Mexico - Rural Finance Technical Assistance and Pilot Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 21319-ME IMPLEMENTATION COMPLETION NOTE MEXICO RURAL FINANCE TECHNICAL ASSISTANCE AND PILOT PROJECT (LOAN 4101-ME) November 28, 2000 Environmentally and Socially Sustainable Development Unit Mexico, Colombia & Venezuela, Management Unit Latin America and the Caribbean Regional Office I This doument has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not othcrwise be disclosed without World Bank authorization.  CURRENCY EOUIVALENTS Exchange rate as of October 5, 2000 Currency Unit = Mexican Peso (N$) Mexican Pesos = US$0.105 US$1= Mexican Pesos 9.5 BORROWER FISCAL YEAR January 1 - December 31 ABBREVIATIONS AND ACRONYMS DGDBD - General Directorate for Development Banks (Direcci6n General de la Banca de Desarrollo) FIMER - Rural Financial Markets Development Fund (Fideicomiso para el Desarrollo de Mercados Financieros Rurales) MOF - Ministry of Finance NGO - Non-Governmental Organizations RE - Rural Entrepreneur RFM - Rural Financial Market RFTAP - Rural Finance Technical Assistance and Pilot Project SMU - Specialized Management Unit TA - Technical Assistance TFI - Technologies of Financial Intermediation Vice President: David de Ferranti County Director: Olivier Lafourcade Sector Director: John Redwood Sector Leader: Adolfo Brizzi Team Leader: Rodrigo Chaves  FOR OFFICIAL USE ONLY IMPLEMENTATION COMPLETION NOTE MEXICO RURAL FINANCE TECHNICAL ASSISTANCE AND PILOT PROJECT (LOAN 4101-ME) TABLE OF CONTENTS: Introduction ..........................................1 Project Objectives and Description ..........................................2 Initial Implementation Experience ..........................................4 Reasons for Cancellation .......6..................................6 Prospects for Future Operations in Rural Financial Markets ......................................7 Lessons Learned ..........................................8 Comments on behalf of the Government .........................................9 ANNEX I ......................................... 11 ANNEX2 ......................................... 12 ANNEX3 ......................................... 13 Planned ......................................... 13 Actual ......................................... 13 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not be otherwise disclosed without World Bank authorization.   IMPLEMENTATION COMPLETION NOTE MEXICO RURAL FINANCE TECHNICAL ASSISTANCE AND PILOT PROJECT (Loan 4101-ME) 1. This Project Completion Note (ICN) is for the Rural Finance Technical Assistance and Pilot Project (RFTAP) in Mexico, for which a loan of US$30.0 million equivalent was approved on October 10, 1996. The loan became effective on January 14, 1998 and was cancelled on December 15, 1999. Out of a total of US$30 million, $1.3 million were disbursed. This ICN was prepared by Manuel Lasaga (Consultant) and reviewed by Rodrigo Chaves (Lead Financial Economist, ECSSD) and Adolfo Brizzi (ESSD Country Sector Leader, LCC1C). Introduction 2. The government of Mexico has long recognized that the functioning of rural financial markets greatly affects agricultural growth, income distribution, rural poverty, and the management and conservation of natural resources. During the early 1990's, the policy dialogue between the Bank and the government, concerning the rural financial sector led to a joint effort to assess the financial services provided to rural entrepreneurs (REs) in the formal and informal sectors in three regions of Mexico. A survey of REs, of non-bank financial intermediaries, and a review of the legal framework for these interrnediaries were conducted during 1994. This study became the framework for the development of the Rural Finance Technical Assistance and Pilot Project (RFTAP).' 3. The study provided a very accurate diagnosis of the problems facing the development of financial services in rural markets. The principal conclusion of the study was that Rural Financial Markets (RFMs) were poor in terms of aggregate efficiency and fairness. Some of the key findings were: (i) RFMs were highlv segmented with very limited flow of funds across regions or groups of individuals; (ii) there was minimal participation by commercial banks in this sector - a condition aggravated by the government's withdrawal from financial intermediation; (iii) lending was hampered by inadequate legal mechanisms for dealing with collateral and loan guarantees; (iv) operating costs for financial intermediation were very high, specially credit analysis and supervision; (v) government participation in RFMs had engendered the use of sizeable subsidies and encouraged defaults through periodic debt-forgiveness programs; (vi) REs were willing to pay high rates of interest; and (vii) there was evidence of a significant propensity to save - albeit in terms of non-monetary instruments. The Bank and the Borrower felt that these problems could be overcome through the development of new financial intermediation methodologies, to be applied by commercial banks, which emphasized lower costs through the introduction of new technologies. ' Mexico: Rural Financial Markets, Report No. 14599-ME, August 25, 1995.  