Colombia - Fourth Small-scale Industry Project

Document of The World Bank FOR OFFMCIAL USE ONLY i; } > j't' jj: ! 4 ' ReportNo. 10152 PROJECT COMPLETION REPORT COLOM-BIA FOURTH SMALL SCALE INDUSTRY PROJECT (LOAN 2464-CO) DECEMBER 9, 1991 Trade, Finance and Industry Division Country Department III Latin America and the Caribbean Region This document has a restricted distribution and may be used by recipients only In the performoance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Currency Unit - Colombian Peso (Col$) Average Rate During Year of: 1984 US$1 - 100.8 Col$ 1985 US$1 - 142.3 Col$ 1986 US$1 - 194.3 Col$ 1987 US$1 - 242.6 Col$ 1988 US$1 - 299.2 Col$ 1989 US$1 - 375.1 Col$ 1990 US$1 - 502.3 Col$ GLOSSARY OF ABBREVIATIONS BR - Central Bank of Colombia CDT - Term Certificate of Deposit CF - Private Development Bank CFP - Public Small Scale Enterlrise Development Bank DFC - Development Finance Company DTF - Average 90-day Time Deposit Rate FFI = Industrial Finance Fund PCR - Project Completion Report SAR - Staff Appraisal Report SSI - Small and Medium Scale Industries SSI3 Third Small Scale Industry Project SSI4 Fourth Small Scale Industry Project SMES - Fifth Small and Medium Scale Enterprise Project SW Staff-weeks FISCAL YEAR OF THE GOVERNMENT OF COLOMBIA January 1 - December 31  FOR OFFICAL UJSE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A Office of Dirootor-Gonoral Operaions Evaluation December 9, 1991 NEMORANDUN TO THE EXECUTIVE DIRECTOR. AND THE PRESIDENT SUBJECT: Project Completion Report on Colombia - Fourth Small Scale Industry Project (Loan 2464-CO) Attached, for information, is a copy of a report entitled "Project Completion Report on Colombia - Fourth Small Scale Industry Project (Loan 2464-CO)" prepared by the Latin America and the Caribbean Regional Office. No audit of this project has been made by the Operations Evaluation Department at this time. Attacbment ThS decumnt m a rmb4ebd dsdhbutea =A ma be ud by Icpkta o*b tIn d fpoaace of t&ek oelI dis. Its cottldDa no ot eiwis. be d _cha without Wold Bank authorftioa.  FOR OFFICIAL USE ONLY PROJECT COMPLETION REPORT COLOMSIA FOURTH SMALL SCALE INDUSTRY PROJECT (LOAN 2464-CO) TABLE OF CONTENTS Pate No, PREFACE .. ............................................... EVALUATCON SUMHARY ....................... 9*99** **99*4 *** 999999 9Al PART I - PROJECT REVIEW FROM TilE BANK'S PERSPECTIVE ..............1 1.1 Project Identity ............. 1 1.2 Barkground 1 1.3 Project Objectives and Desription..cription........... 2 1.4 Project Design and Or gan lzation 3 1.5 Project Implementation.......................... . ....... 5 1.6 Project Results and Sun tainabll4ty 7 1.7 Bank and CFP ...... ..... .. 13 1.8 Conclusions.... 15 Table 1.1 Profile of All Subloansb....... 17 Table 1.2 Profile of Subloan Sample Used in Ex Post Evaluation ....... v .............................. 21 Table 1.3 Key Performance Indicators of CFP.P........... 23 PART II - PROJECT REVIEW FROM CFP'S PERSPECTIVE................. 25 2.1 S y.. .. .. 25 2.2 Application of Loan Proceeds.o ce......... e ds.......... 25 2.3 Action Programo....... 25 2.4 Evaluation of Performance............................ 28 PART III -STATISTICAL INFORMATION ......................... 29 Table 1: Related Bank Loans ... 29 Table 2: Project Timetable ............... ... 29 Table 3: Estimated and Actual Disbursements.......*..... 30 Table 4: Project Costs and Financing..................... 31 Table 5: Project Result..s............... ............... 32 Table 6: Status of Loan Covenants........venant.s...... 33 Table 7: Use of Bank Resourceso................. 35 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.  PRO.JECT COMPLETION REPORT COLOMBIA FOURTH SMALL SCALE INDUSTRY PROJECT (LOAN 2464-CO) PREFACE This is a Project Completion Report (PCR) for the Fourth Small Scale Industry Project (SSI4) in Colombia for which Loan 2464-CO of US$40.0 million was approved by the Executive Directors on July 10, 1984. US$38.2 million was disbursed by thi original closing date of December 31, 1988 and the loan account closed in August 1989 with total loan disbursements reaching US$38.7 million. US$6.1 million had been repaid as at February 28, 1991 by the Corporacion Financiera Popular (CFP), the only financial intermediary for the Project, through the Borrower, the Colombian Central Bark (Banco de la Republica, or BR). The PCR was prepared by the Trade, Finance and Industry Division of Country Department III of the Latin America and Caribbean Region. The CFP has submitted its own assessment of the Project through BR, which has been translated from Spanish to English by the Bank. Excerpts of its report, which exclude only factual data found elsewhere in the PCR, is presented as Part II of this Report. The PCR was compiled on the basis of information available in the LAC Information Center, the Staff Appraisal Report, reports produced by CFP, a report o a consultant engaged by the Bank whose selection and terms of reference were agreed upon with BR, correspondence with the Bank, and staff Supervision Reports.  - ill - PROJECT COMPLETION REPORT COLOMBIA FOURTH SMALL SCALE INDUSTRY PROJECT (LOAN 2464. C0) EVALUATION SUMMARY Proiect Obiectives 1. The Fourth Small Scale Industry Project (SSI4) was prepared in 1983 following four years of marginal growth in Colombia's industrial output and of declining exports. The average performance of SSIs'/ over this period was generally comparable to overall trends and they accounted for about two fifths of the industr.a6al labor force and produced nearly one fourth of the manufac- turing value added. The Bank had made three previous loans totaling US$52.5 million to the Central Bank of Colombia (BR), which on-lent the funds to the Government's special SSI lending institution established in 1967, the Corporacion Financiera Popular (CFM). These loans supported a total of about 2,300 SSI investments and institutional improvements in CFP. Though CFP had made progress in its own development, it continued to suffer particularly from erratic growth, high portfolio arrears, high administrative costs and low return on equity (paras. 1-4). 2. The Project aimed to accelerate employment creation and value added by SSIs in Colombia, mainly by increasing the availability of term credit and technical assistance to SSI enterprises. Its specific objectives were to: (a) provide the SSI subsector with an adequate amouat of medium-term investment credit; (b) help to strengthen the quality and delivery of technical assistance services to SSIs; (c) strengthen the institutional, technical and operating capabilities of CFP, in particular the regional offices; and (d) assist the Government to design and coordinate policies affecting SSI development. Other purposes of Bank intervention mentioned were to support the recently adopted system of "market-related" onlending interest rates by CFP to SSIs and to provide financing in the absence of alternative sources of long-term funds for CFP. The :redit component of the Project was supported by US$39.5 million of the US$40.0 million Pank loan. The remainder was to finance a portion of an estimated US$1.1 million in technical assistance to support an Action XI Notwithstanding the Project's title, small and medium scale industries were included in SSI4 and were defined as firms with total assets not exceeding US$750,000 equivalent.  - iv - Program designed to strengthen the institutional and organizational structure of CFP (paras. 5-7) Imilementation and Results 3. Loan commitments and disbursements for subprojects went substantially according to appraisal estimates. This was facilitated in particular by generally favorable macroeconomic management leading inter alia to significant aggregate demand in the economy, the well-established institutional capability of CFP to reach SSIs, and the relatively streamlined mechanism for subloan commitments and disbursements of the Bank's loan (para. 15). 4. The Project had a measurable effect on the availability of credit to Colombia's smaller business market. The loan financed 2,211 subprojects and increased loan volume by about 129% in real terms over the previous project. Real effective interest rates for fixed asset financing averaged 3.7-5.72 per annum (depending upon business location) and more than 902 of the subloan commitment volume provided repayment of 4 or more years. Significant progress in interest rate reform was made toward the end of the Project as Project and directed credit subloan rates were fully linked to the variable deposit rate on commercial 90-day deposits (the DTF) and set at DTF + 1 percentage point to SSIs (paras. 21-22). 5. About 95% of the subprojects (85% by loan volume) were undertaken by enterprises defined in Colombia as "small" (under 50 employees). Over half of loan commitments were to industrial sectors with moderate to high average labor intensity, with subprojects concentrated most heavily among firms operating in agro-industry (24.5%), metal products and machinery (21.7%) and textiles and clothing (18.4Z). A large share of the loan volumie went to SSIs outside Colombia's three principal cities, reflecting the substantially greater geographical dispersion of SSIs across Colombia than for industry overall. Support for new enterprises exceeded expectation (25% actual vs. 152 estimated) which, though evidence is not available to determine how many of such investments were by new entrepreneurs, very probably served to help build and further develop such human skills in the economy. Those subprojects supporting export promotion were very few (2%), reconfirming much evidence which shows that access to capital is necessary but not sufficient to overcome adverse trade and industrial policies and an absence of support programs (para. 23). 