Sri Lanka - Fourth Small and Medium Industries Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 17272 IMPLEMENTATION COMPLETION REPORT SRI LANKA FOURTH SMALL AND MEDIUM INDUSTRIES PROJECT (CREDIT 2250-CE) January 7,1998 Finance and Private Sector Development Division South Asia Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENT Currency Unit - Sri Lanka Rupee (annual average) Rs. Per US $1.00 US$ Per Rs. 1.00 1990 40.063 0.0249 1991 41.372 0.0241 1992 43.830 0.0228 1993 48.322 0.0207 1994 49.415 0.0202 1995 51.252 0.0195 1996 55.271 0.0181 1997 (Jan-Aug) 58.371 0.0171 ABBREVIATIONS ADB - Asian Development Bank BOC - Bank of Ceylon CBSL - Central Bank of Sri Lanka CEA - Central Environmental Authority CIBSL Credit Information Bureau of Sri Lanka CISIR - Ceylon Institute of Scientific & Industrial Research CBOC Commercial Bank of Ceylon DFIs - Development Finance Institutions DFCC - Development Finance Corporation of Ceylon EDB - Export Development Board GOSL - Govemment of Sri Lanka HNB Hatton National Bank IDA International Development Association IPS - Industrial Policy Statement NDB - National Development Bank of Sri Lanka PB - Peoples Bank PCIs - Participating Credit Institutions PFDP - Private Finance Development Project RRDBs Regional Rural Development Banks SCBs State Commercial Banks SLSI - Sri Lanka Standardization Institute SMIs - Small and Medium Industries TA Technical Assistance TAF - Technical Assistance Fund FISCAL YEARS GOSL, Commercial Banks = January 1 to December 31 Development Finance Corporation of Ceylon (DFCC) = April 1 to March 31 National Development Bank of Sri Lanka (NDB) = January I to December 31 Vice President: Ms. Mieko Nishimizu Country Director - Sri Lanka: Mr. Roberto Bentjerodt Sector Manager: Ms. Marilou Uy Task Manager/Sr. Operations Analyst: Ms. Shideh Hadian  FOR OFFICIAL USE ONLY Imolementation Comrletion Report (Credit 2250-CE) Table of Contents Pase No. PREFACE EVALUATION SUMMARY ......................................... i PART I: PROJECT MIPLEMENTATION ASSESSMENT ..........................................1 A. Statement/Evaluation of Objectives .......................................... 1 B. Achievement of Objectives ..............................2 C. Major Factors Affecting the Project ............................7 D. Project Sustainability ..8.........................8 E. IDA's Performance ............................9 F. Borrower's Performance ............................ 0 G. Assessment of Outcome ...1........................ 1 H. Follow-Up/Future Operations ............................11 I. Main Lessons Learned ............................11 STATISTICAL TABLES 1. Summary of Assessments ............................ 13 2. Related Bank Loans/Credits ............................ 14 3. Project Timetable ............................ 15 4. Loan/Credit Disbursements: Cumulative Estimated and Actual 15 5. Key Indicators for Project Implementation 16 6. Key Indicators for Project Operation ..17 7. Studies Included in Project ..17 8 (A & B). Project Cost and Project Financing . . .18 9. Economic Costs and Benefits ..18 10. Status of Legal Covenants ..19 11. Compliance with Operational Manual Statements ..20 12. Bank Resources: Staff Inputs ..20 13. Bank Resources: Mssions ..21 STATISTICAL ANNEXES I. Total Subloan Approvals by PCIs - IDA and ADB Credit Lines .22 II. Annual Subloan Approvals by PCIs - IDA Credit Line .23 IU. Breakdown of Approved Subloans by Size - IDA and ADB Credit Lines .24 IV. Approved Subloans for New and Expansion Projects - IDA and ADB Credit Lines .24 V. Employment Generation - IDA and ADB Credit Lines .24 VI. Sectoral Distribution of Approved Subloans - IDA and ADB Credit Lines .25 VII. Geographical Distribution of Approved Subloans - IDA and ADB Credit Lines .26 VIII. PCIs' Loan Collection and Profitability Performance. 27 IX Allocation and Utilization of Technical Assistance Component .28 X Allocation and Utilization of Environmental Component .29 XI. Composition of Sectoral Growth to the Economy .30 APPENDIXES A. Implementation Assessment of Environmental Component .31 B. Aide Memoire of July 1997 ICR Mission .36 C. Borrower's Contribution to ICR .43 Map of Sri Lanka This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  I I  IMPLEMENTATION COMPLETION REPORT SRI LANKA FOURTH SMALL AND MEDIUM INDUSTRIES PROJECT (SMI-IV1 (CREDIT 2250-CE) Preface' 1. This is the Implementation Completion Report (ICR) for the Fourth Small and Medium Industries Project (Credit 2250-CE) in the amount of SDR 33.3 million (US$ 45 million equivalent). Of this amount, SDR 3.7 million (US$ 5 million equivalent) was the technical assistance component. The Credit was approved by IDA Board on May 28, 1991 and became effective on September 30, 1991. The original project closing (October 31, 1996) was extended for eight months to June 30, 1997 due to slow implementation of the credit component. Final disbursements took place on Sepiember 1 1,1997 at which time a balance of SDR 546,493.14 (about 1.6 percent of the original credit amount) was canceled. The Asian Development Bank cofinanced the project by providing a total of US$ 30 million for subprojectfinancing and an additional US$ 350,000 for technical assistance. 2. The ICR was prepared by Shideh Hadian, Finance and Private Sector Development Unit of South Asia Region and reviewed by Ms. Marilou Uy (Sector Manager). Assistance was provided by Ms. Kumi Kitamori, Environmental Unit of South Asia Region for the preparation of the environmental component (presented in Appendix A and Annex X). The Borrower contribution included as Appendix C and was prepared by the National Development Bank of Sri Lanka (the Apex for the project) on behalf of the Government. 3 . Preparation of this ICR began during the IDA's final supervision/completion mission in July 1997. It is based on Staff Appraisal Report, Loan Agreement, Supervision Reports and other material in the project file as well as discussions with the participating institutions in the project. The National Development Bank also contributed to the preparation of this ICR by providing the data on different components of the project.  I  FOURTH SMALL AND MEDIUM INDUSTRIES PROJECT (SMI-IV) (CREDIT 2250-CE) SRI LANKA Evaluation Summary1 1. Introduction - Six financial/industrial intermediation projects, totaling US$ 152.8 million, were approved by IDA for the development of small, medium, and large scale industries in Sri Lanka during the 11 years preceding this project. An adjustment credit of US$ 120 million to support the reform program to modernize/privatize public manufacturing enterprises was also approved one year before the SMI-IV project (Table 2). 2. Proiect Obiectives - SMI-IV project had two main components: (i) a credit component of US$ 40 million for providing credit to the SMI sector; and (ii) a technical assistance (TA) component of US$ 5 million for institutional and policy development and subproject implementation support. Asian Development Bank (ADB) co-financed the project by providing US$ 30 million for subproject financing and US$ 350,000 TA grant for related reform programs. Specifically, the project objectives were to: * Deepen IDA's Dialogue with Government of Sri Lanka (GOSL) on issues affecting the financial sector efficiency in particular to: (a) continue support for strengthening of the financial institutions' operations; and (b) improve implementation of debt recovery policies and procedures including support for the Credit Information Bureau of Sri Lanka (this was part of ADBs TA program). * Provide support to GOSL to: (a) improve Customs and Excise policies/procedures; (b) reorient export development policies/procedures; (c) remove constraints to establish new enterprises and formulate appropriate SMI development policies and programs (part of ADB funded programs); and (d) assist the Central Environmental Authority (CEA) and Participating Credit Institutions (PCIs) 2 to introduce sub- sectoral industry environmental impact assessment guidelines in the investment approval process. * Provide credit, through the banking system, for long-term capital investment in eligible productive enterprises. 3. The SMI-IV objectives were in line with the Bank's assistance strategy in Sri Lanka and were important and timely for the Government's priority reform programs. GOSL gave high priority to industrial development (in particular the SMI sector) in its Industrial Policy Statement of 1989 and requested IDA to advance the fourth SMI line of credit (originally planned for Fiscal Year 94). The size of SMI-IV credit line was geared to the term lending capacity of PCIs and estimated SMIs' demand for long- term investment funds. The risks identified at the time of Appraisal (political instability and possible ineffectiveness of institutions in administrating/implementing the credit component) did not materialize and the project was implemented without any major problem (paras. 2, 3, & 4). 4. Implementation ExDerience and Results of the Credit Comnonent - The credit component was implemented highly satisfactorily and achieved its development objectives. It was fully committed as of March 31, 1996 (against the original deadline of September 30, 1995). The delay in the utilization of the credit component was due to ineligibility of the two State-Owned Commercial Banks (SCBs), BOC and PB, to participate in the project one year after effectiveness. The 1991 international financial audits of the 1 Provides cross-references to paragraphs from Part I which provide more details. 