India - Housing Development Finance Corporation Project

Documet of The World Bank IFO OMCL USE ONLY Reot No. 12021 PERFORMANCE ALDIT REPORT INDIA HOUSING DEVELOPMENT FINANCE CORPORATION PROJECT (LOAN 2929-IN) JUNE 17, 1993 MICROGRAPHICS Report No: 12021 operations Evaluation Department This document has a restricted distribution an# may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. COUNTRY EXCHANGE RATES Indian Rupee (Rs/US$) Appraisal Estimate (1988) $1.00 = 13 25 Re Actual: Average 1988 = 13.917 Average 1989 * 16.226 Average 1990 = 17.504 Average 1991 = 22.742 ACRONYMS AND ABBREVIATIONS ALL - Average Loan Life GOI - Government of India HFIs - Housing Finance Institutions HFS - Housing Finance System IFC - International Finance Corporation LIC - Life Insurance Company NHB - National Housing Bank MF - Ministry of Finance PAR - Performance Audit Report PCR - Project Completion Report RBI - Resezve Bank of India SAR - Staff Appraisal Report SOE - Statements of Expenditures FISCAL YEAR April 1 to March 31 FOR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C 20433 U.S.A. Office of Director-General Operations Evaluation June 17, 1993 =MHRAMDUM TO THE EXE9UTIVE DIRECTOR6 MD THE PRESIDET SUBJECT: Performance Audit Report on India Housing Development Finance Corporation Proiect (Loan 2929-IN) Attached .. the report entitled "Performance Audit Report on India - Housing Development Finance Corporation Project (Loan 2929-IN)" prepared by the Operations Evaluation Department. The Audit covers the only loan for US$250.00 million equivalent made directly to the Housing Development Finance Corporation (HDFC), 4 private company launched with IFC's assistance in 1978. The audited project helped HDFC to expand its mortgage lending and to pursue its lower-income targeting by earmarking half the loan to borrowers below the median urban income. The demonstration effect looked for by the project was achieved: mortgage len6ing by private specialized housing finance institutions is viable and sustainable. The improvements to the legal and regulatory framework for housing finance were elusive. The central-government apex institution created on the eve of the project had no success pushing several law changes, but it introduced a contractual savings scheme and provided all housing finance institutions with refinancing facilities. Overall, the choice of the Borrower made difficult the implementation of sector-level reforms. The Audit concludes that the project was successful and that housing finance is a good transfer avenue when the intermediary is well managed. This document has a restricted distribution and may be used by recipients only in the p, rmance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. FOR OFFICIAL USE ONLY PERFORMANCE AUDIT REPORT INDIA HOUSING DEVELOPMENT FINANCE CORPORATION PROJECT (LOAN 2929-IN) TABLE OF CONTENTS Page No. Preface.................................................................. i Basic Data Sheet......................................................... ii Evaluation............................................................... iv I. BACKGROUND...................................................... II. A. Objectives................................................. I B. Disbursements............................................... 2 C. Beneficiary Targeting....................................... 4 III. HOUSING DEVELOPMENT FINANCE CORPORATION............................ 5 A. Financial Policy.............................................. 5 B. Financial Covenants........................................... 7 C. Portfolio Quality............................................. 8 D. Bank Loan's Maturity, Grace Period and Exchange Risk.. ........ 9 E. Organization .................................................. 10 F. Accounting and External Audit................................. 10 G. Sustainabillty................................................. 11 IV. SECTOR POLICY CHANGES................................ ............. 11 A. Role of NHB ................................................... 11 B. Secondary Mortgage Market..................................... 12 V. CONCLUSIONS AND LESSONS ........................................... 13 A. HDFC Performance.............................................. 13 B. Bank Performance.............................................. 14 C. Project Rating................................................ 15 ANNEX Comments from the Borrower........................................ 16 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. PERFORMANCE AUDIT REPORT INDIA HOUSING DEVELOPMENT FINANCE CORPORATION PROJECT (LOAN 2929-IN) PREFACE 1. This is a Performance Audit Report (PAR) on the First Housing Finance Project in India. involving a World Bank loan amounting to US$250.0 million equivalent, to the Housing Divelopment Finance Corporation (HDFC) with the guarantee of the Government of India (GOI) to finance mortgage loans for the purchase or construction of houses or apartments generally in the urban areas of the country. The loan was approved on March 31, 1988; it was entirely disbursed before the closing date. 2. The PAR is based on the Project Completion Report (PCF) prepared by the South Asia Regional Office and issued on June 29, 1992, (Report No. 11453) the Staff Appraisal Report (SAR), the loan documents, and a study of the project files. An OED mission visited India in October 1992. The excellont cooperation and valuable assistance provided by HDFC in the preparation of this report is gratefully acknowledged. 3. The PCR provides a comprehensive account of the project experience, is informative on the achievements and allows checking on the compliance with covenants. The PAR elaborates on selected aspects of proj ct implementation, in particular the financial performance of HDFC, the affordability of its mortgage loans to households, and the relative progress made to put in place a regulatory framework favorable to the growth of housing finance institutions. The PAR concurs with the PCR satisfactory rating of the Project. 4. Following standard OED procedures, copies of the draft PAR were sent to the Borrower and to the Executing Agency for comments. Couments received from the Housing Development Finance Corporation have been reproduced as an Annex to the PAR. PERFORMANCE AUDIT REPORT INDIA HOUSING DEVELOPHENT FINANCE CORPORATION PROJECT (LOAN 2929-IN) BASIC DATA SHEET KEY PROJECT DATA Appraisal Actual Actual as % ungg Expectations Estimates of AVRr. st. Total Project Cost (US$ million) 840.8 851.4 111.3 Loan Amount (") 250.0 250.0 100.0 Date Physical Components Completed 9/91 2/91 Economic Rate of Return (%) N.A. N.A. CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS FY ended June 30, EM FY89 = Y91 Appraisal Estimate (US$ million) 25.0 101.0 182.5 250.0 Actual (US$4 million) 0.0 103.3 195.5 250.0 Actual as of Z Estimate 0.0 41.3 78.2 100.0 nate of Final Disbursement: 2/13/1991 PROJECT DATES Items Orizinal Plan Actual First Mention in Files 01/9/86 Appraisal 08/87 10/87 Negotiations 02/88 02/88 Board Approval 04/88 3/31/88 Loan Agreement Date 05/88 4/21/88 Effectiveness Date 05/88 5/18/88 Loan Completion 09/91 2/13/91 Closing Date 09/30/91 2/13/91 STAFF INPUT (Staff-Weeks) 7JY8j FY90 FY91 FY9 Tta Preappraisal 9.0 9.1 18.1 Appraisal 20.4 20.4 Negotiations 3.0 3.0 Supervision 8.3 9.9 16.4 6.4 6.7 47.7 Other 3.6 4.4 8.0 Total 12.6 40.8 9.9 20.8 6.4 6.7 97.2 MISSION DATA Stage of Month No. of Days in Performance Proiect Cycle Yea Persons Field Rating' Identification 12/86 2 1.6 NA Preparation 03/87 3 4.8 NA Pre-appraisal 06/87 3 9.6 Na Appraisal 10/87 3 9.6 NA Supervision I 09/88 1 3.0 1 Supervision II 03/89 1 0.8 1 Supervision III 08/89 4 5.6 1 Supervision IV 03/90 4 6.4 1 Supervision V 12/90 2 3.6 1 Completion 10/91 2 6.0 1 OTHER PROJECT DATA Borrower: Housing Development Finance Corporation (HDFC) Executing Agency: Housing Development Finance Corporation Fiscal Year of Borrower: April 1 - March 31 Follow-up Project: none 1. Problem Free; 2. Minor P-oblems; 3. Major Problems. iv PERFORMANCE AUDIT REPORT INDIA HOUSING DEVELOPMENT FINANCE CORPORATION PROJECT (LOAN 2929-IN) EVALUATION SUMMARY Introductin to improve sector incentives. The 1. At the time of appraisal in first two objectives we,z to be 1988. the organized -apital market of controlled through the loan made India vas largely ignori-g the directly to HDFC and the third one funding needs of housing investments. through trenching (PAR, para. 4). At present, the housing finance The main component of the project was system (HPS) is still not an conceived as a time-slice of HDFC's integrated part of the financial lending over the period October 1987 system. Furthermore, the informal (appraisal date) to March 1991 (end housing sector continues to be the of FY91). US$248 million of the loan main source of supply. The Housing were to refinance 30% of all Development Finance Corporation disbursements made by HDFC, with at (HDFC) established in 1978, for least half earmarked to sub-borrowers several years was the only financial with incomes below the median urban institution, yet private, dedicated income and the other half without any to providing mortgages (PAR, para. conditionality (PAR, para. 5). 