6. An ex post evaluation of a sample of 22 subprojects gives the impression thaL there was a wide variety of performance with respect to the frequency of design modifications, implementation delays, and the longer run impact on enterprise efficiency and income. Where information was available, installed capacity increases were estimated at between 45-100% for about one- half of the sample; re-estimated IRRs were 12-45% for about half of the investments and negative for the remainder, and the employment generated resulted in a growth in total enterprise employment of about 22Z. However, uniform information was frequently not available and sales were considered by interviewers to be frequently substantially understated, understating the revised IRRs and making these performance indicators not very reliable (paras. 24-28).  - v - 7. For the most part, the programs and studies included in CFP's Action Plan proceeded satisfactorily throughout the project implementation period. Monitoring and management of regional offices was improved with better trained personnel and more specific performance programs for each office regarding lending, supervision, portfolio management and technical assistance. This "institutional strengthening" was expected to manifest itself mainly in higher portfolio growth, better portfolio quality and lower unit costs. By the end of the implementation period in 1988, however, the net effect was between marginally positive to substantially negative in these areas, yielding no material progress toward the objective of CFP becoming an autonomous, financially self-sustaining institution to serve SSIs. CFP's annual lending in real terms virtually stagnated over 1984-88 (0.6% average per year). Though CFP's administrative costs in terms of average total assets showed improvement over 1984-88, this did not improve its already low after-tax profitability due mainly to the effects of increasing arrears. In addition, it did not reduce its high dependence on the Government to replenish its capital base and on domestic directed credit and external credit lines (paras. 31-39). Findings and Lessons Learned 8. The Project's design appears to have been well understood and accepted by those primarily concerned, the CFP and BR. This was due undoubtedly to experience acquired from the substantial number of previous DFC projects involving both entities as well as to the considerable Colombian and Bank staff time invested prior to Project start-up. While Project objectives included support for SSI technical assistance and broad public policy changes. No such significant measures were identified as taken by the Government or under the Project itself during project implementation. Project design and resources focused almost exclusively on narrower financial sector issues. Intensive attention was given to increasing the portion of the subloan interest rate to SSIs which was variable. This was a relatively new concept for such credit in Colombia and was justifiably the most important and innovative measure to be taken at that time. The Project also focused heavily on CFP's institutional development and was motivated by the Government's success through CFP in increasing term lending to SSIs with a satisfactory economic impact, the gradual development of CFP's lending and supervision capabilities, and the need to reduce the unit cost of CFP lending (paras. 8-11). 9. Prolect design did not, however, adequately address the issue of the sustainability of the CFP as a financial institution. CFP's profitability ree(rd, which did not enable it to maintain the real value of its equity, and its growing dependence on Bank resources were recognized during Project appraisal. However, no effective mechanism to bring about fundamental institutional change was incorporated in the Project given its overriding objective of providing finance to SSIs. The result was that while a long- lasting impact was achieved under many SSI investments, the Project did not increase the country's institutional ability to sustain financial intermedia- tion to support such investments. In addition, the Project did not substan- tially deal with the main market failures in SSI lending, which perhaps both the Borrower and the Bank became more sensitive to later in the 1980s: the absence of financial margins for SSIs lending at least equivalent to the opportunity cost of not lending to large scale clients; the need for credit access rather than interest rate subsidies through the involvement of a larger segment of the financial sector in SSI lending; and the negative effects on  - vi - access, credit processing time and cost created by highly restrictive directed credit regulations. A subsequent SSI project (SME5) and the Government's financial sector reform prepared in consultation with the Bank are modifying the policy framework in these areas. This project is incorporating SME lending programs into a multitude of more diversified financial institutions and initial indications are that they are responding well and increasing SSI credit access. Further restructuring or privatization may, however, be required to achieve sustainable CFP operations (paras. 