2 See footnote number 1 in Part I (page 1).  i. SCBs revealed that these banks had inadequate capital base and could not meet the participation criteria under SMI-IV. This resulted in a substantial reduction of the commitment level in the period 1992-94. Other PCIs could not immediately fill the gap because of their smaller branch network and limited term lending capacity. Although, in later years, the eligible PCIs increased their lending activities, the credit component could not be fully committed on time. The international-financial audits of the two SCBs for 1993-1994 indicated that the state banks met the participation criteria under SMI-IV; hence, they were reinstated as PCIs by both IDA and ADB during the course of 1995. Per request of the Government (which was elected in late 1994 and continued its strong support for SMI sector) and in view of satisfactory utilization of credit component by the private banks and DFCC, the SCBs' efforts to once again become eligible to participate in the project, and the importance of SMI sector in Sri Lanka's economy, IDA extended the commitment deadline by six months to March 31, 1996 and the project closing by eight months to June 30, 1997 (paras. 6 & 22). 5. The IDA's SMI projects made significant contributions to integrating the SMI sector with the economy, generating large number of employment, and increasing production and exports in particular of non-traditional products (para. 7; also recent OED study on World Bank Support for Small and Medium Industries in Sri Lanka: An Impact Evaluation; June 1997). The credit funds (IDA + ADB) provided loans to about 6,000 small and medium industrial/services units and created aboi.t 43,000 jobs, with incremental fixed investment per job of about US$ 3,300. The comparable figures estimated at the Appraisal were 25,000 jobs and US$ 6,200 per job investment. The projects financed under SMI-IV were mostly in the labor intensive subsectors (textiles, food processing, garments, construction materials) with the average employment per job of about 7. In addition, with the inadequate technology involved, the employment number increased. The subprojects were mainly concentrated in Western and North Westem Provinces because of continuing ethnic disturbances in the Eastern and Northern Provinces (para. 8). Annexes I - VIII provide details of the utilization of the credit component. The PCIs greatly benefited from the project and are now committed to support this vital sector. The perfomiance of PCIs with respect to maintaining the eligibility criteria (profitability and loan recovery rate) was satisfactory (paras. 7 and 21). The Central Bank Credit Guarantee Scheme was not useful for many of the PCIs. The premium paid by PCIs was, in most cases, more than the expected benefits. The PCIs believe that this guarantee scheme should hence be made optional (para. 9). The market determined interest rate charged on the subloans was based on the Average Weighted Deposit Rate (para. 8). 6. Imnlementation Experience and Results of the TA component - Overall, TA component was implemented satisfactorily and achieved, for the most parts, its objectives. TA component was designed to support: (i) institutional development of PCIs; (ii) policy development in the areas of trade, non-bank financial sector (specifically insurance industry), and environment; and (iii) the SMIs' entrepreneurs in their subproject implementation (paras. 10 & 11). Institutional Development component aimed at assisting the PCIs/NDB build their capacity in project finance as well as supporting the SCBs in developing their operations and preparing them for possible privatization. The PCIs upgrading and staff training program was highly satisfactory and achieved its objectives (para. 12). The consultancies for operational review and strengthening the two SCBs' operations, which started under SMI-IlI in 1988 and continued under IDP-II & III projects, were completed in 1992 under SMI-IV. The studies/consultancies made a series of recommendation on many aspects of the banks' operations (credit policy, Managemnent Information System, organization structure, technology, employment policy, etc.) and were the SCBs' first exposure to techniques used in other countries. The SCBs implemented part of the reconmuendations and improved their operations in some areas., however, the weaknesses remained and restructuring of the SCBs has not yet been achieved (para. 13).  iii 7. The result of TA programs for the Policy Development was mixed and varied by implementing institutions and projects. Modemization of Department of Customs and Excise (DCE) declaration procedures was implemented satisfactorily and had a high impact on upgrading the institution's capacity and increasing revenue collection (Paras. 15 & 16). Allocation of TA fund for institutional upgrading of Export Development Board (EDB) was not, however, utilized due to lack of commitment of management to change and constraints by political environment (para. 14). Allocation of TA for strengthening the insurance industry was not utilized either. However, broader issues with respect to the non-bank financial sector (insurance industries and other contractual savings institutions) were further addressed by IDA's Private Finance Development Project which was prepared in 1992 (para. 17). The Credit Information Bureau of Sri Lanka (CIBSL) was established in 1990 under the guidance of IDA and ADB. ADB further provided technical and financial assistance under its credit line to strengthen the advisory and management capability of the .CIBSL and appointed a short term consultant to assess the institution's potential for expanding credit information and recommending a medium-term operational policies/strategies and business development (para. 19). The SMI-IV project coincided with the first intervention ever made by GOSL on environmental protection and management in the country. The project played a catalyst and pioneering role in strengthening environmental regulatory framework capacity of GOSL. TA was provided for a number of studies, consultancies, training and hardware for laboratory facilities and in general, the result and impact of TA for environmental management was satisfactory (para. 18, Appendix A, and Annex X). 8. The TechnicalAssistance Fund (TAF) - A partial (50/50) grant of US$ 500,000 was allocated for capacity building of the SMI entrepreneurs in implementing their respective subprojects. This component was implemented highly satisfactorily and benefited over 200 small and medium industrial and services units. SMIs found this scheme very attractive and useful for their business and products improvement. Due to their strong demand, the allocation was increased by an additional US$ 50,000 in 1996 (para. 20). 9. Proiect cost. Financins, and Time Schedules - The project was implemented satisfactorily without any major problem. Project cost was estimated at US$ 145 million by the Appraisal and the actual cost was about US$ 140 million (Tables 8A & 8B). SMI-IV was prepared, signed, and became effective as originally planned. Due to slow implementation of the credit component, the project was closed eight months later than the original closing date which was October 31, 1996 (para. 6). 10. Proiect Sustainabilitv - The benefits and impacts of the SMI-IV project with respect to both credit and TA components would most likely be sustained. The commitment of the Government (both the previous Government and the one elected in late 1994) to develop the private industrial sector, in particular the SMI sector, suggest that the benefits of the project with respect to SMI financing will be sustained. In its 1994 industrialization strategy statement and its subsequent policy reform programs, the Government has given high priority and special attention to the development of SMI sector through rural/regional industrialization and increasing industries' access to marketing and technical advisory facilities as well as short/long term investment credits.(para. 23). The internal efficiency and strong financial performance of the participating private commercial banks and DFCC and their increased lending activities during the life of the project (in particular after SCBs declared ineligible in 1992) were instrumental to the successful/smooth implementation of the credit component and a key factor for project sustainability in future. The private domestic banks have continued to further develop their institutions' capacity and project lending operations. Most PCIs are now committed to support the SMI sector and have continued lending to SMIs from their own resources after the closing of SMI-IV project (para. 24).  