Document of The World Bank

Transcription

1 Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Document of The World Bank FOR OFFICIAL USE ONLY IMPLEMENTATION COMPLETION REPORT (SCL-44430) ON A LOAN IN THE AMOUNT OF US$ MILLION TO BANCO NACIONAL DE OBRAS Y SERVICIOS PUBLICOS, S.N.C. WITH THE GUARANTEE OF THE UNITED MEXICAN STATES FOR THE FOVI RESTRUCTURING PROJECT December 21, 2005 Colombia and Mexico Country Management Unit Finance, Private Sector and Infrastructure Department Latin America and the Caribbean Region Report No: 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.

2 CURRENCY EQUIVALENTS (Exchange Rate Effective March 31st, 2005) Currency Unit = Mexican Peso (MXP) 1 MXP = US$ US$ 1 = MXP FISCAL YEAR January 1st December 31st ABBREVIATIONS AND ACRONYMS AFORES BANCO DE MEXICO BANOBRAS BANXICO CAS CETES CNBV CONAFOVI CSDP L, Il DIM FICOVI FIDEC FONHAPO FOVI FOVISSSTE FSA FSAP FSRL I, II FTAL GOM GPI GPO HUDPL HUTAL IDB ICB ICR INFONAVIT IRR MBS MXP MMWs Administradoras de Fondos para el Retiro (Retirement Trust Funds) Mexico's Central Bank Banco Nacional de Obras y Sericios Públicos, S.N.C. (National Development Bank for Public Works and Services) Banco de México Country Assistance Strategy Certificados de la Tesorería de la Federación (Treasury Bills) Comisión Nacional Bancaria y de Valores (National Banking and Securities Commission) Comisión Nacional de Vivienda (National Housing Commission) Contractual Savings Development Program I and II Double-Index Mortgage Fideicomiso de Cobertura para la Vivienda (Housing Insurance Trust Fund) Fondo para el Desarrollo Comercial (Comercial Development Fund) Fondo Nacional de Habitaciones Populares (Government Very Low-Income Housing Fund) Fondo de Operación y Financiamiento Bancario a la Vivienda (Banking Finance Fund for Housing) Fondo de la Vivienda del Instituto del Seguro Social al Servicio de los Trabajadores del Estado (Housing Fund for Public Sector Workers) Financial Sector Assessment Financial Sector Assessment Program Financial Sector Restructuring Loans Financial Technical Assistance Loan Government of Mexico Garantía por Incumplimiento de Pago (Mortage Insurance) Garantía de Pago Oportuno (Timely Payment Guarantee) Affordable Housing and Urban Poverty Reduction Development Policy Loan Housing and Urban Technical Assistance Loan Inter-American Development Bank International Competitive Bidding Implementation Completion Report Instituto del Fondo Nacional de la Vivienda para los Trabajadores (National Housing Fund for Private Sector Workers) Internal Rate of Return Mortgage-Backed Securities Mexican Peso Multiples of Monthly Minimum Wage

3 NGO PLAM PMI PROSAVI PSR QAG SECAL SEDESOL SHCP SHF SOFOLES SMM UDI US GAAP Non-Govermnental Organization Price Level Adjusted Mortgage Primary Mortgage Insurance Programa Especial de Crédito y Subsidio a la Vivienda (Special Program for Housing Loans and Subsidies) Project Status Report Quality Assurance Group Sector Adjustment Loan Secretaría de Desarrollo Social (Ministry of Social Development) Secretaría de Hacienda y Crédito Público (Ministry of Finance and Public Credit) Sociedad Hipotecaria Federal (Federal Mortgage Society) Sociedad Financiera de Objetivo Limitado (Limited Purpose Financial Institutions) Salario Mínimo Mensual (Monthly Minimum Wage) Unidades de Inversión (Units of Investment) US Generally Accepted Accounting Principles Vice President: Country Director Sector Manager Task Team Leader/Task Manager: Pamela Cox Isabel Guerrero Makhtar Diop Mario Guadamillas

4 MEXICO MX FOVI RESTRUCTURING CONTENTS Page No. 1. Project Data 1 2. Principal Performance Ratings 1 3. Assessment of Development Objective and Design, and of Quality at Entry 2 4. Achievement of Objective and Outputs Major Factors Affecting Implementation and Outcome Sustainability Bank and Borrower Performance Lessons Learned Partner Comments Additional Information 24 Annex 1. Key Performance Indicators/Log Frame Matrix 25 Annex 2. Project Costs and Financing 27 Annex 3. Economic Costs and Benefits 29 Annex 4. Bank Inputs 30 Annex 5. Ratings for Achievement of Objectives/Outputs of Components 32 Annex 6. Ratings of Bank and Borrower Performance 33 Annex 7. List of Supporting Documents 34 Annex 8. Final Assessment Report Submitted by the Borrower (and English Translation) 36 Annex 9. Final QAG Panel Report 46