12-14 and paras. 47-51).  PROJECT COMPLETION REPORT COLOMBIA FOURTH SMALL SCALE INDUSTRY PROJECT (LOAN 2464-CO) PART I - PROJECT REVIEW FROM THE BANK'S PERSPECTIVE 1.1 Project Identity Project Name: Fourth Small Scale Industry Loan Number : 2464-CO Loaa Amount : US$'O.O million equivalent RVP Unit : Trade, Finance and Industry Division Latin .merica and the Caribbean Region Country * Republic of Colombia Subsector : Finance and Industry 1.2 Background 1. The Fourth Small Scale Industry Project (SSI4) was prepared in 1983 following four yeara of marginal growth in Colombia's industrial output (12 average over 1978-82) and declining exports, owing ostensibly to weak domestic and international demand, an overvalued Peso, increased quota restrictions and tariff rates, and competition from contraband. The average performance of SSIsB/ over this period was generally comparable to overall trends and they accounted for about two fifths of the industrial labor force and produced nearly one fourth of the manufacturing value added. 2. Moderate optimism underpinned Project preparation, as the overall environment for SSI development was considered to be improving: the rate of Peso devaluation and explicit export incentives were increasing, protection was expected to be reversed, and the Government was giving particular attention to the construction, assembly, and capital goods industries. Also, existing public instruments for SSI development -- mainly a directed credit fund, a specialized public development bank (the Corporacion Financiera Popular, or CFP), and several public technical assistance agencies, were to be reinforced by a new Government initiative. A newly created Advisory Council composed of public and private entities concerned with SSIs was to prepare guidelines for financial and technical assistance to SSIs and to review the potential for subcontracting mechanisms between larger and smaller enterprises, Government procurement practices, industrial parks and policies toward five subsectors in which SSIs predominated. 3. Enterprise-specific Bank support to Colombian industry up to 1983 was associated with steel and nickel exploitation enterprises, an export processing Notwithstanding the Project's title, small and medium scale industries were included in SSI4 and were defined as firms with total assets not exceeding US$750,000 equivalent.  -2- zone, and urban microenterprises. The Bank also had provided eight credit linas totaling US$493 million to the Central Bank of Colombia (Banco de la RepAblica, or BR) to finance larger industry through private development finance companies (DFCs) and had made three loans totaling US$52.5 million to BR which on-lent the funds to thte Government's special SSI lending institution established in 1967, thb CFP. These latter loans had supported a total of about 2,300 SSI investments at a total cost of about US$130 million. CFP's institutional performance had improved under previous projects, though it continued to suffer particularly from erratic growth, high portfolio arrears, high administrative costs and low return on equity. 4. Like its predecessors, SSI4 set out to address the development constraints perceived to be of particular importance to Colombian SSIs. The foremost concern was their limited access to "conventional" commercial financing. In the early 1980s, most SSI financing was sourced from Government directed credit, the Industrial Finance Fund (FFI), and about two-thirds of all (mostly long-term) formal banking credit to SSIs in Colombia was intermediated by the CFP. Also, technical assistance programs for SSIs were con3idered to be inadequate, supplies of raw materials insecure, and enterprise technology and management systems inappropriate. 1.3 rLoject Obiectives and Description 5. Against this background, the Project was viewed as one instrument to help to raise the contribution of SSIs to employment creation and value added in Colombia. It mainly employed a strategy of increasing the availability of term credit and technical assistance to SSI enterprises. Four specific Project objectives were highlighted at the time, namely to: (a) provide the SSI subsector with an adequate amount of medium-term investment credit; (b) help to strengthen the quality and delivery of technical assistance services to SSIs; (c) strengthen the institutional, technical and operating capabilities of CFP, in particular the regional offices; and (d) assist the Governmerc to design and coordinate