iv 11. With the GOSL's attention to this sector and availability of funds from other sources (donors' funds, NDB commercial loans) which provide various forms of investment financing (medium/long-term - leasing/loans), the SMI firms will continue to benefit from having access to investment credits (para. 25). The impacts/benefits of the TA component in the areas of PCIs' institution building and staff training will also be sustained, as these institutions have realized the value/benefits of trained and qualified staff. The impact and benefits of the TA for Department of Customs is likely to be sustained, as the institution has continued further modernization of the computer system and is planning to restructure/reform the organization. The TAF scheme gave SMIs' entrepreneurs the opportunity to further upgrade their business/products. SMIs who benefited from the TAF under SMI-IV were willing to invest their own funds for improving their business activities and have realized the value of training and new/advanced production techniques (through consulting services) and would likely continue their business improvement (para. 26). 12. IDA and Borrower Performance - The performance of IDA was satisfactory throughout the project cycle for the preparation, appraisal, and supervision (paras. 27, 28, & 29). SMI-IV complemented other IDA's financial/industrial sector operations and its preparation was timely and in line with the GOSL's comprehensive development programs. The project objectives were generally realistic. The Appraisal, however, overestimated the ability and commitment of some institutions in implementing the TA programs in particular considering their track record under previous IDA projects (e.g., TA for EDB and SCBs). 13. The project administration and monitoring by NDB (the apex) was satisfactory. The PCIs performance in the utilization of the credit component and increasing their lending activities during 1992- 94, when the SCBs were ineligible to participate, was very good and contributed to the timely and successful project completion. The PCIs complied with the participation criteria and it was only in the last year of the project life when their loan collection was slightly affected by the external factors such as escalating civil disturbances and severe power cuts in the Island (paras. 21 & 30). In light of this situation, the Government took some measures and requested the banks and DFIs to provide some sort of relief to affected industrial and services units. As the situation started to improve in 1997, the industrial activities have gone back on track and repayments of the financed subloan have improved. While NDB's data base on the credit component includes basic information (commitment, disbursement, geographical/sectoral distribution of subloans) on the financed subprojects, it lacks data on SMI subprojects with respect to output and profitability and does not have the capacity for a systematic follow-up and monitoring of financed subprojects. In light of the continued supports for SMIs and their importance in the economic development and exports, it is crucial to evaluate the economic and social benefits/impacts of the SMI projects in a more methodical and systematic way. 14. Government closely coordinated with IDA in the preparation and appraisal of the project and was cooperative with the IDA's supervision missions during the project implementation. GOSL's commitment to improve the DCE administration is commendable, while its performance in implementing the TA in two areas was not satisfactory mainly due to political constraints (in the case of restructuring the two SCBs) and lack of commitment and incentives for change (in the case of EDB). The submission of audited financial reports of the public PCIs (two SCBs and the three RRDBs) were continuously delayed while those of the private commercial banks and the DFIs were always timely. Overall, the performances of the borrower and implementing agencies were satisfactory, considering continued civil disturbances and some economic problems (paras. 30, 31, & 32). 15. Future/Follow-ug Overations. Findines, and Lessons Learned - To further support the GOSL in its efforts to strengthen the financial sector efficiency and develop industrial activities, IDA Board approved the Private Finance Development Project (PFDP) for US$ 60 million in April 1993. The  v objectives of PFDP were to: (i) improve the efficiency of financial intermediation; (ii) assist in domestic resource mobilization for long-term investment; (iii) provide investment finance to industries; and (iv) help deepen the financial system. This project is under implementation and is scheduled to close in June 1999. In March 1997, IDA initiated a review of the financial sector for the preparation of Financial Sector Reform project, scheduled for preparation and Board approval in year 2000. In addition, a number of credit lines for project lending to micro, small and medium industrial and services sectors are provided to Sri Lanka by other donors (paras. 34 & 35). 16. SMI-IV implementation progress was smooth and satisfactory and its development objectives, for the most parts, were achieved. The main lessons Learned from the SMI-IV, with implications for other IDA-financed projects, are summarized below (para. 36): I. Financial viability and internal efficiency of the participating financial institutions, by and large, is the most important factor in the satisfactory implementation of the intermediation operations and sustainability of the project. II. An effective and strong Apex institution is important and instrumental for successful project implementation. III. Government direct involvement in and thorough knowledge of all aspects of the project is crucial for better coordination among stakeholders. IV. With increasing support for the SMI sector and its importance in economic development of the country, it would be beneficial to invest in upgrading the database of and monitoring system for the SMI sector. V. The design of and need for Central Bank Credit Guarantee Scheme should be examined by the Government. The experience of SMI-IV suggests that this scheme should, at least, be optional. VI. The TA program and its goals/objectives should be realistic and achievable. The capacity of implementing agency and Government's commitment and willingness to change should be genuine. The implementation experience of TA programs for the two SCBs and the EDB support this argument. VII. Implementation of TAF scheme under this project and previous IDA projects was satisfactory and suggests that direct assistance to the private sector can have higher pay-off than providing indirect assistance through the publicly funded research, advisory, and technology agencies (such as Industrial Development Board) which are established to provide TA services to the industrial/services units. The implementation experience of TA to these public agencies, under several IDA intermediation operations, proved to be ineffective. VIII. The experience of severe power cuts and its impacts on the economy, calls for an urgent need to develop the country's infrastructures. Sri Lanka's economy is heavily dependent on hydro power and therefore is vulnerable to natural hazards such as drought. In addition, concentration of industries in Colombo area and few Western provinces is partly due to inadequate infrastructure in the country.   IMPLEMENTATION COMPLETION REPORT SRI LANKA FOURTH SMALL AND MEDIUM INDUSTRIES PROJECT (SMI-IV) (CREDIT 2250-CE) PART I: PROJECT IMPLEMENTATION ASSESSMENT A. Statement/Evaluation of Obiectives 1. Project Objectives - The Fourth Small and Medium Industries Project (SMI-IV) was designed to complement and expand on the earlier industrial and financial sector operations (some of which were on- going at that time). The main objectives of the project were: V To deepen IDA's Dialogue with Government of Sri Lanka (GOSL) on issues affecting the financial sector efficiency in particular to: (a) continue support for institutional upgrading of Development Finance Institutions (DFIs), the commercial banks, and Regional Rural Development Banks (RRDBs); and (b) improve implementation of debt recovery policies and procedures including support for the Credit Information Bureau of Sri Lanka which was established at that time (this program was part of the Asian Development Bank's (ADB) program; see paras. 2 & 5). * To provide support to GOSL to: (a) improve Customs and Excise policies and procedures; (b) reorient export development policies and procedures; (c) remove bureaucratic constraints to establish new enterprises and formulate appropriate SMI development policies and programs (this program was also part of ADB funded programs); and (d) assist the Central Environmental Authority (CEA) and Participating Credit Institutions (PCIs) 1 to introduce sub-sectoral industry environmental impact assessment guidelines in the investment approval process. * To provide credit, through the banking system, for long-term capital investment in eligible productive enterprises. 