5 Summary Lexicon of Social Interest Housing Finance in Mexico CONAFOVI (National Housing Commission) federal institution created in 200l that is part of SEDESOL and reports to I he President of Mexico. It develops housing policies and coordinates national housing institutions, particularly at the federal level (such as SHF and FONHAPO), as well as INFONAVIT and FOVISSSTE. CONAVI (National Council of Housing) forum for public and private input on housing policies and programs; managed by and reports to CONAFOVI. FONHAPO (Low-Income Housing Fund) trust fund that serves as the main source of federal support for low-income housing. Poor repayments on below-market-rate loans made to state and local housing institutes which on-lent the funds to low-income households nearly led to FONHAPO s bankruptcy, from which it has recovered. FONHAFO. through FONAEVI, has a mandate to develop a unified system of housing subsidies. FONHAPO s main programs are Tu Casa and PROSAVI. FOVI (Banking Finance Fund for Housing) second-tier institution that provided market-rate mortgages until SHF was created in FOVI is now administered by SHF. FOVISSSTE (Housing Fund for Public Sector Workers) fund receiving 5 percent federal workers salaries, used to provide below-market mortgages, with lower rates for poorer households. It provided about l0 percent of mortgage loans in Mexico in It is also part of the national pension system (lssste). INFONAVIT (National Housing Fund for Private Sector Workers) pension fund and housing lender funded by a compulsory contribution of 5 percent of the salaries of workers in the formal private sector. It provides below-market-rate mortgages, with lower rates for poorer households. It is governed by representatives of workers, employers and government. It provided 50 percent of all mortgages in It is also part of the national pension system (ISSTE). Minimum Wage an annually set income level negotiated by labor and business organizations with the government which is widely used as a unit of measurement for a variety of social programs, including those focusing on socially-oriented housing. The current monthly minimum wage is about US$128. PROSAVI (Special Program for Housing Loans and Subsidies) this subsidy program was originally created by FOVI, which transferred it to FONHAPO. It currently provides housing grants of about US$5,000 to households earning three to five minimum wages. Beneficiary households combine the grants with accumulated prior family savings and used them for the downpayment on a house. The rest of the housing finance normally comes from a mortgage granted by one of the SOFOLES (these subsidized dwellings have on average a surface of some 40 square meters and cost about US$16,000). SEDESOL (Ministry of Social Development) this is the federal ministry responsible, among other things, for urban development. SHF (Federal Mortgage Society) it was created by the Federal Government under an Organic Law enacted in It operates under the charter of a state development bank. Its main objective is to develop housing financing markets in Mexico (primary and secondary markets) and the promotion of social interest housing which it supports with credit guarantees for the benefit of long-term credit providers in the medium term, it is expected to continue to provide long-term funding to mortgage originators servicing low to middle-income households as FOVI did in the past (mainly by on-lending to SOFOLES). In 2002, SHF took over the trustee functions from the Bank of Mexico and absorbed FOVI. SHF funds about 11 percent of all mortgages in the Mexican market. S0FOLES (Limited Purpose Financial Institutions) are non-bank financial institutions. Mortgage-oriented SOFOLES lend to developers and provide mortgages to households, principally to workers in the formal labor market (low and middle income families, mostly). They were set up as specialized lenders a result of the North American Free Trade Agreement (NAFTA) and following the Tequila crisis in 1995 became Mexico s main source of private housing lending. These institutions can issue debt in the securities market, borrow from other financial institutions, but cannot accept deposits from the public. (There are also other SOFOLES which are legally entitled to finance activities such as consumer goods, education and health services, which are not relevant for this ICR.)

6 Project ID: P Team Leader: Mario Guadamillas Project Name: MX FOVI RESTRUCTURING TL Unit: LCSFF ICR Type: Core ICR Report Date: December 21, Project Data Name: MX FOVI RESTRUCTURING L/C/TF Number: SCL Country/Department: MEXICO Region: Latin America and the Caribbean Region Sector/subsector: Housing finance and real estate markets (100%) Theme: Access to urban services and housing (P); Other financial and private sector development (P) KEY DATES Original Revised/Actual PCD: 01/07/1997 Effective: 01/11/ /11/2000 Appraisal: 02/12/1998 MTR: 09/12/ /12/2002 Approval: 03/04/1999 Closing: 06/30/ /30/2005 Borrower/Implementing Agency: Other Partners: BANOBRAS/FOVI STAFF Current At Appraisal Vice President: Pamela Cox Shahid Javed Burki Country Director: Isabel M. Guerrero Olivier Lafourcade Sector Manager: Susan G. Goldmark Augusto de la Torre Team Leader at ICR: Mario Guadamillas Richard L. Clifford ICR Primary Author: Claudio Pardo; Mario Guadamillas 2. Principal Performance Ratings (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HL=Highly Likely, L=Likely, UN=Unlikely, HUN=Highly Unlikely, HU=Highly Unsatisfactory, H=High, SU=Substantial, M=Modest, N=Negligible) Outcome: Sustainability: Institutional Development Impact: Bank Performance: Borrower Performance: S L H S S QAG (if available) ICR Quality at Entry: U U Project at Risk at Any Time: Yes The project was at risk, mostly during 2001, when Project implementation was progressing unsatisfactorily and Project reformulation had not been finalized. The ICR considers the Quality at Entry moderately unsatifactory.