1). The 7th Plan, which gave priority to meeting the housing Project Implementation demand, called for changes in the aector, among which a much greater 3. The loan to HDFC was fully emphasis on private finance for disbursed 32 months before the end of housing. To control the expected the grace period, and over 3 years multiplication of housing finance before a traditional urban project. institutions (HFIs), GOI proposed the This illustrates the resources creation of an apex institution, the transfer advantage of housing finance National Housing Bank (NHB) (PAR, projects over the traditional sites- para. 2). For the Bank, this and-services projects (PAR, para. 6). operation was a major departure from The concept of trenching was the tested strategy of financing introduced prior to appraisal co give specific investment projects, mostly the Bank the means to enact the of the sites-and-services type (PAR, institutional changes in the sector para. 3). which were under GOI's control (PAR, para. 7). Prior to negotiations, the Project Objectives and Components first two conditions were met and the Bank kept only the last tranche, yet, 2. The project's main objective revised to require NHB to implement was to endorse the private sector specific proposals tc regulate HFIs approach to housing finance as and to promote resource mobilization carried out by HDFC. to encourage for the sector. Yet, the first dissemination of its experience, and mission following loan signing stated V that only "the minimum requirements offering training and consulting for the second tranche release were assistance (PAR. para. 12). satisfied". This Audit concurs with the PCR vhich states that: "In retrospect, ne.ther the timing of the tranche conditions not the actions 6. HDFC has succeeded in remaining specified ptovei to be effective in in control of a very rapid ensuring consistency of NEB's actions development. Its financial situation with the project'e aims." during the period FY87-92 has been Retroactive financing (US$25 million characteriztd by solid results due in or 10% of the loaa amount) back to large part to a prudent financial October 1987 (appraisal) although it policy. Growth by any measure has was not really necessary with been rapid. Revenues (made up for financial intermediaries such as HDFC about 95Z of irterest payments) which are permanently refinancing increased by 36.9% compounded p.a. past loans to offset the usual betwern YY86 and FY92 (PAR. para. mismat3h of maturities. Retroac4ive 15). To tap domestic savings, HDFC financing allowed to stretch the had to diversify its products. HDFC time-slice eligible for reimbursement launched a successful deposit ,.y six months and thus to increase mobilization drive. Trust bonds and the loan amount at negotiations (PAR, corporate bonds were issued. The para. 8). contractual savings scheme introduced in 1985 proved to be premature. In 4. It was required of HDFC that at 1989 NHB introduced a better scheme least 50% of the loan finance (PAR, para. 17). HDFC has been an borrowers below the median urban innovator in zany ways. It has income estimated at Rsl,700 (US$131) offered variants of mortgage loans to in 198. Since HDFC's borrowers with make the debt-servicing more income lower than the median affordable. The "Step-up Repayment accounted for 34.7% of loans in FY87, Facility" is a graduated payment the Bank zriterion (being equivalent loan; the "Telescopic Loan Plan" is to 15% of lending) did not put an initial 30-year loan where the pressure on HDFC to lend more to debt-servicing capacity of the lower-income groups. Ninety-two borrower is reviewed after 5 years. percent of the Bank disbursements in these products are, however, not FY91 went to loans granted to popular (less than 4% of outstanding borrowers below the median income loans). HDFC's most rewarding (PAR, para. 10). However, the PCR innovation has been the "Lines of overstates the number of housing Credit" which essentially use units financed with the Bank loan selected corporations to distribute (300,000). This Audit estimates loans to their employees and to that, in the best scenario (no more collect payments from them (about 13% than 50% allocated to lower income of loans are made this way) (PAR, borrowers), 164.500 units were para. 18). financed directly. This would still be more than twice the SAR target 7. The Bank projent retained the (PAR, para. 11). financial covenants which had been set by IFC since 1978 when it war, 5. Te:hnical ussistance was not an associated with HDFC: debt-equity important component of the project. ratio, interest coverage ratio, Actually HDFC has become a resource administrative costs, and provisions institution for HFIs in Asia, for bad loans. HDFC has been able to vi net these covenants without 1986). It cannot be emphasized difficulty (PAR, para. 19). EDFC is enough that HDFC's record on this key using additional ratios to monitor issue of housing finance is its performances capital adequacy, impressive by any standard (PAR. return on equity, average cost of pats. 24). debt, and average yield on loans. HDFC has been succesaful in avotding 9. Loan 2929-IN had an expected these two pitfalls of specialized average loan life of 10.23 years at financial intermediaries. Ensuring a the time of appraisal. The close to positive margin between the average expected disbursements kept it at portfolio yield and the average cost 10.16 years, which is a very good of funds has been a steady item of match with the averaZe duration of EDFC's strategy, even during the loans. Since the exchange risk was recent run of interest rates toward hedged, the 10-year figure will 20%. Ensuring a positive match remain iegardless of the evolution of between the average portfolio the ..e. Ia other words, the loan maturity and the average debt to Ok is a rare case where the Bank maturity is more difficult given that loan has indeed contributed to a few lenders offer the 20-year loan positive match of maturities (PAR, maturity needed to match of mortgage para. 25). The technique used t- loans as the Bunk loan does. The elim-nate the exchange risk was a spread between cost of funds and series of swaps. This removed the yield has, however, decreased below r..sk element in exchange for a flat two points (1.3 in FY92), which is a fee. set such, however, as not to mi-imum for an autonomous financial bring the -ffective intereSL rate intermediary (PAR, para. 21). above HDFC's overall rate on long- term borrowing (12.5%). The fact 8. HDFC takes third-party that HDFC has made a policy of guarantees on 80% of the loans to hedging the exchange risk on its supplement the cumbersome legal foreign borrowings has cont:ibuted in hurdles of registering mortgage a substantial way to its viability as collaterals (it takes up to 5 years a financial intermediary (PAR, para. to complete deed recording in some 26). states and the foreclosure laws are difficult to enforce). When payments 10. As HDFC was already an are lat., HDFC threatens to collect efficient organization, the project from the guarantors. The quality of did not address any organization collaterals is also maintained by issue beyond the decentraiization HDFC's policy of not generally effort to set up additional branches financing more than 802 of the where there is a concentration of appraised value of the dwelling demand for housing loans. HDFC has (actual average around 55%). Yet, been operating at a high level of there is a tendency for HDFC's staff efficiency as can be expected borrowers to over-extend themselves, from a private company. Low staff especially in high-price metropolis mobility is also a noted feature of such as Bombay (PAR, para. 23). The HDFC's performance (PAR. paras. 27- most important demonstration effect 28). expected of HDFC vas that high loan recovery can be achieved in the 11. HDFC's accounts have never been housing mortgage lending sector. qualified (PAR, para. 30). This was the first pi ority laid out in the initials Project Brief (June vii 12. The nature of the project is FIs, while good for the sector, had such that sustainability hinges on an adverse effect on HDFC because the collection of principal and several vere promoted by c-tercial interest payments. HDFC has a very banks which diverted funding good record on loan recovery. As a previously earmarked for HDFC to result, the project benefits are their nw subsidiaries (PMR par. likely to be sustainable (PAR. pa,a. 39). HDFC has been able to keep all 32). the financial perame:srs, which are often the cause of financial distress Setor_Polinr Change& mong financial intermediaries. under control. The interest spread is 13. The Bank questioned early on positive. The duration of funds is the advantage of creating another well matched. The exchange risk is central government financial not an issue. Arrears are institution. NHB is claiming that it insignificant. All these translated is not doing more regulations or. the into good profitability as reflected NFIl than the Resevve Bank of india by the increasing valuation given to (REI) is imposing on non-bank HDFC share by the stock exchange companies. Yet, NHB has strengthen-ad (PAR, para. 40). restrictive regulations and not removed recognized constraints. 15. This project confirmed, Several laws have beea for some time however, the Bank's inability to tie in process of being updated to meet macro reforms with micro the requirements of a growing improvements. By project completion, mortgage portfolio. No legislation the housing finance system was still had been passed at the time of the not integrated into the banking field Audit (PAR, paras. 