Mexico - Second Agricultural Marketing Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 14669 IMPLEMENTATION COMPLETION REPORT MlEXICO SECOND AGRICULTURAL MARKETING PROJECT (LOAN 3141-ME) JUNE 8, 1995 Natural Resources and Rural Poverty Operations Division Country Department 11 Latin America and the Caribbean Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Currency Unit = Mexican New Peso (N$) US$1 = N$3.45 N$1 million = US$289,855 (December 1994) FISCAL YEAR January I - December 31 WEIGHTS AND MEASURES 1 meter (m) = 3.28 feet (ft) I kilometer (km) = 0.62 mile (mi) 1 hectare (ha) = 10,000 m2 = 2.47 acres 1 square kilometer (km2) = 0.38 square miles (mi2) = 100 ha I metric ton (m ton) = 2,205 pounds ABBREVIATIONS AND ACRONYMS ASERCA Services for Commercialization of Agriculture (Apoyos y Servicios a la Comercializacion Agricola) CETES Treasury Bonds (Certificados del Tesoro) CONASUPO National Commission for Nutrition (Comisi6n Nacional de Alimentaci6n) CPP Costo Promedio Porcentual FAO/CP Food and Agriculture Organization, United Nations Agency FIDEC Trust Fund for Commercial Development (Fondo para el Desarrollo del Comercio) FSAL Financial Sector Adjustment Loan GIRA General Rates Interest Agreement GOM Government of Mexico ICR Implementation Completion Report PCR Project Completion Report SAR Staff Appraisal Report SARH Secretariat of Agriculture and Hydraulic Resources (Secretaria de Agricultura y Recursos Hydraulicos) SECOFI Ministry of Commerce and Industrial Development (Secretaria de Comercio y Fomento Industrial) SEPESCA Secretariat of Fisheries SNIM National Market Information Service  FOR OFFICIAL USE ONLY IMPLEMENTATION COMPLETION REPORT MEXICO SECOND AGRICULTURAL MARKETING PROJECT Loan 3141-ME TABLE OF CONTENTS Page No. Preface ............................ i Evaluation Summary ................................ ii PART I PROJECT IMPLEMENTATION ASSESSMENT. 1 Background. 1 Project Objectives. 2 Achievement of Project Objectives .3 Macroeonomic policies .3 Financial and Disbursement Objectives. 3 Food Marketing Policies .4 Physical Objective .6 Institutional Development Objectives. 6 Implementation Record and Major Factors Affecting the Project .9 Factors not Subject to Government Control. 9 Factors Subject to Government Control. 9 Factors Subject to FIDEC Control. 9 Project Sustainability ........................ 10 Bank Performance ........................ 10 PreparationlAppraisal .10 Supervision .13 Borrower Performance ...................... 13 Assessment of Outcome .14 Future Operations .15 Key Lessons Learned .17 PART II STATISTICAL TABLES . .................... 18 APPENDICES ............................ 43 A. ICR Mission Aide-Memoire B. Borrower's Contribution C. Map (IBRD 23547) This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.   IMPLEMENTATION COMPLETION REPORT MEXICO SECOND AGRICULTURAL MARKETING PROJECT Loan 3141-ME Preface This is the Implementation Completion Report (ICR) for the Second Agricultural Marketing Project in Mexico, for which Loan 3414-ME in the amount of US$100 million equivalent, was approved on December 12, 1989 and made effective on April 12, 1990. The Loan closed on December 15, 1994, compared to the original target closing date of June 30, 1995. Final disbursement took place on December 15, 1994, at which time a balance of US$157,363.76 was canceled. This Implementation Completion Report was prepared by Jose Maria Caballero of the Cooperative Program of the Fc"od and Agriculture Organization of the United Nations (FAO/CP), and Louise Cord of the Natural Resources and Rural Poverty Operations Division of Country Department II of the Latin America and Caribbean Regional Office. It was reviewed by Michael Baxter, Division Chief and Theodore Nkodo, Projects Adviser, and cleared by Rudy Van Puymbroeck of the Legal Department. The ICR is based on the findings of field visits, discussions with government officials and a review of material in the project file. The field visits and discussions with government officials were carried out by Jose Maria Caballero from September 7 through 23, 1994. A copy of his Aide Memoire is attached as Appendix A. The Borrower commented on the draft report, and prepared its own evaluation of the project, which is attached as Appendix B to this report.   ii MEXICO SECOND AGRICULTURAL MARKETING PROJECT Loan 3141-ME Evaluation Summary Introduction 1. The Second Agricultural Marketing Project (Loan 3141-ME) was designed to consolidate the momentum achieved under the First Agricultural Marketing Project (Loan 2262-ME), which had been approved in 1983 to provide support to the entire marketing system from producers to retailers in an integrated project. The Second Agricultural Marketing Project was implemented by the Trust Fund for Commercial Development (FIDEC). It was seen as a transitional step within the context of on-going, broad-based, financial sector reforms that were being supported by the Bank's Financial Sector Adjustment Loan (FSAL, Loan 3085-ME) and which were expected to be deepened by the proposed Rural Sector Finance Project, which was appraised but not further processed. Project Objectives 2. As identified by the Staff Appraisal Report (SAR), project objectives were to help improve the efficiency and the transparency of the food marketing system and to consolidate the institutional development of FIDEC. To achieve these objectives, the project had two components. The first component was a line of credit for US$176.1 million (99% of project costs) to be rediscounted by FIDEC to participating commercial banks. The second was a US$1.5 million institutional strengthening component which would procure training, studies, and technical assistance for FIDEC, as well as computer and telecommunications equipment for FIDEC and the National Market Information Service (SNIM). The Loan amount was US$100 million. Implementation Experience and Results 3. The line of credit was fully and rapidly disbursed during the first two years of project implementation (1990 and 1991), assisting in the modernization of the food marketing sector through the construction and rehabilitation of commercial infrastructure. Nonetheless, the project did not help FIDEC become a long-term financing institution. FIDEC charged to the project most of its lending for investment subprojects (thereby complying with the agreed split between working and investment capital), while reserving other sources of funds for working capital subloans. As a result, between 1989 and 1994, slightly under 80% of FIDEC's total lending program was for working capital.  iii 4. The overall implementation of the institutional strengthening component was mixed. Many activities were not fully implemented, such as the studies, technical assistance and training programs for industry personnel. Moreover, many of the activities implemented, such as some of the studies, had little impact. A principal explanation for the poor implementation of this component was probably that FIDEC staff did not believe that FIDEC was the appropriate institution to implement it. Although the full amount allocated to SNIM was disbursed for new computer equipment, the extent to which this support improved the transparency of food marketing was never evaluated. 5. FIDEC is a solid institution, which helps to ensure the sustainability of the project. Nonetheless, several factors create doubts about the sustainability of the project and the implementing institution. First, the absence of subproject monitoring and evaluation data makes it difficult to evaluate the impact of the project's investments. Second, FIDEC has not evolved into a long-term financing institution and relies mainly on working capital lending. Third, the future of FIDEC is not clear, given the 1993 banking law, which affects all trust funds under the Bank of Mexico, and given the possibility of major reforms in Mexico's rural financial sector. 