Philippines - Cottage Enterprise Finance Project

Document of The World Bank FOR OFFl(IAI. lSE ONLY ., ~ ~ ~ ~ ~ . Report No. 9() WU-1'1H STAFF APPRAISAL REPORT PHILIPPINES COTTAGE ENTERPRISE FINANCE PROJECT DECEMBER 21, 1990 Industry and Energy Operations Division Country Department II Asia Region This document has a restricted distribution and mas be used by recipients only in the performance of their official duties. Its contents rnay not other%ise be disclosed without %%orld Bank authorization.  CURRENCY EOUIVALENTS (as of November 20, 1990) Currency - Peso US$1.0 - P 28 P 1.0 - US$ .036 P 1 million - US$ 36 thousand P 1 billion - US$ 36 million US$1.0 - DM 1.5 ABBREVIATIONS CB - Central Bank of the Philippines CSMI - Cottage, Small and Medium Industry DBP - Development Bank of the Philippines DOF - Department of Finance DTI - Department of Trade and Industry ESAL - Financial Sector Adjustment Loan FX - Foreign Exchange GF - Guarantee Fund IGLF - Industrial Guarantee and Loan Fund IICP - Industrial Investment Credit Project IS - Initial Subscription KfW - Kreditanstalt fur Wiederaufbau MGA - Mutual Guarantee Association MOU - Memorandum of Understanding NEDA - National Economic and Development Authority OPG - Operating Policy Guidelines PFI - Participating Financial Institution SMI - Small and Medium Industry TA - Technical Assistance TOR - Terms of Reference FISCAL YEAR Government: January 1 to December 31 DBP: January 1 to December 31  FOR OFFICIAL USE ONLY PHILIPPINES COTTAGE ENTERPRISE FINANCE PROJECT Loan and Project Summary -rower: Development Bank of the Philippines (DBP) irantor: Republic of the Philippines ieficiaries: Accredited participating financial institutions (PFIs) and private sector cottage enterprises )unt: US$15.0 million equivalent -ms: 20 years including 5 years of grace at the Bank's standard variable interest Lending Terms: The Development Bank of the Philippines would use the proceeds of the Bank loan to rediscount subloans made by PFIs to members of approved Mutual Guarantee Associations (MGAs). Loans from DBP to the PFIs would be made in domestic currency at market based rates. Government would bear the foreign exchange risk in exchange for a market- related fee payable by DBP. The PFIs would onlend the funds to their borrowers at prevailing market interest rates. Financiers: Kreditanstalt fur Wiederaufbau (KfW) ? 1ect iectives: The Project's objective is to help improve access to credit by cottage enterprises that typically employ 10-20 persons and have total assets of about US$15,000. )iect 3cription: The proposed Project would involve onlending by DBP to PFIs to rediscount subloans made to cottage enterprises to finance their investment and working capital needs during the 1991-94 period. The central feature of the proposed Project is that groups of cottage firms voluntarily would set up MGAs to whiich they would contribute to form a guarantee fund. This guarantee fund would be deposited with a participating bank, which would then lend some multiple of the fund to the individual members of the MGA. To initiate the scheme with adequate leverage, the members' subscriptions to the MGAs would be supplemented with a matching loan advanced and held in trust by DBP. As the MGAs' resources increase from new member subscriptions and -his document has a restricted distribution and may be used by recipients only in the performance *f their official duties Its contents maN not otherwisC be disclosed %ithout World Bank authorization  - ii - from interest earned on their equity, the matching loan would be repaid and the funds used to assist new MGAs. The average size of subloans to indivLdual enterprises is expected to be about US$10,000. Technical assistance under the Project would be provided to help establish and monitor the MGAs, train key staff of the MGAs, provide advice on waste minimization and treatment, oversee the implementation of the program; and help train personnel in the retail banks to deal with the MGAs under the program. Benefits: The Bank loan would increase the volume of credit for lending to cottage industry and would reduce its cost. The Project would assist DBP and the PFIs in gaining experience and knowledge about lending to cottage industry and in establisning an effective finance facility over the long term to improve the access of cottage firms to bank lending. It is expected that about 4,200 cottage firms would be financed under the Project with subproject costs of about US$33 million. The Project would cover every region in the country and the vast majority of loans would be outside Metro Manila. It is also expected that the project design would enhance the ability of women entrepreneurs to obtain formal credit. Risks: The program is an innovative one which does not currently exist in the Philippines. Accordingly, there are the risks associated with any new program. One risk is that there will be significant defaults under the program that will reduce the guarantee funds. This risk will be dealt with by starting with moderate levels of leverage, by careful selection and licensing of the MOAs and by the exercise of peer pressure to repay, inherent in the project design.  - iii - Estimated Cost: Local Foreig Total ----- (US$ million) Subproject Cost 21.4 11.6 33.0 Technical Assistance (Project Support) 0.9 0.8 1.7 Guarantee Fund 10.0 - 10.0 Total 32.3 ILA_ 44.7 Financing ]SLgn: Local Foreign Total ------(US$ million)----- MGA Subscriptions and Fees 4.4 4.4 Subborrowers (Equity) 3.4 3.4 Participating Financial Institutions 2.9 - 2.9 Kreditanstalt fur Wiederaufbau 12.6 0.8 13.4 IBRD 3.4 11.6 15.0 DBP 5.6 5.6 Total 2 3 12 4 44.7 Estimated Disbursements IBRD FY 1991 1992 1993 1994 1995 1996 -----(US$ million)------------ Annual 1.5 1.2 3.6 4.2 2.9 1.6 Cumulative 1.5 2.7 6.3 10.5 13.4 15.0 Rate of Return: Not Applicable  - iv - PHILIPPINES COTTAGE ENTERPRISE FINANCE PROJECT Table of Contents Page No. I. SECTORAL BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . 1 A. The Industrial and Financial Sectors . . . . . . . . . . . . . . . 1 B. The Cottage, Small and Medium Industries (CSMI) Sector . . . . . . . . . . . . . . . . . . . . . . . 2 C. Access to Financing . . . . . . . . . . . . . . . . . . . . . . . 4 D. Financing Requirements . . . . . . . . . . . . . . . . . . . . . . 5 E. Bank Objectives and Role in the Sector . . . . . . . . . . . . . 6 II. THE DEVELOPMENT BANK OF THE PHILIPPINES . . . . . . . . . . . . . . . 7 A. DBP's Background and Function . . . . . . . . . . . . . . . . . . 7 B. Organization and Management . . . . . . . . . . . . . . . . . . . 7 C. Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 D. DBP's Financial Performance . . . . . . . . . . . . . . . . . . . 9 III. MUTUAI, GUARANTEE ASSOCIATIONS . . . . . . . . . . . . . . . . . . . . 10 A. The Concept .............. ..... ...... . . 10 B. Mechanics of the MGA Scheme . . . . . . . . . . . . . . . . . . . 11 C. Actual MGAs .............. ..... ...... . . 13 IV. THE PROJECT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 A. Project Objectives ........ ... .. ... .. ... .. . 