ASIAN DEVELOPMENT BANK PPA: PHI 24112

Transcription

1 ASIAN DEVELOPMENT BANK PPA: PHI PROJECT PERFORMANCE AUDIT REPORT ON THE THIRD DEVELOPMENT BANK OF THE PHILIPPINES (Loan 1088-PHI) I N THE PHILIPPINES December 1999

2 CURRENCY EQUIVALENTS Currency Unit Peso (P) At Appraisal At Project Completion At Operations Evaluation (15 June 1991) (30 April 1996) (21 May 1999) P1.00 = $0.036 $ $ $1.00 = P P P38.00 ABBREVIATIONS ADB Asian Development Bank BSP? Bangko Sentral ng Pilipinas (Central Bank of the Government of the Philippines) DBP? Development Bank of the Philippines DENR? Department of Environment and Natural Resources DTI? Department of Trade and Industry ECC? environmental compliance certificate EIRR? economic internal rate of return IGLF? Industrial Guarantee Loan Fund ISO? Internationale Standards Organisation (International Organization for Standardization) J-EXIM? Export-Import Bank of Japan KfW? Kreditanstalt für Wideraufbau NCR? National Capital Region OECF? Overseas Economic Cooperation Fund PCR? project completion report PFI? participating financial institution PPAR? project performance audit report SME? small and medium enterprise TA? technical assistance NOTES (i) (ii) The fiscal year (FY) of the Government ends on 31 December. In this report, $ refers to US dollars.

3 2 Operations Evaluation Office, PE-5 3 9

4 CONTENTS P a g e BASIC DATA EXECUTIVE SUMMARY ii iii I. BACKGROUND 1 A. Rationale 1 B. Formulation 1 C. Objectives and Scope at Appraisal 2 D. Financing Arrangements 2 E. Completion 3 F. Operations Evaluation 3 II. IMPLEMENTATION PERFORMANCE 4 A. Design 4 B. Loan Utilization 6 C. Subproject Characteristics 7 D. Modality of Lending 7 E. Monitoring and Supervision 8 F. Technical Assistance 9 G. Compliance with Loan Covenants 10 III. PROJECT RESULTS 11 A. Performance of Subprojects 11 B. Socioeconomic and Sociocultural Impact 11 C. Environmental Impacts and Control 12 D. Operational Performance 13 E. Institutional Development 14 F. Financial Performance 14 G. Sustainability 15 IV. KEY ISSUES FOR THE FUTURE 16 A. Constraints in SME Development 16 B. DBP s Future Role in SME Lending 17 C. Commitment of Financial Institutions to SME Lending 17 V. CONCLUSIONS 18 A. Overall Assessment 18 B. Lessons Learned 18 C. Follow-Up Actions 19 APPENDIXES 20

5 BASIC DATA Third Development Bank of the Philippines Project (Loan 1088-PHI) PROJECT PREPARATION/INSTITUTION BUILDING TA No PHI TA Project Name Institutional Strengthening of Financial Institutions Type Person-M o n t h s Amount Approval Date A&O 3 $115, Jul 1991 KEY PROJECT DATA ($ million) As per ADB Actual Loan Documents ADB Loan Amount/Utilization ADB Loan Amount/Cancellation KEY DATES Expected Actual Fact-Finding 5 Sep Sep 1990 Appraisal 26 Mar Mar 1991 Loan Negotiations Jun Jun 1991 Board Approval 16 Jul Jul 1991 Loan Agreement 24 Sep Sep 1991 Loan Effectiveness 23 Dec Dec 1991 Commitment Date 23 Dec Dec 1994 Project Completion 18 Mar 1996 Loan Closing 23 Dec Mar 1996 Loan Extension 23 Dec 1996 Months (effectiveness to completion) 51 BORROWER GUARANTOR Development Bank of the Philippines Republic of the Philippines EXECUTING AGENCY Development Bank of the Philippines MISSION DATA Type of Mission M i s s i o n s (no.) Person-Days (no.)

6 2 Fact-Finding and Preappraisal 1 50 Appraisal and Loan Negotiations 1 40 Project Administration Review 1 16 Disbursement 1 2 Project Completion 1 18 Operations Evaluation 1 35 A&O = advisory and operational, ADB = Asian Development Bank, TA = technical assistance.