6. Bank Performance. The following flaws are identified in the project concept and design. (a) An inconsistency between the project objectives and the project design and implementation capacity of the principal project agency. The main objective of the project was to promote the modernization of the food marketing sector. The principal mechanism provided by the project to achieve this goal was a line of credit (which was 99% of the project costs). However, promotion, organization, research, training and technical assistance activities are also required. While some of these ancillary activities were included in the project, their overall importance was relatively small and the capacity of FIDEC to implement them was relatively limited. (b) Inadequate attention to poverty alleviation objectives. Although not specifically listed as a project objective in the SAR, given the Bank's poverty alleviation objectives, the project could have focussed more on the role of small and micro traders in the food marketing sector. (c) No timetable for the disbursement of project funds. This allowed the Second Agricultural Marketing Project to be fully disbursed in less than two years. Had the project design considered disbursement by tranche, monitoring on the part of the Bank would have been easier, offering more opportunity to observe, and correct if required, the orientation of the project.  iv (d) Poor specification of how FIDEC was to monitor and evaluate the subloans. (e) The imposition of a restriction on the use of subloan resources that could not be implemented. The limitation of the project to support only food marketing activities could not be fully implemented, since a large part of commerce, especially at the retail level, combines the sale of food with that of non-food items. 7. Project supervision on the part of the Bank was also deficient. In nearly five years there were only five supervision missions, all led by different people. Between December 1990 and July 1992, when the project disbursed nearly two-thirds of the Loan, no supervision missions were fielded. Moreover, aside from the issue of infrequency, the overall impact of the supervision missions on the implementation of the project is questionable. 8. Borrower Performance. FIDEC's performance was highly satisfactory as far as the disbursement of the credit line was concerned, but only partly satisfactory in the implementation of the institutional strengthening component of the project. The overall performance of FIDEC can be considered satisfactory. Compliance with the legal covenants was in general satisfactory. Nonetheless, there were some delays in the submission of audit reports, some relatively minor procurement problems, inadequate implementation of the covenants calling for FIDEC monitoring and evaluation of project subloans. Summary of Findings, Future Operations, and Key Lessons Learned 9. The outcome of the project can be rated as satisfactory, as it was reasonably successful in achieving its objectives. At this time, a future operation with FIDEC does not seem justified for two reasons. First, the financial markets have become more efficient and the need for government second tier financial institutions which provide targeted finance is questionable. In other words, the justification for the Second Agricultural Marketing Project, which was to provide transition support to a financial sector undergoing fundamental changes is now largely removed. Second, while FIDEC does have some targeted programs towards small and low income producers and traders, the majority of its lending is for medium-sized traders and producers. Depending on the pace of future reforms in Mexico's food and agricultural marketing policy, there could be future justification for Bank support in the sector, but most probably through a different mechanism than the one currently used by FIDEC. 10. The Bank should focus future discussions with FIDEC on: assessing the principal constraints of Mexico's food marketing sector and the role of the key government and non-government agencies in lifting these constraints; the extent to which FIDEC should continue to concentrate most of its resources on working capital subloans; and, improving the monitoring and evaluation of subloans by FIDEC and commercial banks.  v 11. Key Lessons Learned. Three main lessons can be singled out. (a) First, project objectives should be clearly in line with the mandate and the capacity of the implementing agencies. If project objectives require an expanded or modified mandate for the implementing agencies, then there should be formal government support for this change and the project should provide sufficient resources for the implementing agency to carry-out its new role. (b) Second, frequent Bank supervision mission are critical, especially in the early years of project implementation, to confirm the validity of project objectives and the appropriateness of project design to achieve these objectives. (c) Third, the use of a credit line disbursed through a variety of commercial banks is difficult to monitor and evaluate by the central responsible agency. Monitoring and evaluation activities require the cooperation of the commercial banks, which are the only ones effectively able to follow-up the implementation of subloans.  IMPLEMENTATION COMPLETION REPORT SECOND AGRICULTURAL MARKETING PROJECT Loan 3141-ME MEXICO PART I PROJECT IMPLEMENTATION ASSESSMENT Background 1. The Second Agricultural Marketing Project (Loan 3141-ME) was designed to consolidate the momentum achieved under the First Agricultural Marketing Project (Loan 2262-ME), which had been approved in 1983 to support the entire marketing system from producers to retailers in an integrated project. The first project had channelled resources to three trust funds, the Trust Fund for Commercial Development (FIDEC), the Trust Fund for Credit in Irrigated and Rainfed Areas (FICART), and the Agricultural Trust Funds in the Bank of Mexico (FIRA). The Second Agricultural Marketing Project would concentrate on FIDEC, considered to be the most efficient trust fund for directing resources to the marketing sector. 2. The Second Agricultural Marketing Project was seen as a transitional step within the context of on-going, broad-based, financial sector reforms The latter sought to privatize commercial banks and improve their supervision activities, reduce the multiplicity of credit support schemes and simplify the Government's development banking system. During the preparation and appraisal of the Second Agricultural Marketing Project in 1988 and 1989, the Bank was supporting the implementation of these financial reforms with the Financial Sector Adjustment Loan (FSAL - Loan 3085-ME). The reform process was also expected to be deepened under the Rural Finance Subsector Project, which was appraised in 1992, but not further processed. 3. In particular, the Second Agricultural Marketing Project was expected to support the on-going reform process in the financial sector in four areas. First, it was to provide resources for long-term lending by commercial banks, which was expected to remain limited until the restructuring measures had time to affect the financial sector, and in light of the significant economic uncertainty that existed at the time the project was prepared. Second, FIDEC's lending policies were to be consistent with the on-going policy dialogue under the FSAL, which called for new loans through development banks and trust funds to be channeled without interest subsidies (i.e., at interest rates above the institution's average cost of funds), except for low- income producers in agriculture. Third, the project was consistent with the policy agreements made under the FSAL, which required the Government to redefine the role of each development bank and trust fund. Regular consultations were expected to occur between the Govermment and the Bank to discuss FIDEC's future, once long-term financial markets re-emerged. Fourth, FIDEC would assist the commercial banks improve their performance in loan evaluation and supervision activities.  - 2 - Project Objectives 4. The Staff Appraisal Report (SAR) of the Second Agricultural Marketing Project identifies three specific objectives of the project. They were to: (a) help improve the efficiency of the food marketing system through the adoption of modem technology and improved marketing practices; (b) improve the transparency of the system and thus competition through a better dissemination of information on price and volume of product traded; and, (c) consolidate the institutional development of FIDEC. These objectives were to be achieved through two components: a line of credit for FIDEC, 80% of which was to be used to finance investment capital to assist FIDEC become a long-term financial institution; and an institutional strengthening component to finance studies, training and technical assistance for FIDEC and equipment for FIDEC and the National Market Information Service (SNIM). 5. These objectives were not modified during the disbursement period. They were reasonably clear and realistic, although somewhat general (e.g., the SAR contains no precise definition of "modem technology" and "improved marketing practices"). 6. The decision, however, of having an second-tier bank like FIDEC, a purely financial institution with very few staff, as the only implementing agency entirely responsible of achieving objective (a), which was the core of the project, was inopportune. FIDEC could only promote the modernization of food marketing activities with credit, not with other important inputs such as technical assistance, training, technical studies, research, demonstrations, and promotion of marketing organizations. The project did provide FIDEC with technical assistance and training support to develop a methodology to monitor and evaluate the food marketing sector and the distribution system of selected products, as well as to improve the organizational structure and technical expertise of FIDEC. However, FIDEC lacked the physical and personnel resources to have the broad based role in the food marketing sector implied by such project activities.' Moreover, given the Government's and the Bank's focus at the time of project appraisal on the downsizing of government financial institutions, it was hardly likely that FIDEC's resources could have significantly expanded during the implementation of the project. Finally, the objective of improving the transparency of the marketing system, although relevant and worthy, was hardly likely to be achieved with US$100,000 of equipment. 1.FIDEC, unlike other second tier financial institutions under the Bank of Mexico (such as FIRA), has only five regional offices and a staff of only eighty people focussed on the management of its credit operations.  - 3 - Achievement of Project Objectives Macroeconomic Policies 7. The project did not directly pursue any macroeconomic objectives and the volume of the loan was too small to have direct macroeconomic consequences. Financial and Disbursement Objectives 8. The financial objective of the project consisted in the disbursement of US$176.1 million (of which US$98.6 was from the Loan) to finance investment subprojects (US$79 million from the Loan) and working capital loans (US$19.6 million from the Loan) during a disbursement period of five years. This objective was fully met during the first two years of implementation, which is an impressive achievement. 9. It was expected at appraisal that the cost of the subprojects would be shared in the following way: 20 the final borrowers, 10% tb'e participating banks, 56% FIDEC from the Loan, and 14% FIDEC from ot.ner sources. In fact, the contributions of FIDEC and the participating banks were as expected, but those of the final borrowers were much larger than anticipated: from the appraisal estimate of US$35.2 million they rose to US$100.6 million, raising their contribution to 41 % of total project costs. Thus, the total cost of the project, initially estimated at US$177.6 million, was US$242.7 million. Approximately one third of this difference is explained by the cost of the building sites contributed by final borrowers in construction projects, and the balance by other higher than expected indirect costs (taxes, feasibility studies, advertising, etc.). 10. Notwithstanding FIDEC's compliance with the agreed split of the Bank's credit line between investment project subloans and incremental working capital subloans, one of the sub-objectives of the project "to help FIDEC to turn principally into a long-term financial institution" was not achieved (para. 4). Originally, the project envisaged that only 20% of the part of the Loan apportioned to the credit line would be used to finance incremental working capital.2 Table 1 indicates that the project did not change FIDEC's policy of focusing its lending on working capital; FIDEC charged to the project most of its lending for investment subprojects (thereby complying with 2.Due to the high demand for this type of subloans, this was revised in June 1991 when, at the request of Government, the Bank authorized a trans42r to working capital subloans of US$ 8.5 million allocated to investment projects, thus raising to 28.3% the participation of working capital subloans in the credit line. In January 1992, at the request of Government and in view of the difficulties and slow progress in carrying out some of the studies envisaged at appraisal, the Bank authorized the transfer of US$ 0.55 million from studies to investment projects subloans.  - 4 - the agreed split between working and investment capital), and reserved other sources of funds for working capital subloans. Table 1 Percentage of all FIDEC Subloan Disbursements for Working Capital* 1989 1990 j 1991 1992 .1973 1994 78% 72% 77% 81% 77% 78% e Includes subloans financed from World Bank credit line and other sources. 11. There are two reasons for the high share of working capital in FIDEC's lending: (a) the demand for working capital is much higher than for investment loans; and (b) FIDEC believes that (i) investment subloans usually need to be complemented with working capital, (ii) many working capital loans can contribute in their own right to the modernization of the sector, and (iii) banks cannot meet all credit demand for working capital from their own resources alone. There is substantial truth in (b) (i) and (b) (ii), although they still do not negate the importance of investment capital. On the other hand, (b) (iii) is to be explained by the interest rate advantage of FIDEC's subloans. If this advantage did not exist, the competitive conditions of Mexico's credit markets would force the amount of credit demanded for working capital to contract in response to the tougher conditions of the loans provided with commercial banks' own funds. Food Marketing Objectives 12. The project had two specific objectives for the food marketing sector. The first was to improve the efficiency of the sector through the adoption of modem technology and improved marketing practices. This was to be achieved through the line of credit to be administered by FIDEC and the institutional strengthening of the trust fund. To the extent that the line of credit was fully disbursed, the project did improve the efficiency of food marketing through the construction and rehabilitation of infrastructure. However, the institutional strengthening component was only partially implemented, which limited FIDEC's ability to have a broad based impact on the sector's efficiency (paras 18-24). Had the training, technical assistance and studies called for under the institutional strengthening component been fully carried out, FIDEC would have been able to anticipate, orient and accelerate the modernization process in the food marketing sector. 13. The contribution of the project centered mostly on the construction of infrastructure, especially wholesale and retail markets, which absorbed 41 % of the Loan. To this, can be added another 15 % in subloans to allow retail merchants and wholesalers buy stalls and storage space in the wholesale and retail markets constructed with project funds. Altogether, these two types of  - 5 - subloans accounted for 56% of the Loan and 78% of the Loan allocated to financing investment subprojects. 