Sri Lanka - Second Small and Medium Industries Project

Document of The World Bank FOR OFFICIAL USE ONLY p B k t Report No. 3480-CE STAFF APPRAISAL REPORT SRI LANKA SECOND SMALL AND MEDIUM INDUSTRIES PROJECT September 14, 1981 Industrial Deve'lopment and Finance Division South Asia Projects 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 EQUIVALENTS Currency Unit = Sri Lankan Rupee Sri Lanka has a floating exchange rate: on December 31, 1979, the rate was US$1 = Rs 15.45; on December 31, 1980, US$1 = Rs 18.00; and on August 31, 1981, US$1 = Rs 19.97. For purposes of the first Small and Medium Industries Project, effective since October 1979, a rate of US$1 = Rs 17.5 is used. For the proposed Second Project, the rate below is used in this report. US$1 = Rs 20.00 Rs 1 US$0.04 Rs 1 million US$50,000 PRINCIPAL ABBREVIATIONS BOC - Bank of Ceylon CBOC - Commercial Bank of Ceylon DFCC - Development Finance Corporation of Ceylon DSI - Department of Small Industries DTI - Department of Textile Industries EDB - Export Development Board EPR - Effective Protection Rates FIAC - Foreign Investment Advisory Committee GCEC - Greater Colombo Economic Commission GOSL - Government of Sri Lanka hNB - Hatton National Bank ICICI - Industrial Credit and Investment Corporation of India IDB - Industrial Development Board LIAC - Local Investment Advisory Committee MTI - Ministry of Textile Industries NDB - National Development Bank NIBM - National Institute of Business Management PCI - Participating Credit Institution PDCP - Private Development Corporation of the Philippines SMI - Small and Medium Industries UNDP - United Nations Development Programme FISCAL YEARS Government of Sri Lanka January 1 to December 31 Commercial Banks = January 1 to December 31 DFCC = April 1 to March 31  FOR OFFICIAL USE ONLY SRI LANKA APPRAISAL OF SECOND SMALL AND MEDIUM INDUSTRY PROJECT Table of Contents Page No. I. INTRODUCTION .................................... ......... I II. SECTORAL BACKGROUND ....................... ... too ... 3 A. Economic Setting . . . ... . ................... *....... 3 B. Industrial Structure and Performance . 3 C. Characteristics and Role of SMI ............... 4 D. Industrial Policy Framework . . 6 E. Financing SMI .............0... ....7 F. Technical and Marketing Services . . .9 III. STRATEGY IN DEVELOPING LIGHT INDUSTRIES... 10 A. Expanding Exports of Light Industrial Products ....... 11 B. Building Private Technical and Marketing Services .... 15 C. Improving Management in Institutions Serving SMIs .... 17 IV. THE PROPOSED SECOND SMI PROJECT .......................... 17 A. Objectives, Scope and Institutional Arrangements 17 B. Lending Component ...............19 1. Small and Medium Industry Fund - NDB 19 2. Credit Guarantee Scheme. 25 3. Participating Credit Institutions 25 C. Technical and Marketing Service Components . .30 1. Industrial Development Board 30 2. Export Development Board 33 3. Coordination and Monitoring .34 D. Industrial Policies and Incentives . .35 V. THE CREDIT . 36 VI. PROJECT BENEFITS AND RISKS . ................. 39 VII. RECOMMENDATIONS ................... .... o .................. 41 This report is based on the findings of an appraisal mission which visited Sri Lanka during February 1981. The mission comprised N. Barry, C. Bam, and J. Pernia. 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.  - ii - ANNEXES Annex 1 - Profile of Subprojects Financed under the First Small and Medium Industry Project Annex 2 - Supporting Tables Table 1: Indicators of Performance in Large, Medium and Small Industries Table 2: SMI Projected Demand and Supply of Credit Table 3: Interest Rate Structure of Credit Institutions Table 4: Financial Projections - SMI Fund Table 5: Costs and Sources of Finance for Technical Service Components Table 6: Projected Disbursement Schedule Annex 3 - Supporting Documents Available in Project File Annex 4 - Map No. IBRD 15823: Sri Lanka - Second Small and Medium Industries Project  SRI LANKA SECOND SMALL AND MEDIUM INDUSTRIES STAFF APPRAISAL REPORT I. INTRODUCTION 1.01 The Government of Sri Lanka (GOSL) is giving priority to private industrial development, with special attention to small and medium industries (SMI). The first SMI Project (Credit 942-CE), made effective on October 24, 1979, has included US$16.0 million in IDA funds for: refinance by the new National Development Bank (NDB) of eligible loans by participating credit institutions (PCIs); assistance in developing operations of NDB; technical and marketing services in key SMI subsectors; and policy studies linked to reforms in tariffs and export incentives. The proposed Second SMI Project would build on these institutional arrangements, while increasing emphasis on export-oriented industries. In this report, an IDA credit of SDR 26.7 million (US$30.0 million) in support of a US$52.5 million project is recommended. 1.02 First SMI Project. The first Project has contributed substantially to the sound initial organization and operations of NDB, providing: a general advisor and institutional assistance in support of NDB'S direct financing activities; and an advisor, training, and frequent IDA supervision missions to build SMI refinance operations. Training of SMI officers of the PCIs also has been provided; over 200 officers have received local or foreign SMI training prior to or during the first year of implementation, and 150 SMI officers now are engaged full time in SRiI lending operations at head office departments and key branches. 1.03 NDB and the PCIs have performed well under the first Project. Com- mitment progress has been about one year ahead of schedule; NDB's refinance approvals totalled Rs 192.0 million (US$11.0 million) as of July 31, 1981, with full commitment of the US$12 million expected by late October 1981. Disbursements by NDB totalled Rs 102.9 million (US$5.9 million) and by IDA, US$5.2 million; full disbursement is expected by the second half of CY82. As of July 31, 1981 about 1,500 subloans averaging US$9,100 had been approved, with fixed costs per job (excluding land) of about US$1,000. Major subsectors * financed have been building materials, agro-industries, and light engineering. Over 80% of the subprojects financed are totally dependent upon local raw mate- rials. Random samples of subprojects indicate that all but a few are operating without notable problems. Annex 1 provides profiles of the S14 projects financed. 1.04 Progress under the technical and marketing service component has been slower, as expected, due to the need to locate suitable consultants, construct facilities, and build staff for the subsector schemes. Steps have been taken to ensure that the rubber products service center, light engineer- ing common facilities and subcontracting exchange become fully operational during CY81. The private handloom export company has been established and consultancy to launch the Garment Training Institute has been completed. Full disbursement of the technical and marketing service component is expected by the end of CY82.  -2- 1.05 The Second SMI Project would accelerate growth and productivity improvements in SMIs, tapping the sector's scope for value added, low cost employment generation, and export earnings. IDA would provide US$28 million equivalent for partial refinance by the NDB of SMI term loans made by the credit institutions. The US$2 million technical and marketing services compo- nent would include improvements in SMI extension services by the Industrial Development Board (IDB), with field-level coaching of regional extension officers, an SMI consultancy fund to tap knowhow of the local private sector in solving problems of SMIs, and expansion of service facilities for light engineering SMIs. The Second SMI Project would give increased emphasis to exports and related supply development, financing the Export Development Board (EDB) program for practical consultancy and training, sales trips, policy studies, and pilot projects involving exporters and export-oriented manufac- turers in key light industrial products. The Second Credit also would finance training and consultancy to improve SMI lending capabilities and overall operations of NDB and the banks. 1.06 Other Bank Group Financing of the industrial sector in Sri Lanka has comprised: two loans and two credits to the Development Finance Corpor- ation of Ceylon (DFCC), totalling US$24.5 million equivalent to finance larger enterprises; IFC equity in DFCC and the Lanka Orient Leasing Company; two IFC industrial operations; a training project for the construction indus- try; and two IFC credits, totalling US$7 million equivalent, to the Bank of Ceylon (BOC). The first IFC loan to BOC, made effective in 1979, consisted of US$2 million for term loans to firms with fixed assets of below Rs 6 million (US$385,000); the second IFC loan is geared to larger enterprises with eligible subloans ranging from 83,000 to US$1 million. Under its second loan, IFC plans to continue its institution-building efforts with BOC, with consultancy financed by UNDP. IDA and ADB credits and loans to DFCC have focussed mainly on term financing of larger industries, although recent loans have included small allocations for smaller enterprises. ADB recently has appraised a US$1.0 million loan to NDB for its direct lending to large industry; an IDA credit for direct NDB finance of large public and private enterprises is under preparation. A 1979 Industrial Sector mission high- lighted the need to improve the tariff system and export incentives; finance for studies in these areas was provided under SMI I; related policy reforms are expected during the implementation period of the Second SMI Project. The proposed Second SMI Project provides IDA with the opportunity to continue supporting GOSL in its objectives of developing SMIs, which represent the major segment of private industry.  - 3 - II. SECTORAL BACKGROUND A. Economic Setting 2.01 The private sector has responded favorably to the post-1977 liberali- zation measures; private investment was high in 1980 and GDP grew by nearly 6%, more than double the annual growth rates during the 1970-77 period. Imports have grown substantially from 1978-80, with the composition shifting markedly from consumer goods to intermediate and investment goods. Investment in export industries has increased although actual export growth has been disappointing and concentrated in a few lines. In spite of these favorable developments, major public spending on infrastructure combined with unsatis- factory performance in the tree crop sector, public corporations and exports threaten the soundness of the economy. Inflation, trade, and budget deficits increased dramatically in 1980 and foreign exchange reserves have declined to precariously low levels. GOSL has taken some steps to address these problems; further measures to fuel private sector output and exports while controlling public expenditures will be of critical importance. B. Industrial Structure and Performance 2.02 Manufacturing industries consist of: 29 public sector corporations; about 9,000 registered private factories composed largely of small and medium industry firms with fixed assets of under Rs 1 million; 1/ and over 20,000 unregistered small and cottage industries. In 1978, these three groups had roughly equal shares in industrial value added. Growth rates in industry 2/ have slowed from 7.8% in 1978 to 4.6% in 1979, and to only about 0.8% in 1980. However, this poor overall performance hides contrasting patterns in different industrial segments. In 1980, value added in processing of traditional tree crops, representing about 35% of industry, declined by nearly 10% due to drought and resultant raw material shortages. Output by public corporations, excluding the petroleum refinery, declined by 6%. In contrast, output value in private factory industries 3/ grew by an average of 7.2% in the 1977-80 period and small cottage industries grew by 9.4%. Related sectors, construction and mining, grew by about 20% and 9% annually during the 1977-80 period. In 1980, manufacturing, mining and construction represented about 29% of GDP, slightly more than agriculture's 24% share. 4/ 1/ Fixed assets here are defined as plant and equipment at original cost; land and buildings are excluded. 2/ Industrial value added represents about 14% of GDP. 31 Excluding firms under the Greater Colombo Economic Commission (GCEC) on and off the export processing zone (para. 2.11). 4/ Agriculture includes also forestry and fishing; mining refers to mining and quarrying.  -4- 2.03 Manufacturing exports, excluding petroleum, nearly doubled from 1978 to 1979, slowing to 31% in 1980. While growth in garment exports dominated in both years, 1/ exports of seafood, ceramics, rubber and metal products also showed substantial increases in 1979; however in 1980, seafood exports declined, and exports in the smaller product groups stagnated. The pattern of export industries approved by the Greater Colombo Economic Commission (GCEC) reflect this concentration in textiles: over 40% of the approvals and 50% of the signed contracts from 1978-1980 have been for wearing apparel and textiles. Likewise, nearly 30% of the approvals by the Foreign Investment Advisory Committee (FIAC) has been in garments. Fabricated metal products, rubber products, and food processing together represent about 45% of FIAC and about 30% of GCEC approvals. However, these promising light industrial product groups have been characterized by longer gestation periods and higher dropout rates between approval and contract signing than in the case of "easier" garment establishments. While garment industries can be launched with short- term training and limited local inputs or infrastructure, major increases in other key export industries with higher local value-added will require: more systematic promotion to attract appropriate partners; technical assistance to upgrade designs, skills and technology; improvements in linkage industries such as engineering; and policy modifications to make exporting at least as attractive as import substitution. 2.04 During the post-1977 period, GOSL has replaced quantitative import controls with tariffs. Local industries, some still being protected by sub- stantial tariffs, have faced increased competition from imports. The rapid increase in approvals for establishing industries and demand for term credit by SMIs (para 2.16) indicates that local industries in most product groups have benefitted from liberalized access to imported inputs and equipment, and have shown a buoyancy in responding to competition. Growth has been notable in building materials and construction, several metal and food product lines, and certain rubber and plastics products. Problems with unemployment and low capacity utilization have been eased. However, adjustments in the industrial structure have occurred with some of the least efficient firms suffering as a result of import liberalization. Measures to increase finance, technical services, and market assistance in promising light industrial product groups will be necessary to sustain growth in private manufacturing and to promote accelerated exports in areas with high value added. Special attention to the needs of SMI is warranted in view of their predominance, potential to manufac- ture competitive products for local and export markets, and need for special credit, technical and marketing services. C. Characteristics and Role of SMI 2.05 Shares. Small and medium industries (SMI) continue to dominate pri- vate manufacturing. 2/ In 1978, only 5% of registered private firms had fixed 1/ Exports grew from 22% to 35% of the value of total industrial output from 1978 to 1979, largely as a result of growth in garment exports; garments represented 20% of industrial exports in 1978 and 26% in 1979. 2/ Small and medium industries are defined by GOSL as firms with plant and equipment (original cost) excluding land and buildings of less than Rs 2 million.  -5- assets exceeding Rs 1 million; none of the unregistered firms had fixed assets exceeding this investment amount. In the 1978-80 period, the Local Investment Advisory Committee approved nearly 5,000 industrial investments, averaging Rs 600,000 (US$34,000) per unit. SMIs account for over 90% of private indus- trial units, over 70% of employment and over 55% of value added in private manufacturing (Annex 2, Table 1). 1/ The only substantial increases in larger industrial investment and output during 1978-80 have been among firms involving foreign capital, approved either by the GCEC or FIAC. Together, about 300 manufacturing firms have been approved during 1978-80; if implemented, these approvals would represent Rs 7.4 billion (US$370.0 million) in total investments, with GCEC firms averaging Rs 33.1 million and FIAC firms averaging Rs 17.2 million. However, only about 25% of these approved firms are in commercial operation, and the number to be implemented is uncertain. 2.06 Geographical Distribution. About 65% of registered SMIs are in and around Colombo, with secondary concentrations in the Kalutara, Kandy, Matara, jaffna, Galle and Kurunegala districts (Map, Annex 4). Unregistered small and cottage industries are more widely dispersed among small towns and villages, the Colombo district representing only about 30% of the units. 2/ While unregistered small firms are characterized by wide dispersal, the major product lines are concentrated in varying districts, depending upon location of raw materials or market access. In each of the major ten SCI product lines, which account for 60% of the 15,500 units in the 300 product lines defined by IDB, three districts contain over 50% of the units. 2.07 Subsectoral Distribution. The main SMI product groups, constituting about 90% of the employment and 85% of the investment in registered SMIs, are textiles and garments, light engineering, rubber and chemical goods, wood and paper products (Annex 1, Table 1). Most LIAC approvals during 1978-80 have been in these groups, as well as building materials in response to rapid growth in public and private construction. Product concentrations among unregistered SCIs are found in related lines: handlooms, pottery, jaggery, blacksmithing, sheet rubber, rice mills, and bricks. However, several of these traditional products have been unsuccessful in competing with local or imported items manufactured by labor-intensive modern firms. 2.08 Prospects for Local Markets. Imports of consumer, intermediate and investment goods all have increased rapidly during the post-1977 period, growing by 36%, 30%, 3/ and 39% respectively in the 1978-80 period. Import substitution prospects in capital goods probably are limited to a portion of machinery, equipment, spares and building materials which constituted 65% of the US$492 million in capital goods imported in 1980. Intermediate goods consist mainly of petroleum, fertilizer and chemicals, in which StMs have no role. Most of the consumer goods imported during this period consisted of food products and textiles. Agroindustries have significant scope for SMI development with growth depending largely upon expansion of agricultural output. I/ Based upoIn a 40% sample of registered firms in 1977. 2/ Source: Industrial Development Board, Survey of Unregistered Manufacturing Establishments in Sri Lanka, 1976. 3/ Excluding petroleum imports.  6- D. Industrial Policy Framework 2.09 GOSL has made major revisions in industrial policies since 1977, replacing most import quotas and licencing with tariffs, and improving invest- ment and export incentives for foreign and local entrepreneurs. However, while recent modifications in both export incentives and import tariffs have been made, improved measures are needed to structure incentives and tariffs on the basis of comparative advantage and local value added. 