2. Evaluation of Objectives - The SMI-IV objectives were in line with the Bank's assistance strategy in Sri Lanka and were important and timely for the Government's priority reform programs. In 1989, the elected Government initiated an adjustment program aimed at stabilizing the economy, reducing unemployment level, and focusing on measures to reduce the size of the public sector and further development of the private sector. The Government's Industrial Policy Statement (IPS) of 1989 gave high priority to industrial development (in particular the SMI sector) as a vehicle for job creation and export-led industrialization. The GOSL's policies in the areas of trade reform, export promotion, industrial development, financial sector reform, and private sector development through strengthening institutional/legal/regulatory framework have been supported by IDA, ADB, and IMF under several investment credits and technical assistance programs since 1977 (Table 2). With an absence of long term investment funding in the financial sector, rapid utilization of previous industrial credit lines, and strong demand for long-term investment fund by the SMI sector, GOSL requested IDA for the fourth SMI credit I The PCIs under the project were Bank of Ceylon (BOC); Peoples' Bank (PB); Commercial Bank of Ceylon (CBOC); Hatton National Bank (HNB); Sampath Bank; Seylan Bank; Regional Rural Development Banks (RRDBs) of Kandy, Kurunegala, Kegalle; and Development Finance Corporation of Ceylon (DFCC).  2 line. SMI-IV which was originally planned for Fiscal Year 94 was advanced, at the request of GOSL, to FY92 in order to link it with the two IDA adjustment operations which were being prepared during the course of 1990 2, These two adjustrnent credits called for significant cuts in government expenditures through privatization of public manufacturing industries and retrenchment of public sector employees during FY91-93. Therefore, SMI-IV could provide technical and term credit support for those affected employees who wanted to utilize their retrenchment gratuities to establish small scale industrial or service operations. 3. The Financial Institutions Review by IDA in 1990 revealed some of the deficiencies in the financial sector (distorted interest rate structure, inadequate capital market for equity funding, institutional weaknesses, inappropriate financial accounting policies, and poor credit discipline). The GOSL envisioned that the health of financial system is the key to industrial financing and country's development program and requested assistance from IDA and ADB to reform the financial sector. However, because the system did not have the capacity to cope with such a comprehensive program in the areas mentioned above, the Government proposed a phased implementation plan, taking up a number of issues under ADB's financial sector program loan (approved in late 1990), addressing some issues under the IDA and ADB SMI operations, and taking actions on broader issues affecting the financial sector under the IDA's Banking Sector Reform Project planned, at the time, for 1992 (see Part H - follow-up operations). 4. The size of SMI-IV credit line was geared to the term lending capacity of PCIs and their subproject pipeline as well as estimated SMIs' demand for long-term investment funds. The FY91-93 demand for investment funding was estimated at US$ 200 million, of which about 50 percent was to be provided by the sponsors and PCIs. The balance of about US$ 100 million was to be mobilized from the multilateral institutions. A total of US$ 70 million was provided by IDA and ADB for SMI subproject financing. Political instability in the country and effective operations of the project by the National Development Bank (NDB, the Apex for the project) were identified as the major risks to the implementation of lending component. With its experience as the Apex under previous SMI credit lines, NDB managed the project satisfactorily. In addition, prevailing ethnic disturbances did not affect project implementation as much as it was expected at the time of project appraisal. B. Achievement of Objectives 5. SMI-IV project of SDR 33.3 million had two components: (i) a credit component of SDR 29.6 million (US$ 40 million equivalent) for providing term loans to the SMI sector; and (ii) a technical assistance (TA) component of SDR 3.7 million (US$ 5 million equivalent) for institutional and policy development and subproject implementation support. ADB provided US$ 30 million for subproject financing through a parallel line of credit (Second Development Finance Loan) with the same terms and conditions as those of IDA's and an additional US$ 350,000 TA grant for related reform programs. The actual utilization of the IDA credit line was SDR 32.75 million (about 98.4 percent); the undisbursed balance was canceled as of September 11, 97 when the final disbursement was made. The credit account was closed on October 31, 1997. 6. Implementation of the Credit Component - The credit component was implemented highly satisfactorily and achieved its development objectives. It was fully committed as of March 31, 1996 (against the original deadline of September 30, 1995). The delay in the utilization of the credit component was mainly because the two State-Owned Commercial Banks (SCBs), BOC and PB, found ineligible to 2 Economic Restructuring Credit (ERC, approved in May 1990) and Public Manufacturing Enterprises Adjustment Credit (PMEAC, approved in November 1990).  3 participate in the project one year after effectiveness (details in Para. 22). In the first year of the project operations, BOC and PB accounted for about 58 percent by amount and 70 percent by number of subloan approvals (Annex II) and their ineligibility to participate in the project substantially reduced the commitnent level in the period 1992-94. Other PCIs could not immediately fill the gap because of their smaller branch network and limited term lending capacity. However, in later years, the eligible PCIs increased their lending activities through aggressive marketing and by strengthening their SMI project finance unit. While the SCBs suspension from participation delayed the project closing, it had a positive impact on other PCIs' lending operations. The SCBs were reinstated as PCIs by both IDA and ADB during 1995 when they met the participation criteria under SMI-IV. In July 1995, when over US$ 5 million of the credit component was still uncommitted, the Government (which was elected in late 1994 and continued its strong support for SMI sector) requested IDA to extend the commitment deadline and the project closing date for one year. In view of satisfactory utilization of credit component by the private banks and DFCC, eligibility of the SCBs as PCIs in 1995, the importance of SMI sector in Sri Lanka's economy, and industries great need for investment funds, IDA responded favorably and established March 31, 1996 for commitment deadline (a six months extension) and June 30, 1997 for the project closing date (an eight months extension). 7. IDA's SMI scheme was well known and very popular in Sri Lanka. The SMI projects made significant contributions to integrating the SMI sector with the economy, generating large number of employment, and increasing production and exports in particular of non-traditional products. Manufacturing industries grew by about 8.6 percent during 1990-96 (GDP growth was 5.3% in the same period) and its share in total exports increased from 54 percent in 1990 to 75 percent in 1996 [also see study by G. Wignaraja, 1997. "Trade Policy: Technology and Manufactured Exports: Sri Lanka's Liberalization Experience". Basingstoke, UK: McMillan Press]. Small and medium industries grew by about 6 percent during the same period and their output level have been positively correlated with larger industrial output and GDP, indicating integration of this sector into the economy (Annex XI). The recent OED study [World Bank Support for Small and Medium Industries in Sri Lanka: An Inpact Evaluation; June 1997] provides more detailed statistics and analysis on the impact of all four SMI credit lines. These IDA projects set the trend for SMI lending in Sri Lanka and the commercial banks and the DFIs are now committed to lend to this vital sector because of strong demand and its potential for growth. The PCIs highly value the experiences and benefits they gained through the SMI projects. They believe that it was the SMI scheme that created awareness and opportunity for them to learn systematic and methodical credit lending and helped them improve their project appraisal and supervision capabilities. The PCIs established a separate "credit and project finance" department with trained and qualified staff to better service the growing demand by their clients in productive sector. The performance of PCIs was closely monitored by the Bank and NDB to ensure their eligibility with respect to participation criteria such as profitability and loan recovery rate (Annex VIII). PCIs' performance was satisfactory (see para. 21 for more details). 