2 4. Not long after the completion of the survey, the Mexican economy was rocked by a major devaluation crisis in December 1994. At the time the final RFM report was issued in August 1995, the financial sector was in the midst of a serious crisis, which necessitated a maj or rescue operation by the government in order to prevent the collapse of the financial markets. While the survey data pre-dated the financial crisis, the final report did raise concerns that the crisis had aggravated the problems faced by RFMs through negative income shocks and the exit of financial intermediaries from these markets. Project Objectives and Description 5. Based on the President's Report and the Loan Agreement, the objectives of the project were to: (i) augment the participation of REs, especially the poor, in rural financial markets as depositors and borrowers; (ii) overcome segmentation of those markets; and (iii) increase competition in the rural financial markets. These objectives were consistent with the fundamental conclusions of the RFM study. Because it was a pilot project, the loan was wisely limited to US$30 million, with the expectation that if it was successful, it could then be followed up with a larger operation. 6. In order to achieve its objectives, the project was designed with two components: - Part A: Technical Assistance (TA) and training in support of rural financial market initiatives. The project was expected to assist the government in undertaking the analytical work to develop policies towards RFMs, to test such policies, and to underpin the sector's improvement. Specifically, the studies would address a broad scope of sectoral issues, which included: (I) improving the business environment for financial transactions in rural areas through product innovations, enhanced efficiency, and greater market information; (2) improving the government's performance in rural financial markets through its two principal rural development banks, Banrural and FIRA; (3) improving public policies regarding REMs; (4) improving government regulations and supervisory framework in the rural sector; and (5) improving the availability of financial services provided to the poor in rural areas by non-governmental organizations (NGOs). * Part B: Pilot Projects in RFMs: Provide resources for the establishment of private commercial banking branches in rural areas, which would benefit from new financial technologies and sector policies developed in Part A. The project would also extend to these newly established branches additional funding for on-lending to REs to finance productive investments. 7. The TA component was to provide the research and development work that would support the development of rural banking branches. It was expected to act as an incubator of new technologies that would reduce transactions costs associated with financial intermediation in rural markets. Of the US$30 million from the World Bank's loan,  3 US$11 million was allocated to Part A, US$15 million to Part B, and the remainder was unallocated. 8. While the project was highly innovative, it was also a high-risk venture. The basic premise was that the TA component would generate the required banking technologies early on, and that their application would be almost immediate as banks established their rural banking network. The pilot component called for the selection of four banks, based on a competitive bidding process. One of the World Bank's concerns was the likelihood that participating banks would experience financial distress due to the aftershocks of the 1995 crisis, and for that reason strict criteria were established for participation. 9. The basic structure of the project was sound, it was consistent with the goals of deepening financial intermediation in the rural markets through innovative technologies. However, there were several design features that may have detracted from the project's ultimate goals. Apparently there was some discussion during the early phase of project design regarding the possibility of including non-bank financial intermediaries in the pilot project. Non-banks such as Credit Unions and Savings and Loan Associations are active in rural markets. However, their financial performance has been very disappointing. This is due in large part to inappropriate operating procedures that have been allowed by an inadequate regulatory framework. Despite these shortcomings, the non-bank intermediaries have developed an extensive franchise throughout the rural areas of the country. Perhaps a more aggressive component of TA for non-bank intermediaries, which provided similar types of technologies as developed for the banks, could have been considered. This component could have also incorporated basic training in banking practices and risk management for those intermediaries that were yet unprepared for the new technologies developed by this project. 