2.10 Import Liberalization and Tariff System. The original tariff struc- ture, introduced in late 1977, was based upon: a low duty band covering essen- tial consumer goods, most raw materials, spares and machinery; a higher duty band for those goods considered to be non-essentials or for which adequate local capacity exists to meet domestic demand; and a prohibitive rate on certain luxury items. Subsequently, revisions have been made by a Tariff Review Committee chaired by the additional Secretary of Trade, largely based upon representations by local manufacturers. A top-level Presidential Tariff Review Commission recently has been established to replace the Committee. Initially, the Commission showed tendencies to favor arguments to increase protective tariffs, in spite of evidence that most private industrial firms have responded well to liberalization (para 2.04). A recently completed study of the tariff system, with financing for consultants provided under the first SMI Project, indicates a wide variance in rates of effective protection, ranging from -50% to over 1,000%. The study team is preparing recommendations for phased revisions in the tariff system to reduce the variation in effective protection rates (EPR), beginning with areas in which rates are at the extremes. These recommended changes would incorporate GOSL's revenue considerations and protection priorities for key subsectors. During negotiations for the proposed Second SMI Project, the status of the EPR study and plans to use the findings in the Government's tariff reform program were discussed, with understandings reached regarding a phased action program and continuing dialogue with IDA (para 4.37). 2.11 Export Incentives and Promotion. GOSL has made major moves to pro- vide attractive incentives for export-oriented industries, particularly those involving foreign investment. 1/ However, investment approvals have been dominated by firms fabricating wearing apparel, since other light industries require systematic assistance in product adaptation, supply development, and market promotion (para 2.03). The Export Development Board, established in August 1979 as the executive arm of the Export Development Council of Ministers, is charged with recommending and implementing export development policies and programs (para 2.20). While EDB gained credibility by providing ad hoc incen- tives and grants to selected export industries, the Board has recognized the need to develop a more systematic program of export incentives, based upon 1I The specific incentive packages for enterprises under the Greater Colombo Economic Commission (GCEC) and Foreign Investment Advisory Service (FIAC) are outlined in Chapter V of the 1981 Country Economic Memorandum for Sri Lanka. Firms approved by GCEC for location on or off Investment Promo- tion Zones receive somewhat better tax benefits than FIAC-approved units, but GCEC firms must commit to export 100% of their output.  - 7 - value added. An Export Expansion Scheme has been approved which would provide cash payments on increases in exports of non-traditional products; these payments would defray part of the product and market development costs in expanding exports. However, problems exist in the rough means of determining value added, which is intended as the basis for the Scheme, and in the funding source for the cash payments. The effective protection rates, now available (para 2.10), could provide a more accurate basis for determining value added in non-traditional export industries. Also, GOSL intends to finance the scheme by adding 10% to import tariffs on items for which tariffs of at least 50% already are charged. This method could cause further distortions in protection. It would be preferable to link revenue generation to a tariff reform program which could reduce variations in effective protection rates while increasing revenues which could be used for export bonuses. These issues have been raised with EDB and other GOSL officials; during negotiations of SNI II, agreement was reached that, by June 30, 1982, EDB would furnish to the Association for comment, a report analyzing the appropriateness of using effective protective rates to calculate value added under its Export Expansion Scheme. 2.12 Public Sector Enterprises other than the Petroleum Corporation have caused reduction in the overall growth in the industrial sector (para 2.02). GOSL has maintained a posture of encouraging private industry while reducing public financing available for public sector enterprises. However, until recently, public corporations have been able to escape this "financial dis- cipline" by increasing commercial bank borrowings and suppliers' credits. Performance in a few public corporations has improved through management and technology, and, in a few cases through private management contracts. However, overall performance has been poor. With these public corporations constitut- ing over 40% of industrial investment and 33% of value added, more concerted efforts to improve financial discipline and performance are needed. GOSL now has taken measures to control non-government financing of the public corpora- tions. Under a UNDP-financed, IDA executed project preparation facility for the National Development Bank, preparation of some projects involving the pub- lic corporations is anticipated. An IDA credit involving NDB direct financing of industrial projects is being prepared; under this project, it is anticipated that public corporations would be financed with conditions applied to help ensure improved performance. E. Financing SMI 2.13 Since liberalization, credit has expanded in all sectors. From June 1979 to June 1980, institutional credit to the private sector increased by nearly 70% in nominal terms. 1/ However, a large portion of private sector credit went for trade, particularly imports. 2/ Industrial credit represented 1/ Public development projects were funded largely through the budget, while public corporations, until recent curbs, relied largely on suppliers' credits and other forms of external commercial financing. 2/ Advances for commercial purposes represented 50.5% of total advances at June 30,1980.  -8- about 25% of total commercial bank advances at June 1980; while short-term advances continued to dominate, medium- and long-term credit showed a modest increase in share, from 19.6% of industrial advances at June 1979 to 21.4% by June 1980. Loans to textile, food processing, chemical, and metal industries constitute over 90% of private industrial credit. 2.14 The present Government has initiated several new measures to faci- litate growth in industrial and export credit. The National Development Bank launched operations in late 1979. NDB has established sound direct lending and SMI refinance operations; however, NDB has not yet played a major role in financing public sector corporations. During 1980, operations of the Sri Lanka Export Credit Insurance Corporation were launched to facilitate pre- and post-shipment credit. Also, the Central Bank has initiated a refinance scheme at subsidized interest for term loans to enterprises which would export at least 20% of their output value. Difficulties exist with this scheme, in using interest rates as an incentive and in differentiating between good intentions and actual ability to export; recently, Central Bank refinance under this scheme has been curbed. During 1980, GOSL approved opening of Sri Lankan operations by foreign banks; to date these banks have concentrated on trade financing, foreign exchange operations, 1/ and term financing of joint ven- tures. These measures are geared to increase availability of short and long term finance, particularly for non-traditional export enterprises. 2.15 With credit effectiveness of the first SMI Project (SMI I), the IDB- Commercial Bank Joint Scheme was replaced by an arrangement under which the credit institutions assumed full responsibility for term lending to SMIs. Previously, IDB formulated and appraised SMI projects and recommended them for financing by one of the two public commercial banks. This scheme did not work well. Project appraisal by IDB was weak and People's Bank and Bank of Ceylon, not willing to rely on project evaluation by another institution, rejected the majority of IDB's proposals. Responsibilities for supervision were unclear, while the banks took the repayment risks. Now, these two public banks, the two local private banks and DFCC are responsible for appraisal, disbursements, supervision and collections of SMI loans. Performance by these five credit institutions has been satisfactory. Term credit approvals under the first SMI Project has totalled about US$14 million during the first two years of operation, and the credit institutions now have about 150 trained SMI officers working full time on SMI lending. 2.16 Credit for SMI, which accounts for about 10% of lending to private industry, has increased in response to the favorable investment climate and the improved facilities for term credit. Commercial banks still lend most of their own resources in short term advances; however, provision of refinance by NDB, effective incentives, and training have resulted in substantial increases in the banks' term lending to SMIs. Demand by existing and new SMIs for term loans has been high, with only initial advertisement by NDB and the PCIs. With sizeable growth prospects in agroindustries, light engineering, and 1/ Of the 21 commercial banks operating in Sri Lanka, 18 have established Foreign Currency Banking Units; total assets/liabilities increased from US$22 million at December 31, 1979 to US$211 million at December 31, 1980. Of the US$180 million in advances, 89% were for under 6 months, with less than 6% having maturities of over 3 years.  - 9 - construction as well as in a range of export-related lines, demand for credit by SMIs is expected to remain strong. As Annex 2, Table 2 indicates, demand for term credit by SMIs during the 1981-84 period is estimated to be roughly US$370 million equivalent. However, the credit institutions still are build- ing their SMI project appraisal and supervision capabilities. About US$107 million of this potential demand for credit is expected to be met from insti- tutional sources; the Second SMI Project would constitute US$35 million, with US$28 million in IDA financing for 2,000 subloans. Other institutional sources would include: NDB refinance using repayments from the first and second SMI Projects, commercial bank credit, mainly for working capital, using internal resources; and Central Bank refinance. The size of the lending component of the Second SMI Project is geared to the capacities of the credit institutions; the Project would involve substantial, continuous institution-building elements, and if commitments and disbursements are more rapid than projected, processing of a third project could be accelerated. 