loan. In 1986, the margins were pegged to the ACF; they peaked at a high of about 9.5 points in 1987, and began to drop (as ACP dropped) in 1988. These high margin rates were granted at the expense of further decapitalization of FIRA. As of 1989, the margins were again provided at a fixed rate, and currently they are established between 5 and 6 percent points with an average bf 5.66 percentege points. It is important to note that FIRA also partially funds the operating costs (or technical assistance costs) of the coamercial banks through reimbursements made by FEGA. These spreads and FIRA's reimbursement of commercial bank's costs might seem hiih relative to commercial banks' intermediation costs but low compared to these institutions' returns on alternative investments. There is a need to review these margins in light of the Rural Financial Subsector Study being undertaken by the Secretaria de Hacienda as part of the Ninth Agricultural Credit Project to properly compensate commercial banks without eroding FIRA's capital. During the life of the Eighth and Interim Loans, the margins provided to intermediaries were in accordance with the loan covenants as adjusted from time to time with approval from the Bank. 1.5.12 Operating Costs: FIRA's operation costs, which are comprised of direct expenditure for administration, technical assistance and training, plus the reimbursements to commercial banks and producer groups for techni- cal assistance, have sharply increased over the recent past as percentage of average loan portfolio. They were between 3 to 4 percent for the perjAd 1981-85 but have surpassed 5 percent since 1987, as the following ta4Ae indicates.  _ 10 - Operating Costs as I of Average Portfolio 1980 2.79 1981 3.84 1982 3.68 1983 3.70 1984 4.20 1985 3.72 1986 4.55 1987 5.09 1988 8.66 1.5.13 These costs would be excessive for pure second-tier rediscounting institution. However, FIRA has been more than that. FIRK had channelled increasing portions of its lending operations to low-income farmers. Therefore, FIRA had performed not only its rediscounting functions but also had monitored every loan rediscounted, the income of every borrower, and. had to perform research and extension functions. It is unfortunate that FIRA could not break down its expenditures between its purely banking role and those of technical assistance and loan guarantee fund. It is evident, however, that since 1981, FIRA's total operating costs have consistently exceeded the 3.5 percent margin recetved by FIRA on Banxico rediscounts. These operational deficits have been eroding FIRA's capital compensated for by (i) the 'cost-free' funds (proceeds of IBRDIIDB loans) which GOM has been contributing to FIRA as equity since December 1983, and (ii) the fis- cal transfers to specifically cover -technical assistance costs. However, the latter have been practically eliminated since 1987 and the former would be virtually eliminated because,' as of 1990, all foreign loans would be lent to FIRA at CETES rates while fiscal transfers to cover other costs fell substantially. It should be noted that since 1987 actual transfers have consistently been substantially below budgeted amounts. Therefore, FIRA must now contain these costs and pass some of them on to the producers and commercial banks. Not to do so would result in further decapitaliza- tion of FIRA and would limit the catalytic role of this institution to induw'e commercial lending to the agricultural sector. 1.5.14 During the project life, the loan guarantee program administered through the FEGA Fund continued to function normally. Even though loss rates continued to be exceptionally low, averaging between 1-2 percent of portfolio covered, starting in 1986, the FONDO and FEFA. Funds have begun to replace the Central Government as -the primary fundlng source. Combining the insurance payments with the technical assistance reimbursements paid- out by FEGA to producer groups and participating banks, FIRA's funding share has increased to about 50 percent, with a corres ponding drop in the share provided by Central Government through budgetary) support. In 1988 FIRA started to charge beneficiaries for the cost of the guarantee fund %2-- to 3 percent according to the amount guaranteed), however, this premium was not sufficient to fully cover administrative expenses. FIRA should review premiums so as to base them solely on pri6r claim experience for individual  - 11 - producers so that premiums received and indemnities paid balance out over a period of years for different types of beneficiaries. 1.5.15 Deczpitalization of FIRAs In spite of the application of Project proceeds as a direct infusion to FIRA's. equity account, the overall equity position of FIRA increased only marginally during the period 1982-1987. The outstanding portfolio expressed in constant terms dropped by over a third during the same period. Viewing this financial situation in ito most favorable light, the result of the Project was to cushion the financial fall of FIRA, leaving it in a position to maintain its infrastructure for future growth when inflation and ACF drop to a level whereby positive inte- rest rates are feasible. This occurred in mid-1988 and in fact FIRA's financial position has substantially stabilized since that date. Viewed in the most unfavorable light, the Project might have assisted the Govern- ment in perpetuating major subsidies to borrowers, contrary to its design, with no effective leverage for eliminating this Government policy. 