15 B. Project Description ........ .. ... .. ... .. .. . 15 C. Beneficiaries and Subloan Size . . . . . . . . . . . . . . . . . . 17 D. Project Implementation . . . . . . . . . . . . . . . . . . . . . . 18 E. Overall Project Costs and Financing . . . . . . . . . . . . . . . 21 V. THE PROPOSED BANK LOAN . . . . . . . . . . . . . . . . . . . . . . . . 23 A. Features of the Bank Loan ...... .. . .. . .. .. . .. . 23 B. Relending Mechanism, Terms and Conditions . . . . . . . . . . . . 23 C. Procurement .... . . . . . . . . . . . . . . . . . . . . . . . 24 D. Disbursements .... . . . . . . . . . . . . . . . . . . . . . . 24 E. Monitoring, Reporting and Auditing ... . . . . . . . . . . . . . 25 F. Environmental Aspects .... . . . . . . . . . . . . . . . . . . 26 G. Benefits and Risks .... . . . . . . . . . . . . . . . . . . . . 26 VI. AGREEMENTS AND UNDERSTANDINGS AND RECOMMENDATION . . . . . . . . . . . 27 A. Agreements and Understandings . . . . . . . . . . . . . . . . . . 27 B. Conditions of Effectiveness ...... . . .. . .. . . .. . . 27 C. Recommendation .... . . . . . . . . . . . . . . . . . . . . . . 28  -V ANNEXES 1. Enterprise Characteristics by Sector 2. Regional Distribution of Enterprises 3. Employment by Cottage Enterprises 4. DBP Historical Financial Statements 5. Firm Characteristics of Initial MGAs 6. Consolidated MGA Guarantee Fund and Loan Accounts 7. Cost Estimates of Technical Assistance Component 8. Accredited Participating Financial Institutions 9. Disbursement Schedule .0. Supervision Plan 1-. Documents in the Project File Ha;< IBRD No. 22459R This report was prepared by Mr. K. Young, Ms. M. Hanson (AS2IE), and Mr. T. Masuoka (YPP) following a field appraisal of the Project in July 1990. The Peer Reviewers were Ms. L. Webster (IENIN) and Mr. C. Magnus (AS5IE). The Project was cleared by Mr. G. Kaji (AS2DR) and Mr. I. Sud (AS2IE).  PHILIPPINES STAFF APPRAISAL REPORT COTTAGE EFTERPRISL FINANCE PROJECT I. SECTORAL BACKGROUND A. The Industrial and Financial Sectors 1.01 Rapid changes in the international environment since 1980, as well as high interest rates and a foreign exchange crisis in the Philippines, led to forced austerity measures which hit hard an industrial base that was brittle and largely uncompetitive. Non-traditional manufacturing exports had been narrowly concentrated in product mix (semi-conductors and garment-s) and destination (the USA). The remainder of industry was dependent on domestic demand, which was artificially inflated by protection, by heavy foreign borrowing and by infusions of credit and transfers from the government. The scope for further expansion of import substitution under heavy protection had been exhausted in the late 70's and technological weaknesses and other inefficiencies fostered by the lack of competition had begun to take their toll. Industrial output declined by 25% during 1982 through 1986. Many firms ceased operation and declared bankruptcy. Even firms in operation limped along at 50-60% capacity utilization. Since 1987, however, there has been a significant recovery in industry which has been growing on average at 8% a year. 1.02 Since 1980, the Government has undertaken a process of gradually removing major sources of distortions within the system. Although the reform process has been far from smooth, much has been accomplished during the period. Paradoxically, the economic and financial crisis which derailed the trade reforms during 1983-85 provided the momentum for drastically reducing some of the more intractable government interventions in the industrial sector. Following the change in government in February 1986, the pace of reform quickened and, with the resumption of import liberalization in 1986, significant progress has now been achieved in all of the key areas of needed reform including tariffs, incentives, foreign exchange and exchange rate management and privatization of public industrial assets. although policy-induced distortions still remain, the degree of transformation is striking and represents a major achievement of the policy reform process. 1.03 Industrial policy reforms in the Philippines have been accompanied by financial sector reforms. Following the 1983-85 recession and resulting distress in the financial system, the Government successfully undertook measures to rehabilitate The Philippine National Bank and DBP, in the context of the Bank-financed Economic Recovery Loan (ERL; Loan No. 2787-PH, 1987). Since 1989 the Government has embarked on a comprehensive program of financial sector reforms to strengthen further the banking system and make it more competitive and responsive to the needs of the real sectors. These reforms are being supported by a Bank Financial Sector Adjustment Loan (FSAL; Loan No. 3049-PH, 1989). The main objectives of the reform program are to: (a) strengthen the supervision and regulatory framework for commercial banks; (b) improve the  institutional arrangaments for deposit protection; (c) reduce intermediation costs in the banking system; and (d) improve the institutional framework for mobilization and delivery of term credit. B. The Cottage. Small and Medium Industries (CSMI) Sector 1.04 The CSMI sector in the Philippines includes cottage, small-scale and medium-scale industries, which may be def'ned either in terms of asset size or number of employees. The asset-based definition is used by the banking sector in determining eligibility for finance; the Census of Manufacturers, however, categorizes industries based on number of employees. Thus, cottage industries may have total assets before financing ranging between P 50,000 - 499,999 (about US$2,100-21,000), or less than 10 employees; small industries may have total assets ranging between P 500,000-4,999,999 (about US$21,000-210,000), or 10-99 employees; and medium industries have total assets ranging between P 5-20 million (about US$210,000-830,000), or 100-199 employees. Based on fixed asset data provided in the Census of Manufacturers, there appears to be strong correlation between the asset- and employment-based CSMI definitions; average fixed assets of the various CSMI subsectors, defined in terms of employment, fall well within the total asset cut-off points for CSMI. 1.05 The Cottage Industries Subsector. Cottage enterprises comprise a large and important subsector, especially in terms of employment impact. The "formal" cottage sector, as reported in the 1983 Census of Establishments, constitutes some 300,000 or 94% of all formal enterprises. These cottage firms employ over 1 million people and accounted 'or 40% of formal non-agricultural employment and 10% of value added (See Table 1.1 below and Annex 1 for greater detail).!, Table 1.1 NUMBER, EMPLOYMENT AND VALUE ADDED OF COTTAGE AND LARGE FIRMS: 1983 (Values in 1989 US$) Cottage Firms Small/Medium & Total Large Firms Firms Firms:Number 301,009 20,738 321,747 % of total 94% 6% 100% Employment: Number 1,020,786 1,543,018 2,563,804 X of total 40% 60% 100% Value Added (billion $) 1.4 12.7 14.1 % of total 10% 90% 100% 1/ It should be noted that this definition includes only formally established enterprises and does not include the large informal household sector.  