10. Some of the criteria recommended by the Bank regarding commercial bank participation, which were subsequently adopted by the Borrower in the Operational Manual, might have been more flexible. For example, each of the selected participating banks were expected to establish at least 30 branches. Each branch had to be located in a community of not more than 20,000 residents. The location had to be at least 30 kilometers from the nearest banking center. There were also constraints on the type of furniture used in the branches. While the feasibility studies justified these parameters, this was an experimental project. Therefore, banks should have been given more flexibility in determining the number of branches, location and their physical characteristics, since they were the ones taking on the business risk. Studies and recommendations made under the project did provide valuable market information, which the banks could then have incorporated into their own branch selection process. In other words, the techniques derived in Part A of the project would have been made available to the banks to support their business development strategies, but the actual choice of techniques should have been up to the banks. 11. Government regulations represented another impediment to the success of the project. Banks were required to submit to the Ministry of Finance (MOF) for approval an annual plan for the local branches they were going to open. In some cases, banks showed  4 interest in this project thinking that their participation in RFMs might result in a more favorable response by the MOF to their annual branch plan. On the other hand, since MOF approval is also required to close down a branch, banks may have hesitated to develop rural branches if the perception was that regulators may have not allowed them to close down unprofitable branches. Perhaps FIMER could have negotiated more flexible conditions regarding the opening of rural branches. 12. In the design of the TA component emphasis was placed on the development of alternative Technologies of Financial Intermediation (TFI) with the goal of identifying techniques that would help reduce operational costs for branches in rural areas. While this type of research and development activity is valuable, it may have been more useful for application in a subsequent project, or once the branching network had been established. For example, it would be difficult to introduce "intelligent cards" in the rural sector due to the lack of familiarity in rural communities with this type of technology. At the same time, the application of these technologies would entail substantial additional investments by the banks, which might not have been justified by market conditions. Initial Implementation Experience 13. The implementation of the RFTAP was short-lived, it was made effective on January 14, 1998 and canceled on December 15, 1999. Because of the pre-mature cancellation, none of the objectives were achieved. Nevertheless, the initial implementation experience provides useful insights into the reasons why the project failed. 14. The time when the project was identified, particularly the diagnostic work based on the 1994 RFM survey, and the time when it was approved and made effective were two very different periods. The financial sector crisis that followed the maxi-devaluation of the peso in December 1994 undermined the financial integrity of the banks. Some banks became insolvent while others were merged with other banks. The seriousness of the situation was already evident at the time of Board approval; however, the government was apparently optimistic with respect to the ability of the banks to recover. Nevertheless, the MOP addressed these risks, and concluded that supporting of banking operations in the rural sector could facilitate the banks' recovery process. 15. The project was carried out under the overall responsibility of MOF in very close collaboration with other government agencies and institutions. Responsibilities for the management of the project were assigned to the General Directorate for Development Banks (DGDBD). A Specialized Management Unit (SMU) was established to operate the project within that division. 