7 EXECUTIVE SUMMARY The Third Development Bank of the Philippines (DBP) Project was intended to stimulate private sector growth in small and medium enterprises (SMEs). The project objectives were to channel long-term industrial credit to SMEs and strengthen the development finance activities of participating financial institutions (PFIs) by upgrading their institutional capabilities. These objectives were a means to achieve the goals of generating low-cost employment, promoting regional dispersal of industries, and generating foreign exchange savings. These goals coincide with the focus of the Asian Development Bank (ADB) to reduce poverty, with the lessons learned in the Project relevant to current efforts. The Project comprised a loan for $100 million to DBP under the guarantee of the Republic of the Philippines and an advisory technical assistance (TA) grant for $115,000 to train PFI staff to better evaluate viable SMEs. With the flexible definition of SMEs adopted, at least half of the loan proceeds were to be channeled only to SMEs and the remainder to much larger enterprises. Due to poor utilization (42 percent of the total loan amount), the loan closing was extended by one year but then canceled three months later without further disbursement. At loan cancellation, only about 6 percent had been used from the portion allocated exclusively for SMEs. The subprojects approved were mostly in the National Capital Region and southern Luzon. Most of the subprojects were larger enterprises, some were market leaders, and a few were SMEs. Because of the fundamental problems in project design, the Project did not fully achieve its objectives of promoting rural SMEs and generating low-cost employment as envisaged at appraisal. The main design weaknesses were the inappropriate assessment of the nature and extent of market demand; duality of the loan, which catered to both SME and larger clients; noncompetitive nature of the loan compared with the parallel facilities available; and lack of SME financing skills in the PFIs. Supervision by ADB and DBP was weak in monitoring the goals and developmental impact that the Project was expected to achieve. PFIs assessed and monitored subborrowers based on their creditworthiness and repayment capability and were vigilant during implementation about repayments. The economic crisis that began in 1997 generally slowed the performance of the subprojects in terms of sales, capacity use, collections, and profits. Yet the subprojects repaid the subloans satisfactorily. Being large and experienced, the subprojects used funds from other sources to service the subloans. The Project achieved, to a limited extent, its objectives of creating jobs (although not at a low cost) and financing a few export generating or export substitution activities. The Project is rated partly successful because of its mixed performance. The TA accompanying the loan was expected to train PFI staff to assess eligible and viable SMEs. However, the efforts would have been more effective if the trained staff members had been in place and performing prior to the credit influx. The TA is also rated partly successful because it created an opportunity for DBP officers to train. However, the TA did not create a sustainable training program to enhance the skills of the PFI staff members, especially those in rural areas, as intended.

8 2 Access to credit (both working capital and investment capital) rather than the availability of credit is one of the constraints to SME development. The other constraints to SME development are lack of (i) information flow on technologies, markets, business linkages, and government policies; and (ii) management and technical skills. In designing a project that addresses only one of the constraints for SME development, ADB should, through a policy dialogue with the Government, ensure that other constraints are also being adequately addressed. The lessons learned from the Project highlight the importance of (i) requiring more rigorous subsector analysis and sound project formulation, particularly in assessing the extent and nature of demand for credit by the subborrowers that are expected to be financed; (ii) ensuring that appropriate banking skills are in place and institutional capacities are established through adequate monitoring of related TA prior to extending substantial credit; (iii) obtaining a clear understanding of the other constraints to SME development and addressing them through policy dialogue with the Government; and (iv) establishing appropriate indicators at the design stage to evaluate whether the Project is achieving the intended results.

9 I. BACKGROUND A. Rationale 1. The Third Development Bank of the Philippines (DBP) Project 1 was intended to stimulate private sector growth in small and medium enterprises (SMEs), a goal of the Government. SME development was to be used as a vehicle to create labor-intensive jobs, develop general and entrepreneurial skills, promote regional dispersal of industries, and reduce income inequality. The Asian Development Bank (ADB) loan was to provide medium- to long-term industrial credit to SMEs to achieve these goals. 2. DBP s charter was revised in as part of the effort to rehabilitate the bank from a state of technical bankruptcy. Under the rehabilitation plan, DBP s focus changed in 1989 from a retail lending to wholesaling long-term credit to private enterprises through a network of participating financial institutions (PFIs). The ADB loan for the Project in 1991 reflected ADB s support for DBP s new role. The loan also expanded ADB s previous assistance directed at SMEs under the Government s Industrial Guarantee Loan Fund (IGLF), which was being successfully disbursed. 3 B. Formulation 3. In May 1990, DBP s request for a loan of $100 million to meet the strong demand for investment funds from large industrial enterprises was endorsed by the Government and forwarded to ADB. Subsequent to the Fact-Finding Mission, ADB staff recommended a $100 million loan to DBP and an attached technical assistance (TA) of $300,000. In April 1991, in accordance with management guidelines, 4 the Appraisal Mission modified the scope of the loan to on-lending mainly to SMEs. As the Government s definition of SMEs was deemed restrictive, the Appraisal Mission adopted a flexible definition (para. 6) of SMEs to cover all important subsectors while, at the same time, aiming to leave out large projects in individual subsectors. 5 The appraisal report noted that lack of long-term funds constrained SME growth. The report cited the limited term lending activities among commercial banks and showed the 1 Loan 1088-PHI: Third Development Bank of the Philippines Project, for $100 million; and TA 1535-PHI: Institutional Strengthening of Financial Intermediaries, for $115,000, both approved on 16 July The TA and the Project became effective on September 1991 and December 1991, respectively. 2 By the Government s Executive Order 81 of 3 December Loans 944-PHI and 945-PHI(SF): Small and Medium Industry Project, for $100 million, approved in January 1989, contributed to the IGLF that was previously managed by the Central Bank and handed over to DBP in Management Review Meeting minutes of 21 February 1991 indicate that ADB s operational strategy for the Philippines clearly supported SMEs and not large industries, therefore, ADB decided that the loan would concentrate on a line of credit for SMEs. From ADB s perspective, supporting industrial development was not a priority area for Philippines, but poverty alleviation and balanced regional development were. 5 This enabled enterprises more than 15 times the legal definition of SMEs (P0.5 million-p20 million) to be eligible as well. Some Board members were not convinced that such an increase in the upper limit was needed and believed that the enterprises excluded by this definition, especially outside of Metro Manila, would be negligible.