8. The GOSL lent the credit proceed to NDB for 15 years including a five years grace period at the Average Weighted Deposit Rate (AWDR) 3. NDB, in turn, onlent the proceed to PCIs at AWDR and received a fee of 1 percent on the credit amount disbursed and outstanding. The PCIs relent the proceed to sub-borrowers and were free to charge a market determined interest rate (fixed or variable) to cover their cost of operations and the project risk. The lending rates to the sub-borrowers was ranging from 19 to 22 percent. The credit component, together with ADB credit line, provided loans to about 6,000 small and 3The AWDR is defined as the weighted average of interest-bearing term deposits of the Greater Colombo branches of domestic commercial banks (excluding the National Savings Bank). A rolling six months AWDR is calculated by the CBSL for the period April 1 - September 30 and October 1 - March 31, with adjustment to the lending rate on January 1 and July 1 of each year.  4 medium industrial and service units and created about 43,000 jobs, with incremental fixed investment per job of about US$ 3,300 (Annex V). The comparable figures estimated at the Appraisal were 25,000 jobs and US$ 6,200 per job investment. It is not, however, clear from the Appraisal whether the ADB credit line was included in these calculations. The projects financed under SMI-IV were mostly in the labor intensive subsectors (e.g., Textiles, food processing, Garments, construction materials) with the average employment per job of about 7. In addition, with the inadequate technology involved, the employment number increased. Over 53 percent of subloans were equal to or smaller than Rs 0.5 million and the average subloan size was about Rs 0.64 million (approximately US$ 13,400) against appraisal estimate of Rs 0.9 million (Annex I). This was mainly due to active SMI lending by SCBs which have very large branch network in rural areas and mostly catered smaller industries. Despite their suspension for about 3 years, the two SCBs accounted for about 33 percent of the subloan approvals by number but only 25 percent of those by amount (Annex I), resulting in less than average subloan size of Rs 0.49 million. The average subloans would have been even smaller if the SCBs were not suspended. About 40 percent of subloans were made for financing new industrial establishments and 60 percent for expansion of existing firms (Anmex IV). The sectoral composition of financed subprojects was consistent with the expected growth subsectors (Annex VI). The subprojects were mainly concentrated in Colombo, Kurunegala, Puttalam, and Gampaha Districts in Western and North Western Provinces (about 61 percent by number and 71 percent by amount) because of continuing ethnic disturbances in the Eastern and Northern Provinces (Annex VII) and inadequate infrastructure in other provinces. This pattern of geographical distribution of subprojects was similar to that of previous three SMI projects. 9. The Central Bank Credit Guarantee Scheme started under SMI-I project. At that time, lack of collateral was one of the main hindrance to the provision of term loans to the SMI sector; therefor, the Government introduced this scheme which offered securities to the PCIs for term lending to SMIs. PCIs, unanimously, believe that while credit guarantee was important in the past under the first two SMI credit lines, its continuation under SMI-IH & IV was not necessary because PCIs' capacity for methodical and systematic project appraisal and supervision was substantially strengthened, their loan collection performance has been satisfactory, and they are now committed to support SMI sector. All the subloans were automatically covered under the credit guarantee scheme and the PCIs were obliged to pay the premium. The PCIs rarely filed claims for their subloans in arrears because the payments by the Central Bank was not guaranteed and the cost of processing these claims outweighed the benefits, if any. Despite the PCIs' views of the scheme, the Government plans to continue this scheme for other donor funded projects. 10. Implementation of Technical Assistance Component - The SDR 3.7 million TA fund was allocated for: (i) institutional development - SDR 2.59 million; (ii) policy development; specifically in the areas of trade, non-bank financial sector (insurance industry), and enviromnent - SDR 0.74 million; and (iii) project implementation support - SDR 37,000 (Annex IX). Overall, the TA component was implemented satisfactorily and achieved its development objectives. Its impact in few areas was not as expected by the Appraisal. In 1994 and 1995 part of the unused/unallocated TA components were reallocated to finance eligible projects, mostly on environment. 11. The adjustment program that was initiated by the Government in 1989 (para. 2) was supported by the IMF's Structural Adjustment Facility and subsequently by IDA under two adjustment loans (ERC and PMEAC) and several financial/industrial intermediation operations. ADB was simultaneously active in dialogue and financing. Therefore, it is not easy to exactly evaluate the impact of one project on the achievement of the policy reform programs. Since 1989, considerable efforts have been made to improve the financial sector and develop a prudent environment for the private sector and each of these donor funded projects have had a share in facilitating the implementation of the elements of the GOSL's reform  5 program. The following paragraphs (12-20) provide the implementation experience and results of the specific TA projects proposed for funding in the SMI-IV Staff Appraisal Report. 12. Institutional Development component was proposed to: (i) assist the PCIs and NDB in building their capability in project finance through staff training and system/operations improvement; and (ii) continue support for developing the operations of the two SCBs. 7he PCIs upgrading and staff training program was highly satisfactory and achieved its objectives. The success of the program is attributed to several factors such as NDB's effective monitoring and management of the training programs; design and content of the courses; clear objective of the courses; proper organization and transparency of selection of staff, frequency of courses; availability of funds to benefit large number of staff at various levels; commitment of PCIs to improve their human resources; and staff motivation and career development objectives. 13. Strengthening the two SCBs' operations was initiated under the Third SMI project which was approved in 1988. IDA allocated substantial amount of TA funds under the previous three intermediary operations (SMI-III, Industrial Development Projects II & III) for phased operational review of all aspects of these banks in order to prepare them for possible restructuring/privatization. Despite these efforts, the 1990 financial sector study identified weaknesses in various aspects of the banks' operations (such as inadequate capital base, high non-performing loans, high administrative costs, etc.). Therefore, with the objective of continuing its support for improving the financial sector in Sri Lanka, IDA allocated additional TA fund under the SMI-IV project for further review of the SCBs' operations and required that the financial accounts of these banks be audited by qualified international auditors. The studies under all IDA projects were carried out by two foreign consultancy firms and completed in 1992. BOC has taken appropriate actions to implement the diagnostic information in some areas (such as credit policies and procedures, Management Information System, organization structure). In other areas, such as human resources and technology, the impact was marginal. As for the PB, the impact of the TA was low because of: (i) unsatisfactory performance of the consultants (lack of continuity in the task teams, their little involvement in the actual implementation, performing the tasks mostly from offices outside Colombo); (ii) unclarity regarding ownership of the studies and lack of close collaboration between PB and the consultants; and (iii) lack of commitment and interest from the PB's management to accept and implement the consultant's recommendations. Initially, less than 40 percent of the recommendations were accepted and implemented which were mainly in the areas of branch reorganization/operations and to a lesser degree in the credit policies. The new PB's management has recently been looking at the recommendations again and has considered to use the recommendations, as applied, and further strengthen the bank. In general, the SCBs improved their operations in some areas and found these studies and consultancies a beneficial exposure to the techniques used in other countries. Some weaknesses, however, remained and restructuring of the SCBs has not materialized. 14. The result of TA on Policy Development was mixed and varied by implementing institutions and projects. To assist the GOSL in its efforts to develop and promote exports, IDA provided TA funds to the Export Development Board (EDB) of Sri Lanka under SMI-IV (US$ 250,000 TA fund) and previous projects (SMI-II, IDP II & III) for policy improvements and EDB's institutional strengthening. EDB is responsible for the formulation of Government's export policies and promoting exports through appropriate incentives and marketing as well as providing exporters with subsectoral information. Three studies, financed by previous IDA projects, were carried out to review the organizational structure and strategies of EDB. The reconmmendations of these studies were, more or less, along the same line. But the recommendations (except for staff training) were not implemented due to lack of commitments of management to restructuring and constraints by political environment. In view of this, the 1994 supervision mission and the GOSL concluded that another similar study (proposed by SMI-IV) would not  6 be a good use of the TA fund and would not benefit the project implementation; therefore, agreed to cancel the EDB's allocation under SMI-IV. This fund was transferred to the pool for other eligible projects in potential areas. 