policies affecting SSI development. Other purposes of Bank intervention mentioned were to support the recently adopted system of "market-related" onlending interest rates by CFP to SSIs and to provide funding in the absence of -lternative sources of long-term funds for CFP. 6. The credit component of the Project was supported by US$39.5 million of the US$40.0 million Bank loan. It was to finance SSI investment expendi- tures on machinery, equipment, spare parts and related services; the purchase of industrial buildings in industrial parks and construction of industrial facilities, civil works, and related permanent working capital needs. Subloans carrier a blend of two interest rates to SSIs: 15% of a subloan would carry a rate of the average local 90-day market time deposit (the "DTF" rate) plus three percentage points, and the remainder would carry a 24Z-26X nominal  - 3 - (^g2-30% effective)&' fixed. The former rate was applied to borrowers outside of Colombia's three main cities and the latter to those within them. The subloan share to which the variable rate applied was to be renlegotiated with the aim of increasing it on January 1 of 1986 and 1987 and the fixed rate was to be reviewed semi-annually with the aim cf ensuring that it was positive in real terms and compatible with other local rates (particularly those on FFI funds). The loan carried a financial margin for CFP of 4.5-5.0 nominal percentage points (8.52-9.0% effective). 7. US$0.4 million of the Bank loan (US$0.1 million financed the capitalized front-end fee) was to finance a portion of an estimated US$1.1 million in technical assistance to support an Action Program agreed upon with the Bank during loan negotiations and approved by CFP's Board of Directors as a condition of loan effectiveness. The Program was designed to strengthen the institutional and organizational structure of CFP by building its regional office network, carrying out a plan to improve subproject appraisal and supervision, carrying out a feasibility study on mobilization of resources by CFP from the market, reducing the unit cost of lending and increasing recovery of delinquencies in the portfolio. The loan proceeds (at a cost of 122 interest charged to CFP) were to finance the purchase of computer terminals for CFP's regional offices, and CFP was to cover the remaining costs of the technical assistance. 1.4 Prolect Design and Organization 8. From all available iniormation, the Project's design appears to have been well understood and accepted by those primarily concerned, namely the CFP and BR. This was due undoubtedly to experience acquired from the substantial number of previous DFC projects involving both entities as well as to the Colombian and Bank staff time invested prior to Project start-up (43.7 Bank staff weeks (SW) through Project appraisal according to the MIS and 5.3 SW thereafter through Board approval). 9. While Project objectives included support for the Government's SSI technical assistance and broad public policy changes (see para. 2), Project design focused almost exclusively on narrower financial sector issues. While the Project Staff Appraisal Report (SAR) briefly described the former areas, there were no Project-related resources or activities dedicated to them. Those entities involved in the Project (both in Colombia and the Bank) had responsi- bilities and orientations mainly associated with the financial sector. To have seriously addressed these non-financial issues in a fashion which today's current Bank standards require would have required substantiully more time and effort on the part of both parties as well as other institutions (e.g., Ministry of Labor, National Planning Department, SSI producers association, etc.) and would have resulted in a far more complex Project design. 21 Interest rates in Colombia are quoted in nominal and effective terms. As interest on loans is paid at the beginning rather than at the end of the period by deducting it from the proceeds of the loan, the "effective" interest cost on funds actually received by the borrower is higher than the "nominal" rate charged.  -4- 10. Within the finar.cial sector, intensive attention was given to increasing the portion of the subloan iTt:arest rate to SSIs which was variable. This was a relatively new concept for such credit in Colombia and was justifiably the most important and innovative measure to be taken at that time. Viewed by today's benchmark for Colombia, the target for the level of interest rates was low -- the real positive fixed rate of 24-26% charged in 1984 is at least 10 percentage points lower in real terms than the market rates charged today for Bank funds under the Fifth Small and Medium Scale Enterprise Project (SME5). 