2.17 Interest Rates and Inflation. Annex 2, Table 3 contains the interest rate levels in Sri Lanka at the time of negotiations. Lending rates increased during 1979-80, reflecting accelerated inflation. Commercial bank interest rates to prime industrial borrowers increased from 15% in late 1979 to 18% to 19% for the same group. NDB and DFCC have kept their rates relatively low, at 17%. Inflation measured by the GDP deflator was 24% in 1980. GOSL has taken several measures to curb inflation, cutting or deferring expenditures on public development projects, curbing spending by public corporations, and reducing investment levels in relation to GDP. Provided that continued measures are taken, inflation could decline over the next four years, averaging 16%-17%. Given this situation, the initial interest rate for SMI term loans would be set at 18%. These rates would be reviewed regularly and revised automatically, according to changes in commercial bank rates to industry, to maintain positive real interest rates in relation to medium-term inflation projections (para 4.13). F. Technical and Marketing Services 2.18 The Industrial Development Board, an agency of the Ministry of Industries and Scientific Affairs, is the principal SMI development institution, responsible for promoting and assisting all SMI subsectors except textiles and cottage industries. Under the first SMI Project, IDB has been responsible for implementing subsector-specific schemes in rubber products, building materials and light engineering; also, IDB is establishing a subcontracting exchange to facilitate purchasing links between small and larger firms. IDB has faced problems with the exodus of officers to private as well as other public insti- tutions. IDB management plans to upgrade its performance by improving career paths for extension officers, strengthening its regional services through intensive coaching, and tapping private sector know-how in addressing tech- nical needs of its SMI clients; IDB would be assisted in these steps by the Second SlI Project (para. 4.29). 2.19 The Ministry of Textile Industries (MTI) formulates policies and implements programs for the textile and garment subsector; under the first SMI Project, MTI is establishing a garment training institute with outside collaboration and participation of the Garment Exporters' Association. The  - 10 - Department of Textile Industries, under MTI, has been successful in transfer- ring most of its operations to the private sector, selling its powerlooms and, under SMI I, fomenting the establishment of Lanka Handlooms Export Private Ltd. to organize decentralized production and exports of handloom products. The Department of Small Industries (DSI), under the Ministry of Rural Indus- trial Development, is responsible for services to cottage industries; while DSI has made some progress in developing the white coir industry, most of its programs deal with promotion of handicrafts for local sale. 2.20 The Export Development Board (EDB) is the executive arm of the Export Development Council of Ministers, headed by the President. EDB was established to: advise GOSL on formulation of export development policies and programs; do research and development on export products and markets; and act as the central agency for export and related supply development, coordinating the activities of key public and private agents. The EDB has been meeting since its legal establishment in August 1979 2/ but major staff recruitment and operations were launched in early 1980. The strong management and financial position of the EDB have enabled it to perform well during its first two years of operation. Management of the EDB places strong emphasis on mobilizing private sector exporters and related manufacturers in expanding exports of light industrial products. Several private associations and advisory committees of exporters and manufacturers in major product groups have been organized; EDB intends to use these private groups to implement subsector development schemes. III. STRATEGY IN DEVELOPING LIGHT INDUSTRIES 3.01 Export and local market potential exists in several light industrial product groups. However, the industrial base in Sri Lanka is thin, and several key constraints will have to be addressed if this potential is to be exploited. In non-traditional exports, products with high value added need to be adapted, quality improved, effective market links established, and operations expanded to meet volume and price requirements. Management in light industries and the institutions serving them needs to be strengthened, with management more than money hindering growth and productivity improvements. GOSL has taken several steps to address these constraints but Government agencies face increased difficulties in attracting and retaining staff to provide technical, managerial and marketing services to SMIs and other key light industrial groups; many of the best personnel are hired into the growing private sector. The Second SMI Project would assist in: promoting exports and supply development of selected light industrial products; improving management of SMIs and the institutions serving them; and mobilizing private knowhow in providing technical and marketing services to SMIs. 1/ In addition, the National Institute of Business Management offers problem- oriented seminars for SMIs in key subsectors; these courses are financed partially under SMI I. The National Apprenticeship Board arranges super- vised, paid apprenticeships with private industries in a range of trades. Several research organizations exist, but the orientation and experience of most staff is academic and practical output is limited. 2/ Under the Sri Lanka Export Development Act No. 40 of 1979.  - 11 - A. Expanding Exports of Light Industrial Products 3.02 Based upon analyses of markets and supply capabilities for non- traditional exports, the EDB has identified several light industrial product groups with significant export potential, provided that improvements in qual- ity, design and organization are made. Major lines include: rubber products; wood items; gems and jewelry; handloom textile goods; frozen seafood; white coir items; light engineering products; cut flowers and plants. These product groups are characterized by high value added; most depend upon local raw materials and large labor inputs. By addressing the constraints blocking growth, export potential could be tapped in meeting employment, output and foreign exchange objectives. Rubber Products 3.03 Characteristics. Sri Lanka exports a sizeable quantity of raw rubber, in sheet and block form; rubber exports are second only to tea, representing 20% of foreign exchange earnings and employing 500,000. However, less than 5% of this rubber is used by local manufacturers. There are about 200 registered private rubber products firms, of which all but 10 are SMIs, and several hundred unregistered units. Most units have less than Rs 200,000 in plant and equip- ment; fixed costs per job (book value) average about Rs 8,000 (US %30); and value added/capital averages 2.1 vs. 1.2 in larger units (Annex 1, Table 1). Registered private units produce mainly tubes, footwear, rubber toys, rubber- ized coir products, gloves and elastics. Export totalled about Rs 14.0 mil- lion in 1980 vs Rs 0.2 million in 1975. Sri Lanka has a comparative advantage in manufacturing latex and dry rubber products; labor intensive processes are suitable for most products 1/ and raw materials represent 55% to 60% of costs. 2/ While rubber must compete with synthetics in some product groups, rubber has superior characteristics for several items, including sporting goods, gloves, and rubber bands. Also, increased costs of petroleum-based synthetics give advantages to natural rubber. SMIs can be competitive in several product groups, including footwear, toys, sporting goods and rice rollers. However, importers require sizeable volumes and strong assurances of delivery reliabil- ity and quality; SMIs may be successful subcontractors, if larger scale export- oriented rubber products units can be developed. 3/ To date, only about ten joint ventures in rubber products have been approved by GCEC or FIAC, with only three in or nearing commercial production. Credit for rubber products 1/ Sri Lanka has labor cost advantages, even relative to Malaysia which has been successful in expanding its rubber product exports. With a 4:1 ratio of unskilled to skilled workers, absorption of unskilled workers is high. 2/ Dry rubber and latex, which constitute 60% to 85% of raw material costs are available locally at about 50% of international prices and with liberalization non-rubber inputs can be imported with 5% duty. 3/ The minimum economic size for units making most rubber products is under Rs 1 million.  - 12 - SMIs under the first IDA project totalled Rs 5.0 million for 17 units at December 31, 1980. Major expansion of rubber products manufacturing capacity and much needed replacement of equipment are expected to take place only once effective technical and marketing services have been provided. 