Colombia - Small-scale Industry Project

Document of The World Bank FOR OFFICIAL USE ONLY FILE COPy Report No. 2645 THE WORLD BANK PROJECT PERFORMANCE AUDIT REPORT COLOMBIA - FIRST SMALL-SCALE INDUSTRY PROJECT (LOAN 1071-CO) August 30, 1979 Operations Evaluation Department 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.  FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT COLOMBIA - FIRST SMALL-SCALE INDUSTRY PROJECT (LOAN 1071-CO) TABLE OF CONTENTS Page No. Preface Basic Data Sheet Highlights iii - iv PROJECT PERFORMANCE AUDIT MEMORANDUM 1 - 17 I. Use of Bank Funds 2 - 5 II. Corporacion Financiera Popular 6 - 12 Organization 6 CFP's Overall Operations, Resources and Profitability 8 Control Procedures and Quality of Portfolio 11 Appraisal and Supervision Procedures 11 III. Technical Assistance 12 - 15 IV. Government Policies 15 - 16 V. Conelusions 16 - 17 Attachment: Project Completion Report I. Introduction 18 II. Objectives and Expectations of the Bank 19 - 20 III. Utilization of Loan Proceeds 20 - 30 Economic Environment 20 Commitments and Disbursements 22 Characteristics of Subprojects Financed 23 Results of Subloans 29 IV. Institution Building 30 - 40 Management and Staff 30 Organization 31 Systems and Procedures 32 Resources 34 Operations 35 Financial Position and Results 38 Quality of Portfolio 39 V. Sectoral Impact 40 - 41 VI. Conclusions 41 - 43 This document has a restricted distribution and may be used by recipients only in the performance of their oficial duties. Its contents may not otherwise be disclosed without World Bank authorization. -2- Page No. ANNEXES: ANNEX 1 Manufacturing Performance Indicators 44 ANNEX 2 Projected and Actual Cumulative Disbursements - Loan 1071-CO 45 ANNEX 3 Analysis of Subloan Approvals 46 - 47 ANNEX 4 Evolution of the Corporacion Financiera Popular's Staff at Headquarters and in the Regional Offices during 1976 through June 1978 48 ANNEX 5 Projected and Actual Loan Approvals and Disbursements for 1974-June 1978 49 ANNEX 6 Projected and Actual Balance Sheets as of December 31, for the Period 1974-78 50 ANNEX 7 Projected and Actual Income Statements at December 31, for 1974-78 51 ANNEX 8 Projected and Actual Financial Ratios for 1974-78 52 ANNEX 9 Aging of Arrears - Evolution between December 31, 1973 and June 30, 1978 53 ANNEX 10 Comparison of Projected and Actual Performance of Sub-borrowers 54 ANNEX 11 Comparison of Accuracy of Projections of CFP Subproject Appraisals 55 LIST OF ABBREVIATIONS ACOPI Asociacion Colombiana Popular de Industria BR Banco de la Republica CFP Corporacion Financiera Popular DANE Departamento Nacional de Estadistica FFI Fondo Financiero Industrial PROEXPO Fondo de Promocion de Exportaciones SENA Servicio Nacional de Aprendizaje SMI Small and Medium Scale Industry SPSS Statistical Package for Social Sciences SSE Small Scale Enterprise SSI Small Scale Industry USAID United States Agency for International Development PROJECT PERFORMANCE AUDIT REPORT COLOMBIA - FIRST SMALL-SCALE INDUSTRY PROJECT (LOAN 1071-CO) PREFACE This report covers an audit of achievement under Loan 1071-CO to the Banco de la Republica of Colombia for on-lending to the Corporacion Financiera Popular (CFP). The loan, in an amount of US$5.5 million, was intended to assist CFP's lending operations with the.small- and medium- scale industrial sector. It was approved in January 1975, declared effective in May 1975 and closed, disbursed at 98%, in November 1978, little less than a year after the original closing date. An OED mission visited Colombia in February/March 1979 and held discussions with officials of CFP and Banco de la Republica, as well as representatives of technical assistance agencies and business organizations. Assistance rendered to the mission during its visit is gratefully acknowledged. The audit memorandum is based on the attached Project Completion Report (PCR) prepared by the Bank's Latin America and the Caribbean Regional Office, file review and discussions with Bank staff, as well as various project and other officials during the OED mission. It reviews the level of achievement under the loan's direct and sectoral objectives, discussing in particular the policies which have been followed so far to promote the extension of technical assistance to small-scale enterprises and the relevance of the experience gained under the project in improving the provision of such services. While the provisions of technical assist- ance for SMI was given a lower priority in designing this project than credit and institutional aspects, this audit considers such assistance essential for the fixed asset financing component of Bank loans to have a lasting impact on the country's industrial development; accordingly, this issue is given particular prominence in the audit memorandum. Moreover, while the PCR deals with the period covered by the loan (1975 to mid- 1978), the audit memorandum, in accordance with standard practice, brings some aspects of the institution's situation up to date. No written comments on the report were received from the borrower.  - ii - PROJECT PERFORMANCE AUDIT REPORT COLOMBIA - FIRST SMALL-SCALE INDUSTRY PROJECT (LOAN 1071-CO) BASIC DATA SHEET Amounts (in US$ mln) As of 2/28/79 Original Disbursed Cancelled Repaid Outstanding Loan 1071-CO 5.50 5.43 - 0.23 5.20 CUMULATIVE LOAN DISBURSEMENT 1975 1976 1977 1978 (i) Planned 1.5 4.0 5.5 5.5 (ii) Actual .6 3.3 5.1 5.5 % of (ii) to (i) 40 85 91 100 PROJECT DATA Original Actual Plan Revisions Est. Actual Board Approval Sept/74 1/14/75 1/14/75 Loan Agreement 1/16/75 - 1/16/75 Effectiveness 4/16/75 5/20/75 5/20/75 Loan Clossing 12/31/77 6/30/78 11/30/78 Total Project Cost (US$ m1n) 10.2 13.2 MISSION DATA Item Sent Month No. of No. of Date of By Year Weeks Persons Manweeks Report IDF Identification N/A Preparation N/A 0 Preappraisal 6/73 0 1 .0 7/73 Follow-up 7/73 1.0 2 2.0 8/73 Appraisal 11/73 4.0 4 16.0 12/11/74 TOTAL 8.0 21.0 Supervision I 1/75 0.8 1 0.8 1/75 Supervision II 4/75 0.4 1 0.4 4/75 Supervision III 10/75 1.0 2 2.0 1/76 Supervision IV 4/76 3.4 3 10.2 5/76 Supervision V 6/76 0.4 2 0.8 6/76 Supervision VI 8/77 0.6 1 0.6 9/77 Supervision VII 1/78 0.6 1 0.6 2/78 Completion 8/78 2.8 3 8.4 1/79 TOTAL 10.0 23.8 COUNTRY EXCHANGE RATES Name of Currency (Abbreviation) Colombian Peso (Col$) Year: Appraisal Year (1974) Average Exchange Rate: US$1 = Col$26 Intervening Years (1975-77) Average US$1 = Col$35 Completion Year (1978) Average US$1 - Col'.3  - iii - PROJECT PERFORMANCE AUDIT REPORT COLOMBIA - FIRST SMALL-SCALE INDUSTRY PROJECT (LOAN 1071-CO) HIGHLIGHTS This report covers an audit of achievement under Loan 1071-CO made by the Bank in 1975 to assist the Corporacion Financiera Popular (CFP) of Colombia in its operations with small- and medium-scale industry (SMI). Reflecting its concern with the two constraints which had been identified as impeding SMI development--lack of access to term credit and insufficient technical assistance--, the Bank allocated US$5 million out of the loan amount, totalling US$5.5 million, to fixed asset financing and reserved the remaining US$0.5 million for term financing of technical assistance services to be provided by external agencies to would-be interested CFP clients. Moreover, while CFP was prevented by its statutory require- ments from lending to enterprises with total assets in excess of Col$ 20 million (approximately US$600,000), the Bank insisted upon having 75% of the loan amount earmarked for firms below Col$ 10 million; this limitation upon CFP's use of Bank funds was specified in the Loan Agreement. The Bank had four basic objectives in making this loan which was its first one to the small-scale industry sector in Latin America: (i) providing term financing to SMI, (ii) strengthening CFP's organization, procedures and financial position, (iii) improving-the_provision of tech- nical assistance to CFP's clients and (iv) initiating with the Government a review of SSI policies. To assess the validity of its rationale for getting involved in this sector which was new to it, the Bank intended to monitor closely the employment impact and growth potential of projects it was to contribute to finance. Bank funds were allocated to enterprises within the population target specified in its Loan Agreement with CFP (paras. 5 and 6 of the PPAM, paras. 3.08, 3.09 and 3.11 of the PCR). Employment generation, at a cost of US$5,000-6,000 per job, was in line with expectations; data on cost per job created were also found to be highly correlated with the size of firms (para. 10 of the PPAM, paras. 3.15 to 3.17 of the PCR). Moreover, there was no evidence of medium-size enterprises benefitting from a comparative advantage, as regards growth potential, over smaller - iv - firms, suggesting that the Bank's focus on the lower range of the SMI spectrum was justified and should be pursued in the future, as long as job creation remains its foremost objective. Twelve percent only of the technical assistance component of the loan was used for the intended purpose of financing advisory services provided to sub-borrowers by local independent consultants. This resulted from the inability of existing agencies to provide SSEs adequate assistance and the relatively high cost of their services. This indicates that the Bank had been insufficiently critical in assessing the existing institutional setting for technical assistance,and its strategy, relying on this particular setting, was inadequate to deal with this major aspect of assistance to SSI. This further points to the need of upgrading the quality and coverage of extension services available to SSEs, and of providing these services at a subsidized cost. CFP's organization was substantially streamlined during the life of the loan. Following the Bank's report, an organizational study of CFP's structure and procedures was conducted and recommendations made by the consultants appointed for this purpose were, by and large, implemented (paras. 