14. Because of the absence of adequate monitoring and evaluation mechanisms by FIDEC and the commercial banks on the subloans, there is no systematic evidence on the quality and impact of the investment subprojects financed by FIDEC. Indirect evidence suggests that the subloan portfolio financed with project funds was reasonably sound. For example, (a) FIDEC has a well running system to appraise loan requests, which is both fast and fairly exact in the screening of information and the application of approval criteria; (b) FIDEC personnel in the regional delegations and at headquarters have a good knowledge of the sector and experience in identifying investments which carry higher risks; (c) larger projects, and those which seem likely to offer special difficulties, have to be directly approved by FIDEC's Technical Committee; (d) commercial banks, which assume the full risk of the subloans made with FIDEC's funds, submit the subloans to their own approval procedures based on market and risk assessment criteria; and, (e) many FIDEC borrowers receive more than one subloan and hence submit information on commercial operations and financial positions on a more or less recurrent basis, which allows FIDEC to carry out some ad hoc infonnal monitoring of the borrower's development and financial standing. 15. The impression on the quality and impact of project-financed investments gathered by Bank supervision missions is mixed. While the first three supervision missions (April 1990, December 1990 and July 1992) did not raise issues of quality of the investments, the fourth mission (in May 1993) observed significant problems in the quality, implementation and profitability of the investment projects in retail and wholesale markets. The fifth mission (in November 1993) concentrated on financial aspects and the compliance of covenants and did not visit the field to assess investment quality issues. The impression obtained by the ICR mission from conversations with FIDEC's field staff, traders, managers of commercial enterprises and staff of the commercial banks involved, as well as scattered visits to subprojects was that: (a) the monitoring by commercial banks of the application of FIDEC funds to the uses earmarked in the subloan agreements was adequate; (b) project selection is demand-led, based on market criteria, although within the eligibility norms of FIDEC, which have become broader over time; and, (c) the quality of the investment subprojects financed with FIDEC funds is equal or superior to that of the normal loan portfolio of the banks in the sector. 16. The second food marketing objective was to improve the transparency of the food marketing system and thus competition through a better dissemination of information on price and volume of product traded. Although the full amount allocated to SNIM was disbursed for new computer equipment, the extent to which this support improved the transparency of food marketing was never evaluated. Conversations during the fourth supervision mission with SNIM staff suggested that project support, while useful, had a relatively  - 6 - small impact on the service's performance and that the equipment purchased with project resources had quickly become outdated. Physical Objectives 17. Since the project essentially consisted of a credit line, no physical objectives were included in the SAR. A summary of the physical achievements of the investment subprojects is provided in the Statistical Annex Tables 5 and 6. Institutional Development Objectives 18. The institutional development objectives of the project were to (SAR paras. 74-79): (a) improve the institutional environment by establishing a methodology for analyzing, monitoring and evaluating the food marketing sector, analyzing the distribution systems of selected products, and strengthening the organizational structure and technical expertise of FIDEC; (b) provide guidance for improving the operation of food marketing through two special studies, one to assess the feasibility of introducing auction systems in selected wholesale markets, and the other to review regulatory issues which could affect modernization and free entry in retail food distribution in Mexico City; (c) eliminate skills gaps through a comprehensive training program for 2,000 participants from FIDEC, SECOFI, SARH, commercial banks and the private sector; (d) upgrade FIDEC's monitoring, evaluation and information systems through the acquisition of computer, statistical and communication equipment; and (e) improve SNIM by increasing its product coverage by including fish and meat, better disseminating information to all regions, providing information on daily volumes of each commodity entering wholesale markets, and increasing the geographical coverage from 14 to 22 wholesale markets. An amount of US$1.5 million was budgeted to achieve all of the above objectives. 19. The fulfillment of the above objectives was mixed. A study was carried out to establish a methodology for analyzing and monitoring the food marketing sector, which was published as a book (para. 18a). The study has  - 7 - analytical quality, but has more a theoretical than a practical orientation, and lacks specific institutional proposals. Its recommendations have not been adopted. Three products--citrus, tomatoes and meat (beef, pork and poultry)-- were selected for studies of the distribution systems (para. 18a). The studies present valuable descriptive materials and carry-out some interesting analyses, but are not action oriented and do not identify specific investment opportunities. The impact of the three distributional studies has been minimal. 20. FIDEC did not carry out two special studies, envisaged at appraisal and included in the Project Agreement 2.05 (b,c,d) and 3.01 (b), mainly because, given their nature, they were not judged by FIDEC to fall under its mandate (para 18b). The study on regulatory issues in retail food distribution in Mexico City was substituted by guidelines prepared in this respect by Secretariat of Commerce and Industrial Promotion (SECOFI), which were sent to the Bank with considerable delay. The May 1993 supervision mission agreed that it would not be effective for FIDEC to carry out the regulatory study in light of the existence of SECOFI's guidelines. Terms of reference for the study on the introduction of auction systems in wholesale markets were prepared, but the study was not carried out. FIDEC maintains it was not the appropriate institution to implement the study given its position as second-tier bank. In September 1994, the Bank agreed with FIDEC's decision not to carry-out the study on wholesale auction markets for two reasons. First, the study was outside the mandate of FIDEC. Second, FIDEC should not have to carry out the study, as the decision to introduce auction markets should be developed endogenously by the private sector, rather than be promoted by a govermment agency. 21. To review the organizational structure of FIDEC and make it compatible with a computerized information and work environment, a diagnostic study, followed recently by an integrated information proposal, was carried out, accompanied by the acquisition of computer and communications equipment (para 18d). Computerization and responsible management have brought substantial organizational improvements to FIDEC and large gains in efficiency. Thus, with approximately the same staff numbers of around 80 people the outstanding loan portfolio has nearly quadrupled in real termns between 1990 and 1994. With its limited staff, FIDEC has programmed to lend an equivalent of US$2 billion in 1994. 22. FIDEC did not significantly strengthen its monitoring and evaluation function, as agreed during project negotiations (SAR, para. 99e). FIDEC has no systematic information on the performance of subprojects financed with its funds, not even on repayment arrears of final borrowers and the default portfolio of the participating banks. Part of the reason for this is the fact that all credit risk is borne by the participating banks and none by FIDEC, which automatically recovers maturing debts from the banks' deposits in the Banco de Mexico. FIDEC, hence, has no incentive to monitor the subloan repayment situation. Nor does FIDEC monitor the employment of the funds to make sure that they are allocated to the uses indicated in the subloan agreements; this is  - 8 - done by participating banks with their own procedures and at their own discretion. Only when irregularities are suspected, does FIDEC ask banks for written evidence of fund use. However, without appropriate monitoring and evaluation of FIDEC subloans, it is impossible to fully assess the impact of FIDEC investments on the sector. Furthermore, a key element in the Bank's dialogue on financial sector issues with the GOM was on improving the supervision and monitoring activities of commercial banks to ensure that only quality loans were authorized (para 3). 