16. The SMU was established in December 1996 as the Fideicomiso Para el Desarrollo de Mercados Financieros Rurales (FIMER). The "Fideicomiso", or Trust, provides a legal framework to overcome the otherwise cumbersome and rigid budgeting procedures applied under Mexican fiscal norms, whereby Bank projects cannot provide additional resources to those already identified in the annual budget cycle. The Bank's funds form  5 part of the foreign currency pool applied to the financing of the overall budget deficit. While not the ideal structure for development financing, the Trust operates from a capital contribution made by the Federal Government. The Trust Agreement between MOF and the Trustees specifies how the project is to be implemented. The Trust was governed by a Technical Committee comprised of representatives from MOF, Banco de Mexico, NAFIN, and FIMER. PartA: TechnicalAssistance 17. The TA component dealt basically with a series of studies, some of which were completed, while others were in the initial phase at the time the project was cancelled. The topics were mostly in accordance with the project description contained in the Loan Agreement. However, some members of the Technical Committee expressed concerns about delays in implementation and relevance of some of the topics to the goals of the project. A number of studies were proposed, some were initiated, but few were completed. The following topics were considered: * Technologies for rural banking: 1) an economic model for rural branch selection to be used by participating banks; 2) a model for customer selection based on artificial intelligence; and 3) an analysis of "intelligent cards" for rural financial market transactions. * Risk management studies: 1) agricultural insurance; 2) agricultural productivity, 3) analysis of bonded warehousing; 4) a model for credit administration; and 5) an income - expenditure model for rural households. * Miscellaneous studies: 1) an analysis of non-bank financial intermediaries; and 2) an evaluation of Banrural based on an analysis of its loan portfolio. 18. A very useful outcome of the research work undertaken by FIMER was an extensive socioeconomic database on rural areas of the country, which assisted banks in identifying the locations for the rural branches. On the other hand, the rationale for the BANRURAL study as mentioned in the above list, was not clear. The Bank had already developed an extensive and critical analysis of that institution, which underscored extensive problems with its loan portfolio. The conclusions of the FIMER study regarding BANRURAL were in contradiction to the Bank's own analysis. FIMER's analysis of BANRURAL's loan portfolio revealed that BANRURAL was highly efficient in financing the rural sector. In view of the divergent conclusions from the earlier Bank study, FIMER should have held discussions with the Bank regarding its conclusions prior to making the results of its study public. For this reason, greater Bank monitoring of the project's activities might have been advisable. Part B: Pilot Experimentation with rural branches 19. Even before Loan effectiveness, the pilot component was not getting a favorable response. The first bidding round for participation by commercial banks was announced  6 in May 1997, but postponed due to lack of response. On the second attempt, only one bank presented a bid, which was subsequently accepted and a participation agreement was signed with Banca Serfin in November 1997. Some banks felt that the RFTAP could have been "marketed" more broadly to the financial community. 20. A second bidding round was eventually cancelled also due to insufficient response. At that point, FIMER proposed opening the pilot project to non-bank financial intermediaries. It was eventually agreed to initiate a working relationship with a Credit Union (UCEPCO) and a Popular Savings Union (Caja Popular Mexicana), but only in terms of TA. 21. Banca Serfin eventually dropped out of the program. After signing the participation agreement, the bank had put together a business plan to develop the rural branches including the relevant site selections. However, senior management delayed its decision for months before canceling the project. The negative outcome in this case was highly influenced by the weak financial condition of the bank. Less than a year after the project had been presented to senior management, the Superintendency of Banks had to intervene to rescue Banca Serfin due to major problems with its loan portfolio. After separating the bank's bad loans and other depreciated assets, the Superintendency sold the remaining portion of the bank to a foreign financial institution. However, the clean up of the bank resulted in a large financial loss for the government, which was only partially recovered from the proceeds of the sale. 22. Interestingly, the development of rural banking methodologies, which consisted of an analytical model for rural branch selection, proved to be a very valuable tool for the non-banks. While commercial banks typically have their own models for branch selection, the non-bank intermediaries lack sophistication for this type of analysis, and in this regard, FIMER could have been a very useful advisor. Greater focus of the TA in the form of banking methodologies for the non-banks could have been very productive in achieving the loan objectives. In this regard, FIMER might have adopted a more proactive role in seeking Bank support to re-allocate project funds to this component. Reasons for Cancellation 23. Timing was a decisive factor in the cancellation of this project. In December 1994, responding to sizeable outflows of capital, the