33-34). The e7stem or so haphazardly (e.g., new idea of developing the housing Uls set up by selected commercial finance sector to a size justifying banks) as to underscore the progress the adjunction of a secondary still needed (PAR. para. 41). The mortgage market has been touted by Bank loan was additional instead of HDFC for several years. This will displacing domestic resources. With not be feasible as long as several hindsight. the Bank's insistence on prerequisites (alleviation of atamp having the exchange risk covered ex duties, simplification of ante proved correct, and explains in foreclosures, etc) are met (PAR, large part why HDFC has been able to para. 36). maintain sound financial ratios (PAR, pare. 44). Given the suc:ess of this Conclusions and Lessons pilot project, the absence of a follow-up loan was regretted by HDFC. 14. The Bank helped increase the scale of HDFC's intervantion. It is 16. froject Rating. This project to the credit of HDFC to have geared was successful in all respects and is the organization to handle the rated, both by the PCR and this additional resour-es. The expectation Audit, as highly satisfactory (PAR, of continued Bank assistance to HDFC para. 46). did not, however, materialize (PAR, para. 38). HDFC has shown a great adaptability tc the changing market conditions, which helped it focus on domestic retail resource mobilization. The multiplication of PERFORMANCE AUDIT REPORT INDIA HOUSING DEVELOPMENT FINANCE CORPORATION PROJECT (LOAN 2929-IN) I. BACKGROUND 1. At the time of appraisal in 1988, the formal capital market of India was largely ignoring the funding needs of housing investments.' Central Bank regulation obliged commercial banks under the directed credit approach to allocate funds to the sector, but the level of transfers was too low to have an impact (0.5% of their total credit advances in 1987). Over 90% of housing investment was financed by household savings and borrowing from the informal sector. The Housing Development Finance Corporation (HDFC) established since 1978 was for several years the only financial institution (furthermo:e private) dedicated to provi1ding mortgages. Despite being very successful in opening a market for formal housing mortgage finance, it still had a limited impact. The perception that the potential market corresponded to the low-income population was a major constraint to the development of an unsubsidized housing finance system. Another factor was GOI policy to limit the availability of formal sector housing finance (SAR, para. 1.01). 2. Past GOI housing strategies depended on direct investments by the public sector with funds mobilized under the directed credit system and recovery being a low priority. The 7th Plan, which gave priority to meeting the housing demand, called for two changes in the sector: (a) a refocusing of public resources on financing land development and shelter for the lowest-income groups, and (b) a much greater emphasis on private finance for housing. The latter involved a major effort to improve institutional finance for housing and a proposal for the creation of an apex institution, the National Housing Bank (NHB). 3. After twelve projects supporting public sector development of urban land, infrastructure and municipal services, this was the first housing finance project. While the shift had been made much earlier in smaller countries, for Bank operations in India, this project was a major departure from the tested strategy of financing specific investment projects, For all its novelty, the Bank chose an established intermediary which had been launched with IFC equity funds ten years earlier. 2/ At present, the housing finance system (HFS) is still not integrated into the financial system. Furthermore, the informal housing sector continues to be the main source of supply. 2 II. PROJECT OBJECTIVES. DESCRIPTION AND IMPLEMENTATION A. Obi*ctives 4. The project's main objective was to support the private sector approach to housing finance as carried out by HDFC. to encourage dissemination of its experience, and to improve sector incentives. As stated in the SAR. the specific objectives wores (a) support of HDFC's lending in a period of transition in the capital markets in order to extend the benefits of market-oriented housing finance to a wider geographical area and to a broader group of middle- and lower- income beneficiaries; (b) further institutional development ot HDFC's roles as an innovator and advocate of market-oriented housing finance, and as sector leader assisting the entry and development of similar institutions; and (c) development of a supportive regulatory framework to ensure the financial integrity of housing finance institutions and their capacity to mobilize resources at mprket rates (SAR, para. 4.01). The first two objectives were to be controlled through the loan made directly to HDFC and the thi rd one through trenching (PAR. para. 7). 5. The main component of the project was conceived as a time-slice of HDFC's lending over the period October 1987 (appraisal date) to March 1991 (end of FY91). US$248 million of the loan were to refinance 30% of all disbursements made by HDFC. with at least half earmarled to sub-borrowers with incomes below the median urban income and the other half without any conditionality. A small technical assistance component (US$2 million) was allocated to assist HDFC in furthering its decentralization, its computerization and its pursuit of innovation. B. Disbursements 6. Th6 Bank loan was disbursed slightly faster than projected in the SAR. Retroactive financing in the amount of US$25 million equivalent had been approved (PAR. para. 8). Table 1 shows a close parallelism between projected and actual disbursements, and the trenching of the Bank loan did not affect the actuals (PAR, para. 7). The comparison with the standard disbursement profile for the urban sector in India illustrates the resources transfer advantage of housing finance projects over the traditional sites-and-services projects: the loan was fully disbursed 32 months before the end of the grace period, and over 3 years earlier than a traditional urban project. This increased considerably the effective availability of Bank funds for HDFC (PAR, para. 25). It also demonstrated the absorptive capacy of the housing finance intermediary since the loan amount was increased from US$100 million prior to appraisal to US$250 million at negotiations. Net transfer to HDFC remained positive during the project implementation period, which is another advantage of housing finance loans. 3 7. Trenchint The concept of trerching was introduced prior to appraisal to give the Bank the means to enact the institutional changes in the sector which were under GOI's control. Initially, three trenches were designed: (first) initiation of an enabling regulatory environment fostering a financially-sound market-oriented approach: to be met by creating NEB; (second) legislation on NHB passed by Parliamentl (third) Bank review of NRB's specific proposals for action.' Within three months, the first two conditions were met and the Bank kept only the last tranche; yet, it revised the conditions to require from NHB the implementation of specific proposals to regulate RFIs and to promote resource mobilization for the eec.or. Bank management noted that Bank review of NHB proposal was not required anymore as a condition and objected. Bank staff responded that th4 appraisal mission had informally assessed - preliminary draft as satisfactory. Yet, the first mission following loan signing stated that only "the minimum requirements for the second tranche release were satisfied" and that "further steps are necessary if the intentions of the conditions are to be effectively carried out".' The distinction between conditions fulfilled in spirit vs in letter was thus made in the record. The following month, the second tranche was nonetheless released because the conditions had been "substantially fulfilled". As a result, the trenching did not affect disbursements to HDFC. This Audit concurs with the PCR which states that: "In retrospect, neither the timing of the tranche conditions no the actions specified proved to be effective in ensuring consistency of NHB's actions with the project's aims." (PCR, para. 17) .s Table 1: Bank Loan Disbursements (US$ million) 1988 1989 1990 1991 1992 1993 1994 1995 Projected in SAR 64.0 76.5 86.4 23.1 Actual 54.0 88.8 93.9 13.3' Urban profile7 2.0 18.0 35.0 40.0 50.0 40.0 40.0 20.0 8. Retroactive Financi ta The Bank agreed to allow retroactive financing (US$25 million or 10% of the loan amount) from October 1987. It was not really necessary. The rationale for choosing the appraisal rather than the more usual negotiation date is that a substantial share of past eligible expenditures would be left unfunded if the Bank did not include them in the project. This rationale 3/ Project Brief (September 2, 1987). 4/ Aide-Memoire (March 1990). 5/ The original english of this sentence is confusing. '/ In terms of HDFC's fiscal year. loan disbursements were Rs. million 1,250, 1,113 and 1,770 in FY8, 1990 and 1991, respectively. 7/ "Standard Disbursement Profiles", June 1992. 