15. The declaration and documentation processes within the Department of Customs and Excise (DCE) was identified, by IDA and UNDP, as a significant bottleneck which constrained the smooth flow of material needed for trade and production. Therefore, rationalization of DEC's procedures was considered essential in facilitating the trade procedures for the private sector and business community. Modernization and computerization of DCE started in 1992 with financial and technical assistance from UNDP and IDA under three operations (ERC, SMI-IV, PMEAC). The project was completed in 1995 at a total cost of about US$ 2 mnillion of which US$ 1.2 million was granted by the UNDP and US$ 200,000 under SMI-IV. The allocation under SMI-IV was not utilized until 1994 because DCE and Ministry of Finance did not know about this allocation under SMI-IV. This is an indication of lack of coordination among the concerned institutions and adequate knowledge of the elements of the approved projects. 16. The ASYCUDA system was used to computerize entire imports declaration procedures (the Long Room), exports, bonding, and air cargo imports/exports units. All the related offices were refurbished and air conditioned, computer furniture were purchased, and a large number (about 350) Customs officers were trained in data processing. In addition, about 2,800 Customs House Agents were trained in the use of new declaration forms. Computerization of DCE was a successful project and resulted in improved statistical and management information, increase of revenue yield (imports revenue recorded an increase of US$ 90 million or 17.5% in the year following computerization), speedier clearance, and improved Customs efficiency. DCE is now in the process of upgrading their computer system to ASYCUDA++ in order to enhance the linkage to other systems such as those of traders. While computerization was successfully completed, DCE was not reorganized as recommended by IDA and UNDP. The management of DCE has recently recognized that certain reforms in the custom procedures and organizational structure are necessary for further modernization programs and smooth implementation of upgraded systems. In view of this need, DCE decided to get further technical assistance from the World Custom Organization (WCO) in this area. The process of this assistance already started and is expected to be completed by end of 1998. 17. The Government's policy and the regulatory framework under which the insurance industry function did not allow their integration into the evolving financial sector. The SMI-IV project proposed to review and improve/strengthen the insurance legislation in order to allow greater competition between the public and private insurance companies and to make the industry responsive to the environment of financial liberalization by pursuing more sophisticated portfolio management. In addition, the project proposed to strengthen the office of the Controller of Insurance. An allocation of US$ 200,000 was made for a brief diagnostic study in this area. Broader issues with respect to the non-bank financial sector, in particular those related to the insurance industries and other contractual savings institutions were further addressed by IDA's Private Finance Development Project (PFDP) which was being prepared in 1992, the first year of SMI-IV project (para. 34). In light of objectives of this project and GOSL's efforts in strengthening the insurance industries, the allocated fund under SMI-IV project was earmarked for possible TA needs in this area. This fund, however, was never used. 18. Environmental Components - SMI-IV project allocated US$ 300,000 TA fund to assist the CEA: (i) develop environmental guidelines and an action plan for industrial subsectors which link the objective of industrial growth and enviromnental protection; and (ii) ensure that the PCIs and NDB incorporate environmental impact assessment and managerial plans into the appraisal process of subprojects. A Japanese Grant Fund (JGF) of Yen 108 million (about US$ 800,000 equivalent) was also provided to further support this component. Subsequently, the Government identified additional and related TA  7 activities which were supported by the 1994 and 1995 supervision missions for financing under SMI-IV (para. 10). The details of objectives, implementation experience, and results of the specific environmental TA projects are provided in Appendix A and Annex X. 19. To further assist the SMI sector, ADB allocated about US$ 260,000 TA fund to review the SMI policies and identify the bottlenecks (policy, infrustractural, managerial, financial) that constrained the development of this sector and to formulate a more prudent SMI assistance program in the medium term. In line with the ADB technical assistance program, IDA allocated US$ 50,000 to GOSL for administrative/technical support to carry out ad hoc short term studies on industrial and financial sector policies. Despite supervision missions' and NDB's efforts to encourage the utilization of this fund, no proposal was formulated by the Government and the allocated fund remained unused. IDA and ADB helped the GOSL establish the Credit Information Bureau of Sri Lanka (CIBSL) in 1990. ADB further provided technical and financial assistance under its credit line to strengthen the advisory and management capability of the CIBSL and appointed a short term consultant to assess the institution's potential for expanding credit information and recommending a medium-term operational policies/strategies and business development. The CIBSL database (about 72,700 loan accounts) are frequently used by the banks and DFIs for evaluation of their clients. The Bureau updates its irregular accounts (minimum of Rs 100,000) on monthly basis and its regular accounts (minimum of Rs 500,000) on quarterly basis. IDA continued its assistance in this area under the PFDP project (paras. 3 and 34). 20. Under SMI-IV, a Technical Assistance Fund (TAF) of US$ 500,000 was allocated for capacity building of the SMI entrepreneurs in implementing their respective subprojects. TAF was a partial (50/50) grant, through PCIs, to the SMI entrepreneurs for technology and product improvement and financing technical assistance such as consulting services for business development, introducing or improving technology, training programs, etc. This scheme was administered by NDB. Utilization of this fund was slow initially because of the availability of similar scheme under the Third SMI project until June 1993 (SMI-I1I closing date). Therefore, NDB financed all the TAF applications from SMI-III project until this project was closed and started to use SMI-IV fund in mid-1993. To improve and broaden the utilization of TAF, NDB increased publicity of the fund and collaborated with other agencies (such as USAID) and local public research/advisory/technology institutions (such as Industrial Development Board, Ceylon Institute of Scientific and Industrial Research, etc.) which provide technical assistance to SMIs. Through these agencies which mobilize smaller industries, NDB could organize several group exposure visits and training programs for small industries (mostly located in remote areas) which could not have otherwise access to information and services available. SMIs found this scheme very attractive and useful and due to their strong demand, the allocation was increased by an additional US$ 50,000 in 1996. The TAF component was satisfactorily utilized by over 200 small and medium entrepreneurs, among which there are many whose business activities were turned around for better as a result of the technical assistance they received. C. Maior Factors Affectina the Project 21. The SMI-IV project was implemented satisfactorily without any major problem. There were, however, factors that delayed the completion of the project and slightly/temporarily affected the PCIs performance in fully meeting the eligibility criteria (collection and portfolio infection ratios). Factors not Generally Subject to Implementing Agencies: The economic and political problems in the country in late 1995 and during 1996 had adversely affected the industrial activities. The industries further suffered because of severe power shortage in the Island during 1996 and many had cash flow shortage which continued until the first half of 1997. As a result, the loan recovery rate of some of the PCIs slightly deteriorated during this period. Severity of this situation (which was observed by the 1996 supervision mission in the field) and its immediate impact on the industries forced the Government to take some  8 measures and provide relief to affected industrial and services units. These measures included, among other things, a 5-point bank strategies. Banks were asked to: (i) exercise some degree of flexibility in their loan recovery; (ii) provide partial waivers in loan approval; (iii) refrain from resorting to debt recovery laws; (iv) capitalize outstanding interest; and (v) provide easy overdraft facilities. Most PCIs provided some sorts of relief to their clients depending on their asset management strategy and mostly on a case by case basis rather than a blanket cover. NDB continued to monitor the PCIs performance. As the situation started to improve in 1997, the industrial activities have gone back on track and repayments of the financed subloan have improved. The 1996 severe power cuts demonstrated that the economy is quite vulnerable to natural hazards (such as drought, etc.) due to its heavy dependence on hydro power for electricity generation and that there is an urgent need to develop the country's basic infrastructure. 