14 to 16 of the PPAM). Portfolio control procedures were also strengthened, leading to some improvement in CFP's arrears situation (paras. 24 and 25 of the PPAM, para. 4.12 of the PCR). Owing to the monetary difficulties which beset the Colombian economy in recent years, the Bank's discussions with the Government have focussed on CFP's resource position and preempted the general review of SSI policies which the Bank intended to initiate. Other points of interest are: - the larger than expected construction element in sub-project costs (para. 8 of the PPAM); - the relatively short maturity offered by CFP on its credits for fixed asset financing (para. 12 of the PPAM and para. 3.09 of the PCR); - the potential advantages of having a single department dealing simultaneously with supervision and technical assistance matters (paras. 16 and 17 of the PPAM); - the relevance of the Bank's covenant regarding sub-borrower asset size (paras. 6 and 10 of the PPAM); and - CFP's resort to short-term financing activities on profit- ability considerations (paras. 21, 22 and 41 of the PPAM). PROJECT PERFORMANCE AUDIT MEMORANDUM COLOMBIA - FIRST SMALL-SCALE INDUSTRY PROJECT (LOAN 1071-CO) 1. Loan 1071-CO, in an amount of US$5.5 million, was the first of two loans made by the Bank to assist the Corporacion Financiera Popular (CFP) in its operations with small- and medium-scale industry (SMI). The loan was approved in January 1975 and closed in November 1978 with an undisbursed portion of US$70,000. It was followed by a second loan (Loan 1451-CO), in an amount of US$15 million, approved in September 1977 and declared effective in February 1978. Moreover, in the context of the first Bank-supported urban project in Colombia, CFP was selected as financial intermediary to channel funds to very small firms located in a number of cities in the country (see para..6.07 of the PCR). A third Bank loan to CFP is now under consideration. 2. The loan under review was designed to address the needs of the Colombian small-scale industry (SSI) and thus, supplement the Bank's emphasis till then, through its lending to a group of private development banks (Financieras), on the higher range of the industrial spectrum. CFP was founded in 1967 as a public sector institution to cater to the needs of small- and medium-size enterprises (SMSEs) and has, thereafter, remained the only financial institution in the country,and one of the few throughout the world,dedicated entirely to assisting SMSEs. Its organization consists of a central office located in Bogota and 14 regional offices scattered around the country, five of them established after the approval of the loan. The choice of CFP as the financial intermediary to channel funds under the loan was therefore logical. 3. The loan was expected to contribute to ameliorate the effects of two major constraints impeding the development of Colombian small- scale industry: lack of access to term credit and insufficient technical assistance. Of the .loan amount, US$5 million was meant for fixed asset financing, the remaining US$500,000 constituting a separate technical assistance component intended to provide CFP's sub-borrowers with addi- tional financing facilities to cover the cost of technical assistance from specialized service agencies. This was the first Bank operation solely to assist SSI in Latin America and a number of complexities, originating in the imposition of a narrow target group as well as in the "software" aspects of the project (such as technical assistance) and its strong employment creating objective, combined to give the project an experimental char- acter. Accordingly, a program of intensified supervision and close monitoring - 2 - of objectives was set up with the intention of assessing the true potential of SMSEs as well as the adequacy of the type of assistance provided to them under the loan. 4. Performance under the loan has to be reviewed in reference to its four-fold objective of (i) making term financing available to SMI, (ii) strengthening CFP's organization, procedures and financial position, (iii) improving the provision of technical assistance to CFP's borrowers and (iv) initiating with the Government a review of SSI policies. Details of the loan characteristics are given in para. 2.02 of the PCR; a notice- able feature was that the target group of sub-borrowers was clearly stipulated in order to concentrate Bank funds on the smaller end of the SNI spectrum. I. Use of Bank Funds 5. The PCR provides a comprehensive discussion of the utilization of funds available under the loan (see paras. 