4 is solid in investment projects with shortage of counterpart funds or badly needed foreign exchange, it is less so with financial intermediaries which are permanently refinancing past loans to offset the usual mismatch of maturities. HDFC's good resource mobilization performance did not justify retroactive financing, except to acknowledge the fact that subprojects were readily eligible (PAR, para. 10). Retroactive financing allowed to stretch the time-slice eligible for reimbursement by six months and thus to increase the loan amount at negotiations. 9. SOecial Account Although HDFC was the Borrower of the Bank loan, it was decided that "it would not be in its best interest to operate the special account".' Fungible retroactive financing eventually served the same purpose. C. Beneficiary Targetin 10. Given the Bank focus on poverty alleviation, it was required from HDFC that at least 50Z of the loan finance bo.rowers below the median urban income estimated at Rsl,700 (US$131) in 1987. Since HDFC's borrowers with inc me lower than the median accounted for 34.7% of loans in FY87, the Bank criterion (being equivalent to 15% of HDFC's lending) did not motivate IDFC to lend more to lower- income groups. Requests for reimbursements by the Bank were supposed to confirm the meeting of the criterion and the corresponding documentation was supposed to be audited (PAR, para. 29). The PCR stated that the target was met at over 30% (PCR, para. 9). Table 2 shows that to be correct every year with respect to the numbers of borrowers, but only two years out of three in terms of loan volumes. Still, it could be considered that 92% of the Bank disbursements in FY91 went to loans granted to borrowers below the median income. Table 2: Targetini of Bank Loan FY 1987 1988 1989 1990 1991 Urban median income (Rs/month) 1700 1860 2000 2110 2450 Z of borrowers below median inc. 34.7 41.6 34.8 37.1 42.4 % of loans to same borrowers 22.6 25.8 26.7 24.3 29.5 Bank target (Re million) - - 625 556 885 Actual approvals by HDFC 424 465 578 969 1623 11. The PCR overstates the number of housing units financed with the Bank loan. From 70,000 in the SAR (p. iii), it is claimed to have reached 300,000 units directly financed (PCR, para. 19).' During the three fiscal years concerned by the above disbursements (1989-1991), total approvals reached 256,075 (PCR, p. 18). Given the need to disburse half of the loan to borrowers below the median income, this Audit estimates that, in the best scenario (no more than 50% allocated to lower income borrowers), 164,500 units were financeu. directly. This 6/ Internal memorandum (February 12, 1988). No verifiable computation is provided. 5 would still be more than twice the SAR target. The median household income of HDFC's borrowers increased from Rs2,300 in FY87 to Rs2,800 in FY91, which is a slower slippage (21.7%) than all India's median urban income (35.3% from Rs1,700 to Rs2,300). Yet, the distribution of loan amounts grew more in favor of higher- income borrowers because the loan ceiling was increased fivefold to RsO.5 million (Table 3). Well after project completion, HDFC is still meeting with a wide margin the allocation criterion set in the Bank project. 12. Technical Assistance. Training and Studies Training abroad and foreign technical assistance was not an important component of the project given the stage of development of HDFC. It did not use foreign consultants as allowed in the component. Instead of purchasing off-the-shelf softwares. HDFC staff developed its own custom software thus gaining in-house capability to keep it tailored to HDFC's operational needs. HDFC has established its own training facility wherr all staff re sent on a rotating basis. Recently it started to offer courses to staff of foreign housing finance institutions with success. It is another indication of HDFC's maturity that it has reached a stage where it can provide training and consultancy abroad. Table 3: Income Distribution of HDFC Loans Rs/month <1000 1000-2000 2000-3000 >3000 in units terms (in % of units approved) September 1988 10.2 32.4 25.5 31.9 March 1992 1.0 6.0 20.0 38.8 in loans terms (in % of "individual loans" approved) September 1988 5.4 22.4 23.8 48.5 March 1992 1.0 3.0 15.0 57.0 13. Procurement Being a first loan to the housing finance sector, procurement followed the pattern established for Bank loans to HFIs made since the early 1980s which essentially adopt the intermediary's procedures. Direct project procurement was used only for the purchase of computers. No problems were reported. 14. Economic Rate of Return The SAR estimated the ERR at 20-30% based on imputed rental values and sales data. Given the pent-up demand for mortgage loans, these rates were expected to remain high for a long period (SAR, para. 5.03). The PCR did not provide a recomputed ERR. This Audit did not attempt either. Given that HDFC borrowers have been able to afford the debt-servicing, the ERR is, however, estimated to be a multiple of the average yield on the loan portfolio (about 14% for the last several years). 6 III. HOUSING DEVELOPMENT FINANCE CORPORATION PERFORMANCE A. Financial Policy 15. HDFC has succeeded in remaining in control of a very rapid development. Its financial situation during the period FY87-9210 has been characterized by solid results due in large part to a prudent financia. policy. Growth by any measure has been rapid. Revenues (made up for about 95% of interest payments) increased by 36.9% compounded p.a. between FY86 and FY92. Profits followed at the same pace (36.8% p.s.), thus showing that growth was not obtained at the expense of margins. The equity base grew mostly out of retained earnings (43.8% p.a.) while the share capital was increased from RS 100 million to 200 in FY87 and again to Re 450 million in FY90. Total debt increased at the same pace as housing loans (33.4% and 34.5%, respectively) which is sound intermediation practice. The Bank ias been the largest single lender (17% of total at the end of FY92) despite granting one loan. Term deposits have been growing at a much slower pace than other borrowed funds (20.6% p.a.) but they still account for the largest share (36%) of total debt. Given the shorter maturity of deposits (actually offset by good stability), this trend is not adverse to the soundness of HDFC's financing plan. It is no coincidence that deposits have financed a decreasing share of housing loans until FY92 when HDFC was forced to launch a deposit collection drive to offset changing market conditions (Table 4): loans from domestic commercial banks, US-AID and the World Bank have levelled off since FY91. IFC (HDFC's initial shareholder) and the National Housing Bank have stepped in, but in limited ways. It is, therefore, significant of HDFC's ability to capitalize on its reputation in a crowded domestic financial market to have raised new debt in FY92 mainly through bond issues (20.6%) and new deposits (47.7%). Table 4: Debt Financing of Housing Loans Fiscal year 1987 1988 1989 1990 1091 1992 debt/housing loans (%) 118.1 114.1 117.1 106.5 107.5 113.5 deposits/same (%) 70.7 64.8 57.2 43.3 34.9 40.9 16. HDFC re.;ource mobilization strategy changed in 1991 due to the downgrading of India's credit rating. No foreign borrowing could be mobilized (the US$60 million syndicated tranche of the last IFC loan was postponed). To tap domestic savings, HDFC had to diversify its funding instruments. HDFC launched a successful deposit mobilization drive with the assistance of a network of brokers paid on commissions. This network is currently bringing Rs600 million per month. The marginal cost has increased for deposits to remain attractive. Although no deposit is remunerated above 15%, the final cost is 17% after brokerage fees. 'o/ While the period covered by loan disbursement is FY89-91, this Audit assesses HDFC's performance over the last six audited years. 7 advertising, and servicing costs. Trust bonds were launched three years ago when HDFC became eligible to receive trusts' investments. They can be subscr!bed for 1-48 months at 13-14.5% interest rate. Corporate bonds have been issued for 10 to 30-year maturities with a buy-back at-par facility to create a resale market for them. Refinancing from NHB has been tapped since made available in FY91 given that its costs of 12.8% is substantially lower than HDFC average cost of funds. 17. HDFC established a contractual savings scheme in 1985 after m"rket surveys concluded that there were savings to be tapped in this form, especially among the middle-income groups. The German model was adopted with some relaxation since HDFC did not insist on regular savings which proved to be a costly loophole because some depositors could claim eligibility for a loan by transferring fundL only at the end of the saving period. Savings were remunerated at one point above deposits with commercial banks (then 5%), and the lending rate was set at 2.5 percentage points above or 8.5% whizh represented a steep discount over the normal mortgage loans at 12.5%. The scheme did not succeed as expected despite an improvement of the loan-to-savings ratio from 1.5 initially to 2.37. As one of the concept's premises (i.e., new savers finance the loans taken by the original savers) did not materialize, HDFC was left with funding the loan commitments with other funds. The scheme was terminated early 1992. The experience proved to be only premature, not wrong for India. In 1989, NHB nad introduced a better scheme called "Home Loan Accounts" which took savings at 10% for 5 years, and granted loans up 4 times the accumulated savings at 10.5-14.5%. Like the other HFIs, HDFC is authorized to keep in its balance sheet the deposits collected under this new contractual savings scheme in exchange for funding out of the same funds the future requests for loans. Only when the original deposits would be exhausted, HDFC could then request NHB to provide complementary financing. 