10 2 importance of DBP s wholesale financing window to channel foreign savings to the private sector, especially SMEs. While DBP s resource gap for wholesale operations was geared toward term lending for enterprises of all sizes and was projected at more than $300 million in , the appraisal report provided no information on the projected demand for long-term funds by SMEs alone. 4. The Appraisal Mission determined that two of the three TAs considered during factfinding were no longer needed. These TAs were for (i) providing consulting services for privatization of the retail entity of DBP, 6 and (ii) financing feasibility studies for consolidation of credit guarantee schemes for SMEs. 7 However, the Appraisal Mission confirmed the need for a TA of $180,000 to train PFIs staff members in project appraisal and monitoring. 8 At approval, the TA amount was reduced, with ADB financing $115,000 to mainly meet the foreign exchange cost, with DBP and the Government financing of $65,000. The loan became effective in December C. Objectives and Scope at Appraisal 5. The objectives of the Project were twofold: (i) support private sector SMEs across a broad spectrum of industries by augmenting the flow of long-term industrial credit, and (ii) strengthen the development finance activities of PFIs and upgrade their institutional capabilities. According to the appraisal report, the Project was to promote SMEs, which were mostly labor intensive and spread across the country. As such, the Project aimed to generate employment opportunities and promote development in urban and rural areas. By blending their funds with those available under the Project, the PFIs were expected to increase at prudent levels the maturity period of their loans. The TA was considered essential to strengthen the PFIs institutional capabilities in project appraisal and loan monitoring. 6. The Government defined SMEs as enterprises with assets of P0.5 million-p20 million. This definition was considered restrictive to disseminate the benefits of the Project to the important industrial subsectors. Therefore, the Project was to channel at least half of the proceeds only to SMEs with a prefinancing asset size of not more than P20 million. In addition, by using a flexible definition, enterprises with a prefinancing asset size of up to $12 million (P333 million) were also eligible for financing under the remaining half of the loan proceeds. To improve use of the loan during implementation, DBP sought ADB s approval in October 1993 to include subborrowers with assets in excess of $12 million, but ADB did not approve this request. D. Financing Arrangements 6 The World Bank was reviewing the privatization plan, and ADB left related work in this area to the World Bank. 7 The Magna Carta for SMEs (Republic Act 6977), which was approved in January 1991 with the aim of rationalizing Government assistance to SMEs, addressed this issue (Appendix 27). 8 The memorandum of understanding of the Appraisal Mission notes that through TA 1113-PHI: Training for Staff of Participating Financial Institutions, for $50,000, approved on 19 January 1989, project monitoring skills of the accredited PFI staff in Metro Manila were upgraded, but an urgent need remained to train the staff of rural branches in project appraisal, monitoring, and DBP s procedures for wholesale lending.

11 3 7. On 16 July 1991, ADB approved a loan for $100 million from ADB s ordinary capital resources to DBP (the Borrower) under the guarantee of the Republic of the Philippines. DBP agreed to provide a repayment guarantee fee of 1 percent per annum to the Government. The amortization period was 15 years, including a grace period of 3 years. The Government also assumed the foreign exchange risk for which DBP paid a fee based on the loan utilization amount and a preagreed formula. 9 The loan was complemented by an advisory TA grant for strengthening the development financing activities of the PFIs, particularly their rural branches (footnote 1). 8. DBP was to relend loan proceeds to the PFIs against their clients applications for financing, which would include a description and summary analysis of the subprojects. The PFIs onlent the ADB funds to finance the direct and indirect 10 foreign exchange costs of development projects undertaken by SMEs for (i) establishing new production facilities; and (ii) modernizing, expanding, or upgrading existing facilities. The loan covenants required that the PFIs finance at least 10 percent of the cost of each subproject from their own resources. Both DBP and the PFIs were granted a subloan free limit of $4 million for any single subproject. 9. DBP s subloans to the PFIs were at market-based variable rates and fixed interest rates that were set quarterly or semiannually based on a formula agreed to with the Department of Finance and ADB. 11 The PFIs onlent the proceeds on fixed and variable rates at existing market rates to private sector subborrowers. Subborrowers and the PFIs had the option to convert the fixed rate to a variable rate or vice versa once during the life of the subloan. The PFIs would determine the repayment schedule within the maximum amortization period of 12 years, considering the projected cash flow of subprojects. E. Completion 10. The loan was closed on March 1996 after one extension, three months later than the original schedule. The loan disbursed $42 million (42 percent utilization), and the remaining $58 million was canceled. The project completion report (PCR) was circulated on 20 March The report concluded that ADB s loan was only partly successful in meeting its objective due to poor loan utilization. The PCR attributed the low loan utilization to the noncompetitiveness of ADB s loan with respect to its lending terms when compared with other facilities, such as those of the World Bank and Overseas Economic Cooperation Fund (OECF). DBP also shared this opinion during and after loan implementation. 11. The PCR indicated that works under 30 subloans (Appendix 1) were completed without significant cost overruns and implementation delays, but the report noted the difficulty of monitoring the data on subloans from the PFIs in a timely manner. The PCR found DBP s 9 The Government made a policy decision to absorb foreign exchange risk to avoid making the institutions like DBP vulnerable to foreign exchange movements. Since the disbursements under the Project was made mainly between , they were not subject to foreign exchange movements resulting from the financial crisis. 10 Direct foreign exchange costs comprised 100 percent of the cost of imported machinery, technology, and services, and the imported component in the initial/additional working capital. Indirect foreign exchange costs comprised up to 60 percent of the cost of locally procured machinery. 11 The formula reflected the weighted average interest rate of 91-day treasury bills for variable rate loans plus a premium. DBP s profit margin took into account the cost of funds, administrative costs, and the credit risk involved.