23. Being a second-tier bank, FIDEC has no capacity to monitor directly the use of the funds, the repayment situation, and the evolution of the subprojects; this can only be done by the first-tier participating banks. There are, however, two things that FIDEC could have done: (a) to collect information from participating banks on the repayments of FIDEC-financed subloans; and (b) to organize a simple system to monitor on a stratified sampling basis some basic parameters concerning the development and impact of FIDEC-financed subprojects. FIDEC did contract a study to review several FIDEC subprojects, to evaluate their impact and to develop a strategy for FIDEC to improve its Management and Evaluation activities. However, the study was not complete at the time of the ICR mission. 24. The ambitious training program envisaged at appraisal was not carried out (para 18c). Training courses to small traders were offered until mid-1991, but were discontinued because the then new FIDEC management estimated that it could not satisfy the demand, it had doubts about the impact of the courses, and considered that, as a banking institution, it was not FIDEC's role to carry out this type of activity. With its own resources, FIDEC has promoted and financed many training activities for its own staff in a number of fields. It has also prepared written and computerized manuals to disseminate FIDEC financial services and to train staff from participating banks in FIDEC's operating procedures. The staff of FIDEC's regional delegations regularly carry out promotion and dissemination activities of FIDEC services among commercial bank staff, traders and staff of commercial firms. 25. During the project period, SNIM expanded and improved considerably its food market information services, making a substantial contribution in this area (para 18e). All the specific objectives proposed for SNIM in the SAR were largely exceeded, with the exception of the provision of information on daily volumes of each commodity entering wholesale markets, which proved to be infeasible. It is not possible to assess how much of the success of SNIM can be assigned to the computing and telecommunication equipment supplied by the project.  -9- Implementation Record and Major Factors Affecting the Project Factors not Subject to Government Control 26. No factors not subject to government control have been identified, other than the performance of the Bank as examined below (para. 33). Factors Subject to Government Control 27. Three major factors subject to government control affected the project positively. The first was the privatization of commercial banks in 1992. This brought about a restructuring of the commercial banking system, broad favorable managerial changes, enhanced competition, and stronger banking dynamism. The second was the negotiation and final signing of the North American Free Trade Agreement with the United States and Canada in 1993. In preparation for this, and due in particular to the fear of competition from foreign commercial chains, there was--and still is--an intense restructuring process in the commercial sector, which has triggered investment demand. The third factor is the decline of inflation rates in the 1990s, which has decreased price uncertainty and improved the general price environment, although interest rates continue to be high. Factors Subject to FIDEC Control 28. Competent FIDEC management and constant revision nf financial policies and procedures in response to rapidly changing market circumstances have been two major elements supporting the development of FIDEC. They have contributed to FIDEC's success in increasing its volume of lending, which allowed disbursement in less than two years of the entire credit line provided by the Loan. This contrasts favorably with the slow start of FIDEC in the 1980s and the delays in the disbursement of the first Loan (Loan 2262- ME). 29. Two accidental circumstances hindered project implementation in mid- 1991: the death of FIDEC's Director, and consequent appointment of new management, which took three months to materialize, and the resignation of the officer in charge of the Institutional Development Office of FIDEC, who, together with the Director, were the persons most familiar with the project. A good part of the project's historic memory was lost by these actions in 1991, which more or less coincided with the completion of disbursement of the Loan's credit line. In view of this and of the increasingly clear difficulties found by FIDEC in carrying out the institutional development objectives outside its institutional mandate, the ICR mission and current FIDEC management believe that the project could have been closed at the end of 1991. However, no Bank supervision missions were fielded during 1990, nor during the first six months of 1991 (para 35). The third supervision mission, fielded in July 1992, discussed with Government the possibility of closing the project, but no action was taken, mainly because it was believed that it would  - 10 - be useful for FIDEC to carry-out the technical assistance activities and studies proposed under the institutional development component. 30. There were implementation delays in the preparation of the studies due to the delays in the submission of terms of reference and the selection of consultants, which could have been avoided. The deeper reason why these delays occurred, and why some of the studies were finally not prepared, was probably that FIDEC staff did not believe in the importance of carrying them out and/or did not believe that FIDEC was the appropriate institution to implement them. Project Sustainability 31. Although FIDEC is a stable institution, several factors raise doubts about the institution's and the project's sustainability. On the positive side, FIDEC has: a competent, well trained and reasonably well paid staff; sound loan appraisal and disbursement mechanisms; a good rapport with commercial banks with which it enjoys good reputation; and, fmancial resources to cover its operating costs with the income from its interest spread. In addition, FIDEC is very sensitive to market signals and has proven its ability to adjust fast to changing policy and market environment. 32. However, four factors could undermine the sustainability of FIDEC and its investments. First, the Bank of Mexico will no longer serve as trustee for any trust funds, including FIDEC after 1995, in light of the new 1993 banking law, which gave the central bank autonomy from the federal government. As a result, the future of FIDEC remains uncertain and there is a possibility that it be eliminated, although conversations with government officials suggest that this does not appear likely. Second, FIDEC could also disappear should there be a major restructuring of the rural sector financial institutions, thereby eliminating the need for specialized second-tier financial institutions such as FIDEC. Third, the focus of FIDEC on working capital loans could also undermine the institution's sustainability. Finally, the sustainability of the investments undertaken by FIDEC has not been fully ascertained; as the quality and impact of FIDEC's subprojects is unclear given the lack of appropriate monitoring and evaluation mechanisms for FIDEC subprojects. Bank Performance Preparation/Appraisal 33. With the benefit of hindsight, it is possible to identify the following flaws in the project concept and design: (a) An inconsistency between the project objectives and the tools provided by the project to implement these objectives. The main objective of the project was to promote the modernization of the food marketing sector through improved marketing  - 11 - practices (para. 4). However, the principal mechanism provided by the project to achieve this