1. 06 Rezional Dispersal. A significant characteristic of the cottage enterprise subsector worldwide is its substantial regional dispersion reflecting small firms' ability to locate in areas where infrastructure is poor and markets are small. This pattern hold3 crue in the Philippines where almost two-thirds of the cottage enterprises' value added comes from outside Metro Manila while for small/medium/large firms only about a third of value added emanates from outside Metro-Maniila (see Chart 1.1 and Annex 2). Given the very large income disparities among regions in the Philippines, the regional dispersion of _ottage enterprises suggests they can be used as important vehicles for improving incomes and employment in the regions. Chart 1.1 ENTERPRISE VALUE ADDED By Region 1983 SMA LL/MEDIUM/ COrTAGE ENTERPRISES LAROL ENTERPRISES Me,ro Manl!a (NCR) 34% + Matro Manila (NCR) Other Regions Other Re9lons 66% 3tR% 1.07 Capital/Labor Ratios. Cottage enterprises in the Philippines are much more labor intensive than large firms. The capital/labor ratio for coctage firms is about one-tenth that of larger firms. This may in part be due not to size per se but to different product areas in which large and small firms specialize. However, to the extent that the cottage subsector is disadvantaged by industrial policies favoring larger, more capital intensive firms or is impeded from obtaining credit by structural rigidities in the financial sector, the labor intensity of the economy is lower than it otherwise would be.  1.08 Sectoral Distribution. In terms of employment, about 57% of cottage inidustry jobs are in wholesale and retail trade; 18% are in manufacturing; 21% in personal arid financial services; and the rest, about 4%, scattered in construction, mining and transport. About 25 product areas (at the 4 digit level) account for three-quarters of cottage enterprise jobs. Of these, food and beverage rctailing is by far the largest, accounting for about one-third of cottage employment (see Annex 3). 1.09 Female Participtaion. A noteworthy characteristic of cottage eniterprises worldwide is that women comprise a significant share of employment, ownership and management. This also seems to be true in the Philippines, where about two-thirds of the labor force in manufacturing in rural areas, where cottage firms predominate, is female. Interviews with firms and producer associations suggest a significant female participation in the cottage sector. 1.10 Firm Expansion. Studies on other countries suggest that cottage firms are an important and possibly the main source of small and medium firms. Interviews in the Philippines also confirm that a significant share of labor absorption in larger scale firms is due to expansion of small firms.91 At the same time, the relative lack of small and medium firms in the Philippines suggests that there are important barriers to the expansion of cottage and very small firms. Some of the important macro barriers imposed by a protectionist trade policy, a capital intensive and large scale bias in fiscal incentives and regulated incerest rates and a protective approach to cottage firms were greatly reduced during the 1980s. At the same time, the ability of cottage firms to expand is still severely constrained by lack of access to adequate finance (paras 1.10 to 1.13). C. Access to Financing 1.11 Interviews with cottage firms, industry associations and banks indicate that cottage firms have almost no access to credit through banking channels. This includes both working capital as well as term financing. The commercial, rural, development and thrift banks interviewed generally indicated that the relatively sm.ill loan size together with a high perceived risk and lack of commercially acceptable collateral deterred them from lending to cottage firms. 1.12 Collateral Deficiency. The main issue in this regard is that most cottage firms lack real estate titles other than for their own homes, which are usually already mortgaged. Thus they frequently have only chattel mortgages to offer banks. In common with banks worldwide, chattel mortgages are considered greatly inferior to real estate mortgages. This inferiority is substantially compounded for cottage sized firms where the chattel may be small machines which can easily be moved and/or substituted, difficult to value and to repossess. Further, such firms usually lack established premises, reputations and banking 1 "Small Enterprise Development Policy in the Philippines: A Case Study", Dennis Anderson, World Bank Staff Working Paper No. 468, 1981.  track records. Hence the smallness of the firms is an important reason deterring banks from accepting their assets as collateral. 1.13 Lack of Special Programs. Existing special credit programs supported hy Gove-nment do not reach the cottage industry sector. The Industrial >uarantee and Loan Fund (IGLF) program, which in principle provides refinancing of term credit for cottage as well as small and medium sized firms, in practice has been ineffective in getting credit to cottage firms. For example, only US$210,000, or two-tenf:hs of one percent of total IGLF loans made over the past 6 years, has gone to cottage firms.V The 'Tulong Sa Tao" program operated by the Department of Trade and Industry (DTI) through private voluntary organizations is focused on micro and livelihood activities. With a maximum loan size of about US$1200 and operating through non-formal credit channels, this program is not able, nor intended, to address the credit needs of the vast majority of cottage firms. 1.14 Other Sources. One result of cottage firms' being shut out of the formal financial system is that they turn to the informal moneylenders. However, interest costs in such markets are extremely high, ranging from 3% up to 17% or more per month. The high cost of such credit clearly limits its use for normal expansion or working capital requirements. Alternatively, firms limit their operations to what can be financed with internally generated funds and fail to expand and seek out new opportunities. D. Financing Reguirements 1.15 The annual iinancing requirements of the cottage sector are quite large. Even with conservative assumptions about turnover period and expansion, it is estimated that cottage enterprises would require ovcr US$500 million per year, comprised of about US$150 million term credit for replacement and expansion and about US$350 million for working capital needs. 