7 3. Assessment of Development Objective and Design, and of Quality at Entry 3.1 Original Objective: As stated in the original Loan Agreement (May 11, 1999), the objective of this Project was to develop the social interest housing sector through: (i) the restructuring and institutional strengthening of the Banking Finance Fund for Housing (Fondo de Operación y Financiamiento Bancario a la Vivienda, FOVI); and (ii) the increase of the flow of funds for social interest housing. These objectives were in line with CAS objectives, as the Project advanced social development and growth with stability based on sound financial and capital market development. FOVI, a public housing finance trust, was managed at the time by the Bank of Mexico on behalf of the Federal Government. Social interest housing in Mexico covers households earning 3 to 8 times the monthly minimum wage (MMW), which at the time of the Project s inception meant the equivalent of some US$100 per month. Basically, FOVI S social housing programs, which supplied both subsidies and credit, were targeting the urban working poor. According to the PAD, the Project was aimed at supporting the sound development of the social interest housing sector through greater private sector participation in its financing and it supported three actions to be undertaken by FOVI: a) an increase in the flow of funds to the sector; b) a restructuring program to strengthen FOVI S products and services for primary and secondary markets, finances and management structure; and c) the implementation of a new subsidy policy. The Project objective of significantly improving the flow of financing for social housing by aggressively tapping private funding sources particularly from the emerging private pension industry, the recently introduced retirement trust funds (Administradoras de Fondos para el Retiro, AFORES) was extremely challenging. It simultaneously required the elimination of a long list of bad credit practices and the introduction of a new, financially non-distortionary housing subsidy policy plus important changes in the institutional framework and macroeconomic environment. Prevailing practices among all public and semi-public agencies at the time including FOVI was the use of below market interest rates to pass a heavy subsidy component via soft long-term mortgage lending. The separation of financing from subsidies, along with improved mortgage underwriting and operational practices, were much needed preconditions for the revival of the housing sector in Mexico via serious access to private funding. All this was clearly understood by the Project team and well articulated in the PAD. The GOM also made its position clear in its Letter of Sector Policy sent to the Bank in January of At the time, aiming for a substantial increase of funds flowing to the social interest housing sector was commendable. In fact, following the Tequila crisis of the mid-nineties, construction of new dwellings and housing prices had suffered a major setback with little hope of a fast recovery, given the numerous obstacles to the revival of mortgage financing by banks and other financial institutions burdened with heavy non-performing loan portfolios and serious capital shortages. Compounding the social stress was the long-standing large housing gap that kept growing at an alarming pace, given the inability of the Mexican housing sector to deliver new housing solutions at a pace even close to the rate of new household formation. A key question was whether the objective of seeking a major flow of private financing for social interest housing and the proposed timing was realistic, given the particularly distorted - 2 -

8 environment of the sector at the time. Indeed, the challenge of meeting that objective was made much harder by the overbearing presence of the National Housing Fund for Private Sector Workers (Instituto del Fondo Nacional de la Vivienda para los Trabajadores, INFONAVIT) a semi-autonomous public entity managing a private-sector workers fund receiving mandatory contributions of 5 percent of salaries in housing finance. This fund which had been created in 1972 to assist in home acquisition by workers employed in the private sector was by far the largest source of mortgage loans in Mexico and its largest potential source of long-term funding as well. However, INFONAVIT lending practices besides raising serious reservations on its social fairness contributed heavily to the fragmentation of the Mexican housing mortgage market with a distortionary mix of heavy subsidies via the interest rate and other financial conditions attached to its lending. INFONAVIT had extremely poor underwriting, back-office and collection practices which added a large and unintended grant component to those fortunate to get their mortgages. To make matters worse, the retreat experienced in mortgage origination by commercial banks after the Tequila crisis left an ever more dominant INFONAVIT. As the housing finance market shrank, INFONAVIT mortgage originations went from 55 percent of the total in 1995 to 73 percent five years later in 2000, when this Loan became effective. The Project s PAD clearly recognized the substantial risk behind the inability to contain the potential negative impact of INFONAVIT in the social interest housing sector. Despite the presence of this high and critical risk, the Project decided to center its reform efforts on restructuring and reforming FOVI a much smaller financial intermediary making the strategic decision of eroding the dominant and distorting role of INFONAVIT from the outside or other means rather than seeking to correct the main obstacle to the modernization of housing finance in Mexico head on. The housing fund for public sector workers (Fondo de la Vivienda del Instituto del Seguro Social al Servicio de los Trabajadores, FOVISSSTE), a smaller mortgage originator than INFONAVIT, was marred by similar problems. Numerous other state agencies, including the Government Very-Low Income Housing Fund (Fondo Nacional de Habitaciones Populares, FONHAPO) providing housing credits and subsidies to the bottom 10 to 15 percent of the formal sector and a few of the large informal economy also contributed to the severe segmentation of the housing finance market with their backward operational practices and products. The high degree of financial uncertainty prevailing in the social housing sector dominated by public and quasi public entities clearly kept innovative private parties on the margins or simply absent of the supply side. Reality was even harder since a solid 45 percent of Mexican families, usually at the bottom of the income pyramid from rural areas and the informal economy did not have and still lack a fluid access to housing finance in the formal market. Thus, in the mortgage market reality at the end of the 1990s, when the Project was designed, financial institutions (mostly public) were generating less than half the mortgages required just to meet the potential demand coming from new family formations (about 700,000 families a year). Faced with this state of affairs, the Government of Mexico (GOM) as part of the Project preparation defined its vision for the housing finance system in its Letter of Sector Policy, where the outlined strategy gave FOVI a leading role to change housing finance in Mexico and bring in - 3 -