government allowed the peso to float, resulting in a maxi-devaluation of the currency. The combination of a major devaluation and a sharp contraction of liquidity in the banking system severely compromised the financial position of many intermediaries. The liquidity crisis was aggravated by a serious deterioration in the banks' loan portfolios. As a result, the Government had to provide substantial financial support to the banking system. In some cases the superintendency was able to facilitate the sale of weak banks, while in others it was necessary to bail them out, resulting in huge losses for the government. Eventually the government opened up the market to foreign investors, thus providing fresh capital and a new start for the failed banks. The aftershocks of this crisis are still being felt today, although the banking system is in a much stronger position.  7 24. The worst period for the banking system was 1996. In that year the banks posted a net loss of NP$ 7.8 billion, or about 1 percent of assets. However, this does not reflect the implicit losses through the government's swap of treasury notes for bad loans under the trust fund FOBAPROA. As a result of the crisis banks began to downsize their operations, including the closure of branches throughout the country. Banks tightened considerably their credit policies, with very limited financing available only to their best clients. The rural markets were thus left out of the financial system loop. 25. With respect to the implementing agency, FIMER's Technical Committee, which acted as the governing board, did not have a clear agenda regarding this project. This was reflected in the long delays in making decisions - although the low frequency of the Committee meetings may have also contributed to this problem, since the Committee met only twice a year. There were extensive discussions on procedural matters and less on strategies for rural financial markets. Thus the lack of consensus as to the implementation of the project may have distracted attention from the operational challenges faced by the project. 26. Cancellation was also influenced by an internal debate within the government and the Technical Committee regarding the transfer of FIMER from DGDBD to another agency. During the second half of 1998, one of the Technical Committee members proposed transferring FIMER to FIRA, the government's general trust fund for financing the rural sector. The reason was that FIRA was a large institution with the administrative and operational capacity to manage FIMER. This would then free up FIMER's staff so as to concentrate on the implementation of the TA. The Committee was particularly divided on this issue. Other members proposed moving FIMER to another trust fund, FOCIR. While the World Bank assured the Committee that it had no objection to any transfer as long as it did not interfere with the implementation of the project objectives, the discussion languished for several months, at which point both the TA and the pilot components had fallen far behind schedule. Once the project had lost its initial momentum, and in view of the continual divergence of views over the future course of FIMER, cancellation became the only option. Prospects for Future Operations in Rural Financial Markets 27. One of the principal reasons for the failure of the RFTAP was the banking crisis of 1995. However, the diagnosis that was made of the problems in rural financial markets is still as relevant today or even more so than when the Bank prepared the 1994 survey. It would be very useful to update the 1994 survey, as this would help to understand the impact of the 1995 crisis on the RFMs. 28. Participants in the RFTAP observed that the TA component should have continued since a number of the studies that had been identified at the time of the project are still highly relevant today. In this regard the Bank is preparing a follow-up operation in the form of a Rural Microfinance Capacity Building TA. This project addresses the lack of access by the poor in rural areas to financial services. It will focus on institution building and the potential for savings mobilization and overall financial intermediation capacity in  8 the rural sector. The target participants for this project will be selected non-bank financial institutions. While the project is small in terms of funding, it should provide very valuable capacity building for non-banks. 