1.16 The Financing G a. Presently, the sources of funds are mainly own funds plus short term high interest loans from moneylenders, which are both expenrive and limited. The data on this are scarce but the 1983 Survey indicates that gross additions to fixed capital by cottage firms were only about a quarter of estimated depreciation, indicating that capital was not even being adequately maintained let alone expanded. In contrast, larger firms were fully offsetting depreciation and expanding their fixed assets at the rate of 9%. It .i. The design of the IGLF program appears inappropriate to address the needs of the cottage enterprise group in that it fails to address the problematic lack of "bankable" collateral of the cottage firms. Furthermore, the IGLF program fails to be effective in the regions, where cottage firms are located. In 1987, only 27% of IGLF loans went to the regions. Tables showing the characteristics of IGLF subloan approvals by size, purpose and maturity, as well as the regional distribution of IGLF subloans, are included in the Project File.  is not suggested that these informal sources be fully replaced with credit provided through formal bank channels. Certainly, for start-up firms with minimal track records, reliance on own funds is reasonable. However, for the top 5-10% of cottage firms which show promise of expansion, continued reliance on informal sources of credit constitutes an effective barrier to their growth which should be remedied. In this regard, a reasonable mediuid term indicative target might be to expand bank credit to this sector by about US$50 million. 1.17 In summary, cottage firms have relatively large financing requirements for both working capital and investment. The banking sector is essentially closed to cottage firms. A significant part of their needs is probably not being met as evidenced by a very low rate of capital formation. What is available has to come from high cost moneylenders on short term and from their own funds. While data are not available to provide precise estimates of the financing gap, it is clear that there is a large role for the banking sector to play to meet the financing requirements of a select group of promising cottage firms at commercially viable rates. E. Bank Objectives and Role in the Sector 1.18 The Bank's emphasis in the financial sector will continue to be on strengthening the supervisory and regulatory framework, and institutional arrangements for depositor protection. The Bank will also help to establish financial markets and instruments to ensure that priority areas of the economy have access to funds at competitive rates, appropriate duration and on a timely basis. 1.19 While the proposed Cottage Enterprise Finance Project is the first Bank project in the Pnilippines to introduce a scheme for cottage industry finance, it would complement four previous projects (Loans 1120-PH, 1727-PH, 2169-PH and 3038-PH) designed to promote the development of the SMI sector. These loans supported the IGLF program which provides mainly term finance for small and medium industry, and provided additional funding for SMI through DBP. The first three of these loans have been fully disbursed, and Project Performance Audit Reports (PPAR# 3696 dated June 16, 1982, PPAR# 7940 dated June 30, 1989 and PPAR #8781 dated June 21, 1990) were distributed to the Board. Under these projects, the Bank approved US$304 million for assistance to the SMI sector. Apart from the reduced demand for IGLF funds in 1986 and 1987, the IGLF components of the SMI projects have generally performed well and achieved their objectives in terms of channeling substantial credit to SMIs, increasing the number of financial institutions participating in the program and operating with low arrears and high collection ratios from both the PFI and end-users. The DBP component of the third loan was largely canceled due to DBP's organizational and financial difficulties. The fourth loan to support IGLF (SMI IV) became effective July 14, 1989. Loan demand has been strong and disbursements exceed projections. Despite an overall good performance, one weakness of the IGLF program has been its inability to address the credit needs of cottage size firms and to extend credit to the various regions of the country (para 1.13).  II. THE DEVELOPMENT BANK OF THE PHILIPPINES 2.01 The Development Bank of the Philippines (DBP) would be the borrower of the proposed Bank loan and would relend the funds in local currency to participating financial institutions (PFIs) for onlending to cottage enterprises. A. DBP's Background and Function 2.02 Although DBP was established in 1958, it has existed in its present form only since the late 1980s, following an extensive rehabilitation and restructuring program. Due largely to dislocations in the industrial sectors in the early 1980s (para 1.01) and a concomitant disarray in the financial sector, and also to problems specific to DBP -- including outside interference in DBP's lending decisions; the undertaking of costly government credit programs through DBP without reimbursement; a lack of internal controls and auditing; poor management practices in terms of lending decisions and portfolio management; and excessive size and diversity of functions -- DBP had serious financial difficulties in the mid-1980s. Its rehabilitation in 1986-1989 included: (a) a new charter and policy statement; (b) installation of a new management team and Board of Directors; (c) transfer of P 74 billion of non- performing accounts to the Asset Privatization Trust created by the government to sell those assets; (d) implementation of a drastic cost reduction program through staff reduction and sale of branches; (e) an internal reorganization and strengthening program focusing on credit policies, legal procedures, financial controls, accounting and internal controls, and personnel management; and (f) phasing out of subsidized government deposits. It left DBP healthy and self- sustaining, with market-oriented lending policies and a new mandate as a predominantly wholesale bank for onlending funds to retail financial institutions rather than to end-users. The proposed Project is consistent with the evolution of DBP as a wholesale lending institution in that it provides funds for relending by DBP to PFIs. B. Organization and Management 2.03 The DBP organizational structure is appropriate for implementing its wholesale banking strategy. DBP has an upgraJed Capital Markets Group (headed by a Vice President) adequate to meet DBP's resource mobilization needs, and has a Wholesale Loans Department headed by an Executive Vice President for Marketing. DBP has a total staff of about 2000, and has in place a job and performance evaluation system as well as a merit oriented financial reward system. DBP supplements its salaries with productivity bonuses in order to offer compensation competitive with the private financial sector. In the last three years, DBP has attracted nine senior officers from leading private sector banks to head its departments. 2.04 DBP has trained its staff in corporate finance, loan packaging, financial analysis and legal issues. To er.sure adequate staff preparation for DBP's new activities in wholesale banking, DBP is undertaking a training program covering (a) its capital markets, merchant banking and resource mobilization  - 8 - actlivities, and (b) wholesale lendir.g, focussing on the criteria, systems and procedures required for dealing with PFIs as clients. 