22. Factors Subject to Implementing Agency Control: On the positive side, the continued GOSL efforts in improving the financial and industrial sector policy with the assistance from IDA/ADB/IMF under other related operations and follow up operations (PFDP), resulted in satisfactory implementation of the TA projects in some areas. In addition, strong performance of NDB in administrating the project components was instrumental to the smooth and successful project implementation. On the negative side, suspension of the two SCBs delayed the project implementation by eight months. The 1991 international financial audits for the SCBs revealed that these banks had inadequate capital base and under-funded pension. Based on these audits, a capital infusion of over Rs 24 billion was required to recapitalize these banks to internationally acceptable standards as well as provide for their pension obligations. Therefore, on February 27, 1992, IDA declared the SCBs ineligible to participate in the credit component and consequently did not cover the subloans made by these banks for funding under the project effective April 1, 1992. ADB also suspended these banks in 1992. The suspension by IDA did not affect disbursements on subloans that had been approved by IDA since the effectiveness and it was not applied to the use of TA fund by the SCBs. In April 1993, the GOSL issued long-term Government Bonds to the PB for Rs 10,541 million and for BOC for Rs 13,547 million in order to ensure that their capital base is at least 8% of their risk-adjusted assets. The Bonds had thirty years term at 12% interest per annum payable semi-annually, and non-transferable. Although the SCBs were recapitalized, IDA could not re-instate their participation because they were not in compliance with the timely audit and reporting requirements and therefore, their financial performance could not be properly assessed. The SCBs' suspension substantially reduced the commitment level during 1992-1994 (para. 6). The international financial audits of the two SCBs for 1993-1994 indicated that the banks met the participation criteria under SMI-IV; hence, they were reinstated as PCIs by both IDA and ADB during the course of 1995. Subloans made by BOC during August 1995 to March 1996 were covered for funding under SMI-IV while those for PB were financed under ADB credit line. D. Project Sustainability 23. SMI Financing - The commitment of the Government (both the previous Government and the one elected in late 1994) to develop the private industrial sector, in particular the SMI sector suggests that the benefits of the IDA's SMI scheme would be sustained. In its 1994 industrialization policy statement, the Government gave high priority and special attention to the development of SMI sector through rural/regional industrialization and increasing industries' access to marketing and technical advisory facilities as well as short/long term investment credits. Therefore, with the closing of IDA and ADB credit lines and strong and growing demand for long term investment by the SMIs, the Government decided to find other quick channels to generate funds for term lending. In 1996, the Government allowed NDB to raise a foreign loan of US$ 20 million specifically for financing the SMIs capital investment. In addition, NDB raised another US$ 50 million commercial loan (from a syndication of Japanese financial institutions), guaranteed by ADB, for long term project lending to SMIs. NDB loan has been supplemented  9 by a US$ 5 million ADB loan (Small and Medium Enterprises Assistance Project). These loans will be made available to SMIs, through the commercial banks, with the terms and conditions close to those of IDA's SMI scheme. Overseas-Economic Co-operation Fund (OECF) of Japan also provided a US$ 46 million loan for small and micro -industries. This loan, however, will carry a concessional rate at about 6.5 percent for onlending to PCIs and maximum of 14 percent for the subborrowers. 24. Participating institutions - The internal efficiency and strong financial performance of the participating private commercial banks and DFCC and their increased lending activities during the life of the project (after suspension of SCBs in 1992) were instrumental to the successful/smooth implementation of the project. IDA had significant impact on imnproving PCIs operations and their project finance activities and it is very likely that this impact will be sustained in the future. PCIs are now committed to the development of the SMI sector and have continued their SMI lending from their own funds after the closing of SMI-IV project. Many financial institutions have designed their own SMI schemes and intend to participate in most of the credit lines mentioned in previous paragraph. In 1996, the HNB introduced a new scheme (Small and Medium Enterprise Credit Scheme, SMECS) to provide loans ranging from Rs 250,000 up to Rs 10 million to SMI sector. PB and BOC have been lending to SMI sector from their own sources under the Industrial Project Loan Scheme and SMI Loan Scheme, respectively. Other PCIs have also been extending credit facilities to SMIs from their own funds. The SMI projects expanded the institutional coverage of lending operations by using the public and private commercial banks as intermediaries for term lending. This impact is likely to be sustained in future, in particular for the private domestic baniks. 25. SM7firms - The SMI projects had significant impact on the development of SMI sector through providing long term investment credit which enabled the entrepreneurs hire more labor and expand and diversify their business activities (paras. 7 & 8). These opportunities would have been lost without availability of such credit line because in Sri Lanka the financial institutions were constrained by insufficient access to the term savings for long term investment to the productive sector. With the continued GOSL's attention to this sector and availability of funds from other sources which provide various forms of investment financing (medium/long-term - leasing/loans), the SMIs will continue to benefit from having access to investment credits. Many industrial units which borrowed under SMI scheme have successfully grown out of this sector and are now eligible for larger loans. 26. Technical Assistance - The impacts and benefits of the TA component in the areas of institution building of PCIs and training would be sustained, as these institutions have realized the value of staff training and the impact of having qualified staff at all level of the business operations. The impact and benefits of TA for DCE is also likely to be sustained, as this institution has continued further modernization of computer system after the initial phase and is planning to make certain reforms in the customs procedures and organization structure. The technical assistance under TAF scheme gave SMIs' entrepreneurs the opportunity to further upgrade their business/products. Many SMIs who benefited from the TAF scheme under SMI-IV and were willing to provide 50 percent of the TA cost from their own resources have realized the value of training and new/advanced production techniques. They envisaged that product improvement are the key to good/profitable business and their survival in the future competitive market (Para. 20); suggesting that the benefits of TAF would likely to be sustained. E. IDA's Performance 27. Against the background given in paras. 2 and 3, preparation of SMI-IV project was timely, was in line with the GOSL's comprehensive development programs, and was based on an assessment of SMIs'  10 potential demand for term investment credit. SMI-IV complemented other IDA's financial and industrial sector operations (ongoing at the time or closed) and was consistent with the IMF adjustment program and IDA's country assistance strategy. The project objectives identified at the Appraisal were generally realistic. The Appraisal, however, overestimated the ability and commitment of some institutions in implementing the TA programs (case of TA for EDB and SCBs). Despite the track record of these public institutions under the previous IDA project, TA supports continued under SMI-IV which, as mentioned above, resulted in less than satisfactory outcome of their respective programs. 28. IDA played a significant role, through four SMI credit lines, in developing SMI sector and brought a general awareness of the great potential and importance of this sector in the economic development and employment generation in the country. Through continued TA programs, the projects had substantial impact on the financial and institutional development of the two DFIs and domestic commercial banks. 