18. HDFC has been an innovator in many ways. It has offered variants of mortgage loans to make the debt-servicing more affordab3. The "Step-up Repayment Facility" is a graduated payment loan; the "Telescopic Loan Plan" Is an initial 30-year loan where the debt-servicing capacity of the borrower is reviewed after 5 years to reduce maturity without change in the interest rate. Although advantageous, these products are not popular (less than 4% of outstanding loans in the Madras branch). It is due in part to the newness and the perceived complexity of the products, which active promotion by HDFC could overcome. It is also that HDFC borrowers have a high propensity to prepay their loans. A study based on HDFC loans granted in the initial years showed a 37%- prepayment level. The reasons are: the Indian dislike for debt, the tax incentive, and the use of refunds given every 5 years on life insurance policies. HDFC's most effective innovation has been the "Lines of Credit" which essentially use selected corporations to distribute loans to their employees and to collect payments from them. From FY87-91 48,200 corporate loans were extended (13.4% of total). 8 Table 5: Financial Covenant. Fiscal year Covenant 1987 1988 1989 1990 1991 1992 Debt,'equity <17.5 13.3 14.1 15.4 15.3 12.1 12.9 Interest coverage >1.20 1.26 1.26 1.25 1.27 1.24 1.26 Administrative costs/total loans outstanding <1.5% 1.11 .93 .90 .76 .73 .74 Arrears/loans <0.4% .01 .01 .01 .04 .06 n.a.1 Provisions/loans >0.4% .68 .90 .89 1.01 1.14 1.21 B. Financial Covenants 19. The Bank project retained the financial covenants which had been set by IFC since 1978 when it was associated with HDFC. They are: (1) the debt-equity ratio designed to limit the leverage derived from a relatively small capital base, (2) the interest coverage to ensure a positive spread between average yield and cost of funds, (3) the administrative cost in percentage of the loan portfolio to maximize the productivity of the institution, and (4) the provisions for bad loa.s also in percentage of outstanding loans to self-insure against irrecoverable loans. HDFC has been able to met these covenants without difficulty or creative acccunting as it is often the case (Table 5). One of the covenants is redundant because the local practice imposes a more stringent ceiling. The debt-equity ratio, while allowed to go up to 17.5:1 with the Bank Group, is limited to 15:1. 20. HDFC is using additional ratios to monitor its performance. In its bid to be considered a full-fledged financial institution, HDFC is also monitoring its capital adequacy (7.3% by mid-1992 on the basis of 100% risk'weighted assets) against the international "Cook Ratio" of 8%, which implies lowering the leverage further.12 The doubling of the share capital carried out at the end of 1992 has brought HDFC above 8%. Additional ratios are: (1) the return on equity (ranging from 25-33% since 1986) because HDFC is a private corporation whose shares are traded on the Bombay stock exchange, (2) the average cost of debt outstanding, and (3) the average yield generated by the loan portfolio. 21. Although SAR projection assumed that the spread between cost of funds and yield would remain above 2.2 percentage points, it has decreased below two points, which is a minimum for an autonomous financial intermediary (Table 6). Only because its administrative costs are kept low, is HDFC able to remain profitable with a reduced margin. It is the increased cost of funds which forced HDFC for the first time in 1992 to pause in its rapid lending growth because of the affordability issue. HDFC made a conscious decision to reduce lending to maintain portfolio quality since there is a correlation between higher interest rates and greater arrears. The same trend also explains why HDFC has been 1/ HDFC disclosed only arrears older than six months iin its FY92 Annual Report. n/ The capital adequacy or Cook ratio requires from financial institutions an equity base equal to at least 8% of their liabilities. 9 reducing its leverage. Beyond these, measures to restore the spread are limited becanse the leneing rates are imposed by NHB if its refinancing is used. Table 6: Yield, Cost of Funds and Spread Fiscal year 1987 1988 1989 1990 1991 1992 Average cost of debt (%) 11.6 12.0 12.5 12 7 12.7 13.3 Average yield of loans (Z)14.1 14.0 14.1 14.1 14.1 14.6 Spread (1 points) 2.5 2.0 1.6 1.4 1.4 1.3 22. The interest risk and the mismatch of tuaturities are two perils which have undone many financial institutions; a notable example in the housing sector is the US savings-and-loans system. The interest risk was flagged during project preparation, but deemed "limited" at a time when inflation rates were single- digit.1 HDFC has been successful in avoiding these two pitfalls of specialized financial intermediaries. Ensuring a positive margin between the average portfolio yield and the average cost of funds has been a stady item of HDFC's strategy, even during the recent run of interest rates toward 20%. Ensuring a positive match between the average portfolio maturity and the average debt maturity is more difficult given that few lenders offer the 20-year loan maturity needed to match mortgage loans as the Bank does. The stability of deposits has been a positive factor as well as the long availability of the Bank loan (PAR, para. 25). C. Portfolio Quality 23. HDFC takes third-party guarantees on 80% of the loans to bypass the cumbersome legal hurdles of registering mortgage collaterals (it takes up to 5 years to complete deed recording in some states and the foreclosure laws are difficult to enforce). When payments are late, HDFC threatens to collect from the guarantors. The pressure is generally sufficient as the impressive recovery record attests. While these guarantees seem superfluous given the overall good debt-servicing record, the practice has contributed to educate mortgage borrowers that loans need to be repaid. The quality of collaterals is also maintained by HDFC's policy of not generally financing more than 80% of the appraised value of the dwelling." This Audit found in a sample of 60 recently-approved loans in Bombay that this ceiling is respected (3 cases over 80% and an average of 55%). Yet, there is a tendency for HDFC's borrowers to over-extend themselves, especially in high-price areas such as Bombay. The already-mentioned sample showed that borrowers purchased housing equal to 4.76 years of income on average, but that many were well above this already-high figure (25% were above 6 years, 10% above 8 years and 2% above 10 years). Even if the current recovery record is good, these figures represent potentially greater risks. This Audit confirmed /3 Issues Paper (January 9. 1986). /I Letter from HDFC (January 14, 1988). 10 that in Madra. for example, the average price-to-income ratio has crept up from around 5 times to 6-7 in about ten years. 24. Loan Recover The most important demonstration effect expected of HDFC was that high loan recovery can be achieved in the housing mortgage lending sector. This was the first priority laid out in the Initial Project Brief (June 1986). It cannot be emphasized enough that HDFC's record on this key issue of housing finance is impressive by any standard. It compares very well with the delinquency ratio observed in the U.S.. for example, where arrears of over three months reached 0.88Z in September 1992. To facilitate payments. HDFC has opened accounts in 200 commercial banks spread in the country. Thirty five percent of payments are received in this fashion. Every borrower over two months in arrears is visited to ascertain the situation and work out affordable debt reduction arrangements. As a result, HDFC has had no need to change its policy of maintaining a loan-loss reserve equal to about 1% of outstanding loans, which is two and a-half time the level required by covenant. D. Bank Loan's Maturity, Grace Period and Exchange Risk 25. Average Loan Life The Bank Loan was granted directly to HDFC for 20 years of maturity, including 5 years and 6 months of grace, in order to help match the long maturities of mortgage loans. From HDFC's viewpoint, the availability over time of borrowed funds can be measured by an indicator called the Average Loan Life (ALL). It is defined as: the ratio of (a) the sum, until maturity, of the loan balances (in US dollars) outstanding at the end of each year over (b) the loan amount not of cancellations. This indicator measures the number of years during which the entire loan proceeds stay effectively at the borrower's disposal assuming up-front disbursement. The faster the disbursements and the slower the repayments, the longer the availability of loan funds. In the case of Loan 2929- IN, it had an expected average loan life of 10.23 years at the time of appraisal. The close to expected disbursements kept it at 10.16 years, which is a very good match with the average duration of loans. Since the exchange risk was hedged (PAR, para. 26). the 10-year figure will remain regardless of the evolution of the Rupee. In other words, the loan to HDFC is a rare case where the Bank loan has indeed contributed to a positive match of maturities. 