2.05 The accounting and management information system is adequate for DBP's new role as a wholesale bank. Liquidity and reserve summaries are generated daily; financial statements of branches and the head office are consolidated monthly. The general ledger is computerized and closed daily. Subsidiary ledgers for loans have been centralized in the Transactions Processing Department. A central mainframe computerized information system, acquired from Citibank, is expected to be operational in 1990. This state-of-the-art system will provide an integrated management information system that allows immediate access to all transaction information which is scored in a centralized data base. The system will integrate on-line authorization and printing of customer advices, and contract administration for loans, placements, bonds, deposits, commitments, receivables and sundries until final maturity. The system will also incorporate an automated accounting service. C. ORerations 2.06 DBP's operations are governed by a set of policies designed to ensure that it avoids the mistakes of the past. On April 5, 1989, DBP's Board approved a new Policy Statement (included in the Project File) which reflects its wholesale lending orientation and which deals with, among other things, DBP's investment policies, financial prudence limits and capital structure. Salient features of the Policy Statement provide for: (a) private sector orientation with autonomy in decision-making; (b) an interest rate and fees policy that is consistent with prevalent market rates and that ensures full recovery of all its direct and indirect costs, including provisions as required, plus an adequate profit margin; (c) prudent interesc rate and maturity match between its assets and liabilities; and (d) a debt-equity ratio not exceeding 5:1. These market- oriented policies should enable the integration of DBP into the market-based financial system of the Philippines. DBP is now ready to undertake more substantial lending activities. 2.07 DBP's transformation to a primarily wholesale bank is being accomplished through a gradual reduction of its retail loan approvals and outstanding retail loans to enterprises, relative to the wholesale loan approvals and outstanding wholesale loans to financial institutions. DBP's wholesale lending has been enhanced since late 1989 with the Industrial Investment Credit Project (IICP) and the IGLF, whose transfer from the Central Bank occurred on July 31, 1990. In addition, the following programs are expected to increase DBP's wholesale lending activities substantially in 1990- 1991: the Asean Japan Development Fund; an Asian Development Bank line of credit similar to the IICP; and the proposed IBRD-financed project for Industrial Restructuring. The strategic transition to wholesale banking also entails gradual disinvestment from DBP's retail operations. This involves the separation of the retail and wholesale operations by end-1990, to be followed by privatization of at least 30% of the equity base of the retail operations by March 1991. By 1993, DBP's equity stake in retail operations would have been reduced by 70%.  -9- 2.08 DBP's lending in the coming years will consist of (a) wholesale loans to PFIs, (b) retail loans to enterprises in areas where the commercial banking system may still be hesitant to enter fully and (c) syndicated loans wherein it will play a catalytic role in mobilizing the participation of private sector banks. Given DBP's predominantly wholesale orientation, with the transfer of the IGLF program to DBP and with DBP's expected role as the main AJDF wholesale conduit, wholesale loans outstanding are projected to grow substantially, from P 5.6 billion in 1990 to P 34 billion in 1995. The bulk of these loans will be for the industrial sector. Retail lending is projected to grow at a more modest pace, from P 6.5 billion in 1990 to P 13 billion in 1995. D. DBP's Financial Performance 2.09 DBP has been solvent and profitable since its reorganization (see Annex 4). Its indicators of capital adequacy, earnings performance and liquidity are sound. Its debt/equity ratio is a conservative 0.85:1 and its earnings indicators are sound: its return on assets is 9.5%, and its return on average net worth is 17.5%. DBP's performance compares favorably with that of leading private financial institutions in the Philippines. DBP is in a sound liquidity position with liquid assets representing 163% of deposits. In the coming years, liquidity will decrease somewhat as (a) interest-free government deposits are reduced and (b) wholesale lending activities grow. The current average maturity of funding liabilities is about 4.3 years, against an average loan maturity of 6 years. 2.10 It should, however, be pointed out that DBP's 1989 earnings performance is not fully indicative of DBP's long-term earning capability on a going concern basis. Of the total operating revenues of P 1.9 billion, only 52% is interest from loans, whereas 25% comes from investment in securities, primarily government issues, 5% from the sale of assets, and the remainder from trust funds, foreign exchange gains and other miscellaneous sources. As for expense, until December 19, 1989, P 1.4 billion of interest-free Government deposits were on DBP's books, comprising 12% of its total funds. On December 20, 1989, these funds were converted into a special time deposit costing DBP 15.8%. As DBP mobilizes higher cost funds, it should expect lower, though still positive, net returns.  - 10 - III. MUTUAL GUARANTEE ASSOCIATIONS A. The Concept 3.01 The main obstacle to improving the access of cottage enterprises to banks is collateral (para 1.12). Banks consider cottage enterprises to be high risk clients because of their weak financial structure and management dependence on a single founder. To hedge their risks, banks require significant collateral, even when that demand amounts to over-covering their exposure to risks. Not being able to meet such collateral requirements, cottage enterprises are often barred from bank credits. 3.02 It would be counterproductive to bridge the gap between cottage enterprises and banks by trying to convince banks to soften their demands. Bank managers are responsible for the financial situation of their institutions. Rules of management, no matter how over-prudent they may appear to outside observers, cannot be influenced without triggering a dangerous interventionism that would be detrimental to the financial sector. The Mutual Guarantee Association (MGA) concept being supported under the proposed Project attempts to narrow the gap between banks and cottage enterprises by trying to meet (rather than ignore) banks' legitimate concern for acceptable collateral. It provides a system that offers banks unquestionable, independent and financially sound guaiantees, with prompt payment whenever a guarantee is called. The credibility of the system is enhanced by the fact that the beneficiaries themselves (i.e. the cottage enterprises benefitting from the system) would have their own money at stake. 