3.04 Some important steps have been launched. IDB has improved its tech- nical services and raw material supplies to rubber products manufacturers; the first SMI Project funded intensive foreign technical training for IDB's key rubber products staff; these officers have conducted diagnostic studies of improvements needed in SMIs and have begun providing needed technical services. IDB is constructing a major Rubber Products Quality Testing and Product Development Center, financed under the first SMI Project; staff trained under the project are providing technical services, and the Center is expected to become fully operational during CY81. In promoting joint ventures, EDB has sponsored the completion of detailed feasibility studies in five promising rubber product lines. These studies need to be followed up with active promotion, identification of appropriate joint venture partners, and technical services to upgrade local capabilities. The Second SMI Project would assist in these measures (paras 4.43 and 4.44). White Coir Products 3.05 Production of white coir fiber, yarn, and woven products is well suited to rural SMIs due to dispersed land, water supply, and raw material availability. White coir fiber has advantage over brown fiber with earnings per metric ton of Rs 20,000 vs. Rs 3,000. White fiber exports totalled about 3,000 tons or Rs 20 million in 1980. Potential is much larger, 1/ with as many as 100 million green husks available which could be collected eocnomi- cally. Major constraints in increasing the output of white coir fiber and yarn are that: (a) the long period required and limited space available for retting (soaking) the husks limit output; (b) the danger and uneconomically high costs to decorticate and defiber the husks make the occupation unpopular and costly; and (c) poor processing and spinning operations make a substantial portion of existing yarn unsuitable for mechanized processing. 3.06 Assistance. If these problems are addressed, the EDB calculates that exports of production using white coir yarn could be increased from 3,000 mt to 15,000 mt by 1985 which would result in incremental rural employment of 10,000, and increased annual foreign exchange earnings of Rs 204 million. The EDB is sponsoring about 5 pilot projects on the west coast to develop this decentralized production. These pilots would be implemented by joint stock companies with equity participation by manufacturers and exporters of white coir fiber and products. Under these pilots, lagoons would be upgraded and crushers added to reduce retting time from nine months to one or two months; simple mechanized equipment would be introduced for decorticating, squeezing and unravelling; and spinning wheels would be improved. About Rs 1 million 1/ Factories constructed to manufacture white coir products are underuti- lized. Mechanized capacity for white coir mats is 5,000 mt with approved projects contributing another 5,000 mt. However, only 50% of existing capacity is utilized due to limited availability of white coir yarn.  - 13 - would be required for fixed assets and permanent working capital for each pilot; the EDB would provide a portion of the equipment costs. The first pilot has been 'Launched, and the major coir companies are preparing proposals for subsequent projects. The Second SMI Project would help finance similar pilots sponsored by EDB (para 4.34). Gems and Jewelry- 3.07 Problems and Prospects. In 1979, official exports of gems and jewelry totalled about Rs 494 million (US$30 million). 1/ While these exports are substantial, significantly larger export earnings from gems and jewelry are possible. Gem mining is done by small operations. Not all of the gems and none of the precious metals needed for jewelry are available locally; the import of gold and silver was banned, until recently, for two decades; and the private sector has depended upon sporadic Government import and supply of basic metals. As a result, black market imports have been rampant. In gem cutting and polishing, quality control is lacking. Cutting techniques are crude; while experts deem that Sri Lankan craftsmen are talented and quick learners, tools are primitive and craftsmen have little knowledge of modern manufacturing methods or designs. As a result, stones are crudely cut which reduces the gem value. Lack of selective mechanization and poor methods have resulted in long labor time on each item which reduces price competitiveness of products. Most designs produced do not suit export market tastes, either for classical or fashion items. Lack of organization among the many small scale miners, craftsmen, jewelry manufacturers and exporters has limited the ability to introduce improvements in equipment, design techniques, and quality control, or to produce the necessary volumes to meet minimum export orders. 3.08 Assistance Needs. Some steps have been taken to address these problems. The EDB has sponsored about three man years of outside technical assistance to the gem and jewelry industry. While a substantial portion of this assistance has gone to the State Gem Corporation (SGC), some measures have been initiated which should help development in the private sector. Most importantly, a Jewelry Manufacturers and Exporters Association has been esta- blished, which was sponsored by EDB to help make the industry more responsive to inputs on improved techniques, product adaptation and marketing. Secondly, the ban on imports of precious metals has been lifted. SGC trainers trained on design and techniques are expected to sponsor short courses for private sector craftsmen. Further assistance is needed. Marginal improvements in equipment could have major impact on quality and productivity; foot- and hand-operated bench and dye presses, and simple cutting machiLies could be manufactured in Sri Lanka. Tool and dye making capacity needs to be expanded; while the gem and jewelry industry would not provide a sufficient base for major expansion, these basic service industries also are needed for metal products, plastics and ceramics. IDB has developed engineering common facil- ities which will provide some of the needed back-up services; expansion of private capacity also is needed. Other major areas requiring immediate atten- tion are: equipment for assaying and quality control; improved purchasing of a range of metals rather than a standard dimension of sheet metal which 1/ These registered exports are thought to represent a small portion of total exports.  - 14 - needs to be recast, milled or drawn by each craftsman; and training of exporters, jewelry manufacturers, and craftsmen on market requirements for a broadened product range. The Second SMI Project will support EDB in meeting marketing and technical service needs, with project elements implemented by a private association (paras 4.33 and 4.34). Also, Stage II of IDB's engineering common facilities would be financed under the Second SMI Project. Small gem and jewelry manufacturers would be eligible for loans under SMI II, to make the necessary improvements in tools and equipment. Handloom Products 3.09 Potential and Needs. The handloom industry has had difficulty in competing with local and imported milled fabrics for standard domestic cloth requirements; local sales have stagnated at about 35 million yards annually. However, export prospects for specialized textiles and simple made-ups are promising. EDB's handloom specialist has identified significant scope for Sri Lanka in exports of: light furnishing fabrics, particulary drapery; table linens such as napkins and placemats; and handkerchiefs of handloom cloth. 1/ In these lines, Sri Lanka would not compete directly with Indian handloom exports but rather would focus on smaller orders of higher quality items. There also is scope in supplying handloom shirting fabric to Sri Lanka's export garment industry; however, competing with large scale Indian handloom factories will be difficult. Market agents, Lanka Handloom Exports Pvt. Ltd., 2/ and a public company are gearing themselves to organize the decen- tralized supply base for exports. These companies as well as production units will need to strengthen their skills in: producing quality products; adapting designs to western fashion trends; and securing small orders which can be increased as production capabilities expand. While exports of handloom fabrics are small, recent sales efforts in Europe focussing on light furnishing fabrics and simple made-ups have met with market interest and initial orders. 3.10 Assistance. EDB plans to assist in developing handloom exports by organizing and channelling funds for training in improved handloom production. The University of Moratuwa has begun providing short training to production supervisors of private and cooperative units, with outside expertise provided by EDB's handloom consultant. The project could support an expansion of this local training activity by helping fund the chief consultant, trainers, and equipment. Also, an international course in handloom design is proposed; a curriculum combining theoretical and practical work with exposure to trade fairs has been developed; and an appropriate sponsoring instutition has been located. About eighteen private sector designers would participate from Sri 1/ More detailed analysis of product/market possibilities is provided in D. Holbourne's report to the Export Development Board, August 1980, (Annex 3). 2/ Lanka Handloom Exports Pvt. Ltd. is a private limited company, composed of major exporters and manufacturers of handloom products; this company has been launched under the First SMI Project.  - 15 - Lanka, Bangladesh and Nepal. 1/ Funding would be available for partial equity financing of commercial dye houses to achieve standardization and color- fastness of dyed yarn and cloth. Finally, partial financing would be provided for exporters' participation in fairs and sales trips, for companies with cap- abilities to produce strong designs, reliable quality and sufficient volumes. Wood Products 3.11 Potential and Constraints. Exports of wood products grew from Rs 700,000 in 1975 to Rs 28.6 million in 1979. A recent study for the EDB 2/ indicates that production capability, suitable for export markets, exists in: quality wooden toys and games; handcarved tableware; kitchenware; masks and figures; colonial furniture; and wood components. Traditional furniture woods such as ebony, mahogany, and nadun are in scarce supply, while other timber also is needed by the construction industry. Competition from con- struction and the scarcity of appropriate woods of wide girth make expanded furniture production difficult. Partial solutions to the timber constraint include: (a) expanded production of boron-treated rubberwood by IDB and others; (b) increased chipboard production by the Plywood Corporation to meet needs of the construction industry; (c) utilization of waste wood from the Mahaweli and other public works projects; and (d) concentration on small items with high value-added such as wooden toys and household items. 