12 4 overall financial position to be satisfactory but indicated that DBP s provision for bad debts was relatively low and required management s attention. This cautionary note may have been recommended due to DBP s experiences. But given DBP s actual wholesale lending repayment experience, the bad debt provision did not appear to be a major concern (para. 61). The PCR noted compliance with all financial covenants, except the covenant that PFIs contribute at least 10 percent of their own resources to the subproject cost. However, ADB had changed this covenant to 10 percent participation of total debt financing in August Although the PCR analyzed the causes for the low utilization of the loan, the report did not address several other issues fundamental to the objective of the loan, including the (i) paucity of small clients among the subborrowers, (ii) DBP s and ADB s lack of monitoring for development impact, and (iii) effectiveness of the TA grant for institutional strengthening. F. Operations Evaluation 13. This project performance audit report (PPAR) evaluates the Project s effectiveness in meeting its objectives. The key issues examined are the (i) appropriateness of the project design; (ii) development impact and sustainability of subprojects; (iii) impact of the economic crisis on DBP, PFIs, and subproject performance; (iv) constraints to SME development; (v) disposition of PFIs to SME lending; and (vi) DBP and its future role in SME lending. The PPAR conclusions are based on the findings of the Operations Evaluation Mission (the Mission) conducted in May 1999, review of ADB documents and files, review of records of DBP and the PFIs, and discussions with ADB staff. The draft PPAR was provided to DBP, Government, and ADB staff concerned for review and comments. 14. Subprojects analysis is partly based on the results of the survey completed by 20 subborrowers. The Mission found it difficult to obtain the survey data from some subprojects, particularly from those that had repaid their subloans. The Mission aimed to visit at least one third of the total PFIs and subprojects, representing several industry categories, prefinancing asset sizes, and dispersed geographical locations. The final site visits depended on the availability of subproject owners to meet with the Mission, with 10 of the 27 subprojects visited (Appendix 2). The Mission also visited 8 of the 21 PFIs that approved subloans to subprojects and a PFI that did not avail of the DBP loan although accredited (Appendix 3).

13 II. IMPLEMENTATION PERFORMANCE A. Design 15. There were several weaknesses in the project design that contributed to the poor use of the funds. 16. First, the objective of the Project and the assessment of demand were mismatched. DBP initiated the loan requesting long-term funds for large enterprises. Because ADB s operational strategy in the Philippines mainly supported SMEs, the focus of the Project was diverted to SME lending without (i) properly estimating the extent and type (fixed or working capital) of credit needed by the SMEs, (ii) evaluating the appropriateness of the lending terms for SMEs, and (iii) reviewing alternative lending programs available in the market. Given the short-term nature of the deposit liabilities in the Philippines financial sector, the demand for long-term funds was evident; but this demand may have been from medium and large enterprises. If the appraisal had been more rigorous in analyzing subproject potential demand, the nature of the actual credit gap for SMEs may have been foreseen. DBP s subsequent request to remove the size limit (para. 6), indicating that the demand for credit was from larger subborrowers, also suggests that insufficient analysis was done on the projected demand for credit by SMEs at project appraisal Second, the lending criteria and terms and conditions of the loan were not carefully reviewed against alternative lending programs in the market. The ADB loan was more costly, restrictive (i.e., shorter predisbursements period, working capital restrictions), and cumbersome than the other facilities available (Appendix 4). The OECF terms were more competitive for interest rates and the World Bank had more flexible and lower commitment fees (Appendix 5). 13 DBP s pass-on (relending) rates to the PFIs indicate that the OECF rate was lower by 2 percentage points than the ADB rate (Appendix 6). The eligible expenditure for the ADB loan was more restrictive than that of alternative credit lines. The Project covered only imported working capital, whereas the World Bank s, IGLF, and OECF credit lines covered imported and local working capital. 14 Unlike the World Bank s, IGLF facility, which could guarantee a PFI s credit risk, no guarantee arrangement was attached to the ADB loan. IGLF did not require an environmental compliance certificate (ECC), which was cumbersome and costly to obtain, or receipts for disbursement. 18. Third, the implementation of the TA (footnote 1) for DBP and PFI staff members to improve their aggressiveness in SME lending and conduct of credit analysis of subprojects was 12 In February 1994, ADB advised DBP that it agreed to raise the preinvestment asset size from P20 million to P40 million to keep in line with the Government s definition of SMEs. 13 During , when bulk of the disbursements were made, the World Bank interest rate and commitment fee together was less costly than ADB. Besides, World Bank had more attractive conditions for subborrowers such as working capital finance and waiver of the environmental compliance certificate. 14 PPAR findings from the Small and Medium Industry Project (footnote 3) also suggest that SMEs seek long-term loans to finance permanent working capital in addition to fixed assets.