3.03 In the MGAs, potential borrowers would group together to establish local, self-managed associations that would issue guarantees to lending banks on their behalf. An MGA would group 40 to 100 members that would contribute financially to a liquid Guarantee Fund (GF) owned by the MGA. The GF would be placed in an account in a bank and pledged as collateral for future individual borrowings by the MGA members. The GF thus would provide a liquid and immediately available guarantee to the bank, should one of the members default on his or her loan. A written agreement authorizing the bank to debit the GF account upon occurrence of specifically designated events would insure the lender that a called guarantee would be payable and paid at first demand. The capacity of the MGA to serve its members would be dependent on the size of its GF: in aggregate, the MGA would safely guarantee loans up to a given multiple of the GF. 3.04 The MGA concept has evolved differently in different countries. Some MGAs have been set up to guarantee individuals (and not only enterprises) when the individuals (such as dentists and radiologists) have professions that demand heavy investment in equipment. Other MGAs guarantee borrowings from individuals who want to set up their own companies or invest in existing companies (venture capital). In Germany, MGAs have even become full commercial banks, thus providing both guarantees and capital to their members. 3.05 The MGA approach differs fundamentally from the traditional approach to collateral in the Philippines. Unlike the "co-maker" system, the MGA does  - 11 - not provide the bank with alternate recourse against other persons or companies; the recourse is against a safe liquid fund held by the lending bank itself. Unlike other cooperative arrangements, the MCA guarantees the individual borrowings of individual members even though it has put personal money of all its members at stake by pledging the GF as collateral. This insures the bank that the traditional relationship with its client will not be disturbed and, at the same time, that due care will be taken by the association of timely repayments of loans by the borrower. Finally, the approach differs from the centralized approach in which all banks are provided with guarantees issued by one centralized guarantee fund. B. Mechanics of the MGA Scheme 3.06 Guarantee Funds and Matching Funds. Under the proposed Project, MGAs would reach agreement with individual banks in which they would deposit the Guarantee Fund (GF), which would serve to guarantee member borrowings. A Memorandum of Understanding (MOU) and supporting legal documents signed by the MGA and the bank would state the conditions under which the bank would be authorized to draw from the GF. 3.07 The personal subscriptions of the MGA members to their GF are key: the guarantee capacity of each MGA depends on the overall size of its GF, and the members' subscriptions represent their financial commitment to the program. In the long term, GFs would increase in size through the inflow of interest income, fees, and subscriptions. At the inception of the proposed Project, however, most GFs would be too small to facilitate significant borrowing if they consisted only of the subscriptions of the initial members of the MGA. In this regard, the proposed Project includes a Matching Loan Fund which would be lent to individual MGAs. For initial members of MGAs established during 1991-94, the matching loan would be twice the members' own contributions. For additional members to existing MGAs, the matching loan ratio would be reduced to 1.5 in the second year and to 1 in the third and fourth years. The Matching Loan Funds would be held in trust by DBP. The share of the Matching Loan Fund in the GF would decline with increases in equity resulting from new members, fees and interest income, and with increases in the banks' lending multiple as they gain experience under the program. 3.08 MGA Revenue. MGAs would require from their members an initial subscription (IS) fee on joining the MGA. Under the proposed Project, the initial subscriptions are expected to average P 20,000. Being set up as limited liability stock companies, MGAs require entering members to pay for a share of the capital of the company: the capital contribution accordingly buys shares in the corporation for the members. Each member would be required to buy one common share, entitling him/her to a corresponding vote. Members, at their option, may also buy one non-voting preferred share. After a 12-month waiting period, members would be free to sell their shares at their discretion, if they have no outstanding default on an MCA-guaranteed loan. After five years, members would be entitled to sell back their shares to the MGA. In addition to the IS, each member is assessed an annual fee of 15% of the amount of the IS. The structure of the MGA and the level of fees and share valuation are outlined  - 12 - in Articles of Incorporation and By-laws, which were agreed at negotiations (para 4.12). The funds flow of the MGAs is reflected in Chart 3.1 below. 3.09 The MGA would also derive income from commissions paid by members on the actual guarantees issued by the MCA on their behalf. Under the proposed Project, the guarantee fee would be set initially at 2.5% p.a. of the outstanding guaranteed portion of the MCA members' loans. Further income for the MGA would be generated from interest earned on the deposit with the local bank. 3.10 Specific rules regarding the use of the GF and member subscriptions are expected to differ from one MGA to another. Each MGA would be expected to define its own precedence order regarding use of the GF among its members. MGAs would also determine for themselves the relationship between the size of a member's equity and his or her use of the GF. 3.11 Guarantees. When an MCA member wants to borrow from the bank, the enterprise will submit a request to the guarantee committee of the MGA. The MGAs would be able to issue guarantees covering up to 80% of each loan granted by the bank to individual members, with the total amount of MGA guarantees being limited by the total loans to CF ratio agreed between the MCA and the bank. Although long-standing MGAs in other countries are highly leveraged (e.g. with loan to GF ratios of 8-10), reflecting the confidence gained from the program's track record, discussions with banks during preparation of the proposed Project suggested that a multiplier of 3 was a likely starting place for the proposed scheme in the Philippines. It is expected that with time this initially low multiplier would be increased as the credibility of the MGA increases. The guarantee issued by the MGA would cover both the principal and interest of the bank loan. Chart 3.1 GUARANTEE FUND: Sources and Uses of Funds Annual Fees Initial Subscriptons Leaving Members ~~~~~ ~~~~Loan Defaults Matching tching Fund Fund Repayment of Loan Matching Fund Operating Expenses Guarantee Fees Interest Revenues  - 13 - 3.12 A guarantee would be payable at first demand, when the bank calls it. This differs significantly from other types of guarantees which are not payable at first demand but only after the beneficiary has shown that it has the right to call the guarantee. Experience in the Philippines under other guarantee programs has shown that this is frequently a highly protracted legal affair ultimately undermining the credibility and effectiveness of the guarantee. 