1.5.16 The Ninth Agricultural Credit Project (which became effective in 1987) was a necessary step to maintain FIRA's financial sustainability. The Bank continued to support the rebuilding of FIRA's financial position with the proceeds of this Ninth Loan being granted by the Government as equity. The results of the Eighth and Interim Project were not at all sustainable; the proceeds of the loans were used to subsidize both FIRA and the borrowers. As the following table shows, in spite of massive inflows durinS the period, at December 31, 1988, FIRA's equity stands at 11 percent above the 1980 level, and its portfolio stands at 27 percent below the 1980 balances Millions Pesos 1988 Constant Equity Portfolio Balance Dec. 31, 1980 1,675,500 3,286,900 Balance Dec. 31, 1988 1,852,797 2,410,585 Net Change + 177,297 (876,315) Percent Change + 11? (27Z) The projects under review, plus the previous project (FIRA VII) and the subsequent project (PIRA IX), have had the effect of maintaining PIRA's equity at approximately status quo as of 1980. This has been accomplished at a substantial loss to the Central Goverment which has absorbed the decapitalization losses over the period. 6. Results 1.6.1 Physical Results: No independent review of physical output was performed by the Bank. Statistics included in Tables 6 and 7 in Part III which present data on direct benefits and rates of return, were provided by FIRA management and are based on data gathered by PIRA as part of its moni-  _ 12 - toring and evaluation system. Based on such information FItA 1/ found that the project outputs in terms of crop yields were marginally below the appraisal estimates for all categories of key products, with the exception of rainfed maize, which exactly matched the original estimate. Apparently, the most positive results were obtained in the Low Income Producer category of producers whereby net income per producer showed a increase of about 10 percent, and the number of work days increased by approximately 50 percent. While the Medium Income Producer and Other Producer groups also showed a large increase in number of work days, their net income per producer actually showed a decrease of about 25 percent. With respect to perennial crops, the acreage in production increased about 25 percent on the average for all categories of producers. These calculations are based on before and after credit activity and are not necessarily caused by credit availa- bility only. 1.6.2 Due to fungibility problems, there are serious problems in trying to identify how much of the investments IBRD financed through FIRA are net increment, a difference between with and without project as compared to before and after project financing. When the conditions in which the sec- tor has to operate discourages investments, as was the case for the Mexican agriculture during the implementation of the two lcans under review, some of the official credit expansion will compensate for sectoral disinvestment by farmers and by the commercial banks. Therefore, even if the investments IBRD financed through FIRA are financially and economically viable as the M&E data suggest they were, they might merely be the instrument that permitted borrowers and commercial banks to withdraw other resources from the sector. 1.6.3 In spite of the above, the following considerations suggest the soundness of FItA financed investments. First, FIRA carries out a close supervision of loans to ensure that all loans rediscounted by participating banks are good agricultural investments. FIRA has a cadre of well quali- fied and motivated technicians who tend to discard bad investments. Second, even though negative interest rates lessen borrower incentives to. invest carefully, they do not remove it altogether. Moreover, sectoral distortions discouraging agricultural investments were so severe that cauld have tended to offset FIRA investment incentive via subsidized credit. Therefore, if some FITRA loans were accompanied by their borrowers' trans- fers of resources out of the sector, not FIRA but the sectoral environment was responsible for that. There was probably net agricultural disinvest- ment in Mexico despite FIRA, but there would have been even more without it. 1.6.4 Rates of Return* Based on FIRA's information, the financial and economic rates of return at the farm level also were satisfactory. Both rainfed and irrigated annual crops exceeded the appraisal estimates. Other categories of activities produced results which, although less than projected, were reasonably positive. However, as mentioned before, while these returns might be correct, one cannot .-onclude that these are returns to the credit component due to fungibility problems. 1/ As stated in the Government's PCR.  1.6.5 aeneficiariess The project anticipated 51.3,percent of the total credits to be channelled to Low Income Producer category of beneficiaries. The actual al1ocation resulted in 54 percent being directed to this group. In terms of number of persons benefited by investment subloans under the Project, the Low Income Producer group represented 56 percent of the total. as the following table indicatess Total No. of Total LIP Other Families Benefi. Benefi. Benefi. Work.Capital 23,907 119,535 Investment 94,853 474,265 266,986 207,279 Total 118,760 593,800 Appraisal 105,000 580,000 7. IzstitutionI Performace 1.7.1 Bank Performance. Throughout the Eighth and Interim Projects, the Bank's efforts were directed to the interest rate issue through the appli- cation of the GIRA accord. Its concern during appraisal and negotiations about the relationship of subloan lending rates to the ACF rates and the general level of inflation, was properly justified. However, when the adjustments stipulated by GIRA came due, and the Government -found them politically difficult to implement, the Bank temporarily stopped disburse- ments, and subsequently resumed them, after protracted negdtiations. Al- though this process was painful and time consuming, and caused delays on meeting GIRA agreements, the end result was a decrease in the gap between ACF and interest rates on agricultural loans. It should be emphasized, however, that because ACP lagged annual inflation rates, interest rates for agricultural activities were highly negative during the project period. 