29. Commercial banks have also expressed an interest in a project similar in scope to the RFTAP, but which offers much greater flexibility in the modus operandi of the branch network. Some intermediaries are already operating in these markets but on a very limited basis. At the same time, BANRURAL has adopted the branch selection model and is also developing a risk management model along the lines developed by FIMER. Some interest was also generated outside Mexico with respect to the results of this project. An NGO in Chile with interest in rural markets contacted FIMER in order to develop a similar program. Based on these considerations, it would appear that the project concept is still relevant; but in the case of the RFTAP, its timing was unfortunate. Lessons Learned * The success of a rural financial markets project that operates through the formal banking sector requires a healthy financial and banking sector. One of the principal reasons for the failure of the RFTAP was the banking crisis of 1995. Rural financial markets are much more vulnerable to the effects of banking sector problems since they are usually not a priority market for the banks. * Good diagnostic work can facilitate the policy dialogue as well as the preparation of the project. Solid identification and quantification of the problem can generate much value-added for the project. One of the outstanding features of project preparation was the quality of the rural financial markets survey. Government officials interviewed for this report commended the survey for its insights into RFMs. * In the case of experimental projects, design needs to emphasize flexibility. One way to smother innovation is to impose rigid rules, especially if they are not well founded on market practices. In the case of the RFTAPP, the Government adopted somewhat inflexible policies regarding branch development that may have discouraged participation. * Analysis of external factors during loan preparation can prevent subsequent disappointments during implementation. The financial crisis of 1995 was in full swing at the time of project preparation. While World Bank staff raised this issue early on, it may have considered postponement of the project, or phasing in the pilot component at a later date. * Borrower ownership of a project is an important factor in determining its success. There was close collaboration between the Bank and the government during the loan preparation phase. However, the governing board of FIMER did not have a clear agenda as to the implementation of the project. A change of senior MOF officials at the time the project had been prepared also impacted negatively on the outcome.  9 * While Borrower ownership is important, inclusion of other stakeholders is critical to a project's success. Some banks felt that the government was not proactive in marketing the project. Private sector participation in the Technical Board may have been useful. * The implementing agency for this type of project should be staffed with individuals who have operational experience. Some of the members of the Technical Committee questioned the delays and higher than expected costs for studies carried out under the TA component. In some cases, the relevance of the studies was challenged. In this regard, the excellent technical quality of the staff at FIMER could have been complemented with greater operational experience that would have given the project a more dynamic focus. Comments on behalf of the Government 2 30. In 1994, the Government of Mexico in collaboration with the World Bank performed a study of financial services available to rural entrepreneurs. The principal conclusions of this study were that rural entrepreneurs had very limited access to financial services and that government activities in rural financing have resulted in economic distortions. 31. The Government of Mexico and the World Bank defined the basis for the Technical Assistance and Pilot Project in order to find solutions to the problems of financial markets in rural areas. The principal innovation of this project was that it would perform an empirical test of various experimental technologies for rural financial intermediation. This was justified by the fact that commercial banks were reluctant to invest in these new technologies, since they could not capture a good part of the social benefits from these innovations. 32. A trust fund was created under the name of Fideicomiso para el Desarrollo de los Mercados Financieros en Regiones Rurales (FIMER) in December 1996. A first auction for participating banks was held in May 1997, and was subsequently postponed after banks had requested additional time to prepare their bids. The bid was reopened on October 2, 1997, and Banca Serfin was qualified as a participant in the project. 33. One of the key issues identified by FIMER regarding the lack of financing in the rural markets was the inability of commercial banks to perform an acceptable credit analysis of prospective borrowers, because of limited information and high transaction costs. For this reason, FIMER focused its research activities in the development of intermediation technologies that would reduce transaction costs in the rural sector. 2 This section is based on an executive summary of the following report prepared by the FIMER: Nota de Termino del Proyecto- Cumplimiento y Alcance de los Objetivos del Proyecto (March 2000). The full document is available upon request.  