26. Exchanie Risk Bank staff have been pro-active on the issue of the exchange risk since project yreparation, requesting "appropriate allowance" for that risk as early as 1986. Yet the Bank estimated later that "HDFC can bear the risk" on the proposed Bank loan while recommending its coverage with GOI assistance."' This meant removing the risk element in exchange for a flat fee, set such, however, as not to bring the effective interest rate above HDFC's overall rate on long-term borrowing. The technique used to eliminate the exchange risk was a series of swaps. While it was originally aimed to swap with other Indian corporations, the magnitude (US$75 million for the first tranche) obliged HDFC to deal exclusively with GOI. This explains why the accounting presentation is confusing. Schedule 14, Note 10 to the FY92 Annual Report /! Letter to the Minister of Finance (July 8, 1986). /, Internal memorandum (January 21, 1988). 11 discloses the interest payments to IBRD and IFC under the label "expenditure in foreign currency". Although HDFC does buy the foreign currencies needed to insure the debt-servicing, it is done on behalf of the swappers (GOI for the Bank loa, the Industrial Finance Corporation of India for the IFC loan). The resulting cost of Bank funds for HDFC was a fixed 12.5%, which was higher than the then-cost of funds, but proved to be a favorable deal since the devaluation of the Rupee since 1991. The fact that HDFC has made a policy of hedging the exchange risk on its foreign borrowings has contributed in a substanteal way to its viability as a financial intermediary (PAR., para. 29). E. Organisation 27. Staffina HDFC has been operating at a hieh level of staff efficiency as can be expected from a private company. Low staff turnover is also a noted feature of EDFC's performance. 28. Organization Chanaes As HDFC was already an efficient organization, the project did not address any organization issue beyond the decentralization of fort to set up additional branches where there is a concentration of demand for housing loans. Ten branches were opened during the project period, bringing the total to 26. They are given large autonomy: they approve all loans to individuals and up to a set amount loans to corporations (for on-lending to their employees). With the check of internal auditing decentralization is considered a good experience. The organization chart has not changed substantially since appraisal, except for the addition of a deputy managing director and the regrouping of the ten additional branches under regions each overseen by a functional director. The introduction of computers and customized software under the technical assistance component of the project contributed to productivity gains and institution efficiency. F. Accounting and External Audit 29. HDFC is being audited by independent auditors acceptable to the Bank. They issued at Bank request separate reports on Statements of Expenses (SOE), and there was no problems with this disbursement procedure. The audit of beneficiaries is part of the normal audit process (through random samples taken at each branch), but only as far as their loan balances vis-A-vis HDFC, not for the affordability of the debt burden. To that extent, the intent stated in the SAR (para. 4.04) of having the external auditors review the documentation kept by HDFC to ensure that at least half of the Loan finance lower-income groups was not fulfilled. However. HDFC met the target (PAR, para. 10). 30. HDFC is operating under the Company Act which says what financial data ought to be disclosed and the Accounting Board set the specific guidelines. Some improvements are possible although HDFC's accounts have never been qualified. For example, the lending comitments accumulated outside balance sheet such a. those linked to the new contractual savings scheme promoto-d NHB are not disclosed because it is not atatutory. They need to be known by the main creditors of HDFC. Similarly the funding requirements of the approved but yet undisbursed loans weighs heavily onto the medium-term financial policy of HDFC and thus ought to be also readily available in the published accounts. Another example deals with the exchange risk. Although HDFC does not bear any, an item 12 in the income statement labelled "Expenditure in foreign currency" might be confusing. Actually HDFC buys the foreign currencies needed to service the Bank loan, for example, and then debits the Central Government's Stabilization Fund of the difference between the nominal interest rate (about 7.7%) and 12.5% (the swap interest rate). The somewhat misleading label should be either changed or clarified in a footnote to the audited accounts. The cost of swapping the foreign-currency resources to avoid the exchange risk is substantial (although less than the annualized exchange losses thus Fsr); it should be shown on a separate line in the income statement. 31. Environment was not a prominent issue in this project. Over 90% of HDFC lending vent to urban areas and 75% to finance apartments. These are modern dwellings, all linked to electricity, water and sewerage networks. G. Sustainability 32. The nature of the project is such that suatainability hinges on the collection if principal and interest payments. HDFC has a very good record on loan recovery. As a result, the project benqiits are likely to be sustainable. IV. SECTOR POLICY CHANGES A. Role of NHB 33. The Bank questioned early on the advantage of creating another central- government financial institution.1 34. NHB is claiming that it is not imposing more regulations on the HFIs than the Reserve Bank of India (RBI) is imposing on non-bank companies. Indeed it promoted the establishment of additional HFIs by providing "refinancing" of selected mortgage loans initiated by them. NHB has become de facto the only large-scale financier of mortgages to low-incoce groups because HFIs without access to low-interest deposits (such as the subsidiaries of coumercial banks) cannot afford lending below the current market cost-of-funds. This refinancing facility is a form of subsidy because, while limited to loan- that finance houses of a certain price or size, it can reach large multiples (up to 5 times the HFI's networth and up to 3 times deposits). NHB relaxed its refinancing criteria in March 1990 by allowing loans for housing larger than 40m2 provided the cost was below Rsl50,000. This easing was partly offset by halving the spread left after refinancing on loans above Rs50,000 (to 1% instead of 2%). In FY90 and FY91, HDFC's borrowing from NHB went from 0 to Rs2,781 million, or exactly 37% HDFC's additions to debt these two years. 35. Several laws have been for some time in the process of being updated to meet the requirements of a growing mortgage portfolio. The industry has been "/ First Issues Paper (January 9, 1986). 13 lobbying since its beginning for a relaxation of the stamp duty and of the Foreclosure Law. Both make the transfer or repossession of mortgages and properties more costly. There is a proposal -o MOF to allow more rapid foreclosure (which is said to take up to ten years) through an amendment to the NHB Act rather than a revision of the original law. The stamp duty is adding so much (up to 131 in one state) that HFIs are not recording the deeds, but instead holding them in safe deposit boxes. They are hindering HDFC's intent to launch a secondary mortgage market based on a free-flow of mortgage-backed securities (PAR, para. 36). The Bank noted in December 1991 that NHB had prepared draft legislation to simplify mortgage registration and foreclosure, but one year later (field Audit mission) it was still not passed despite the need created by the increased number of HFIs whose lending is growing rapidly. A committee set up by RBI to look at the non-banking sector recommended recently several measures to ease other widely acknowledged constraints, inter alia by reducing the minimum deposit period from two years to one year and offering tax parity with commercial banks on deposits. NHB pointed to some advantages enjoyed by HFIs such as the 101 liquidity requirement compared to 15% for other financial institutions. Yet, the Bank noted in January 1991 that NHB had strengthened restrictive regulations, including imposing a new structure of interest rates. 