14 2 done in parallel in with the project loan implementation. Credit analysis skills should have been in place and tested before the large credit influx. Had the intended capacity building efforts (especially in rural branches) been carried out under the TA prior to the Project, the results would have been better. 15 In addition, due to the funding constraints, the coverage of the rural branches under the TA was far more limited than expected (para. 37). 19. Fourth, the design fell short of meeting the Project s objectives. The subprojects financed and the overall project benefits cited at appraisal were never explicitly linked. The subproject eligibility criteria did not specify minimum requirements to be met to ensure their contributions to job creation, export promotion, or foreign exchange savings. Specific sectors were not targeted to promote employment generation activities, and rural development was not promoted. 16 This too indicates that the objectives were not carefully considered during the design stage. 17 Because financing SMEs was a means to achieve other developmental goals, building monitorable indicators (such as employment generation, foreign exchange savings, and location) into the project design and arranging DBP to collect data on their indicators would have been useful. According to the Subsidiary Loan Agreement, DBP could collect data only through PFIs. This could have been addressed at project design stage. 20. Fifth, the channel of distribution was not adequately assessed. Generally, accredited commercial financial institutions (with their profit objectives) may not be the proper vehicles to carry out a developmental mandate. Although expected at appraisal, the requirement for PFIs to contribute 10 percent of the project cost did not motivate them to channel their own funds into development financing. In an environment where their liabilities are predominantly short term, the PFIs used funds obtained through DBP or other sources, such as Land Bank or the Social Security System, for long-term lending. Of the 50 financial institutions DBP accredited at appraisal, only 21 considered subloans under the Project. Some of these PFIs were more aggressive in SME lending, and they have created a niche in the SME market. They use informal networks and training programs for entrepreneurs to seek potentially viable SMEs. Channeling the loan through more active PFIs or using DBP s retail window to lend directly to subprojects may have been effective. But according to the Loan Agreement, the retail window was not eligible to use this line of credit. DBP may have had better results if it had concentrated on building the capacity of only the PFIs that were enthusiastic to serve SME clients. 21. At the Board meeting in July 1991, Board members raised several reservations about the project design. Some argued that ADB should not be limiting itself to addressing only the credit problems of SMEs and that ADB should initiate a policy dialogue on issues constraining the development of SMEs. The Board members expressed that a smaller loan, accompanied by relevant policy dialogue and conditionalities, would have been more appropriate. 18 Some 15 As per DBP, the scheduling of the TA was determined by ADB. 16 During the Staff Review Committee meeting, ADB decided against any regional focus. The Board argued that loan proceeds should be more equitably spread across the country with priority to low-income provincial areas. The project staff responded to Board queries prior to approval and indicated that the IGLF had fairly balanced subsectoral dispersion and that commitments to the National Capital Region were slightly more than 70 percent. Taking this as an indication of the future direction of DBP wholesale lending to SMEs, the staff maintained that sector or geographical guidelines were not required. 17 While the subloan eligibility criteria as covenanted in the Loan Agreement were based on the potential contribution for developmental impact, no minimum targets were specified. Other eligibility criteria were minimum rates of return on investments (i.e., financial internal rate of return and economic internal rate of return), specified leverage and debt service coverage ratios, and compliance with environmental rules and regulations in the Philippines. 18 Staff informed the Board that policy reforms in the financial sector were being addressed by the World Bank s

15 3 members commented that ADB s priorities were not emphasized enough and that the eligibility criteria seemed loose. 19 Concern was also expressed whether the Project would promote balanced regional development, and some Board members opined that loan proceeds should be more equitably spread across the country with special priority to creating jobs in low-income provinces. 20 Another suggestion was to carefully analyze how to directly use commercially oriented intermediaries instead of the standard development finance approach. Finally, some Board members believed that the TA to strengthen the PFIs should have preceded the loan. In hindsight, most of these concerns were valid, and not addressing them at the design stage weakened the loan s chances of achieving its objectives. B. Loan Utilization 22. The loan utilization was low at 42 percent, especially against a backdrop of economic recovery in the Philippines and the full use of the ADB loan for the Small and Medium Industry Project (footnote 3). The Project was the third consecutive ADB loan to DBP that failed to be utilized effectively. 21 In contrast, the credit lines extended to DBP by the World Bank and OECF during the same period were successfully utilized (PCR, para. 34). 23. The ADB loan to DBP for $100 million was signed in September 1991 and became effective in late December The closing date for commitment was scheduled for December 1993 and that for disbursement was scheduled for December At the end of December 1993, midway through the implementation schedule envisaged at appraisal, only $33.5 million (less than 34 percent of the loan amount) had been disbursed (Appendix 7). 24. According to DBP, the reasons for the slow movement of the loan were the (i) availability of financing facilities with more attractive features, and (ii) preference of the PFIs to minimize their credit risks by financing only reputable corporations with assets of more than P300 million. To improve the marketability of the ADB loan, DBP suggested several measures and requested ADB s approval for the following changes: (i) expand the scope to include borrowers with preinvestment asset size in excess of $12 million equivalent, (ii) align the loan features and conditionalities with those of existing wholesale programs for large-scale enterprises, and (iii) extend the commitment and loan closing dates by two years. In December 1993, ADB indicated that it would consider extending the loan commitment and closing dates, but informed DBP that the terms and conditions of the loan could not be changed without Board approval. Subsequently, loan commitment dates and closing dates were extended by one year. No additional disbursements occurred during the extension period, which lasted for three months and ended with the loan s cancellation in March Financial Sector Adjustment Loan approved in May Four years later, ADB approved Loan 1363-PHI: Capital Market Development Program, for $150 million, approved on 22 August Some directors suggested that staff should look into the possibility of allocating 60 percent of the funds to rural subprojects in relatively high value-added and with the potential for high employment generation. 20 ADB staff indicated the lack of loan approval in rural areas was because the PFIs staff members were not trained. By augmenting their training under the TA, this problem was to be addressed. However, implementation problems occurred regarding this issue (para. 37). 21 The PPARs for the first and second DBP projects evaluated them to be partly successful because they failed to meet project objectives arising out of poor performance of subprojects, low utilization, and inadequate ADB supervision.