9 substantial private sector participation. But for that to happen, it was necessary to strengthen FOVI S products and services for the primary and secondary markets. The development of secondary markets was deemed important since, as in more developed economies, the tapping of fixed-income securities market was seen as the key funding source for long-term mortgages. To meet this Project objective, a key pre-condition was to provide FOVI with skills and a financial capacity consistent with the challenge ahead. All this was properly summarized in the Project objectives. The crucial decision of selecting FOVI as a leading institution carrying out a major structural change of housing finance in Mexico was based on its willingness to restructure, its market-friendly approach and capabilities, its professional attitude and its experience as a major provider of long-term liquidity to private lenders targeting the social interest housing segment of the market basically, all SOFOLES (mortgage companies) at the time. For this the Project objective was the correct one. However, the strategy of carrying out a deep reform of the mortgage market using a second-tier player circumscribed to a rather small area of influence in the social housing finance market presented many challenges as the reform to bring the private sector in required much more than just a sound long-term liquidity provider willing to take a high portion of the credit risk involved. In fact, for an enhanced private participation in the origination, funding and risk-taking of mortgages required an all out legal, institutional and financial reform to get rid of financial distortions and establish an efficient and competitive market for housing finance, particularly in the segment providing financing for social housing solutions, which imposed a heavy burden on public resources. A lot was at stake and despite the reservations expressed by the Quality Assurance Group (QAG) panel at the time of Project preparation (May 1998) on the clarity and appropriateness of the chosen objectives and the probability of its achievement, in hindsight the strategic decision of relying on FOVI to create the right market environment and use it as the public vehicle to facilitate the way forward was not misguided. 3.2 Revised Objective: The amendment signed on December 17, 2002 broadened the original Project objective by replacing the term Social Interest Housing Sector for Mexico s housing and housing finance system in the Loan Agreement. This change was part of a long set of amendments submitted to the Bank s Board of Executive Directors for approval. The expansion of the Project objective made sense since most of the financial reforms required in the mortgage market for social interest housing were also quite valid for the housing finance market as a whole. In fact, it was essential for the modernization and integration of the Mexican market for housing finance to take a comprehensive and all encompassing approach, giving due consideration to both the social and financial dimensions behind most issues. Only by having a holistic approach was there a chance to resolve the complex set of issues being faced and achieve a successful transformation. The recent successful history of Project implementation, which has been supported by a variety of parallel interventions supported by the Bank on pension fund and social housing reform, strongly argues for such an approach. Thus, the decision to broaden the Project objective and include the whole of Mexico s housing and housing finance system was the - 4 -

10 right one. Broadening the Project objective also made sense since the December 2002 amendments made Sociedad Hipotecaria Federal (SHF), the newly created state development bank that absorbed FOVI, the executing agency of the Project. The Organic Law gave SHF the objective of impelling the development of the primary and secondary credit housing markets in general, although the emphasis was supposed to be on social housing. The original objective of transforming and restructuring FOVI then became the urgent need to make the new SHF a modern second-tier financial institution ready to provide credit enhancing and temporary funding support to private participants in the housing finance market. 3.3 Original Components: According to the PAD, the Project had three components: (i) the FOVI lending plan, an investment component, (ii) the restructuring and institutional development of FOVI, a technical assistance component and (iii) the provision of up-front housing subsidies by the Special Program for Housing Loans and Subsidies (Programa Especial de Crédito y Subsidio a la Vivienda, PROSAVI), which the Bank did not finance. The total cost of the Project, including contingencies, was estimated at US$7,310 million, of which the Bank loan financed US$ million (including the front-end fee), that is, 7 percent of the total. A. FOVI Lending Plan: The FOVI lending plan represented the bulk of the Project s cost (US$6,548 million, of which the Bank was supposed to finance US$477 million, though actual financing was US$495.6 million) and most of the funds were expected to come from private flows financing social housing mortgages. The Bank s previous experience with FOVI as executing agency had been good, so the Bank relied on the best option in the public sector and its experience with social housing finance to carry out the Project. The Bank loan provided fresh credit resources to FOVI (US$477 million) for on-lending to primary lenders engaged in social housing financing, in the same way the already closed Bank loan that supported the Housing Market Development Project had done until recently it closed on December 31, It was recognized from the start that by providing investment financing, the Bank was not addressing directly the important issue of institutional segmentation in the social housing sector. For that purpose, a Sector Adjustment Loan (SECAL) operation would have been more appropriate. The Bank investment resources were meant to supply FOVI with transitional financing to give time for a major restructuring of the way business was conducted, so as to create the conditions and incentives for a robust private sector participation in the financing of social housing mortgages. There was urgency in accelerating private sector lending and in the short-term FOVI was urgently in need of keeping the supply of long-term funds to SOFOLES via open-market credit auctions. SOFOLES, specialized mortgage companies, were the only primary lenders targeting the social interest housing sector, but they were 100 percent dependent for their long-term liquidity from FOVI, so a continued supply of investment funds and credit enhancements was essential to keep FOVI and SOFOLES focused and engaged in the reform process. There were also concerns that a major financial crisis could affect the SOFOLES industry and developers that relied on them with loans for new housing construction. In the end, some US$6.5 billion of future FOVI-supported lending under the Project was expected to come from sources other than the Bank, which to some extent mitigated valid concerns with additional - 5 -