11. Unlike the trends in past DFC projects but in step with current expectations described by the Levy Report and more recent Bank guidelines, the Project also focused heavily on CFP's institutional development. In fact, while there are a total of six pages of SAR devoted to rSI characteristics, Government policies, the overall financial sector and technical assistance, another twelve pages present a detail4od analysis of CFP's structure and systems, prc edures and operations, resources and financial performance. Attention to institutional dettil was high and preparation work was thorough in these areas. This design approach, as well as overall Project design, seems to have been motivated mainly by three lessons from previous SSI Projects (which were duly noted in the SAR): (a) the Government's success through CFP in increasing term lending to SSIs with a satisfactory economic impact, (b) the gradual development of CFP's lending and supervision capabilities; and (c) the need to reduce the unit cost of CFP lending, in particular by expanding the role of its regional offices. 12. Project design did not, however, seriously address two important areas which were beyond these confines: the sustainability of CFP as a financial institution and other alternatives to address the market failure of lending to SSIs. While the SAR recognized that CFP's profitability levels did not enable it to maintain the real value of its equity, it nevertheless assessed its financial performance as "reasonably satisfactory" given its focus on smaller SSIs and development activities for which it was not remunerated. It also assessed the quantitative mobilization of resources by CFP as adequate, though noted concern over the significant fluctuations in their availability and a continuing dependence on public domestic agencies and the Bank for them. 13. Specific consideration was apparently given to channelling Bank funds through commercial banks and private development banks (CFs). It was considered that the National Guaraustee Fund, which had been established recently to help overcome insufficient collateral problems, and increasing financial margins under the Project and FFI-supported lending might make SSI lending more attractive. Nevertheless, unlike the trend across the Bank which had been employed significantly since 1978,'1 inclusion of commercial banks as financial intermediaries in the Project was rejected because of (a) a trend toward specialization in a recent Colombian financial sector reform, (b) the limited branch network of private CFs, and (c) CFs' reluctance to finance smaller SSIs. It See "World Bank Support for Small and Medium Industry in Selected Countries" (page 9), Report No. 9530 of April 22, 1991 by the Operations Evaluation Department.  14. The final design of SSI4 was thus shaped largely by an acceptance of the specialization concept in the financial sector, a policy focus almost entirely on interest rates and an expectation of sustainable progress in CFP's institutional development given especially the positive growth and portfolio quality trends in the years immediately preceding preparation of SSI4. However, preparation work did not substantially address the main market failures in SSI lending which perhaps both the Borrower and the Bank became more sensitive to later in the decade: the absence of financial margins for SSIs lending at least equivalent to the opportunity cost of not lending to large scale clients; Lhe need for credit access rather than interest rate subsidies through the involvement of a larger segment of the financial sector in SSI lending; and the substantial negative effecta on access, credit processing time and cost created by highly restrictive directed credit regulations. 1.5 Proiect Implementation 15. The loan was signed about two months after Board approval and made effective about four months thereafter. Loan commitments and disbursements for subprojects went substantially according to appraisal estimates. The Bank's loan was fully committed in September 1987, several months ahead of the December 31, 198? Commitment Closing Date. This was facilitated in particular by generally favorable macroeconomic management leading inter alia to significant aggregate demand in the economy, the well-established institutional capability of CFP to reach SSIs, and the relatively streamlined mechanism for subloan commitments and disbursements of the Bank's loan. Disbursement of the loan also ran smoothly and, for most of the implementation period, slightly ahead of appraisal estimates. The original Closing Date was December 31, 1988 and final loan disbursement was made in March 1989. 