16 4 C. Subproject Characteristics 25. The number of SME loans financed under the Project confirms the inadequate evaluation of the demand for credit by SMEs. Twenty-one PFIs extended 37 loans to 30 subprojects, 22 of which only 3 were SMEs. 23 The SMEs accounted for about 6 percent of the $50 million allocated specifically for them. Two subprojects exceeded the eligible size limit of $12 million (about P333 million). 24 More than half of the loans were in excess of $1 million (more than P27 million) each, and these accounted for about 87 percent of actual loan utilization. Only one subloan exceeded the $4 million free limit approval, but finally used only P3.9 million. As expected, given a choice, most PFIs preferred to lend to larger subborrowers to avoid the perceived higher risk and the greater transaction costs of lending to SMEs. 26. A variety of industries benefited from the ADB loan, mostly through manufacturing and assembly subprojects. Several subloans were also used for purchasing vehicles (Appendix 8). About half the subprojects financed were in the National Capital Region (NCR), comprising 48 percent of the actual loan utilization. Seven were in southern Luzon, a few were in the rural areas (Bulacan, Pampanga, Ilocos Norte), and only two were outside of Luzon (Davao del Sur, Negros Occidental). Nine subprojects were new investments, while the rest were for expanding and modernizing existing operations. D. Modality of Lending 27. Channeling funds through DBP for relending to PFIs and on-lending to SMEs and other subborrowers was effective. This modality supported prudent lending in two ways. First, the modality allowed PFIs to select subprojects and exercise due diligence in lending because they bear the credit risk. Second, the modality gave importance to DBP s accreditation process (para. 31) in selecting PFIs as on-lenders of ADB funds. The Mission believes that DBP s accreditation process was effective. 25 The modality also allowed lending rates to be determined by the market rather than to be distorted through interest subsidies. 28. Despite the advantages of this lending modality, a few implementation problems occurred. The PFIs practicing aggressive SME lending generally reached their maximum credit limit with DBP. Therefore, the main constraint for DBP in this regard was its single borrower s limit (of about P3 billion) which, under the Central Bank s statutory requirement, is at 25 percent of its unpaid capital and surplus resources. DBP requested the Central Bank s approval to increase this limit, but without success. Alternatively, DBP and the Government were discussing increasing DBP s capitalization at the time of the Mission. 22 The 30 subprojects included 2 subprojects that had 2 loans in the same facility and 3 other subprojects that had multiple agreements with syndicate loans (Appendix 1). 23 This is based on the definition used by ADB for SMEs at the project commitment date, which was P0.5 million- P20 million until February 1994 and up to P40 million after that. 24 These were precleared with ADB. Their prefinancing asset sizes were P365 million and P458 million. 25 However, after the 1997 economic crisis, DBP s reporting requirements increased, much to the exasperation of the PFIs.