goal was a line of credit, which accounted for 99 % of the project costs at the time of appraisal. As highlighted in the Project Completion Report of the First Agricultural Marketing Project, it is not possible to modernize the food marketing sector with purely credit instruments; promotion, organization, research, training and technical assistance activities are also required.3 However, these activities were given relatively little weight in the design of the Second Agricultural Marketing Project and were allocated only US$1.5 million (1 %) of project expenditures. Moreover, to the extent that the above inputs were included in the project, they were entrusted to FIDEC, which given its reduced staff and its purely financial mandate had difficulties carrying them out. The Bank's appraisal team did expect that FIDEC's resources would be increased during the life of the project, although the SAR did not detail the specific staff, financial or physical inputs necessary for ELDEC to have an expanded role in the sector.4 Meanwhile, in a letter to the Bank at the time of appraisal, it appears that the Government saw the trust fund as specialized financial institution that would support long-term investments in the marketing sector (SAR, para 68). This suggests that the Government and the Bank did not fully agree on what should be the role of FIDEC at the time of appraisal. The Government's reduced role for FIDEC was more in keeping with its and the Bank's overall objective of reducing the size of government trust funds and agencies in the agricultural sector. One alternative could have been to downsize the project's objectives and link them more closely to FIDEC's financial role in the marketing sector. Perhaps some pilot initiatives to provide research, technical assistance and training activities to the participants in the marketing sector could have been tested by some other more appropriate organizations (e.g. SARH, SECOFI, state agencies or civic groups). (b) Inadequate attention to poverty alleviation objectives. Although not specifically listed as a project objective in the 3.The report was completed in December 1991, and only a draft version was available at the time of appraisal. 4.The background section of the project's SAR underscores FIDEC's limited resources (small organizational units, inadequate staff mix, etc.) to carry out an expanded role in the marketing sector and the SAR comments that "these deficiencies would be addressed under the proposed project (SAR, para. 60)."  - 12 - SAR, given the Bank's poverty alleviation objectives, the project could have focussed more on the role of small and micro traders in the food marketing sector. No provisions were included in the SAR or in the Loan Agreements to ensure that these groups received priority attention from project resources. For instance, no specific credit instruments were incorporated in the project design to facilitate the access of small and micro traders to the formal credit mechanisms supported under the project. (c) The imposition of a restriction on the use of subloan resources that could not be implemented. The limitation of the project to food marketing could not be fully maintained in practice, since a large part of commerce, especially at the retail level, combines the sale of food with that of non-food items. More operational limitations would have been possible, for instance--and this is what prevailed in practice--to restrict subloans to investments related to trading in a combination of food and other basic consumer products. (d) No timetable for the disbursement of project funds, most likely due to the slow disbursement experience of the First Agricultural Marketing Project. This allowed the Second Agricultural Marketing Project to be fully disbursed in less than two years. Had the project design considered disbursement by tranche, project monitoring on the part of the Bank would have been easier, offering more opportunity to observe and correct if required the orientation of subloans. It should be acknowledged, however, that a disbursement by tranche could have jeopardized a rapid response to market pressures. (e) Poor specification of how FIDEC was to monitor and evaluate the subloans. Although the SAR (SAR paras. 82 and 99e) indicate that FIDEC should carry out monitoring and evaluation activities of the investments financed with project resources, the variables to be monitored, the concrete ways and means to monitor them, and the agreements that FIDEC should reach with the participating banks were not spelt out. The latter was particularly important in view of the difficulty FIDEC had in monitoring and evaluating the investments in the absence of specific agreements with the first-tier banks. 34. In view of these flaws, Bank performance at preparation/appraisal can only be rated as unsatisfactory.  - 13 - Supervision 35. Project supervision on the part of the Bank was deficient. In nearly five years there were only five supervision missions, all led by different people. During supervision, the project had four task managers. Between December 1990 and July 1992, when the project disbursed nearly two-thirds of the Loan, no supervision missions were fielded. Aside from the issue of infrequency, the missions did not transfer know-how on financing the modernization of the commerce sector, as FIDEC would have liked. Nor did they manage to reorient a project showing increasingly evident design flaws. Supervision missions instead emphasized deviations in the compliance of legal agreements, some of which FIDEC had structural difficulties to abide by (Statistical Appendix Table 10). The last two supervision missions (in May and November 1993) rightly insisted, however, on the need to organize a system to monitor subloan investments. Unfortunately, they did not get to the bottom of the problem, which was to ascertain the responsibility that first-tier commercial banks ought to have in this and to identify practical ways to carry it out. Borrower Performance 36. In general terms, FIDEC's performance was highly satisfactory as far as the disbursement of the credit line was concerned, but only partly satisfactory in the implementation of the institutional strengthening component of the project. Moreover, it was deficient in the monitoring and evaluation of the subloans. The overall performance of FIDEC can be considered satisfactory. 37. Compliance with the legal covenants was in general satisfactory. Nonetheless, there were some delays in the submission of audit reports as well as some problems in the application of agreed procurement procedures for the recruitment of consultants and for contracts that included civil works and goods and equipment, although on the whole these problems were not significant. Moreover, FIDEC, as a second tier financial institution, really did not have the mechanisms to ensure that the final borrowers followed the procurement procedures agreed upon with the commercial banks. This should be paid attention to in the design of other credit projects. 38. Conditionality with respect to the grace and repayment periods to be applied to subloans were respected by FIDEC. On the other hand, compliance with agreements relative to interest rates was controversial and was the object of attention from supervision missions. The controversy refers to two issues: the structure of interest rates of FIDEC loans, i.e., the difference in rates to be applied to different types of subloans; and the interest rate level to be charged by FIDEC to the participating banks. These issues are examined below.  - 14 - 39. According to the Project Agreement, the interest charged to the project's final borrowers should not be less than the rates established in the General Interest Rate Agreement (GIRA). Under this agreement, two type of interest rates were authorized: 0.95 of CPP5 for low income producers and CPP + 2 points for other producers.6 7 During Loan negotiations, however, FIDEC was authorized to allow participating banks to differentiate rates to final borrowers according to the norms of FIDEC's Operating Manual. The latter also separated loan rates into two types (called Schedule I and Schedule 1I), using different criteria than the GIRA. The Operating Manual distinguished rates according to the potential profitability and social priority of the proposed investment. As observed by the November 1993 supervision mission, which examined this issue in depth, by approving FlDEC's Operating Manual, the Bank implicitly waived GIRA. Nonetheless, the average level of interest rates charged to final borrowers on FIDEC funded subloans remained in conformity with GIRA. 40. With respect to the interest rate charged by FIDEC to the participating banks, the Project Agreement established that the weighted average rate charged by FIDEC to the banks should be two points above FfDEC's weighted average cost of funds. This was understood by Bank supervision missions to mean that lending to participating banks would be at the CETES rate plus two points, since the only two sources of borrowed funds of FIDEC, the Loan and the Banco de Mexico financial facility, charged a CETES rate to FIDEC. Under this interpretation, FIDEC did not abide by the agreement, for the rates it charged to the banks were on average two to three points below the agreed CETES plus two. FIDEC, however, put forward a different interpretation based on the fact that the Project Agreement indicated as reference the "weighted average cost of funds" and not the weighted average cost of borrowed funds. Hence, FIDEC included in the weighted average cost of funds the cost of its own resources, which of course was zero (excluding their opportunity cost). This interpretation was discussed with the Legal Department of the Bank, which considered it acceptable. FIDEC, therefore, was technically in compliance of the covenant. Assessment of Outcome 41. The outcome of the project can be rated as satisfactory. The project was reasonably successful in achieving its objectives. The project made a contribution to the modernization of the food marketing sector (via the provision of commercial infrastructure), which was a central concern of the 5.CPP (Costo Promedio Porcentual) is a reference rate published monthly by the Banco de Mexico, which reflects the average cost of money in the economy. 6.The reference rate of GIRA was later changed to CETES. 7.Low income producers are those producers with net annual family income (from all sources) below 1,000 times the minimum regional daily wage.  - 15 - Government's sectoral policy. The project was also extremely successful in disbursing the project's line of credit to FIDEC. The achievement of the institutional objectives was mixed (paras 4 and 18-24). FIDEC did consolidate its institutional structure during the project. The areas where the project has been the least successful are transforming FIDEC into a provider of long-term investment capital and in supporting FIDEC's own as well as the monitoring and evaluation activities of the commercial banks. The sustainability of project achievements is linked to that of FIDEC as an institution, which is not fully certain in light of: recent changes in the banking law; the institution's focus on working capital loans; possible policy reforms; and, the lack of a clear assessment of the impact of subproject investments (para. 32). Future Operations 42. The issue of future operations can be broken down into two questions. First, should the Bank have another project with FIDEC to support financial investments in the food marketing sector? Second, should the Bank have another project to support food marketing activities in a broader sense? 43. The answer to the first question cannot be separated from the issue of the whole operational program and future orientation of FIDEC. A plan of operations containing FIDEC management views on the future evolution of the institution, will be presented to the Bank together with its own evaluation report of the project.' However, at this time, a future operation with FIDEIC does not seem justified for several reasons. First, as the financial sector becomes more efficient, given the liberalization of interest rates for commercial banks in 1989, the privatization of the commercial banks in 1992, and the decision to allow foreign banks enter the sector in 1993, the need for government second tier financial institutions which discount to all actors in the sector is questionable. In other words, the justification for the Second Agricultural Marketing Project to provide transition support to a sector undergoing fundamental change (para.3) is now largely removed. 44. Second, while FIDEC does have some targeted programs to reach low income producers and traders, it remains focussed on medium sized traders and small commercial chains. The institution's present plan is to expand its lending operations to middle and small traders, as well as shops and commercial firms, which aim to modernize in response to increasing pressure from internationally-based firms. FIDEC's ability to reach small and micro traders will also be limited by the fact this group of traders are often not served by commercial banks, which are FIDEC's principal vehicle for channeling resources to the sector.9 8.Guidelines for the preparation of both documents were prepared by the ICR mission and discussed with FIDEC management. 9.FIDEC does support some non-bank institutions, such as credit unions (as a second tier bank), although the share of these institutions in its total portfolio remains small.  - 16 - 45. Finally, no new IBRD loans are envisaged to be needed by FIDEC at this time: it is a mature institution, due in part to the two IBRD-fmanced projects; and, it has been able to raise, during the last three years, enough resources from the Bank of Mexico to widely increase its operations, and this support is likely to continue. 46. The answer to the second question remains open and will depend to a large extent on the pace of future reforms in food and agricultural marketing policy in Mexico and on government interest in continued Bank involvement in the sector. Should there be another Bank operation in the sector, it could focus more on the provision of support services for marketing primary products. It could assist the Government create the appropriate legal and regulatory framework for storage, spot and forward markets and could also provide technical assistance for small producers and traders to participate in commodity markets. 47. The Bank should focus future discussions with FIDEC on three areas. (a) Assessing the principal constraints of Mexico's food marketing sector and the role of the key government (ASERCA, SECOFI, FIDEC, CONASUPO, Solidarity, SARH) and non-government (commercial banks, civic organizations, semi-fonnalfinancial institution, etc.) agencies in lifting these constraints. In particular, the Government should review how its programs are serving the needs of the smaller and more marginal sectors of the trading community. (b) The extent to which FIDEC should continue to concentrate most of its resources on working capital subloans. The intention of the Government when FIDEC was created, and during the early years of its operation, as well as the intention of the two Bank projects, was that FIDEC would complement the resources of commercial banks to facilitate investment projects requiring medium- and long-term financing. This rationale for the existence of FIDEC seems to be still valid. It is true that, as repeatedly argued by FIDEC, a mix of investment and working capital loans is required, but the present balance is too tilted in favor of the latter. (c) Improving the monitoring and evaluation of subloans. FIDEC should aim to develop a simple, but reasonably effective, monitoring and evaluation system that could be implemented by the participating banks to review the impact of their loans and to monitor repayment rates for the subloans.  - 17 - Key Lessons Learned 48. Three main lessons can be singled out. (a) First, project objectives should be clearly in line with the mandate and the capacity of the implementing agencies. If the project objectives require an expanded or modified mandate for the implementing agencies, then there should be formal government support for this change and the project should provide sufficient resources for the implementing agency to carry-out its new role. FIDEC effectively specializes in rediscounting subloans to the commerce sector. In the project it was asked to carry out training, technical assistance, studies and organizational activities to support modernization of the commerce sector, which were outside its mandate and for which it did not have the resources to implement. (b) Second, frequent Bank supervision missions are critical, especially in the early years of project implementation, to confirm the validity of project objectives and the appropriateness of project design to achieve these objectives. (c) Third, the use of a credit line disbursed through a variety of commercial banks is difficult to monitor and evaluate by the central responsible agency. Monitoring and evaluation activities require the cooperation of the commercial banks, which are the only ones in a position to follow up the subloans. Consequently, monitoring and evaluation responsibilities should be discussed and agreed upon with commercial banks during project preparation and appraisal, and assurances received from the banks that they will carry out these activities.  - 18 - STATISTICAL TABLES  - 19 - MEXICO: SECOND AGRICULTURAL MARKETING (Loan 3141-ME) Annex A: Statistical Tables Table 1: Summary of Assessments Substantial Partial Negligible Not applicable A. Achievement of objectives Macroeconomic policies Sector policies Financial objectives Institutional development Physical objectives Poverty reduction Gender concerns Other social objectives Environmental objectives Public sector management Private sector development Other (specify) Likely Unlikely Uncertain B. Project sustainability Highly Satisfactory Deficient satisfactory C. Bank performance Identification Preparation assistance Appraisal Supervision D. Borrower performance Preparation Implementation Covenant compliance Operation (if applicable) Highly Satisfactory Unsatisfactory Highly satisfactory unsatisfactory E. Assessment of outcome ****  - 20 - MEXICO: SECOND AGRICULTURAL MARKETING (Loan 3141-ME) Annex A: Statistical Tables Table 2: Related Bank Loans Loan/credit title Purpose Year of Status approval Preceding operations 1. Agricultural Marketing Credit line to support private 1983 Closed in Project marketing investments at the December 1989 (Loan 2262-ME) producer, w'.olesale and ictail levels, plus some funds for institutional development 2. Ninth Agricultural A continuation of the credit 1987 Closed in Credit Project program managed by FIRA June 1992 (Loan 2837-ME) consisting of short, medium and long-term subloans for crops, livestock and agro-industry, plus some funds for productive support programs 3. Financial Sector Project aimed at Restructuring Adjustment Project Mexico's highly controlled (Loan 3085-ME) financial system, moving it toward a market-based framework, while at the same time reforming the corporate and income tax system and tax administration. 1989 June 30, 1993 Following operations None  - 21 - MEXICO: SECOND AGRICULTURAL MARKETING (Loan 3141-ME) Annex A: Statistical Tables Table 3: Project Timetable Steps in project cycle Date planned Date actual/latest estimate Identification N/A June 1987 Preparation July 1987 - February July 1987 - 1st Mission 1989 September 1988 - 2nd Mission October 1988 - 3rd Mission Appraisal February 15, 1989 April 1989 (departure) Negotiations August 15, 1989 Octobel 29 - November 4, 1989 (start) Board presentation October 15, 1989 December 12, 1989 Signing N/A January 23, 1990 Effectiveness N/A April 12, 1990 Project completion December 31, 1994 Loan closing (latest estimate) June 30, 1995 December 31, 1994  - 22 - MEXICO: SECOND AGRICULTURAL MARKETING (Loan 3141-ME) Annex A: Statistical Tables Table 4: Loan Disbursements: Cumulative Estimated and Actual (US$ thousands) FY 1990 FY 1991 FY 1992 FY 1993 FY 1994 Appraisal 26,290 43,390 61,340 80,200 100,000 estimate Actual 9,970 - 940 49,380 99,430 99,532 Actual as % of 37.9 186.5 162.0 124.0 99.5 estimate Date of final disbursement - Last disbursement 11 May 1994  - 23 - MEXICO: SECOND AGRICULTURAL MARKETING (Loan 3141-ME) Annex A: Statistical Tables Tables 5 and 6: Key Indicators for Project Implementation and OperatioiP' 1. Key Implementation/Operation Indicators in SAR Report Estimated Actual A. Credit Line 1. Subloans for investment projects (US$ million) 71.05 2' 71.09 2. Subloans for working capital (US$ million) 28.10 2 28.09 B. Institutional Development 3. Study on the methodology for analyzing, monitoring and evaluating the food marketing sector 100% 3l 4. Study of the distribution systems of selected products 100% 3 5. Review of the organizational structure and technical expertise of FIDEC 100% 3l 6. Suport to a permanent advisory function in FIDEC related to institutional development 60% 7. Study on the feasibility of introducing an auction system in selected wholesale markets not done 8. Study and action plan on regulatory issues that may create barriers to entry and delay the modernization of retail food distribution in Mexico City 20% 9. Training of FIDEC, SECOFI, SARH and commercial banks staff and the private sector on various marketing and financing areas 20% 10. Skill upgrading of FIDEC's staff to improve marketing expertise 20% 11. Acquisition of computers, statistical and telecopier systems by FIDEC 100% 12. Increased product and geographycal coverage of SNIM, better disemination of information to all regions and provision of information on daily volumes on each commodity entering wholesale markets 90% 13. Acquisition of computer and telecopier system by SNIM 100% Tables 5 and 6 have been combined because key indicators were not grouped in this maner in the SAR. US$8.5 million were transfered with WB authorization from credit for investment projects to credit for working capital in June 1991, and US$ 0.55 million from studies to credit for investment projects in January 1992. See comments on the impact of the study in Table 7.  - 24 - MEXICO: SECOND AGRICULTURAL MARKETING (Loan 3141-ME) Annex A: Statistical Tables Tables 5 and 6: Key Indicators for Project Implementation and Operator (continued) III. Other Indicators A. Investment Projects Loans Amount1' No of Subloans US$000 % N' % A. Investment Projects Loans 71.058 71.7 2,068 59.4 1. Construction of Facilities 43,743 61.6 350 16.9 Type of works/facilities Supply centres 26,128 59.7 Retail markets 14,571 33.3 Self-service stores 615 1.4 Shopping centres 512 1.2 Others 1,917 4.4 Loanees Promoters(developers) 32,016 73.2 Wholesalers 5,057 11.6 Merchant associations 4,571 10.4 Merchants 1,283 2.9 Self-service stores 615 1.4 Agricultural producers 201 0.5 Range of subloans 0 to 83,000 US$ 4,024 9.2 150 46.2 83,000 to 333,000 US$ 12,664 29.0 119 36.6 more than 333,000 US$ 27,056 61.8 56 17.2 2. Acquisition of facilities 15,506 21.8 1,159 56.0 Loanees Merchants 8,058 52.0 Wholesalers 7,181 46.3 Others 267 1.7 Range of subloans 0 to 83,000 USS 11,710 75.5 1,118 96.5 83,000 to 333,000 US$ 3,796 24.5 41 3.5 more than 333,000 USS l 3. Operation equipment 5.750 8.1 157 7.6 Loanees Merchants 4.034 70.2 Self-service stores 403 7.0 Wholesalers 677 11.8 Agricultural producers 636 11.1 Range of subloans 0 to 83,000 US$ 1,599 27.8 145 92.4 83,000 to 333.000 US$ 1,011 17.6 9 5.7 more than 333,000 US$ 3,141 54.6 3 1.9 4. Transport equipment 5,340 7.5 380 18.4 Loanees Merchants associations 17 0.3 Merchants 2.681 50.2 Agroprocessors 19 0.4 Wholesalers 2,526 47.3 Agricultural producers 97 1.8 Range of subloans 0 to 83,000 US$ 4,070 76.2 367 96.6 83,000 to 333.000 USS 987 18.5 12 3.2 more than 333,000 US$ 283 5.3 0 0.2 5. Improvement of facilities 719 1.0 22 1.I Loanees Merchants 203 Self-service stores 204 Wholesalers 312 28.2 Range of subloans 28.4 0 to 83,000 US$ 719 43.4 22 100.0 100.0 Only the amount rediscounmed by FIDBC and charged to the WB loan included. Since only up to 80% of the subloans could be financed from the WB loan, the difference being met from FIDEC and commercial banks own funds, the size of the subloans is bigger than indicated by the figures.  - 25 - MEXICO: SECOND AGRICULTURAL MARKETING (Loan 3141-ME) Annex A: Statistical Tables III. Other Indicators B. Working Capital Loans Amount I' N