1.7.2 The Bank continued to be.sensitive to Mexico's difficult financial problems. It maintained the higher level of disbursement percentages esta- blished by SAP during the latter stages of the Seventh Project and into the Eighth and Interim Projects, and it transferred in excess of US$ 40 million of investment funds to short-term credit to meet the increased demand for working capital support. The Bank also quickly responded to the highly unusual rapid disbursement of the Eighth Loan by processing and approving the Interim Loan, representing a sixty percent increase in funds availabi- lity. 1.7.3 Bank supervision missions were frequent and timely, although they were necessarily directed principally to financial matters. 1.7.4 Borrower Performanee. Similar to the situation of the Bank, the Government had few options to deal with its non-complianc4 with the GIM& accord. The solution to the difficulties in implementing GIRA was townego- tiate modifications. This was successfully accomplished by the Government. The, side effect was to prolong subsidies to producers yia negative interest  - 14 - rates, and to allow for further erosion of FIRA's financial assets. To its credit, Central Government absorbed the majority of FIRA's losses by providing relatively cheap credit through the BANXICO rediscounts, and by granting the proceeds of external loans to FIRA as equity. The estimated subsidy via cheap sources of funds (mainly through BANXICO) provided by the GOM is as follows: Billions of Pesos 1988 Constant Average Actual Interest Balance of Interest Paid Amo:t't Borrowed Paid Calculated of Funds by FVIA at ACF Subsidy 1980 2,354.2 300.2 487.3 187.1 1981 2,845.6 468.6 813.8 345.2 1982 3,803.9 437.5 1,536.8 1,099.3 1983 3.026.0 452.5 1,712.7 1,260.2 1984 2,422.4 4.95.2 1,237.9 742.7 1985 2,503.2 447.8 1,404.3 956.5 1986 2,467.0 459.6 1,995.8 1,536.2 1987 1,819.5 564.6 1,721.3 1,156.7 1988 1.280.9 499.9 865.9 366.0 Total Subsidy 7,649.9 1.7.5 This amount represents the equivalent of US$ 3,354 millions (December 31, 1988 exchange rate fff 2,28111) of accumui.lated subsidy from this source over the period 1980-1988, which, of course, has been trans- ferred to the end borrowers. In addition to this subsidy transfer, the GOM grants to FIRA as equity contributions,.were as follows: Billions Pesos 1988 Constant GOM ContributLons (mainly IBRDIIDB Loans) 1981 54.9 1982 40.6 1983. 55.6 1984 461.0 1985 666.4 1986 791.8 1987 458.6 1988 700.3 Total 3,229.2 1.7.6 This amount represents the equivalent of US$ 1,416 millions (December 31, 1988 exchange rate of 2,28111) of accumulated subsidy from  - 15 - this source over the period 1981-1988, which had the result of directly reinforcing the financial position of FIRA. 1.7.7 These two subsidies from GOM have allowed FIRA to emerge from this economic crisis relatively intact in spite of highly negative interest rates. FIRA's infrastructure and administrative and operational capacity, in particular, its technical assistance capacity, remains strong. 1.7.8 During this period of declining resources and shrinking lending program (in real terms), FIRA has maintained a high level of technical assistance and training targeted towards both the producers and interme- diary bank staff. In addition they successfully transferred their head- quarters to Horelia, Michoacan and procured and installed a sophisticated computer network, financed by the IDB, which greatly enhances their capa- city to support their decentraUzation str!t"gy. 8. Proiect Relationship 1.8.1 The Project did not go according to plan. It was characterized by unexpected events and continual modifications. Throughout all of this, accommodation was reached. The borrower, pushed by its very difficult economic crisis, sought changes and postponements. The Bank performed its role by negotiating delays and modifications to GIRA against reduction of the gap between ACP and interest rates to agriculture, in particular to low-income producers. Throughout this difficult and rather protracted negotiation, both parties acted cordially and professionally. In the final analysis, the Government has paid a very high price by postponing subsidy reductions, but in the process it has protected FIRA as an institution.  _ 17 - PROJECT CO M LTION REPORT uimzco EIGHT AND IRTERII AU!CULTURA CREIT PROJECTS (LOAS 2454-M AND 2610tER) PART 1I: PROJECT REV1 FWMlil THE BORROWER'S PERSPECTIVE I~~~~~~~~~  - 19 - TiASLAION BANK OF MEXICO Trustee of the Agricultural Guaranty and Development Fund Office of the Assistant Director of Agricultural Credit Morelia, Nichoacan, April 27, 1990 Mrs. Graciela Lituaa International Bank of Reconstruction and Development 1818 H Street N.W. Washington,DC 20433 USA Dear Mrs. Lituma: In response to your request of March 19, please find enclosed a copy of part II of the Eighth Agricultural Credit Project (Loans 2454-ME and 2610- ME) "'troject Completion Report" entitled: Project Review from the Borrower's Standpoint. Sincerely yours, Bank of Mexico FIRA. l~ ~ ~ ~~~FE ce: Nacional Financiera, SNC-Insurgentes Sur #1971, Torre Norte 8