17 5 E. Monitoring and Supervision 29. Monitoring and supervision was done by (i) ADB on DBP and on the Project, (ii) DBP on the PFIs and the Project, and (iii) the PFIs on the subprojects. ADB s monitoring of DBP and the Project mostly focused on addressing requests by DBP to orient the loan to larger clients (para. 40). ADB s supervision of the Project, with respect to the Project s key objectives of providing financing to SMEs (especially those in rural areas), and monitoring the progress of the TA (footnote 1) was weak. There was no inception mission in 1991 to develop a strategy to focus on the goals ADB wanted to achieve. ADB s first review mission was not until October 1993, which was 22 months after loan effectiveness and just 2 months prior to the deadline for closing loan commitments. The PCR notes informal contacts were facilitated by the proximity of the two institutions. 30. While the noncompetitive features of the ADB loan were first mentioned in the phase 1 TA final report in 1992, 26 no action was taken. DBP informed ADB in October 1993 of the slow response on the loan and the reasons for it. ADB did not take a proactive stance in addressing the underlying causes as it would have necessitated changing the project scope with Board approval, indicating the original project design was flawed. Subsequently, DBP requested a cancellation of the Project in November 1995 to avoid the mounting commitment fees for a facility that was not being fully used by PFIs. Following the Government s concurrence, ADB closed the loan in March 1996 with a remaining balance of $58 million. At that point, 76 percent of the loan available for larger enterprises was used, but only about 6 percent of the portion allocated to SMEs was used. 31. DBP supervised the PFIs effectively. DBP s role in wholesale banking is highly regarded by the PFIs. DBP s Wholesale Banking Unit is structured to focus on accredited financial institutions, which allows DBP to carefully monitor them and focus on their specific needs. 27 The PFIs consider DBP to be technically competent and professional. DBP often assists the PFIs in computing economic internal rates of returns (EIRRs) and other technicalities. DBP s loan processing time, on average, improved from about three weeks at PCR to about two weeks at the time of the Mission. 32. The PFIs are generally driven by profit motivation and cannot be expected to pay much attention on their own to monitor the development impact of development financing. The PFIs generally monitor the financial viability and the repayment performance of subprojects through their banking activities and annual site visits while the subloan is outstanding. In doing so, the PFIs were quick to respond to their subborrowers needs and were concerned about the credit risk involved. Upon repayment of a subloan, a PFI s continues to interact with a subproject through the normal banking relationship. 33. DBP received financial reports on the subprojects while the subloans were outstanding. Financial viability and repayments were monitored closely. DBP generally visited a subproject 26 Final Report TA 1535-PHI phase I. October Submitted by SEA BMB Consultants, Inc. 27 Accreditation of financial institutions is generally based on their (i) records of profitability, (ii) ability to manage a sound portfolio, (iii) managerial skills, (iv) project analysis capability and experience in corporate finance, and (v) compliance with relevant laws and regulations.

18 6 once for end-user verification. DBP also required the PFIs to submit a subproject completion report six months after the final release of the subloan, indicating that DBP and the PFIs were vigilant about the financial viability of subprojects. After the submission of the subproject completion report, which provides some data on a subproject s development impact, neither DBP nor the PFI does any regular reporting or monitoring of the socioeconomic and developmental aspects of the subproject. ADB and DBP need to be more proactive in these efforts (para. 68). If at appraisal, development impacts are expected to be monitored, then mechanisms to monitor them should be carefully built into the project design. 34. During project processing (on DBP s request), the requirement for EIRR computation was waived for subloans below P40 million, 28 which effectively refers to SMEs. Although the EIRR is difficult to quantify, the waiver discouraged monitoring of the Project s socioeconomic impact with respect to SMEs. If the EIRR was too complicated to monitor, other development indicators could have been substituted instead. If socioeconomic impact needs to be evaluated and does not coincide with the commercial objectives of the PFIs, then DBP should be encouraged to lead this task. 29 F. T e c hnical Assistance 35. The associated TA (footnote 1) of $115,000 was granted for institutional strengthening of financial intermediaries aiming to promote active participation of the PFIs in SME lending and efficient delivery of project funds to rural areas. 30 The TA was to develop a comprehensive training program comprising a series of six seminars in Manila and a series of 24 seminars at various rural locations. The TA was designed in two phases. In the first phase, training materials were to be developed and resource persons trained for the next phase. The second phase aimed to conduct many training programs throughout the country. Although three personmonths of consulting time was allocated to do the needs analysis, design training materials, and conduct the trainers training program, DBP indicated that this period was insufficient to do a thorough analysis. 31 Twenty-five resource persons, of whom 15 were from DBP, were trained in October 1992 in Manila. ADB s TA funds provided the equipment necessary to implement phase Trained local professionals (funded by DBP), were to carry out phase 2, mainly in the rural branches. Using the course material and training design developed in phase 1, two sessions of the management workshop were conducted in 1993 in Baguio and Cagayan de Oro to train bank managers and PFI regional executives in strategic directions for project financing focused on SMEs. Seven out of the planned 30 modules of the project financing course were conducted in in various locations in and outside of Manila 32 to provide PFI technical 28 It was not clear why the waiver was done. Presumably, the PFIs requested this (through DBP) to eliminate the requirements for technical calculations for small subloans. DBP may have also agreed to this to conform with procedures on its other financing facilities, e.g., OECF. 29 The lesson that it is difficult for projects with lines of credit to achieve commercial and development objectives simultaneously was highlighted in the PPAR of the Small and Medium Industry Project, which was completed in 1996 (footnote 3). 30 TA 1113-PHI (footnote 8), implemented earlier, mostly benefited staff of the PFIs in the NCR. 31 The balance in the TA budget at closing on March 1993 was $52,760, largely due to the hiring of a Philippine consultant instead of an international consultant. 32 Status Report TA 1535-PHI phase 2. October Submitted by SEA BMB Consultants, Inc.