29. With NDB's satisfactory administration of the project and timely quarterly progress reports, IDA's field supervision of the project once every year was considered sufficient to ensure smooth implementation. IDA sent eight supervision missions with the first one in September 1991 just before the effectiveness and the last one (ICR mission) in July 1997 right after the project closing. The supervision missions carefully reviewed the progress of the credit and TA components and PCIs' performance and were diligent in recognizing the shortcomings in the implementation of TA in some areas. The supervision missions were open to the Government's ideas in financing other potential projects in line with objectives of SMI-IV and identified potential projects for financing. IDA's performance was satisfactory. F. Borrower's Performance 30. The project administration and monitoring by NDB (the apex) was satisfactory. All the PCIs highly value the assistance they received from NDB and believe that the success of the project was partly due to the strong and capable apex. The PCIs performance in the utilization of the credit component and increasing their lending activities during the suspension of the SCBs was very good and contributed to the timely and successful project completion. PCIs complied with and maintained the participation criteria and it was only in the last year of the project life that their loan collection was slightly affected by the external factors such as escalating civil disturbances and sever power cuts (para. 21). 31. NDB's data base on the credit component includes basic information on the commitment, disbursements, and classifications of the subloans by size, sector and geographical aspects. However, it lacks data on SMI subprojects with respect to output and business activities/profitability. It also lacks a systematic follow-up monitoring mechanism which is necessary for an in-depth evaluation and analysis of the financed subprojects. This weakness of NDB's monitoring capacity was highlighted in the previous completion reports. In light of the continued supports for SMIs and their importance in the economic development and exports, it is crucial to evaluate the economic and social benefits/impacts of the SMI projects in a more methodical and systematic way. 32. The Government closely coordinated with IDA in the preparation and appraisal of the project and was cooperative with the IDA's supervision missions during the project implementation. GOSL's performance in implementing the TA in some areas was not satisfactory mainly due to political constraints and lack of commitment and incentives for change (case of TA for EDB and restructuring of the two SCBs). However, GOSL's commitment to improve the customs administration and its cooperation with IDA and UNDP resulted in satisfactory project implementation in this area. The audits of the public PCIs (two SCBs and the three RRDBs) were continuously delayed while those of the private commercial banks and the DFIs were always timely. Overall, the performances of the borrower and implementing agencies  11 were satisfactory and shortcomings in few areas were the result of some external factors outside the control of the Govermment. G. Assessment of outcome 33. Based on the assessment of the elements of the project (paras. 5-26), the outcome of the project was satisfactory with respect to both the implementation progress and achievement of development objectives. The credit component was utilized satisfactorily and its impact was high. The TA components for institutional development, computerization of the Customs, environment, CIBSL, and project implementation support (TAF) were implemented satisfactorily and achieved their respective objectives. H. Follow-up/Future Operations 34. At the time of SMI-IV preparation (1990/9 1), it was agreed that part of the Government reform program in the areas of financial sector and private sector development would be addressed in an IDA project planned for 1992 (para. 3). This project was prepared as planned under the name Private Finance Development Project, approved by the Board for US$ 60 million in April 1993, and became effective in August 1993. The broad objectives of this project were to: (i) improve the efficiency of financial intermediation by supporting the regulatory reforms in the financial sector; (ii) assist in domestic resource mobilization for long-term investment by developing the local bond markets; (iii) provide investment finance to small, medium, and large scale industries; and (iv) help deepen the financial system and strengthen the key players including the contractual savings institutions (insurance sector, provident/pension funds), Central Bank of Sri Lanka, Institute of Chartered Accountants, and the PCIs. This project is under implementation and is scheduled to close in June 1999. In March 1997, IDA initiated a fresh review of the financial sector comprising the banking sector, the capital markets, and contractual savings industry. A Financial Sector Reform project is scheduled for the preparation and Board approval in year 2000. 35. In addition to the above projects, a number of credit lines for project lending to micro, small and medium industrial and services sectors are provided to Sri Lanka by other donors (OECF, ADB). The two DFIs (DFCC, NDB) are also active in raising funds from international financial markets for short/long term investment lending to the productive sectors (para. 23). I. Main Lessons Learned 36. Based on the implementation experience of this project and similar previous IDA's operations, the main lessons learned are summarized as follows: I. Financial viability and internal efficiency of the participating financial institutions, by and large, is the most important factor in the satisfactory implementation of the intermediation operations and sustainability of the project. In case of Sri Lanka, a comparison of the performance of the two SCBs with the private domestic banks and DFCC under SMI-IV supports this argument. If it was not for the sound financial performance of the private domestic banks and DFCC and their efficient and effective SMI lending, the credit component would not have been utilized when the two SCBs were suspended due to their capital deficiency (paras. 6, 30). II. Satisfactory and smooth implementation progress of SMI-IV credit component was, in part, due to effective project administration by the Apex institution. The implementation experience of several  12 financial and industrial intermediation operations in Sri Lanka shows that projects with a good and effective Apex system were implemented more successfully/smoothly than projects which had either a weak or no Apex system (paras. 4, 29, 30). III. The Government, as the borrower, should be directly involved in and have a thorough knowledge of all aspects of the project in order to be able to better coordinate among stakeholders and among various donors' funded projects, to avoid duplication of projects, and timely use of funds. The experience of TA for Customs supports this argument and a lesson to learn for future operations (para. 15). IV. An in-depth evaluation of the benefits and impacts of the financed SMI subprojects requires proper sectoral data on output and subprojects' performance after the loan is made as well as a systematic follow up and monitoring mechanism. With increasing support for SMI sector and its importance in economic development of the country, it would be beneficial to invest in upgrading the data base of and monitoring system for the SMI sector (para. 31). V. PCIs' experience with the Central Bank Credit Guarantee Scheme was not positive. Despite this, the scheme is again mandatory for other donors funded projects. The GOSL should examine the design of and need for this scheme or at least take into account the preference of the participating credit institutions (para. 9). VI. The TA program and its goals/objectives should be realistic and achievable. The capacity of implementing agency and Government's commitment and willingness to change should be genuine. Without these considerations, the TA fund is either wasted/misused or never utilized and would be a net loss for the Government. The case of TA programs for the two SCBs and EDB supports this argument, a lesson to learn in future operations (paras. 13 and 14). VII. The implementation experience of TAF (direct assistance to the SMI entrepreneurs on a partial grant basis) under this project and previous IDA operations (IDP-III and SMI-III) was satisfactory and its impact was high. IDA, under previous projects, also provided TA for institution upgrading of the public research, advisory, and technology agencies which were established to assist the industrial/services units. The results of these TA were highly unsatisfactory and despite continuos efforts by IDA and other donors, most of these institutions have remained weak and ineffective. This suggests that future TA projects, if any, should be directly provided to the entrepreneurs rather than through the publicly funded agencies (para. 20 and the implementation experience of previous IDA credit lines in Sri Lanka). VIII. The experience of severe power cuts and its impacts on the economy, calls for an urgent need to develop the country's infrastructures. Sri Lanka's economy is heavily dependent on hydro power and therefore is vulnerable to natural hazards such as drought (para. 21). In addition, concentration of industries in Colombo area and few Western provinces is partly due to inadequate infrastructure in the country (para. 8). December 12, 1997  13 Table 1: Summary of Assessments A. Achievement of Objectives Substantial Partial Negligible Not applicable Macro Policies E E E 0 Sector Policies [I [El Cl Financial Objectives El O O a Institutional Development El E I El Physical Objectives