16. The loan financed 2,211 subprojects, noticeably above the 1,500-1,700 estimated at appraisal (Table 1.1). Average subproject size was markedly below appraisal estimates (US$46,500 vs. US$65,000 equivalent at 1983 prices). The average subloan was US$25,600 (1983 prices) with about 25Z of the loan volume composed of subloans of less than US$12,500 each. Tlwo-thirds of the Bank loan finenced machinery and equipment, another 172 supported construction costs, and the remaining 17% covered primarily related permanent working capital needs. Subloans above the free limit (US$100,000 equivalent) were actually more than double appraisal estimates: 81 subprojects accounting for 221 of the loan vs. the estimated 35 projects accounting for about 102. Ex ante review of subprojects under the free limit averaged an estimated 10.8 SW per year and processing of subprojects within the free limit averaged 1.0 SW per year of the responsible division in the Projects Department (now LA3TF) during the loan commitment per:5od, or about twice the Bank staff time which would have resulted had the appraisal estimate been met. 17. Subloan terms effectively softened in relation to those of the previous Bank-financed Project (SS13), given a ripe in inflation and an extension of average maturities. Real effective interest rates under SS14 for fixed asset financing averaged 3.7-5.71 per annum (depending upon business location), a decline from 7.71 under SSI3. Although BR did not agree to increase the variable portion of the subloan interest rate according to the schedule expected, significant progress was made toward the end of the Project with the Government's 1987 reform of interest rates for directed credit.  - 6 - Project and FFI subloan rates were pegged fully to the DTF and set at DTF + 1 percentage point to SSIs. Subloaa maturities were ample, particularly in response to the effects on SSI finances of the recessionary years and to the rising level of payment arrears. More than 90% of the subloan commitment volume under SSI4 provided repayment of 4 or more years (65Z under SSI3), and about one-third exceeded 6 years (11% under SSI3). 18. While frequent CFP management turnover was identified as a Project risk, management continuity was assured throughout most of Project implementa- tion with the same General Manager heading CFP from the period of Project preparation in 1983 through the end of 1986 when a new Government was elected. For the most part, the programs and studies included in the Action Plan proceeded satisfactorily throughout the loan period. Delays in efforts to strengthen CFP's regional offices, another Project risk identified by the Bank, also did not materialize. Monitoring and management of regional offices was improved with better trained personnel and more specific performance programs for each office regarding lending, supervision, portfolio management and technical assistance. CFP used the US$400,000 provided under the loan to acquire 21 personal computers, development tools, and communications hardware and software for establishing the information system now used for decentralized data gathering. Some concern was expressed early in Project implementation about subproject appraisal, especially in the areas of market analysis, repayment capacity and procurement. Later supervision missions, however, noted improvements in CFP's appraisal and supervision activities. 19. CFP met accounting and audit report requirements, which reflected excellent presentations of financial statements and footnotes, and were qualified in only two (1985-86) of the five-year implementation period due to portfolio arrears above the guidelines of the Superintendency of Banks. However, the major variance against Project expectations was in CFP's financial performance, which was manifested inter alia through repeated non-compliance with financial covenants in the Project and Guarantee Agreements: (i) CFP's administrative expenses as a percentage of average total assets exceeded targets over 1984-86 (7.3% vs. 6.7%, 6.8% vs. 6.5%, and 7.32 vs. 6.0%, respectively); (ii) CFP's debt:equity ratio exceeded substantially the stipulated 6:1 limit over 1985-88 (6.2, 13.5, 16.7 and 8.6, respectively) and only thereafter fell within the limit following negotiation of SME5; and (iii) CFP's lending for special programs entrusted by the Government for CFP's administration, representing 3.3% of total 1984 CFP's lending and nearly 52 in 1986, exceeding the 2.5% stipulated in the Project Agreement. 20. Annual loan supervision work by the Bank was relatively high (averaging 19.1 SW over about 3.5 years of active supervision), and concern was expressed about these problems by the Bank on a number of occasions. Its efforts were focused mainly on encouraging CFP to establish more explicit operational performance targets for recovery of portfolio arrears and to increase its overall lending operations as well as on pressing the Government to meet its commitments to provide additional equity. However, given the Project's rapid loan commitment rate and the satisfactory execution of the  - 7 - Action Program, attention to CFP's overall financial performance progressively became an issue to be dealt with in the anticipated follow-up Project. 