36. NHB is an agency of GOI. Its financing plan calls for substantial support from the Government. The central bank provided equity (16%) and soft loans (11%); GOI provided guarantee on debt (16%) and channeled the tax-deferring of capital gains in the business community through NHB-issued bonds (10%), thus bringing total GOI support to 53% of total resources mobilized. The insurance monopoly (Life Insurance Company) contributed 25%, foreign aid agencies (6%), and various schemes and retained earnings the remainder. The average duration of NHB borrowings is 12 years, their average cost is 12.5%. With the average return on loans at 13.7%, NHB has a small but positive spread. NHB does not compete with the HFIs for deposits. The Home Loan Account scheme (PAR, para. 17) is only a contra account (i.e., outside NHB's balance sheet) because funds mobilized under this contractual savings scheme launched by NHB in 1990 are left with the HFIs. The multiplication of HFIs on the HDFC model will, however, create additional ,mand for refinancing which would stretch NHB's capability. B. Secondary Mortgage Market 37. The idea of developing the housing finance sector to a size justifying the adjunction of a secondary mortgage market has been touted by HDFC fir several years. It is the corner stone of every fully-matured mortgage markets such as in the U.S.A. or France. While understanding that it can be a goal for HDFC's founders, it remains that securitizing housing portfolios in India would require several prerequisites to be met for the market to serve its purpose and which are yet to be met (PAR, para. 34). The practice of limiting the loan-to-value ratio at 80% and actually recording a 55% average (PAR, para. 23) is a positive step toward the necessary standardization of mortgages for suitable refinancing. 14 IV. COVCLUSIONS AND LESSONS 38. A distinction must be made between the tasks which were to be carr-.ed out by the borrover (HDFC) And those by the guarantor (GOI). Given the legal set-up (e.g.. loan made directiy to a private corporation), the Bank did not have the same leverage to achieve the sector-wide objectives as under traditional urban projects. This only confirms what ought to be an hypothesis in any risk analysis at appraisal time, that is, sector changes are best achieved if the Bank lcan is granted to the government, hence, if the dialogue is with the agencies concerned. A. DEC Performance 39. The project met its main objective of demonstrating that housing finance is a viable alternative in India. While it can be argued that HDFC had already proven the point in the early 1980s, it was more of a pilot phase given the low volume of loans involved. The Bark helped increase the scale of HDFC's intervention. In the process, however, a "bubble effect" was created because HDFC h 4 to absorb much larger resources in a short time while the lack of a follow-up loan left HDFC with the challenging task of finding alternative resources of same magnitude. It is significant that the issue of HDFC's absorption capacity was raised when a loan of US$100 million was considered.'8 When the amount was eventually increased .o US$250 million, the issue did not appear in the risk assessment in the SAR. Ihere was indeed no problem for HDFC to onlend what represented then 64% of its outstanding portfolio. It is to the credit of RDFC to have geared the organization to handle the additional resources. The expectation of continued Bank assistance to HDFC did not, however, materialize. 40. HDFC has shown a great adaptability to the changing market conditions. Following the downgrading of India's credit rating. HDFC changed its funding licy and focused on domestic retail resource mobilization. The Bank always ialt that it was HDFC's biggest challenge to sustain future growth." "Deteriorating prospects for resource mobilization" casted doubts on the sustainability of growth.2 The multiplication of HFIe, while good for the sector, had an adverse effect on HDFC because several were promoted by commercial banks which diverted to their new subsidiaries funding previously earmarked for HDFC. As a result, term loans from banks and insurance companies, which accounted for over 50% of resources in 1987, contribute now an insignificant sharb of HDFC's sources of funds. 41. HDFC has been able to keep all the financial parameters, which often the cause of financial distress among financial intermediaries, under control. The interest spread is positive. The duration of funds is well matched. The 1/ Initial Project Brief (June 24, 1986). 19/ Aide-memoire (March 1990). 20/ Supervision report (January 23, 1991). 15 exchange risk is not an issue. Arrears are insignificant. All these translated into good profitability as reflected by the increasing valuation given to HDFC shares by the stock exchange. B. Bank Performance 42. This project confirmed, however, the Bank's inability to tie macro reforms with micro improvements. By project completion, the housing finance system was still not integrated into the banking system or so haphazardly (e.g., new HFIs set up by selected commercial banks) as to underscore the progress still needed. Although the project's achievements should not be discounted, the sector impact was disappointing. The Bank believed that the project offered an opportunity for influencing the direction of NHB policies. This did not materialize in a significan* way. To this extent, the Bank performance is rated mixed. 43. During project preparation, it was stressed that the Bank loan should not displace domestic resource mobilization." This objective, although ambitious given the loan amount, was met overall. While the Bank disbursed Rs 4,133 million, HDFC raised Ra 6,372 million from deposits, bonds and NHB. Yet the PCR acknowledges that the loan increase at negotiations and the retroactive financing "allowed HDFC to choose to replace deposit mobilization by the foreign funds to a substantial extent" (PCR, para. 14). 44. A premise of the Bank assistance was that high interest rates "especially discourage long-term borreing for durable goods such as housing"" The recent years have demonstrated this to be incorrect in the case of India. While HDFC has slowed down its psce of growth, the new HFIs have stepped in. The Bank project, which rewarded HDFC with visibility, in turn encouraged the entry of other financial institutions (commercial banks and insurance companies) into the sector. 45. The dank's insistence on having the exchange risk covered ex ante was correct, and explains in large part why HDFC has been able to maintain sound financial ratios. The Bank position was not based on hindsight only given the reasonable expectation of depreciation of the Rupee vis-a-vis the hard currencies made available by the foreign lenders. When the foreign funds are being recovered only in local currency, the exchange risk associated with them should be hedged through financial techniques (when available and mastered) or insured with the Guarantor's assistance (against preset flat fees to be passed on). 46. Given the success of this pilot project, the absence of a follow-up ioan to HDFC and to the new HFIs is an anomaly in view of the Bank's tenets of country's assistance. All the required ingredients for effective development lending are met: quick disbursing loan, acceptable funds targeting below the median urban income, above average cost recovery, viability and sustainability of the financial intermediary. The inability of achieving needed regulatory changes highlighted the limitation of the direct loan format. 21/ Letter to the Ministry of Finance (July 8, 1986). 22/ Housing Finance Sector Review (December 22, 1986). 16 C. Project Ratina 47. This project was successful in all respects. It confirms the viability of a private specialized housing finance institution to help address the permanent issue of providing housing to the rapidly increasing urban population. The Loan was disbursed on schedule. By its nature, the project had no cost overrun. The institution-building targets were met and the loan recovery ratio was very high. This Audit, therefore, concurs with the PCR-based assessment of a highly satisfactory project with substantial institutional achievements and a high likelihood for sustained benefits. 48. Main Leesone To pick a private comparv as the Borrower increases the chance of success given the many other market-oriented reasons company management has to survive and to succeed. The drawback of this choice is likely a low leverage on the Guarantor to enact far reaching reforms of the legal and regulatory framework. Specialized financial intermediation is a worthwhile objective for a government to encourage and for the Bank to support if prerequisites (such as sector size and appropriate financial approach) are met. Positive spread, positive match of maturities and highest possible loan recovery are the ingredients of success in fInancial i.termediation such as housing finance. 17 Page I of 5 CObMNTS FROM THE BORROWER Coqqnts to the draft Performance Audit Report concerning the Housint Finance Proiect 2929-IN ANNEXURE I THE WORLD BANK'S FUTURE HOUSING STRATEGY FOR THE 1990S IS BEING RECAST IN TERMS OF PROGRAMMES NOT PROJECTS? MACRO NOT MICRO? ENABLING NOT INTERVENTIONIST IN THE HOUSING SECTOR WITH FOCUS OF ATTENTION ON THE EFFICIENCY OF HOUSING MARKETS (BOTH DEMAND AND SUPPLY), CAREFUL SELECTION OF INSTITUTIONS AND INNOVATIVE INSTRUMENTS FOR RESOURCE MOBIL-SATION AS WELL AS MORTGAGE LENDING. THE EVOLUTION OF THE INDIAN HOUSING FINANCE SYSTEM IS AT A STAGE WHERE CRITICAL CONCEPTUAL AS WELL AS STRTEGIC INPUOPIY REQUIRED PRECISELY IN THIS DIRECTION AT THE PRESENT TIME. THE INSTITUTIONAL BASE IS IN PLACE AND THE FINANCIAL SYSTEM IS PRESENTLY IN TRANSITION FROM CONTROLS TO MARKET PRICING? THE LEGAL ISSUES NEED TO BE URGENTLY ADDRESSED. 