11 government-induced funding interventions in the social interest housing market. There were big challenges involved in the chosen strategy since much depended on the capacity of FOVI to carry out a deep market reform, including its own restructuring, and on the Project s capacity to assist effectively in the process. The Bank contribution to the investment component was front loaded, with three tranches to be disbursed in a period of 27 months by July 31, The conditionality attached to this component was tied to progress in the reform program, with the first US$150 million dependant on meeting Effectiveness Conditions, which were rather light from the reform viewpoint. Even so, it took exactly 8 months for Loan Effectiveness following Loan signing. This gave a first clear signal that the reform process was not moving forward as rapidly as originally envisaged. The investment component, which provided up to 40 percent of the funding of each mortgage originated by participating primary lenders, was not expected to be difficult to disburse given the large unsatisfied potential demand for mortgages confronted by the SOFOLES. The difficulties were with the capacity of FOVI to carry out the reforms in a timely fashion and meet the attached conditionality. (Because of the delay in declaring Loan Effectiveness, the Borrower requested and got an amendment postponing the first tranche release for three months, to September ) B. Institutional Strengthening: (Total estimated cost of US$29.7 million, of which the Bank was supposed to finance US$23 million, though actual financing was US$4.4 million). A pre-condition for a lasting reform of the sector started with the transformation of FOVI. The Project aimed at putting FOVI on a path of accelerated restructuring, that to some extent was already underway supported by direct technical assistance from the Bank and resources from a Japanese PHRD grant. This technical assistance component was planned to support five sub-components on which the restructuring and institutional development of FOVI rested: primary and secondary market development; implementation of a new subsidy policy; development of a long-term financial strategy; carrying out FOVI s corporate reorganization; and implementation of a Strategic Plan and project management capabilities. The supervision mission of May 2000, four months after Loan Effectiveness, already estimated a six-month delay in Project implementation and unevenness across its various sub-components. This in part was the result of the difficulties of separating FOVI S on-lending business activities from its role as a regulator and supervisor of its client base, the SOFOLES. There was significant progress with actions in the primary market. However, time-bound actions, expected to take place by January 31, 2000 had not begun, like the contracting out of consultants for restructuring primary mortgage insurance (PMI) or the development of the information and risk management systems, required to support secondary market operations. Necessary actions to strengthen FOVI S corporate management were also delayed. On balance, the May 2000 mission concluded that there was insufficient progress on meeting second tranche expected actions and the Project was declared unsatisfactory overall. The supervision mission of November 2000 re-emphasized the earlier assessment of unsatisfactory performance. However, important progress was recognized in primary market strengthening and product line and corporate management restructuring. Nonetheless, there were serious concerns - 6 -

12 that SOFOLES might suffer a liquidity crunch due to reduced flows from FOVI and their lack of access to long-term funding in the marketplace. Unfavorable market conditions reentry of banks to mortgage lending and the aggressive commercial policy pursued by INFONAVIT also contributed to the weakening position of the SOFOLES. It added to this perception FOVI S difficulties managing a complex market reform, its own restructuring process and the slowness in the absorption of the Commercial Development Fund (Fondo para el Desarrollo Comercial, FIDEC) meant to improve FOVI S balance sheet and its on-lending capacity. However, the mission recognized the positive impact expected from the just-approved Inter-American Development Bank loan for US$505 million, which once it became effective was going to provide additional on-lending resources to FOVI (US$420 million) and allow a reactivation of PROSAVI, FOVI s subsidy program, with US$75 million for funding up-front subsidies. All told, second tranche actions, originally expected to be met by January 31, 2000, were far from being satisfactorily met, leaving the Project seriously behind schedule. The November 2000 mission also met with the Transition Team of the incoming Fox Administration, which had a goal of constructing 750,000 new dwellings per year by Meeting this goal was a big challenge since most mortgage loans at the time were granted at below-market rates. By early 2001, the new administration, which had come to power in December 2000, undertook an IMF/WB Financial Sector Assessment Program (FSAP) that included a complete assessment of the housing finance sector. A key conclusion, as explained in the Financial Sector Assessment (FSA) note of October 2002, was that the existing housing finance system would be unable to support significant progress towards the stated government s goal. As first policy steps, the new administration created the new National Housing Commission (Comisión Nacional de Vivienda, CONAFOVI) in charge of formulating and implementing Government policy in the sector, and moved ahead with the creation of SHF, the new development bank servicing the housing sector. Establishing Sociedad Hipotecaria Federal (SHF), which eventually took over FOVI from the Bank of Mexico, was a key decision for the Project s future. Its Organic Law was enacted in October The FSAP report greeted the creation of the SHF as positive on balance, while emphasizing the need to accelerate complementary reforms. It also recognized the crucial need to restructure INFONAVIT, but it considered it a daunting challenge. In fact, the FSAP recommended a reform strategy capable of addressing interrelated challenges covering subsidy reform, institutional restructuring and private sector development of mortgage origination and securitization. On the basis of all the information gathered, in June 2001 the Country Director sent a letter to the Undersecretary of the Ministry of Finance (SHCP) with proposals for ways forward with the Project, so as to move away from the unsatisfactory state of affairs. In his view, The non-compliance with the loan conditions combined with potential major difficulties in attaining the originally envisaged objectives calls for a joint consideration by the Government and the Bank of options for moderate or substantial restructuring of the loan, and/or partial or full cancellation of the remaining balance. Indeed, all things considered, cancellation of a major portion of the loan may become the preferred option. It was noted that the unsatisfactory outcome had resulted despite valuable efforts by FOVI in many key aspects of the proposed reform. The SHF was not - 7 -