3.06 to 3.25). Following a slow start due to the Government's delay in making the loan effective and CFP's initial organizational problems, commitments and disbursements accelerated to a pace comparable to that originally forecast, indicating that the scarcity of local funds for SSI fixed-asset financing which had been identified at appraisal was real. The loan was closed nine months later than expected, although average delays in disbursements did not exceed six months. Disbursements were, however, somewhat hampered by CFP's liquidity problems compounded by the Bank's time-consuming review process of disbursement requests (five weeks on average). CFP's liquidity difficulties were, in turn, the result of poor financial planning on its part in the past, compounded by the Government's slowness in providing funds to the institution. 6. Bank funds were allocated to 305 companies most of which were at the smaller end of the SNI spectrum. As expected (and as required under Section 3.02 of the Loan Agreement), the loan proceeds went to enterprises with total assets below Col$ 10 million (US$300,000 equivalent) and to enterprises with total,assets between Col$ 10 million and Col$ 20 million in a 75:25 proportion- . Funds allocated to the second category of enterprises were committed somewhat more rapidly, indicating that the clause contained in the Loan Agreement (as well as CFP's statutory requirement to limit its lending operations to enterprises with assets 1/ Interestingly, the original design of the loan restricted CFP's on- lending of Bank funds to firms with assets below Col$ 10 million; it was only at negotiations that the Bank agreed to extend the original limit to Col$ 20 million. -3- 1/ not exceeding Col$ 20 million- ) had been effective in ensuring that Bank finance was channelled to the population. 7. In comparison with estimates made at appraisal, average sub- project costs were higher than expected (by about 33% in U.S. dollar terms) while average sub-loans made out of Bank funds were smaller than forecast (by about 20%). This largely resulted from the sub-borrowers' larger than expected working capital financing requirements (up from 25% to 33% of total project costs) which the Bank loan was not meant to address. As a result, the share of Bank resources in the total cost of sub-projects reached only 41% instead of 54% as forecast. CFP's severe domestic resource constraints limited the amount of financing it could provide to supplement the Bank loan proceeds (8% of total costs instead of 30% forecast) and more than2 alf of sub-project costs were covered by additional local borrowings- and larger contributions from project sponsors. Overall, the capacity of larger (medium-scale) firms to draw resources from the institutional market and the saving capacity of smaller borrowers were found to be better than expected. 8. In estimating the utilization of the loan proceeds, the Bank had made projections of the composition of capital costs of sub-projects to be financed out of the loan. In fact, the construction element of the sub-projects actually financed under the Bank loan was higher than that projected (32% as against projected 25%) and much higher in a number of individual sub-projects. Overall, the Bank financing of the sub-projects were lower than the proportion of the foreign exchange cost of the sub-projects, so that Bank financing can be said to have been used to meet the foreign exchange needs of the sub-projects. However, the larger (than projected) proportion of construction cost (some of which was found to have gone into office buildings) leads towards the conclusion that the availability of Bank finance may have facilitated the financing of the construction element on sub-project costs and that, to that extent, the Bank funds did not help increase directly productive capacity in the country. It is also possible to say that, in most cases, such construction costs could have been met from alternative sources of finance available in the country. While there would be no objection in principle to the use of Bank funds for factory construction, an unduly high proportion of the loan used for this purpose would not be desirable without assuming that the construction is part of an overall 1/ This limit was subsequently raised to Col$ 35 million in response to the highly inflationary trends which prevailed in the country. 2/ The contribution of alternative local lending sources (including the non-bank market) was not taken into account at the time of the first loan but was, however, included in sub-project financing projections made for the appraisal of the second loan. -4- expansion and modernization which would include also equipment purchases and the creation of new employment opportunities. In view of this, the Bank might want to consider specifying limits on the construction element in individual sub-projects to determine their eligibility for access to Bank funds. 