3.12 Assistance Needs. Quality standards, delivery reliability, produc- tion efficiency, and designs suiting export tastes all require improvement. While markets are strong, these aspects need to be tackled before major expansion in wood product exports can take place. Prior to sponsoring an exhibition or participation in fairs, EDB needs to use consultants to advise major manufacturers on wood technology, export quality standards, kiln drying and selected modifications in equipment and methods, to cut costs and improve quality. EDB now has information matching wood product manufacturers in Sri Lanka with interested European buyers; this should be used by the EDB in making contacts to develop joint ventures for marketing and transfer of technical and design know-how. B. Building Private Technical and Marketing Services 3.13 Government institutions have useful roles to play in: providing effective investment and export incentives for industry; identifying and promoting promising product lines; and filling some gaps in technical ser- vices to help improve productivity. However, government institutions have difficulty providing effective consultancy and commercial services needed for SMI development, due to problems in attracting, retaining, and motivating 1/ Participants from Bangladesh and Nepal would be sponsored under IDA- financed small industry projects. 2/ Sri Lanka: Wood Products, Alchemade Product Development Ltd., July 9, 1980.  - 16 - suitable managerial and technical staff. Higher salaries are available in the private and banking sectors and civil service rules make it difficult for a government agency to offer attractive performance incentives. In contrast, private commercial agencies have a direct, concrete interest in maximizing efficiency. The comparative advantage of public institutions is good inten- tions in developing certain target groups for economic and social reasons, rather than focussing only on short term profitability. However, public agencies are likely to continue suffering from the rapid migration of its best talent to the private sector. The challenge then, is to have government agencies tap the know-how available in the private sector, making it profit- able for efficient private organizations to meet development objectives. Some of this can be achieved by "getting the prices right" through export and investment incentives and tariff revisions. However, more direct managerial, marketing and technical services also are needed to address key constraints blocking growth in SMIs and other key export-oriented light industries. The Second SMI Project would assist GOSL in tapping private capabilities to meet these assistance needs. 3.14 SMI Consulting. The buoyant climate in industry, commerce and construction has lured a sizeable number of engineers and other technically trained Sri Lankan personnel to return from abroad. Several consulting firms and individuals, established during the past two years, are providing technical assistance to larger industrial enterprises; others are concentrating on civil engineering activities. The Institute of Engineers has about 6,000 registered engineers working in Sri Lanka, some of which would be interested in full or part time consultancy. The newly established Association of Consulting Engi- neers is an indication of this burgeoning activity. While a large number of the mechanical, electrical, civil and chemical engineers have time and skills relevant to improving productivity and sales of SMIs, many SMIs do not appre- ciate the potential benefits of practical consultancy, and most are still too small to readily pay commercial rates. Under the Second SMI Project, the Industrial Development Board would establish an SMI consultancy fund under which private local consultants would be hired to resolve specific technical problems of small firms. This measure would help channel private know-how into addressing the needs of SMIs and would assist in establishing ongoing commercial links between SMIs and private consultancy (para 4.31). 3.15 Supply Base for Exports. Sri Lanka has made rapid entry into garment exports; substantial potential exists for adding value to traditional tea, rubber and coconut products; and frozen seafood sales will continue to grow if deep sea prospects are exploited and quality standards are improved. However, Sri Lanka's thin industrial base creates constraints in rapid export development in most product lines. Major foreign exchange needs will be met only if Sri Lanka also takes advantage of growth potential in some of the smaller product groups, many of which are based upon production by small enter- prises. Prime examples are: cut flowers and plants; white coir products; spices and essential oils; cotton handloom products; certain handicrafts; gems and jewelry; wood products; and light engineering items. EDB intends to mobilize larger scale manufacturers, commercial agents and other private entities to organize and expand the decentralized production base for exports. In the white coir pilot projects, for example, the EDB is paying part of the costs which qualified private agents incur in making critical investments to modernize and coordinate growth in segments of this decentralized industry.  - 17 - Similar schemes would be appropriate for several of the other above-mentioned products. The Ceylon Tobacco Corporation, as part ot its diversification moves, is providing credit, inputs, and extension services to about 6,000 household orchid growers. Under the Second SMI Project, EDB would continue to promote this "trading company" concept, recognizing the importance of private agents in upgrading processes and products and organizing decentr-l- ized producers. Also the EDB has been successful in fomenting private asso- ciations and steering committees in main industrial export groups, inciuding jewelry, rubber products, white coir fiber, and handloom products. These associations are to be used in channelling most EDB sponsored consulLancy, technical services, and export promotion ventures; active involvement by exporters and manufacturers should help ensure that these inputs address real needs. C. Improving Management of Institutions Serving SMI 3.16 Sri Lanka faces major foreign exchange and budgetary constraints. However, the less obvious but more fundamental problem blocking growth is weakness in management--in public, private, and banking institutions. Local private industry has been tightly held, family owned; this tradition has limited growth and those firms which have grown often have not developed delegation of responsibility or management information systems. Public and private institutions often have impressive, articulate top management, but middle management often is thin and most operating staff members do not have obvious management potential. These weaknesses at the middle ar-d lower levels, most severe in public institutions, are rooted in: the deterioration of the educational system; poor career development policies; and failure to develop the needed management systems. These problems need to be addressed directly if economic objectives are to be achieved. The Second SMI Project would tackle some of the problems of organization and management among institutions serving SMIs. IV. PROPOSED SECOND SMI PROJECT A. Objectives, Scope and Institutional Arrangements 4.01 Objectives and Scope. The objectives of the proposed project would be to increase output, employment, efficiency and exports of viable SMIs, building on institutional capabilities developed under the first Project, while increasing emphasis on light industrial exports. The project would be national in scope; a significant portion of the credit and technical services would be expected to take place in secondary industrial towns. New and existing SMIs would be eligible for loans, provided that their projects are financially and economically sound. Major subsectors demanding credit probably would continue to be agro-industries, light engineering and construction-related enterprises; and, as a result of technical services of SMI I becoming fully operational, lending to rubber products, handloom, and subcontracting firms also is expected to increase. Loans from participating credit institutions and technical services  - 18 - would focus on firms with plant and equipment of less than Rs 4 million (US$200,000) with the majority of firms assisted having fixed assets of under US$20,000 equivalent. Export development elements would be provided in selected light industries, regardless of firm size. The project would tap private knowhow in providing commercial and technical services to SMIs, while building capabilities of banking and government agencies. Also, improvements in the tariff structure and export incentives are expected, using information generated through studies financed under the first SMI Project (para 4.37). 4.02 Components, Costs and Institutional Roles. Major elements of the proposed Second SMI Project would be: (a) term loans to SMIs from participating credit institutions (PCIs) with subproject review and partial refinance by the SMI Fund of the National Development Bank (paras 4.04-4.25); (b) training and consultancy for the SMI Fund of NDB and the PCIs, to improve SMI and overall operations (paras 4.15-4.16); (c) improvements in IDB's promotion and extension services, with coaching of regional extension officers, an SMI consultancy fund to tap private sector knowhow, and expansion of engi- neering service facilities (paras 4.29-4.32); (d) support for EDB's export promotion and supply development pro- grams, including consultancy and training, sales trips, policy studies, and pilot projects involving exporters and manufacturers in key light industrial groups (paras 4.33-4.34); and (e) funds for expansion of successful subsector schemes launched under SMI I and II (para 4.35). 