13 yet created and while the new second-tier bank was a desirable outcome which was expected to incorporate many of the FSAP recommendations, its extended transition period of implementation implied that achieving the institutional objectives contemplated under the Project was going to take much longer than envisioned. 3.4 Revised Components: In July 2001, the new Task Manager sent a follow-up letter with detailed options for restructuring the Project and proposals for the work ahead. It included a thorough market analysis and clearly stated the cost and competition factors facing FOVI/SHF. It also listed FOVI S accomplishments and shortcomings to date and summarized suggested revisions and comments on the TA action plan, pointing out areas where the TA component of the Inter-American Development Bank loan could complement the Bank-supported effort. But perhaps the most striking feature of the proposed revisions outlined in the letter was the proposal for a much more comprehensive strategic and business action plan for the new SHF than originally agreed for FOVI. In hindsight, this outlined a more aggressive approach to develop a strategic and business plan that was crucial to get the Project on the fast lane and right footing. The comprehensive follow-up mission of November 2001 reached a tentative agreement on the loan restructuring proposal and the revised actions associated with tranche release. The proposal included splitting the original second tranche into two sub tranches of US$82.5 million each, with the key action of SHF s Board approval of the comprehensive Strategic and Business Action Plan attached to the second sub tranche. The implementation of the first phase of the approved Plan became a central condition for releasing the third tranche of US$162 million. In addition, two important in-depth studies were added to the Project Description and their completion also became actions due prior to the disbursement of the third tranche. The first of these technical studies (Subsidy Study) was supposed to recommend operational ways to improve the targeting and harmonization of low-income housing subsidies across public entities, without causing significant distortions in the mortgage market. The second one (INFONAVIT/FOVISSSTE Institutional Options Study) was supposed to review issues and options for the reform and restructuring of INFONAVIT and FOVISSSTE aiming at leveraging their own housing finance with additional resources from financial markets. It took over a year to fine-tune the agreement reached, process the paper-work, meet pre-conditions and have the restructuring/amendments approved by the Bank s Board and the Mexican authorities. The amendment to the Loan, Project and Guarantee Agreements and the Supplemental Letter were finally signed by the parties on December 17, Basically, the amendments kept the original Loan amount at US$ million, while extending the Closing Date for two and a half additional years. All disbursement conditions had to be met by December 31, 2004 (originally by July 31, 2001) and the new Loan Closing Date was set for June 30, The revised Project kept the three original components and their respective shares in the Bank loan. The changes were in the timing of the disbursements and the actions associated with them, which were rationalized and simplified to reflect the experience with Project implementation so far, the ambitious goals and policy changes of the new Administration, the improvement in the macroeconomic environment and financial markets, in particular, and the need - 8 -

14 to accommodate the Project to the realities of a new, fully-dedicated implementing agency, SHF. The Project revisions were major and, in hindsight, the changes introduced facilitated pursuing the Project s objectives. In the end, the full Loan amount was disbursed and by and large the revised list of expected actions met, with a few exceptions such as the completion of the two studies (to be explained later). There were no disbursements during 2002, although considerable progress was accomplished to bring the Project back to a satisfactory status. As a result, the objective of the December 2002 supervision mission was to verify that SHF/FOVI had complied with the actions expected prior to releasing the first and second sub tranches of the second tranche of the Bank loan, which were due by December 31 of 2002 and 2003, respectively. The Bank team found that SHF had met all 11 actions attached to the sub-tranches and recommended the release of US$165 million from the Loan account. The mission also reviewed third tranche conditionality and found that good progress had been made in that direction. The list below (Table 1) summarizes the key actions that had been met and which in turn upgraded the Project to a satisfactory status: Table 1. December 2002 Supervision Mission Conclusion: Actions Taken for Second Tranche Release First Sub-tranche due by Dec. 31, 2002 Second Sub-tranche due by Dec. 31, 2003 FOVI covenant obligations transferred to SHF SHF was a fully capitalized, licensed and operational bank Market rates for SHF lending implemented FOVI and SHF financial statements followed US GAAP SHF Board approved property appraisal standards Comprehensive strategic plan delivered to the Bank Local currency AAA rating obtained from S&P and stand-alone local currency. A pre-rating from Fitch in place Time-bound plan to implement results of PROSAVI study done and partially implemented Timely-payment Guarantee product designed and tested used first on a construction loan securitization Update of UDI/SMM-swap Coverage product done TA for risk management and information system contracted out to consultants. The next supervision mission a year later found that SHF, under the leadership of a dynamic new General Director had made considerable strides in the implementation of the SHF s Business Strategy and Action Plan. SHF s newly consolidated and able management team had fully - 9 -