1.6 Prolect Results and Sustainabilitv 21. SSI Investment. The Project had a measurable effect on the avail- ability of credit to Colombia's smaller business market (Table 1.1). About two-thirds of the subprojects (40% by loan volume) supported very small formal sector enterprises (under 9 employees), and 952 (852 by loan volume) were undertaken by enterprises defined in Colombia as "small" (under 50 employees). Over half of loan commitments were to industrial sectors with moderate to high average labor intensity, with subprojects concentrated most heavily among firms operating in agro-industry (24.52), metal products and machinery (21.7Z) and textiles and clothing (18.42). 22. Bank financing under SSI4 increased by &bout 1292 in real terms over SSI3 and the number of subprojects increased by nearly 662 (900 investments) to reach a total of 2,211 subprojects. The Bank's loan was complemented by the mobilization of US$31.6 million equivalent by SSIs themselves and other domestic market sources. While this was a positive development, there is also evidence that the Bank's loan effectively substituted rather than complemented the use of domestic resources.41 Essentially, while the interest rates on the two lines were comparable, the many conditions and administrative requirements of the FFI motivated the banking system to prefer to use Bank resources. 23. A large share of the loan volume went to SSIs outside Colombia's three principal cities, substantially as expected (67% of the total vs. a 60Z appraisal estimate). This reflects the substantially greater geographical dispersion of SSIs across Colombia than for industry overall (two-thirds of all industrial plants are found within the three major metropolitan areas). Support for new enterprises exceeded expectation (252 actual vs. 152 esti- mated). Though evidence is not available to determine how many of such investments were b'y new entrepreneurs, it remains very probable that Project support served to both help build and further develop such human skills in the economy. Those subprojects supporting export promotion were substantially below even the small share expected (2X actual vs. 102 estimated). Quite clearly, and consistent with Bank experience in such projects elsewhere, access to capital is necessary but not sufficient to overcome adverse trade and industrial policies and an absence of support programs. 24. A detailed financial analysis was conducted on a small but representative sample of 22 subprojects, accounting for about 12 of the total number and 52 of the total Bank loan (Table 1.2). Data for this analysis was provided primarily by the CFP as part of its normal supervision process and supplemented by enterprise interviews. In many cases, however, CFP's data reflected mostly overall enterprise data rather than subproject specific 41 See the Staff Appraisal Report of the Colombia Fifth Small and Medium Scale Enterprise Project, paras. 2.12-2.15.  - 8 - monitoring indicators and financial statements as at Dacember 31, 1990 were unavailable.!' 25. The subproject sample reflected average cost overruns of 10.72 over estimated costs and average delays in subproject execution of 4.5 months (excluding one subproject which was delayed nearly 2.5 years). In reality, however, performance varied drastically. Absent more information from the ex post evaluation, it appears that nearly one-third of the sample experienced substantial design changes and/or delays in execution, and another one-quarter did not provide the information necessary to determine these parameters. 26. Twelve of the 22 subprojects in the sample increased installed capacity over a range oE 45% to 100%. About two-thirds of them succeeded in increasing both installed capacity and efficiency. In the cases where both the original and revised IRR were available, six had IRRs of between 12.0% and 43.6% while the remaining six had negative IRRe. These are not likely, however, to be very reliable performance indicators as, according to evaluation interviews, the full investment costs seem to be reported while sales tend to be considerably understated, thereby leading to substantially under re- estimated IRRs. This may be supported by the fact that four of the 22 subprojects displayed repayment problems while the remaining 18 cases had a normal debt service of their financial obligations to CFP. This rate of arrears is roughly consistent with the fact that CFP's overall arrears over 1986-90 averaged 13% of total portfolio. This corresponds approximately to a repayment rate of 872 of total principal and interest due and would rank CFP above the world-wide average for 33 completed Bank-financed SSI projects (80Z). However, this rate is noticeably below the average for such operations in the Latin America and Caribbean Region (92%).