3.13. If a borrower defaults on a payment, the bank would call the guarantee issued by the MGA. Once called by the bank to pay the guarantee, MCAs would pay only the overdue amount for the first three defaulted payments; only after three payments are missed by the borrower would the bank have the right to call the entire loan. This approach would give the MGA time to try to convince the defaulting member to resume payments to the bank or even propose a restructuring of the loan. Once the MGA assumes the full obligation of the loan, it would take action against the defaulting member to recuperate its money. The legal documents of each MGA would provide a detailed procedure to be followed in such occurrences. 3.14 The MGAs when issuing a guarantee on behalf of one of their members would have the member simultaneously issue a counterguarantee to the MGA. This would enable the MGA to recover part of the claim in the case of default. This use of formal legal instruments entitling the MGA to take legal action against the member would be complemented with peer pressure from within the MGA. Since a defaulting member endangers the viability of the whole association, such a member would come immediately under the strong encouragement of his fellow members to repay. C. Actual MGAs 3.15 During project preparation, 25 MCAs, each with at least 40 members, were formed in six regions. These MGAs typically originated as producer associations and were selected for the Project after a review of their management and the degree of discipline existing among members. The MGAs were established independently from the associations: not 811 of the members of the producer associations were interested or qualified to become members of the corresponding MGAs. MGAs elected their own management teams which often were different from those of the producer associations. 3.16 In line with project objectives, the MGAs formed to date have a wide regional distribution covering 15 provinces. Only one MCA is located in Metro Manila. The average firm has assets of P 350,000 (about US$15,000) and 17 employees. The MGAs have a wide sectoral distribution with significant representation among the furniture, food, handicrafts and metal manufacturing subsectors. A summary of key characteristics of the initial firms is presented in Table 3.1 and Annex 5. Detailed firm-by-firm data is available in the Project File. The regional distribution is indicated in the attached map.  - 14 - Table 3.1 CHARACTERISIICS OF THE INITIAL fGA-MEMBER COTTAGE ENTERPRISES 8 Average Asset Size ! 350,000 Average Number of Employees 17 Share of Female Owners/Managers 43% Share of Female Employees 53% Share of Firms Outside Metro Manila 99% Sectoral Distribution Furniture 11% Food Preparation 15% Garments 13% HandicraftW' 21% Metal Manufacturing'/ 17% Cut Flowers 5% Footwear 3% Servicesd/ 9% Other 6% TOTAL 100% a/ Based on questionnaire of initial 1,064 cottage firms (excludes applicants with assets above n 3 million) Includes shellcraft, woodcraft, embroidery, and jewelry C/ Includes machinery and electronics / Includes retail  - 15 - IV. THE PROJECT A. Project Oblectives 4.01 The object of the proposed Project is to facilitate the extension of credit to cottage enterprises through formal banking channels. Under the Project, about 60 MGAs would be established with about 4200 member enterprises with a wide geographical distribution and a significant proportion of women entrepreneurs. The regional phasing is shown in the attached map. B. Project Description 4.02 As discussed above (para 2.01) the proposed Project would involve onlending by DBP to participating financial institutions (PFIs) to rediscount subloans made to cottage enterprises for investment and working capital needs during the 1991-94 period. The proposed Project would have three components: (i) subloan financing to fund projects of MGA member cottage enterprises through accredited banks; (ii) technical assistance to assist the MGA Management Unit to set up, monitor, train and supervise MGAs and provide pollution-abatement assistance to MGA member cottage enterprises; and to assist DBP in providing training to financial institutions; and (iii) a guarantee fund comprised of MGA member subscriptions and supplemental matching loans used to guarantee MGA member subloans. 4.03 Subloan Financing. The line of credit under the proposed Project would provide financing for MGA members whose loans are guaranteed by the MGA. The amount of subloans would be related directly to the size of the GF and the loan multiplier agreed with the bank. The base case assumptions and project parameters discussed in Annex 6 indicate loan demand by 1994 of about US$30 million on an outstanding basis. DBP would act as wholesaler through a network of accredited retail banks. 4.04 Subloan Size. It is expected that the typical individual subloan size would be about P 250,000 (about US$9 thousand), with a minimum of P 50,000 (about US$2 thousand) and a maximum of P 1,000,000 (about US$36 thousand). Loans could be short term (for working capital) or medium term (for capital investment), with the proportion medium term/short term defined by each MGA on the basis of its members' needs. Based on prevailing practice, it is likely that at the outset of the program, about 60 percent of outstanding loans would be short-term working capital and about 40 percent longer term for fixed assets. The longer term share is expected to increase as the program progresses and the MGAs gain experience in allocating access to guarantees among short and longer term users. The terms of the subloans would not exceed 5 years. The minimum equity participation by the subborrower for investment projects would be 20% of total project cost.  - 16 - Technical Assistance. Technical assistance (TA) would be essential ,lement the project. TA would be channeled through DBP to a central MGA zment Unit responsible for developing and ovetseeing the program. The TA would include the following tasks: - establishing the MGAs. This would involve: identifying enterprises that could serve as core members for proposed MGAs; working with these core members to help them structure the MGAs and determine, for example, the initial equity required from each member, minimum and maximum subloan sizes, the rules for allocating guarantees among members, the penalties for delinquent or defaulted loans, and so forth. It would involve helping the members to incorporate their MGA -- assisting with legal documents and registration with the Philippine Securities and Exchange Commission, and drafting contracts for use with banks; - promoting the program among potential women entrepreneurs and providing them training in banking requirements and procedures (para 4.10); - training the executive and guarantee committee members of the MGAs. This would comprise training on: a) their specific duties and responsibilities; b) rules