15 embraced the strategic direction given by the Project to the modernization of the housing finance market and had made an impressive start to thoroughly implement this market-driven strategy. This is clearly articulated in the November 2003 PSR, which also gave due credit to the synergies that had developed between the Bank and SHF teams. No major obstacles to meeting pending conditionality for disbursement of the last tranche were seen. The mission gave the Project a satisfactory performance rating. The May 2004 supervision mission found that SHF had undertaken an extensive and fundamental restructuring of its role, products, risk management and organization. Also, SHF was advanced in meeting third tranche conditionality, although the important Subsidy and INFONAVIT/FOVISSSTE Institutional Options Studies remained outstanding and moving slowly. Delays in the execution of these studies were holding up the last tranche release. The Project was performing satisfactorily. During the mid-2004 mission, SHF s management informed the Bank that they wanted SHF s new research unit to conduct further studies on SHF s business development and that they preferred to use their operational budget for that purpose. Thus, SHF expressed an interest in reallocating unused TA funds to the investment component financing on-lending to SOFOLES. The mission supported these changes. In June 2005, this reallocation was finally carried out and US$18.6 million from the TA component was added to SHF s on-lending program. Thus, the investment component financed by the Bank ended up being US$495.6 million. The reduced amount of the Bank loan used for the TA component, which disbursed only US$4.4 million, is misleading in terms of the decisive importance of TA in the implementation of the Project. Much was spent on consulting and properly equipping SHF, but this new state bank got a generous equity endowment from the government and opted for the flexibility of using its own funds to pay for most of the TA required. Also, the IDB loan provided another TA funding source. In addition, as recognized by the SHF management team during the October 2005 ICR mission, the TA provided by the Bank staff through the supervision phase was of very high relevance. The third tranche was released as scheduled by December 31, The December 2004 supervision mission was satisfied with the progress made and it downgraded several risks listed in the PSR. However, the tranche condition of having completed the two important Subsidy and INFONAVIT/FOVISSSTE Institutional Options Studies was not met. In fact, the consultants for the first study were not hired until January 2005 although significant progress has been made since then. SHF used its own resources to pay for it. Meanwhile, the second study was still being prepared for bidding by the end of October However, overall significant preliminary work and difficult inter-agency agreements have been made, so expectations were high by the time of the ICR mission in October that these two studies were indeed going to provide the basis for an important consensus that could lead to better practices and make the separation of subsidies and financing a reality for all public entities financing social interest housing. In the case of the second study on institutional options, it was decided to use funds from the ongoing Housing & Urban Technical Assistance Loan (HUTAL-P088080). A summary of the actions/indicators that justified the release of the third tranche are summarized below (Table 2):

16 Table 2. Progress Made for Third Tranche Release Action by SHF/Market Indicator 1) FOVI and SHF fund a total of at least 300,000 mortgage credits between 1998 and ) FOVI/SHF has upgraded its mortgage insurance product and implemented in the marketplace 3) FOVI/SHF has issued in the market a total of at least US$30 million in mortgage-backed securities (MBS) 4) SHF has completed the first stage of implementation of the adopted Strategic Business Action Plan Progress SHF figures indicate that total number of housing solutions financed during the seven-year period under the PROSAVI and PROFOVI programs was 376,020 (from 2000 to June 2005 the FOVI/SHF on-lending program supported the origination of 289,268 mortgages equivalent to an investment about US$7 billion) A new first-loss credit insurance (MCI) product was launched, with a pricing structure based on careful research into historic default and loss rates in Mexico, and taking into account best international practices. Coverage was restricted to 25 percent of mortgage amount. The MCI has continued to be improved with the intention of aligning the structure to best U.S. practices Rather than focusing on the direct securitization of mortgages, SHF rightly decided to concentrate on providing Timely-payment Guarantees (its GPO product) for MBS backed by pools of mortgages originated by SOFOLES and securities issued by private special-purpose market conduits. As of September 2005, over US$200 million in MBS had been issued backed by the SHF s GPO product The first stage of the implementation was completed ahead of schedule and by now all hiring for key areas, including business development, risk management and research is in place. SHF continues to make substantial progress in its operations and strategic planning recently, it has formally requested the Bank to continue to provide direct TA to assist with meeting its ambitious future business/developmental goals There was a final supervision mission in April 2005, which defined SHF as a leading housing finance entity with an international reputation. The Project continued to perform satisfactorily