19 7 staff with an understanding of the project cycle and to enhance their skills in project financing analysis of SMEs. Unfortunately, due to a lack of DBP funding, the number of training programs conducted under phase 2 was too limited 33 and did not cover most of the rural branches as envisaged. 37. Most of the resource persons were trainers from phase 1. Overall, the feedback from participants was very satisfactory in terms of relevance of topics discussed and effectiveness of programs conducted. During training, a continuous need for the type of courses conducted was observed, especially in rural banks. At the time of the Mission, most of the PFI staff members trained under the ADB TA grant were no longer working for the PFI or had transferred to a different division. This training was not sustained, and the expected multiplier effect was not generated. 38. Most of the PFIs visited have specialized lending units to coordinate with DBP on programs to source development funds for the PFI s term lending activities and to scrutinize the subprojects. The technical staff members of the PFIs undergo on-the-job training on credit analysis for long-term lending and internal training in general banking operations. Employees of all of the PFIs visited expressed the need for classroom training, particularly on credit analysis and project monitoring, and recalled that since 1995, they had not been informed of such training (except for product briefings). The Mission learned that some PFIs would be willing to share the cost of such training. DBP might consider playing a more active role in organizing and conducting such training for the PFIs staff, especially in rural areas. G. Compliance with Loan Covenants 39. DBP generally complied with the terms and conditions stipulated in the Loan Agreement (Appendix 9). The requirements of a maximum long-term debt-equity ratio of 2 times; a debt service coverage ratio of at least 1.5 times; and the submission of an ECC, waiver, or letter from the Department of Environment and Natural Resources (DENR) were checked by DBP prior to release of a subloan. However, DBP experienced difficulties in obtaining timely information from the PFIs During the loan processing, some covenants were changed with ADB approval. The covenant that the PFIs contribute at least 10 percent of their own resources to the subproject cost was changed to 10 percent of total debt financing and was approved by ADB in August The subloans was disbursed to two subprojects that exceeded the prefinancing asset size limit of $12 million (about P333 million), but these were both precleared with ADB. 36 ADB 33 DBP later combined this training with another program under an OECF TA providing training to PFI (until 1994) and providing grant facilities for industrial associations to conduct training in management and marketing skills and to strengthen the technical capabilities of DBP s staff. 34 The Mission also had similar difficulties in gathering data and arranging site visits, particularly to subprojects that had fully repaid the subloans. 35 DBP indicated this would be consistent with the other wholesale programs. Except for two subprojects, all others complied with the revised requirement. Those two subprojects only had 9 percent participation from the PFI. DBP s records show that in this case, in lieu of the covenant, a minimum of 10 percent of the third ADB loan was allowed, referring to the total loan portfolio. 36 These were loans to two subprojects that had prefinancing asset sizes of $15 million and $18 million, respectively. ADB discouraged DBP from making similar requests in the future.

20 8 also waived in February 1994 the compliance with minimum EIRR for subprojects that had DBP financing of P40 million or less (para. 34). The subprojects with larger loans met the required 15 percent minimum estimated EIRR at appraisal. Neither the DBP nor the PCR subsequently calculated the actual EIRR. The Mission found it difficult to obtain the data for calculating the EIRR, which did not seem relevant because many factors (i.e., new investments and the economic crisis) besides the ADB loan had affected subproject performance since the influx of project funds in

21 9 III. PROJECT RESULTS A. Performance of Subprojects 41. Thirty subprojects were financed through the ADB loan (Appendix 1). These subprojects were generally completed without significant cost overruns and implementation delays. 37 Repayment experience was satisfactory, with 28 of the 30 subloans repaying on schedule. Only two subprojects experienced repayment difficulties. Three others had their loans rescheduled with their PFIs. 38 Three enterprises also experienced negative net profits during , but two made repayments by drawing on external resources (Appendix 10). This initiative indicates the financial strength of the enterprise owners and that they were not typical SMEs. The economic crisis of 1997 had some adverse effects on subproject sales, collections, profits, and capacity use (Appendixes 11-12). Appendix 13 summarizes the impact of the 1997 economic crisis on the subprojects and PFIs. 42. Subprojects financed under the Project created direct job opportunities for 1,340 workers at an average cost of P2 million per worker at the time of the PCR. 39 This increased to 1,968 jobs by the first quarter of 1999, stemming from the growth in all subprojects financed (Appendixes 14-15). Seven out of 30 subprojects were involved in exports resulting in an equivalent of about P1.6 billion in foreign exchange earnings/savings. As of the end of 1998, only two had substantial direct export sales. When compared with the other wholesale lending programs of DBP, the ADB loan has the highest ratio of foreign exchange earnings/savings per peso lent and the second highest ratio of annual taxes generated per peso released. However, the Industrial Investment Credit Project (3.54), IGLF (2.76), and Cottage Enterprise Finance Project (1.78) have better ratios in terms of number of jobs generated per million pesos released than under the Project (1.36). This high foreign exchange earning capability and tax generation ability and lower employment creation tendency may reflect the capital intensity of the larger enterprises financed by the Project despite its original intention of promoting SMEs. B. Socioeconomic and Sociocultural Impact 43. In addition to the direct socioeconomic impacts of the Project discussed elsewhere in the report, this section presents the indirect impacts induced as a result of the Project. Many subprojects visited provide indirect business opportunities and community service within their vicinities. 40 Some are poised for International Organization for Standardization (ISO) 9002 accreditation, which follows best practice concepts in management, and motivate their 37 The PCR indicated that three projects experienced cost overruns of more than 10 percent, and 14 subprojects were delayed by less than one year during implementation. 38 A comparison of percent time elapsed and percent repaid (Appendix 1) indicate wide gaps for three subprojects. 39 According to the appraisal report, capital labor ration of SMEs (P63,500 per worker) was half that of the entire manufacturing sector. The discrepancy arises because the subprojects financed were generally non-smes. 40 One subproject has established a training school within its company to train out-of-school youths in Laguna.