8 NPLUDING THOSE RELATING TO FORECLOSURE AND DEBT SECURITISATION. IT WOULD BE A PITY IF THE FIRST HOUSING FINANCE LOAN TO INDIA WERE NOT TO BE PURSUED VIGOROUSLY, BUILDING ON THE STRENGTHS OF ITS PERFORMANCE AND CORRECTING ITS WEAKNESSES IN THE NEAR FUTURE. MOST OF THE HOUSING FINANCE INSTITUlTIONS BESIDES HDFC HAVE BEEN PROMOTED BY MAJOR COMMERCIAL BANKS OR INSURANCE COMPANIES. THESE INSTITUTIONS HAVE LARGE PUBLIC INSTITUTIONS BACKING THE ACTIVITIES OF THEIR PROMOTEE COMPANIES, TO THE EXCLUSION OF THE INDEPENDENT HOUSING FINANCE INSTITUTIONS. AS COMMERCIAL BANKS AND THE LIFE INSURANCE CORPORATION PROMOTED THEIR OWN INSTITUTIONS. THEIR LENDING TO OTHER HFIS ESPECIALLY HDFC TENDED TO STOP IN FAVOUR OF GREATER SUPPORT TO THEIR OWN INSTITUTIONS. WITH THE EMERGENCE OF NHB, ATTENTION FOCUSSED ON REFINANCE AND THE CREATION OF DETAILED GUIDELINES FOR THE EMERGING HOUSING FINANCE INDUSTRY. AS A RESULT, NHB PLAYED AN INT-ERVENTIONIST ROLE IN THE SYSTEM RATHER THAN AN ENABLING ONE. REFINANCE HAS CREATED DEPENDENCE THUS FORCING THE SMALLER HFIS TO RELY TO AN UNHEALTHY EXTENT ON NHB FUNDING. AS THE DEMAND FOR REFINANCE IS GROWING, NHB HAS TO LOOK FOR ALTERNATIVE FUNDING SOURCES, SPECIALLY AS DIRECTED CREDIT IS EXPECTED TO REDUCE SUBSTANTIALLY IF NOT ALTOGETHER. ANNEX 18 Page 2 of 5 THE KEY REQUIREMENT OF THE GROWING HOUSING FINANCE SYSTEM HAD NOT BEEN MET: ITS RESOURCE BASE. TODAY HOUSING FINANCE INSTITUTIONS ARE REQUIRED TO COMPETE WITH THE BANKS AND MUTUAL FUNDS FOR RETAIL RESOURCE MOBILISATION. ALTHOUGH THERE HAVE BEEN INDICATIONS OF A MOVEMENT TOWARDS A LEVEL PLAYING FIELD, IN REALITY THE SITUATION IS VERY DIFFERENT. FOR EXAMPLE HOUSING FINANCE COMPANIES ARE REQUIRED TO DEDUCT TAX AT SOURCE FROM INTEREST PAYMENTS IN EXCESS OF RS 2500 PER ANNUM. THIS PROVISION DOES NOT APPLY TO BANKS AND MUTUAL FUNDS. FURTHER HOUSING FINANCE INSTITUTIONS CANNOT ACCEPT DEPOSITS FOR PERIODS BELOW 24 MONTHS. IF HOUSING FINANCE IS TO INTEGRATE WITH THE FINANCIAL SECTOR SUCH ANOMALIES SHOULD BE REMOVED. HDFC CONTINUES TO DIVERSIFY ITS RESOURCE BASE. IN EFFECT BUILDING A MULTIFACETED RESOURCE MOBILISATION STRATEGY. NHB IS NOW IN THE PROCESS OF DETERMINING ITS OWN STRATEGY FOR LINKING ITS RESOURCE RAISING TO THE FINANCIAL SYSTEM. HUDCO RELIES ON GUARANTEED DEBT FOR RESOURCE MOBILISATION AS WELL AS FOR CREDIT DISSEMINATION. WHILE THIS IS A SIMPLISTIC ANALYSIS IT DOES HIGHLIGHT THE ESSENTIAL FEATURES OF THE SYSTEM. THE POTENTIAL DANGER OF A FUTURE STRATEGY COULD BE CUTTING OFF DIRECT ACCESS OF HFI'S TO THE FINANCIAL SYSTEM AND THE HOUSEHOLDS AND CREATING ANOTHER APEX SYSTEM OF DIRECTED CREDIT. POSITIONING INSTITUTIONS, LOOKING AT THEIR MAIN OBJECTIVES AND THEN DETERMINING MARKET ORIENTED RESOURCE MOBILISATION STRATEGIES (WHICH WOULD AUTOMATICALLY ENCOURAGE MARKET PRICING) WOULD SEEM TO BE THE DIRECTION IN WHICH THE SYSTEM SHOULD MOVE. A POTENTIAL FUTURE STRATEGY OF HOUSING FINANCE IN INDIA ONE OF HDFC'S MOST IMPORTANT CONTRIBUTIONS TO THE SECTOR HAS BEEN ITS DEMONSTRATION EFFECT, THAT GIVEN SOUND MANAGEMENT AND ORGANISATIONAL STRUCTURING, EVEN MAJOR ENVIRONMENTAL CONSTRAINTS CANNOT HINDER THE DEVELOPMENT OF A SOUND HOUSING FINANCE SYSTEM. IN FACT IT TENDS TO OBVIATE THEM. IT HAS ENCOURAGED OTHERS TO ENTER THE MARKET WHICH HAS GRADUALLY DEVELOPED AN EXTENDED INSTITUTIONAL PRESENCE IN THE SECTOR. THE KEY ISSUE SEEMS TO BE: WHERE DOES ONE GO FROM HERE? A CLEAR METHODOLOGY IS REQUIRED TO ACHIEVE CAREFULLY ESTABLISHED OBJECTIVES WORKING THROUGH THE INSTITUTIONAL STRUCTURE AS IT EXISTS TODAY. HOUSING DEVELOPMENT FINANCE CORPORATION LIMITED .HRFC CONTN_UFS TO BE THE MST TYNAMT- ANr,D iNNOVATIVE RHFI AND A STANDARD BEARER FOR THE INDUSTRY. IT CONTINUES TO BE LOOKED UPTO FOR ITS LEADERSHIP QUALITIES. HDFC COULD BE !LNCOURAGED AND SUPPORTED IN ITS ENDEAVOUR TO 1. PUT IN PLACE AN APPROPRIATELY STRUCTURED INSTRUMENT TO DEVELOP A SECONDARY MARKET IN MORTGAGES. 2. ASSIST IN STANDARDIZING DOCUMENTATION ACROSS THE INDUSTRY TO AID THE DEVELOPMENT OF A SECONDARY MORTGAGE MARKET. ANNEX 19 Paae 3 of 5 3. DEVELOP APPROPRIATE MORTGAGE PRODUCTS SUITABLE TO THE CURRENT ECONOMIC ENVIRONMENT, SUCH AS ADJUSTABLE RATE MORTGAGES. 4. DEVELOP INNOVATIVE RESOURCE MOBILIZATION INSTRUMENTS WHICH WOULD HELP REDUCE TERM TRANSFORMATION RISKS FOR LENDERS. 5. TRANSFORM ITS RESOURCE BASE TO INCLUDE HOUSEHOLDS DIRECTLY. 6. TRY AND DEVELOP AN EFFECTIVE METHODOLOGY FOR LINKING FORMAL AND INFORMAL FINANCE MECHANISMS THROUGH INSTITUTIONAL DEVELOPMENTS AT THE GRASS ROOTS LEVEL. 7. DEVELOP A MENU OF COURSES/PROGRAMS OF INTERNATIONAL STANDARDS THROUGH ITS TRAINING CENTRE -- THE CENTRE FOR HOUSING FINANCE TO BE MADE AVAILABLE TO THE EMERGING HFIS. AS THE ECONOMY OPENS UP HDFC WILL EXPAND ITS FINANCIAL SERVICES TO INCLUDE NEW SERVICES LINKED TO HOUSING INVESTMENTS SUCH AS INSURANCE AND RETIREMENT FUNDS. WHILE THE WORLD BANK'S COMPLETION REPORT ON THE HDFC PROJECT FINDS HDFC'S PERFORMANCE PARTICULARLY ACTIVE AND EFFICIENT. WITH OPERATIONS EXPANDING MUCH FASTER THAN EXPECTED AND BEING UTILISED TO ACHIEVE RESULTS BEYOND EXPECTATIONS. IT IS NOT AS SANGUINE ON THE POLICY RESULTS OF THE PROJECT. WHILE THE AUDIT REPORT BEMOANS POLICY INEFFECTIVENESS, THE NUMBER OF UNITS FINANCED BY THIS LOAN EXCEEDED THE NUMBER OF UNITS AND PLOTS FINANCED BY 10 EARLIER BANK FINANCED PROJECTS IN INDIA. THIS WAS ACHIEVED THROUGH EFFICIENT CREDIT CREATION (MAJOR EXPANSION IN LENDING WITHOUT SACRIFICING PORTFOLIO QUALITY). INNOVATIVE LENDING INSTRUMENTS AND GEJGRAPHICAL SPREAD. WHILE THE PROJECT WAS A SUCCESS FOR HDFC AND THROUGH A DEMONSTRATION EFFECT ON THE DEVELOPMENT OF MARKET ORIENTED HOUSING FINANCE. THE PROJECT DID NOT SUCCEED IN THE DEVELOPMENT OF THE HOUSING SECTOR FRAMEWORK. STRUCTURING A LOAN PROGRAM IT SEEMS THAT THE OPPORTUNITY THAT PRESENTS ITSELF AT THE PRESENT TIME IS ONE WHERE THE SHORTCOMINCS OF THE EARLIER PROJECT CAN BE CORRECTED THROUGH A PROJECT THAT ENCOMPASSES THE ENTIRE SECTOR. IDENTIFIES THE LINKS BETWEEN VARIOUS SEGMENTS AND PUTS IN PLACE A FUNDING MECHANISM THAT IS LINKED TO A RESOURCE BASE ON THE BASIS OF COMMERCIAL PRICING. WHILE THE POLICY AGENDA MAY NOT HAVE BEEN AS SUCCESSFULLY IMPLEMENTED. AS HOPED FOR. IN THE FIRST HOUSING FINANCE LOAN. THERE IS NO DOUBT THAT THE ?DEDNSTRATION EFFECT? THAT.---- HDFC'S SUCCESS HAS HAD IN THE WIDENING AND DEEPENING OF THE HOUSING FINANCE SYSTEM HAS BEEN A MAJOR CONTRIBUTION TO THE DEVELOPMENT OF THIS NASCENT SECTOR. 20 AN"X Pgga 4 of 5 LEADING FROM THE FRONT, SO TO SPEAK. HAS IMMEASURABLE VALUE IN PERSUADING NEWER ENTRANTS TO EMULATE SIMILAR PRACTICES, ESPECIALLY DURING TURBULENT TIMES. THE FINANCIAL SECTOR IN INDIA IS CURRENTLY IN THE MIDST OF MAJOR UPHEAVALS ,BROUGHT ABOUT BY (I) THE TRANSITION FROM A REGULATED ECONOMY TO A MARKET DRIVEN ONE AND (II) THE IRREGULARITIES IN SECURITIES TRADING BROUGHT TO LIGHT RECENTLY. THE NEED TO DEVELOP ADEQUATE REGULATORY AND SUPERVISORY CAPABILITY OF THE APEX INSTITUTIONS CANNOT BE OVEREMPHASIZED. AT THE SAME TIME THERE IS URGENT NEED TO DEVELOP PROPER CREDIT CONTROLS AND INFORMATION FLOWS IN THE INTERMEDIARY INSTITUTIONS. ONE WAY OF TACKLING THIS ISSUE COULD BE TO BRING ABOUT A WRENCHING. ONCE FOR ALL, CHANGE IN THE ENTIRE FINANCIAL SECTOR WITH THE ATTENDANT RISKS AND UPHEAVALS. ALTERNATELY THESE CHANGES COULD BE BROUGHT ABOUT IN A SUB-SECTOR - SAY, HOUSING FINANCE - TO BEGIN WITH, INCLUDING THE ENFORCEMENT OF CAPITAL ADEQUACY NORMS AND A SHIFT AWAY FROM THE DIRECTED CREDIT SYSTEM SO PREVALENT IN INDIA. THE DEMONSTRATION EFFECT OF THESE CHANGES AND THE CONSEQUENTIAL BENEFITS, COULD FACILITATE INITIATING SIMILAR CHANGES IN THE REST OF THE FINANCIAL SECTOR. PREFACE PARA 1 LINE 4 SUGGEST SHOULD READ AS ?...FINANCE MORTGAGE LOANS FOR THE PURCHASE/CONSTRUCTION OF HOUSES OR APARTMENTS GENERALLY IN THE URBAN AREAS OF THE COUNTRY.? PAGE III ?OTHER PROJSCT DATA FISCAL YEAR OF BORROWER : APRIL 1 - DECEMBER 31 EVALUATION SUMMARY PARA 6 LINE 3 ?CORPORATE LOANS? SHOULD BE REPLACED WITH ?LINES OF CREDIT? UNDER THE LINE OF CREDIT FACILITY HDFC LENDS EITHER TO OR THROUGH A CORPORATION FOR 'EMPLOYEE OWNED HOUSING'. UNDER THE CORPORATE LOAN FACILITY HDFC PROVIDES ASSISTANCE TO CORPORATIONS FOR 'EMPLOYER OWNED HOUSING' PARA 8 LINE 3 REPLACE THE WORD 'BYPASS' WITH 'SUPPLEMENTS' PERFQRMANCE AUDIT REPORT PARA 15 LINE 8 HDFC'S SHARE CAPITAL WAS INCREASED FROM RS 100 MILLION TO RS 200 MILLION IN FY 1987 AND WAS FURTHER ENHANCED TO RS 450 MILLION IN FY 1991. PARA 18 LINE 13 21 ?CORPORATE LOANS? SHOULD BE REPLACED WITH ?LINES OF CREDIT? Page 5 of 5 UNDER THE LINE OF CREDIT FACILITY HDFC LENDS EITHER TO OR THROUGH A CORPORATION FOR 'EMPLOYEE OWNED HOUSING'. UNDER THE CORPORATE LOAN FACILITY HDFC PROVIDES ASSISTANCE TO CORPORATIONS FOR 'EMPLOYER OWNED HOUSING' PARA 20 LINE 3 CAPITAL ADEQUACY OF 7.3 INDICATED IN THE PARA IS ON THE BASIS OF 100 RISK WEIGHTED ASSETS. PARA 22 LINE 4 ?AT A TIME WHEN INTEREST RATES? SHOULD READ 'S ?AT A TIME WHEN INFLATION RATES? PARA 23 LAST LINE AFTER 5 DROP THE WORD 'YEARS' PARA 24 LINE 8 SHOULD READ AS ?THIRTY FIVE PERCENT? PARA 26 THE LABEL ?EXPENDITURES IN FOREIGN CURRENCY? INCLUDES THE LOAN OF US DOLLAR 4 MILLION FROM IFC WHICH WAS REPAID IN FY91. THE SECOND IZC LOAN OF US DOLLAR 40 MILLION WHICH WAS SWAPPED WITH INDUSTRIAL FINANCE CORPORATION OF INDIA IS REFERRED TO IN SCHEDULE 14 NOTE 3 TO THE FY92 ANNUAL REPORT. PARA 30 )8,3 5 REPLACE 'NBH' WITH 'NHB' PARA 34 LINE 13 SHOULD READ AS ? HDFC'S BORROWING FROM NHB AMOUNTED TO RS 2,781 MILLTON IN FY92...' NHB FORMULATED ITS REFINANCE SCHEME IN 1989 AND HDFC MADE ITS FIRST DRAWAL OF REFINANCE IN JUNE 1990. THEREFORE THE REFINANCE LEVEL OF RS 2.781 MIILION WAS REACHED OVER A TWO YEAR PERIOD AND NOT ONE YEAR AS INDICATED IN THE REPORT. 1184829 HDFC IN 248423 WORLDBANK =05141254 RCA455 WWB1057 .05140916 =05171054 1Y T.TTT