17 and the final disbursement took place on June 23, 2005, with the Closing at the end of June, as scheduled. 3.5 Quality at Entry: A pre-approval Quality at Entry Report was prepared by a QAG Panel and discussed with regional staff. Its final report was released on May 29, 1998 (see Annex 9 for a copy of the full report). The Assessment Panel found the Quality at Entry of the Project unsatisfactory on four grounds: a) lack of clarity and appropriateness of objectives, b) inconsistencies in the Project design vis-à-vis its objectives, c) seemingly weak government commitment, and d) inadequate attention to risk mitigation measures. The QAG Panel also found the Bank process to be unsatisfactory for accepting the Project as formulated and identified shortcomings in the internal quality assurance/enhancement system. In retrospect, many of the observations and conclusions of the QAG Panel were quite relevant and appropriate although from today s perspective, this ICR is inclined to assess the quality at entry a bit more favorably and rate it as only moderately unsatisfactory. The QAG Panel was right when it stated that the Project s agenda was substantial while the time period to realize it was too short, in particular in view of the complexity involved in developing a modern market-based mortgage finance system, as shown by international experience. Restructuring and transforming institutions like FOVI require plenty of time, the concerted effort to change by many stakeholders, and strong political will at the highest level. Some of these ingredients were not readily available at the outset of the Project s implementation, although the preparatory work for the Project was recognized as extensive and generally sound. As proposed by the QAG Panel, two years to complete the proposed complex market reform indeed indicated that there was a flaw with the initial design. In connection with the logistics, the QAG Panel saw a large and ambitious TA program lacking appropriate sequencing of various activities as well as linkage to clear time-bound output benchmarks. Past Bank experience has shown that TA work plans should be manageable and built gradually based on achievements. It must be said that during appraisal, the size of the TA component was reduced by more than half and the work plan improved. However, the optimistic view of the capacity of the Borrower to carry out these market and institutional reforms in a short time period remained part of the adopted design, even in the face of the medium to substantial overall risk rating given to the Project at appraisal. While agreeing with the perception of an unsatisfactory initial Project design due to overly optimistic expectations about the capacity to meet agreed conditionality and monitoring indicators, this ICR is of the view that the strategic decision of approaching the Mexican housing finance reform using the leadership of FOVI and the example of its capacity to transform itself was a sound one. To put FOVI center stage was risky but to wait for changes in INFONAVIT and FOVISSSTE and the right political climate was clearly unrealistic, particularly with the urgent needs of the housing sector and the weak condition of private mortgage suppliers. In retrospect, much time was gained and the housing finance reform process eventually gained momentum as a result of the strategic choices made. After all, the Bank was engaging the Borrower in a broad action front where the Project was one component in an effort to affect important reforms in financial markets, the private pension arena including INFONAVIT and FOVISSTE and improvements in the living conditions of the very poor, all of which generated externalities and

18 facilitated the modernization of the housing finance system and promoted home ownership and private construction of social interest housing. 4. Achievement of Objective and Outputs 4.1 Outcome/achievement of objective: The Project is rated as satisfactory since it achieved its major relevant objectives and accomplished substantial development results by effectively supporting the process of modernization of the housing finance sector in Mexico. It is true that it took longer than originally planned for the Project to attain its objectives and that the effort required important changes in the Project design and conditionality. But the final outcome went beyond original expectations on: (i) institutional development a well respected and market-engaged SHF; (ii) the pace with which best international practices are being introduced into the Mexican mortgage and structured securities markets; and (iii) the rate of acceptance of the principles of transparent and financially non-distorting housing subsidies, detached from mortgage origination. A much more favorable macroeconomic environment and improved financial sector conditions helped to bring this in-depth and likely sustainable modernization of housing finance in Mexico. The momentum of the reform is strong and gains continue to be made, but there is still much to be done. However, the foundations on which future change will rest have been set, and the Project can claim a determining role in all this. In particular, despite impressive gains in the overall performance and integration of INFONAVIT to the marketplace, changes are still in their infancy in FOVISSSTE and many of the States social housing providers. Moreover, market-friendly housing subsidy policies, which now are more generally accepted, still need to be implemented and practiced by all. Several efforts, many supported by ongoing Bank operations, are being directed to resolve pending issues, which should help to consolidate the gains made under the current Project. Current ICR guidelines require that project outcome be assessed against both the original and revised project objectives. In this case, separate outcome ratings give an unsatisfactory mark to the early period of the Project roughly, mid 1999 to early It covered the period when the executing agency, FOVI, was under Bank of Mexico s management. The fact that US$150 million or 30 percent of the Loan amount had already been disbursed (excluding the up-front Bank fee) when the difficulties with second tranche-related actions were encountered and for which the Project got an unsatisfactory rating weighs significantly on the overall final rating of this Bank operation. In hindsight, this ICR decided to give a satisfactory rating to the Project, despite the burden imposed by an unsatisfactory implementation performance in that first stage of implementation. The perspective here has been forward looking and emphasizes the Project s success in meeting its long-term objectives. The second stage of the implementation was clearly satisfactory. Fortunately, the initial problems were resolved satisfactorily, which was helped by the enhanced priorities on housing and its financing by the new Fox Administration, the creation of SHF and the improved financial and macroeconomic environment in the country. The conditionality was simplified and the Project s main objective broadened by the November 2002 amendments, while