10 34. FIMER identified numerous research projects dealing with the problems of financing in the rural sector. The terms of reference were prepared for a number of these studies, which included the following: * A Financial Analysis of Banrural * Development of a Model of Rural Bank Branching * An Artificial Intelligence Model to Evaluate Credit Risk * Insurance in the Rural Sector 35. Under the technical assistance component FIMER proposed the participation of non-bank financial intermediaries. An agreement was signed in October 1998 with a savings and loan association (Caja Popular Mexicana), and in July 1999, with a credit union ( Uni6n de Credito de Productores de Cafe del Estado de Oaxaca. The agreement was to provide these organizations with the branch selection model developed by FIMER. 36. With respect to the Pilot Component, FIMER worked with Banca Serfin to develop its rural branching network. A business plan was prepared including the location and physical characteristics of each branch. FIMER was able to obtain World Bank agreement that Banca Serfin could use existing branches as part of the network 37. FIMER announced the second auction in January 1998. However this time the auction did not attract interest on the part of the banks, and was subsequently declared vacant. 38. One of the principal causes for the delays and lack of interest in this project was the problems encountered by the financial sector following the December 1994 devaluation of the peso. As a result of this, FIMER proposed to focus more intensely on the technical assistance component of the project. However, a number of administrative problems caused undue delays in the implementation of this component. In June 1998, the Finance Ministry proposed the transfer of FIMER to the government's rural sector financing agency, FIRA. In May 1999, the government announced that the transfer was going to be made to another agency, FOCIR. Long delays associated with these administrative changes adversely impacted project implementation. Eventually the government opted for the cancellation of the project in view of the very limited progress on implementation and the disagreement regarding the project's activities. The project was thus cancelled in December 1999. 39. In view of the government's interest in finding a solution to the problems in the rural financial markets, $30 million from the Federal budget were transferred to FOCIR in December 1999 to finance studies on the development of financing strategies for the rural sector. These activities will be managed by a special unit within FOCIR, Fondo para la Realizaci6n de Estudios que Promuevan el Desarrollo del Sector Financiero Rural (FOFIRE). In this regard, the project's technical assistance initiatives will continue but under different modalities.  11 ANNEX 1 MEXICO RURAL FINANCE. TECHNICAL ASSISTANCE AND PILOT PROJECT ESTIMATED COSTS AND FINANCING PLAN US$ million Foreign Component Project Components Foreign Local Total as % of as % of Base Total Cost Project Costs A. Technical Assistance 1. Business environment for financial 1.4 1.1 2.5 56.0 2.4 transactions in the rural areas 2. Government direct participation in RFMs 0.4 0.9 1.3 31.0 1.3 3. Applied policy analysis 0.4 0.9 1.3 31.0 1.3 4. Regulation of RFMs 0.5 0.4 0.9 55.5 0.9 5. Studies of markets related to RFMs 0.4 1.2 1.6 25.0 1.5 6. Support to financial NGOs 0.3 0.5 0.8 37.5 0.8 7. Policy dialogue and dissemination of research 0.2 1.4 1.6 12.5 1.5 results Subtotal Direct Technical Assistance 3.6 6.4 10.0 36.0 9.7 B. Pilot Experimentation with Financial Technologies 1. Development of financial technologies 0.4 0.1 0.5 80.0 0.5 2. Establishment of experimental bank rural 1.9 1.1 3.0 63.3 2.9 branches 3. Operation and supervision of rural branches 16.0 16.0 0 15.5 4. Establishment and operation of a Specialized 0.1 4.2 4.3 2.3 4.2 Management Unit 5. Initial working capital for experimental 6.3 56.7 63.0 10.0 61.2 branches 6. Dissemination of pilot results 0.2 0.4 0.6 33.3 0.6 Subtotal Pilot Experimentation 8.9 78.5 87.4 10.2 84.8 Contingencies 0.5 5.1 5.6 8.9 5.4 Total Project Costs 13.0 90.0 103.0 12.6 100.0 Proiect Financinz Government 0.0 3.1 3.1 0.0 3.0 Participating private banks 0.0 69.9 69.9 0.2 67.9 IBRD 13.0 17.0 30.0 43.3 29.1 Total 13.0 90.0 103.0 12.8 100.0  12 ANNEX 2 MEXICO RURAL FINANCE TECHNICAL ASSISTANCE AND PILOT PROJECT BANK RESOURCES: STAFF INPUTS Planned Actual Stage of project cycle Weeks US$ ('OOOs) Weeks US$ Preparation to Appraisal 35.20 56.00 25.20 56.40 Appraisal 14.90 38.10 59.90 282.4 Negotiations through Board approval 10.00 18.90 3.10 8.00 Supervision 67.161 194.90 49.76 155.68 Completion 3.00 13.00 2.07 11.43 Total 130.26 320.90 140.03 513.91  13 ANNEX 3 MEXICO RURAL FINANCE TECHNICAL ASSISTANCE AND PILOT PROJECT PROJECT TIMETABLE Planned Actual Time taken to prepare 8 months First Bank Mission January, 1996 January, 1996 Appraisal April 8, 1996 April 8, 1996 Negotiations August 5, 1996 August 5, 1996 Board Approval October 10, 1996 October 10, 1996 Effectiveness November, 1996 January 14, 1998 Cancellation December 15, 1999 Closing June, 2000 NA