4.03 Financing Plan. Total project costs would be about US$52.5 million equivalent over a three year commitment period, with IDA providing about US$28.0 million equivalent for refinance of SMI subloans and US$2.0 million for technical and marketing services. The participating credit institutions would contribute about US$7.0 million; SMI subborrowers, US$15.0 million; and GOSL, US$500,000. NDB would provide temporary bridge financing to cover the lag between SMI I and SMI II (para 5.09). 1/ NDB would not be asked to increase capitalization of the SMI Fund prior to project implementation, but would provide additional capital, if needed, to cover the lag between NDB's disburse- ments and its withdrawals from IDA. The IDA credit would cover: up to 100% of NDB's refinance, or 80% of each eligible SMI subloan made by a participating credit institution; and 100% of eligible expenditures for local and foreign consultants, foreign training and trips, and equipment for pilot projects and common facilities. GOSL's contribution would cover costs of land and buildings, incremental salaries and overheads, and a portion of the export promotion programs. The following table summarizes project costs and sources of funds: 1/ Since bridge financing is not a permanent expenditure, it is not reflected in the cost table, to avoid double counting.  - 19 - Project Financing Plan (US$'000) IDA IDA GOSL PCIs SMIs Total Foreign Local Subloan Component Term Loans for SMIs 17,000 11,000 - 7,000 15,000/a 50,000 Technical Services Industrial Development Board 410 210 200 - - 820 Export Development Board 550 250 200 - - 1,000 Training & Consultancy - NDB, Credit Institutions 190 140 - - - 330 Expansion Fund 200 50 100 - - 350 Subtotal, Technical Services 1,350 650 500 - - 2,500 TOTAL 18,350 lb 11,650 500 7,000 15,000 52,500 /a The US$15 million equivalent in subborrowers' equity to contributions is an estimate, based upon the assumption that sponsorsr equity would represent 30% of subproject costs under SMI I; minimum equity contributions would be 20% of any subproject (para 4.11). /b The foreign and local portions of IDA contributions are estimates and not allocations. B. Lending Component 1. Small and Medium Industry Fund - NDB 4.04 NDB Operations. The National Development Bank of Sri Lanka (NDB) was established in January 1979 to finance industry, agriculture and commercial enterprises, through direct loans, equity, and refinance. NDB issued Rs 600 million in initial share capital: 67% to GOSL, 17% to the Central Bank, and 8% each to the two state-owned commercial banks. 1/ The Ministry of Finance and Planning, the Central Bank and the two state-owned commercial banks are represented on NDB's Board. The organization has six departments: two for direct lending; one for SMI refinance; 2/ and three for support functions. 1/ NDB's authorized share capital is Rs 2 billion. Of the issued share capital of Rs 600 million, Rs 450 million has been paid in, leaving on call Rs 150 million, or 75% of the contributions of the Central Bank and the commercial banks. 2/ NDB management is considering expansion of its refinance operations to incorporate other industry groups, including large scale construction enterprises.  - 20 - Department heads report to the general manager, who is responsible through the chairman to the Board. NDB's management has been assisted by consultants financed under the first SMI Project. A senior advisor, whose two year term expires in October 1981, has helped formulate overall policies, institutional arrangements and operating guidelines. ICICI, under contract until December 1981, will have provided about 48 manmonths of assistance, training NDB staff in appraisal and helping establish operating standards and procedures for NDB's direct lending operations. An advisor to the SMI Fund, whose one year term has been completed, assisted in establishing SMI refinance operations and procedures. These consultants contributed substantially to the initial development of NDB. 4.05 NDB has made significant progress in developing an institutional structure, recruiting staff 1/ and establishing satisfactory operating stan- dards and procedures. By December 31, 1980, NDB had approved direct loans totalling Rs 280 million (US$16.9 million) to sixteen projects, 2/ with equity participation of Rs 6.3 million (US$350,000); Rs 77 million (US$4.7 million) had been disbursed. To date, NDB's direct operations have been financed by equity funds; however, the Asian Development Bank recently signed a US$10 million loan to NDB and an IDA Credit is under preparation. 4.06 NDB commenced refinance operations in November 1979, with the effec- tiveness of the first SMI credit. To date, refinance has been limited to subloans made by the PCIs under the first SMI Project. The SMI Fund was established by NDB with separate policies, staffing, financial resources and accounting. By July 31, 1981, the SMI Fund's refinance approvals totalled Rs 192.0 million (US$7.2 million) to about 1,500 subprojects. About Rs 102.9 million (US$5.9 million) had been disbursed by NDB. During 1981, NDB expects expansion in its SMI refinance operations with funding under the proposed Second SMI Project. Also, NDB may establish a fund for refinance of larger loans for construction equipment. 4.07 Organization and Staffing. The SMI Fund, headed by a manager, is organized into three main sections. The analysis and review section reviews appraisals and refinance applications of the five PCIs. The reporting and liaison section prepares periodic reports on the credit institutions' lending, supervision and collection performance. The finance and disbursement section disburses refinance, collects repayments, administers SMI project funds for the technical service components, and maintains accounts. Professional staff positions of the SMI Fund have consisted of the manager, three section heads, six analysts and three other officers. The SMI Fund has had difficulty recruit- ing and retaining senior staff, due both to shortages in experienced staff in Sri Lanka and the greater appeal of NDB's direct lending operations. Despite staffing problems and higher than expected refinance activity, performance of the SMI Fund has been quite satisfactory under the first SMI Project. As anticipated, the emphasis during the SMI Fund's initial period of operation was on developing suitable procedures, reviewing refinance applications, and 1/ At July 31, 1982, NDB's staff numbered about 40 professionals. 2/ The total project costs were Rs 735.3 million (US$44.6 million) including Rs 302.8 million (US$18.4 million) in foreign exchange.  - 21 - upgrading PCIs' appraisal standards. Now, monitoring and supervision require increased attention. The SMI Fund intends to formalize its training programs for PCIs. Also, the Fund would make subproject review more systematic by: establishing norms for key aspects of appraisal, including appropriate levels of permanent working capital, land, and buildings; preparing subsector analyses to avoid over-investment and to identify market opportunities; and investigating sources of appropriate equipment. Consequently, under the Second SMI Project, three new sections would be added for monitoring and supervision, training, and subsector analysis. 1/ 4.08 With this reorganization, additional professional staff are needed for the new sections and to fill vacancies in the existing sections. Nineteen officers would need to be in place, with qualifications satisfactory to IDA, prior to effectiveness of SMI II: the manager of the Fund; the head of the analysis and review section and seven analysts; a section head and two officers for monitoring and supervision; a section head and two assistants for finance and disbursements; a senior officer and another officer for subsector analysis and two officers for reporting and training. NDB now has taken steps to fill the needed positions: sixteen of the needed nineteen officers have been appointed or are in place. Prior to credit effectiveness NDB needs to retain the services of the existing manager of the SMI Fund or find a suitable replacement with substantial experience in both project appraisal and in management within a commercial bank on industrial finance institution. Also, prior to credit effectiveness, NDB would have an appropriate head and senior officer for the supervision section in place. New staff would be given training and consultancy assistance (para 4.15). 4.09 Policies and Procedures. Prior to effectiveness of the first SMI credit, NDB adopted a statement of policies and operating procedures for the SMI Fund, in a form satisfactory to IDA. The statement, which provides the basis for the SMI Fund's operations, now requires minor modifications to reflect changes in eligibility criteria, terms, and procedures proposed for the Second SMI Project; a draft statement reflecting these changes was dis- cussed during negotiations with agreement in substance reached (Annex 3). The Sri Lanka delegation reported that the revised statement had already been approved in substance by NDB's Board of Directors. Prior to credit effective- ness, NDB's Board would need to ratify the statement, including terms and conditions for refinance and onlending, which were agreed during negotiations; any subsequent modifications would require IDA approval. Procedures for review- ing SMI appraisal reports 1/ in support of the PCIs' refinance applications 1/ Subsequently, the six sections would be grouped under two new divisional managers. One would be responsible for operational aspects: analysis and review; monitoring and supervision; and subsector analysis. The other wouLld oversee administrative aspects: finance and disbursement; reporting; and training. 2/ The two standardized appraisal forms (one for subloans above Rs 500,000 and one for subloans below that amount), agreed with IDA and now in use are satisfactory. In addition, the Fund, in consultation with the credit institutions, will prepare a simplified form for subloans of below Rs 20,000.  - 22 - have been satisfactory. The SMI Fund would continue to grant refinance on subloans of below Rs 50,000 (US$2,500) without a detailed review; the review section would check to see that subprojects meet the eligibility criteria, appear financially viable, and have appropriate terms; more detailed post- approval reviews would be conducted of a sample to ensure maintenance of standards. Subloans of above Rs 50,000 would continue to be subject to a more detailed review of technical, financial, marketing and economic aspects, with the SMI Fund completing the review within two weeks. The Rs 50,000 limit could be increased by the Fund, after consultation with IDA, for those PCIs which demonstrate consistently strong appraisal standards. After reviewing subloans of over Rs 1,000,000 (US$50,000) NDB would submit the appraisals and refinance requests to IDA for review and approval. The existing reporting system would be used to aggregate quarterly reports of the credit and tech- nical service institutions, for submission to IDA; also this section would submit evaluations of SMI operations and impact. The supervision section would make periodic visits to head offices, branches and SMIs to review standards, procedures, staffing and supervision performance. The SMI Fund's revised policy statement also would specify the responsibilities of each section (para 4.07). 4.10 Eligibility for Refinance. PCIs eligible for refinance would have signed participation agreements with the SMI Fund of NDB. The existing par- ticipation agreements would be modified and new agreements signed to reflect changes in the eligibility criteria, terms and conditions, and procedures of the Second SMI Project. The five credit institutions involved in SMI I would participate; of these, DFCC's involvement is expected to remain limited due to alternative financing on more attractive terms, lack of branches, and concen- tration on lending to larger industrial projects. Interested foreign banks also could be eligible, provided that they comply with staffing and other requirements; inclusion of any foreign bank would require agreement by NDB, GOSL and IDA, after an appraisal of the institution's Sri Lankan operations. However, participation of foreign banks is considered unlikely, given their emphasis on trade financing and foreign exchange operations, and their lack of branches. 4.11 Subprojects eligible for refinance, as under SMI I, would involve the creation, modernization or expansion of private or cooperative enter- prises engaged in manufacturing, mining, construction, agro-industry, fish processing, handicraft production or industrial services. 1/ The maximum enterprise, subproject and subloan sizes would be doubled to reflect inflation and enable financing of more substantial SMI subprojects. The SMI Fund would provide 80% refinance of subloans for enterprises with total fixed assets (excluding land and buildings) with original costs of up to Rs 2 million in fixed assets 2/ prior to the subloan; the maximum subproject size would be 1/ Defined as repair workshops and other industrial services which add value to a product, and export enterprises with direct involvement in organizing the decerntralized production base. Transport would be eligible only if it forms part of an industrial enterprise. 2/ Only the portion of land and buildings needed for the SMI subproject would be part of a subloan eligible for refinance; the subsector analysis section of the SMI Fund would develop norms on this aspect.  - 23 - 4 million (US$200,000) excluding permanent working capital. 1/ Sponsors would be required to provide at least 20% of total subproject cost in equity; based upon experience under SMI I, sponsor's contributions are expected to average 30%. The maximum subloan size would be Rs 2 million (US$100,000) including fixed assets and permanent working capital. 2/ All subprojects would need to be financially viable. Regarding economic viability, PCIs would continue to assess whether a subproject would improve quality, price competitiveness, or production of goods for which demand has been established. For more common subprojects, the credit institutions would use standard project profiles for reference. The SMI Fund would do subsector analysis on product group's eco- nomic viability. The credit institutions would be expected to incorporate simple indicators of economic impact within their subproject appraisals, such as expected fixed costs per job and percentage of raw materials of local origin. 4.12 Refinance Arrangements. Equity participation is expected to average 30% of subproject costs; refinance would cover about 56%; and the PCIs would contribute 14%. The PCIs would continue to assume full repayment risk on refinance, on a fixed schedule determined at the time of approval. To cover administrative costs of SMI operations and provide PCIs with an incentive to lend to smaller firms, differential spreads based upon subloan size would continue to be provided. The refinance rate would be adjusted to provide the PCIs with a gross spread 3/ of 5% p.a. for subloans of up to Rs 500,000, and 4% for subloans of Rs 500,000 to Rs 2 million. These spreads, slightly higher than the 3% to 5% spreads under the first project, are justified in view of the added emphasis on subproject supervision and the needed impetus to a further reduction in PCIs' security requirements. NDB would receive a 2% to 3% spread depending upon subloan size (para 5.02). 4.13 Terms and Conditions. Refinance granted by the SMI Fund would con- tinue to be denominated in Rupees and onlent by the credit institutions at a standard interest rate. The initial onlending rate would be 18%, which represents a significant increase from the 15% onlending rate under the first project and which is reasonable in view of medium term inflation projections (para 2.17). The rate would be reviewed every six months, beginning on July 1, 1982, with automatic adjustment by an equal number of percentage points to any 1/ Subprojects would consist of subloan and equity portions (including permanent working capital) to establish an SMI or for balancing, modernization and expansion of existing firms; the subproject amount would not include assets of a firm prior to the subloan. In the case of a Rs 4 million subproject, the subloan could cover up to Rs 2 million, with equity covering the balance. 2/ Since eligible subloan sizes for subprojects financed under IFC's second line of credit to BOC would be US$85,000-US$1 million, the overlap with SMI II would be minimal. 3/ These spreads are prior to the PCIs paying a premium to the Central Bank for coverage under the SMI credit guarantee scheme; the guarantee premium would be 1% on the guaranteed amount outstanding of the subloan (para 4.16).  - 24 - change in the rate which People's Bank charges on its loans to large industry. Only if this "prevailing rate" were to fall below 18% would the SMI onlending rate fall, with parity between the project onlending rate and the prevailing rate achieved,, and maintained for the commitment period of the project. These rates also would apply the 20% of each subloan not eligible for NDB refinance under SMI II. Maturities on eligible SMI subloans would continue to be for 3 to 10 years, including grace periods ranging from 3 months to 2 years. The PCIs would establish appropriate grace periods and maturities, based upon each subproject's debt service ability; the Fund would review these maturity and grace periods. 4.14 Projected Financial Results. The resources of the SMI Fund would comprise: capital and interest from NDB's initial contribution of Rs 20 million, repayments on the US$12 million IDA credit component of SMI I; and withdrawals from IDA for US$28 million in refinance under the Second SMI Project. Largely as a result of repayments under SMI 1, the SMI Fund is expected to have a cumulative surplus of Rs 29.8 million by the end of CY81, and Rs 41.9 million by the end of CY82. With SMI Fund repayments to GOSL beginning only in 1984 (Year 5), the SMI Fund is expected to generate sur- pluses which would be utilized to cover: the lag between refinance by the SMI Fund and withdrawals from IDA; bridge financing requirements if gaps occur in outside funding; and refinance to PCIs for SMI subloans, utilizing resources of the SMI Fund. It is anticipated that the SMI Fund would not require additional capital infusions from NDB during the disbursement period of SMI II; however, NDB would commit to making additional funds available, if required, to cover needs for bridge financing, administrative costs, or the gap between the SMI Fund's disbursements and withdrawals from IDA. Detailed financial projections for the SMI Fund are provided in Annex 2, Table 4. 4.15 Institutional and Training Needs. During negotiations, the Chair- man of NDB indicated that outside training for eight new SMI Fund officers had been arranged with the State Bank of India and would be completed by December 31, 1981. Funding for this initial foreign training is available under the first SMI credit. The technical service component of SMI II would include provisions for: overseas training of three to five SMI Fund staff yearly (para 4.26); and 12 man-months of consultancy for on-the-job training of SMI Fund staff in project review, supervision of banking operations and subsector analysis (US$60,000). This consultancy probably would be under- taken by an institution with extensive experience in these areas. 1/ In addi- tion, assistance would be provided to the subsector analysis section, probably by local consultants (US$50,000). The project also would continue support of NDB's overall operations by providing 18 man-months of assistance by ICICI in technical appraisal, project promotion and other institution-building (US$90,000). 1/ This would be similar to NDB's consultancy arrangement with ICICI for NDB's direct lending operations.  - 25 -- 2. Credit Guarantee Scheme 4.16 Under the Second SMI Project, the Central Bank's SMI Credit Guaran- Lee Scheme would continue to provide partial coverage of eligible dMI suloaI-s; to encourage the PCIs to reduce collateral requirements; coverage would l,e restricted to subloans refinianced under the SMI Project. Guarantee prelviums, which would come from the PCIs' spreads, would be 1% of the outstandi-nog amou;nt guaranteed. I/ Modifications in the scheme would be made prior to creCdt effectiveness to streamline procedures and to reflect the increased s