and regulations of the program; c) financial and accounting procedures and record keeping; and d) bank procedures and negotiations; - training staff of the PFIs in the operation and mechanics of the program; - providing technical advice and assistance on waste minimization and waste treatment for firms in threrA subsectors where pollution might be a significant factor -- electroplating, leather tanning, and food processing; - assisting the MGAs to resolve ad hoc problems as they arise during the initial period of operation; and - providing overall project monitoring, auditing and co- ordination. To accomplish these tasks the TA component would include expert foreign ocal consultants, a small amount of computer, communications and )ortation equipment for the Management Unit, and training expenses. About iths of foreign consultants would be required to provide overall management Unit over 4 years on a declining basis and to organize and structure the rnmental assistance; 226 months of local consultants would be needed to le legal and auditing assistance, training, promotion and development, rision and co-ordination of the program, and technical advice to selected  - 17 - firms on environmental matters. TORs for the technical assistance were agreed during negotiations. Copies are provided in the Project File. 4.07 The MGA Management Unit would be provided external assistance for four years after World Bank loan effectiveness. After that, it is expected that the program will be well established and that the flow of funds with'n the program should provide for funding the Unit on an ongoing basis. Cost estimates for the technical assistance provided under the Project are presented in Annex 7. 08 Guarantee Fund. The Guarantee Fund (GF) comprises the funds used by tt,e MGAs to guarantee loans to their members. The GFs would be establi.s"ed with initial member subscriptions (averaging about P 20,000) and supplemented with a matching loan fund (para 3.07). DBP would extend the matching loans as well as administer the corresponding accounts The MGAs would earn interest on their own funds but would neither pay interest on the matching loan nor earn interest on the corresponding deposit. They would keep the matching loan without repayment for an initial grace period of five years, after which they would repay the principal of the matching loan over a fifteen year period. The GF would earn operating income from members' annual fees and guarantee fees and from interest income, and would reimburse the retail banks for the guaranteed portion of loan defaults. Significantly, the GF wouild continue to grow even after DBP stops making matching loans for new member subscriptions (in 1994). The base case assumptions and parameters are discussed in Annex 6. C. Benefi.ciaries and Subloan Size 4.09 Beneficiaries. The target beneficiaries of the MGA scheme would be enterprises belonging to the cottage industry sector. The typical firm would have assets of about P 350,000 (aLout US$15,000) and about 10-20 employees. However, in order not to constrain the growth of MGA enterprises or to force premature exit from the program, firms with assets up to P 3 million (US$107,000) and full time employees of not more than 50 would be allowed to participate in the program. This size of firm roughly coincides with the lower end of firms financed under the Industrial Guarantee and Loan Fund (IGLF), the major source of financing for small and medium firms. The minimum size firm allowed under the Program would be firms with assets (before financing) of P 50,000 (about US$2 thousanu). The firms may be involved in any commercial activity in industry or services, including agribusiness but excluding agriculture. 4.10 Access by Women. The proposed Project would help to meet a need for credit by women entrepreneurs in the Philippines. A Bank Report (Philippines: Women in Development: Issues for the World Bank; August 23, 1989) indicated that women in the Philippines have a more difficult time than men getting credit because of the (albeit subtle) superiority of husbands' rights as they pertain to family held-property. Since the proposed Project is specifically designed to provide a collateral substitute to collateral-weak firms, it would help to alleviate this important constraint on women entrepreneurs' access to credit. This tends to be borne out by the relatively high female involvement in the initial 1000 cottage firms participating in the Project. Among these firms, 53X of the labor force is female. This compares with overall female participation  - 18 - in the labor force of 47% in 1988. More significantly, 43% of these cottage firms are headed by women either as owners or managers. While there are no overall data on the national average of women entrepreneurs, the 43% share is clearly high by any international standard. To ensure that this unique collateral substitute feature of the Project is used fully to the advantage of women entrepreneurs, under the Project the MGA Management Unit will make specific efforts to promote the program to women owners and managers of cottage enterprises. D. Project Implementation 4.11 Overall Structure. The Project would be managed and funded through a two-tier structure. The loans would flow through DBP to accredited retail banks, and then to MGA member firms. The Matching Fund loans would be advanced and administered in trust by DBP. The overall program would be governed by a Review Committee with members from DTI, DBP, DOF, NEDA and the retail banks. Reporting to the Review Committee would be the MGA Management Unit, which would oversee the MGA portion of the Project. 4.12 The structure of the Project, the legal arrangements, and the guidelines for project operation and management are included in policy manuals and legal documents which were agreed at negotiations. Chart 4.1 shows the institutional structure of the Project. The legal documents comprise the following: Operating Policy Guidelines outline the relationship of DBP with the PFIs and the MGAs under the program. The Policy Manual and Licensing Agreements govern the relationship between the MGA Management Unit and the MGAs. Legal agreements under which the MGAs would be incorporated, governed, and registered with the SEC are the Articles of Incorporation, the By-laws, and the SEC Registration, respectively. Legal agreements regulating the relationship between MGAs and their selected retail banks are the Hemorandum of Understanding, with the underlying Trust and Guaranty Agreement (spelling out the agreements regarding the deposit of MGA funds in the retail bank, the issuance of the guaranty by the MGA to the retail bank, the procedure for the retail bank to call the guaranty, and the actual text of the guaranty). Subsidiary Loan Agreements between DBP and the PFIs would govern refinancing arrangements and Memoranda of Agreement between DBP and the MGAs would regulate Matching Loan Funds.  - 19 - Chort 4.1 CEFP INSTITUTIONAL FRAMEWORK ~~~~~~~~~~~~~~~/e REVIEW commirrEe T __'i OTI I ~ 1 ': -DBP Staff S ..i....Staff I~~~~~ MANAQElfM r..