9. There is little comprehensive (actual) data available on sub- projects performance, as CFP started only recently to conduct systematic supervision of its sub-borrowers. Profit and sales figures indicate however that, on the whole, the performance of sub-projects has been satisfactory,although the PCR underlines the substantial disparities found between projected and actual data regarding sales and employment impact of sub-loans, which points to the difficulty of making accurate projections of SSI operations. 10. The available evidence is much more conclusive as regards employment generation and indicatesthat sub-projects financed under the loan helped generate approximately 2,000 jobs at an average (fixed asset investment) cost -of between US$5,000 and US$6,000. A review of sub-projects carried out as part of the appraisal of the second Bank loan to CFP identified a strong correlation between the average cost per job created with the size of firms; the average cost per job created for medium-size firms (between Col$ 10 million to Col$ 20 million) amounted to US$12,000 equivalent, a figure twice as high as that for small-scale firms (below Col$ 10 million), further supporting the relevance of the Bank's attempt at focussing its lending on the smaller firms in the SMI sector. Assuming a 1:1 incremental capital/output ratio, the sub-projects would, in addition, result in an increase in annual industrial output of approximately US$10 million, thus amply confirming the high potential of SMI in terms of growth and employment generation which had been expected at appraisal and were at the origin of the Bank loan. 11. The regional distribution of the sub-projects was not as wide as anticipated (see para. 3.18 of PCR); 35% of the loan amount only went to enterprises located outside Colombia's three major industrial centers of Bogota, Medellin and Cali, 40% going to the state of Cundinamarca alone (Bogota), as against an initial forecast of 25%. In view of the difficulty it experienced in channelling funds outside the more developed areas of the country, CFP requested the Bank, in January 1976, to introduce regionally differentiated interest rates along the line of the 3% interest rate differential introduced by the Banco de la Republica (the Central Bank) on its industrial financial fund (Fondo Financiero Industrial [FFI]), the resources of which CFP was using for working capital financing purposes. The Bank refused at the time to go along with CFP's proposal because it did not believe that differential interest rates would be conducive - 5 - towards achieving greater regionalization' of credit. A posteriori interest rate differentials proved to have had very little impact on the geographical allocation of industrial credit, and CFP's own pro- motional efforts at regionalizing its operations (para. 18 below), coupled with the Government's fiscal incentives, appeared to have been substantially more effective in bringing about a wider dispersion of industrial investment in later years than interest rate differentials. 12. In view of CFP's conservative lending practices in the past, and to ensure satisfactory lending terms to borrowers, a minimum four- year maturity was applied under the loan. A review of CFP's sub-loan maturity structure reveals that, while this rule was strictly followed and the Bank should be credited for bringing about longer maturities on fixed asset financing loans, the average maturity of sub-loans remained low. Of the loan amount, 89% went to sub-loans with maturities of six years or less, the average being around five years, as against four years for fixed asset financing sub-loans made by CFP before the Bank loan; this appears to have been less than adequate to meeting the sub- borrower's needs, and was partly responsible-for CFP's improved, but still difficult, arrear position. One of the reasons for these low maturities was, for a long time, the nature of the particular financial analysis carried out by CFP,as in most cases, and often for lack of proper account- ing system on the part of sub-borrowers, CFP financial analysts based their recommendations for sub-loan maturity on the ratio of expected annual incremental profit of sub-projects to total investment costs. Only recently has a full flow of funds analysis been introduced as part of CFP's appraisal procedure; the impact of this increased sophistication in appraisal on the choice of sub-loan maturity remains to be demonstrated. Given the still high level of CFP's arrears, which by end-1978 affected 12.1% of its total loan portfolio, a case can be made that the Bank should pursue its efforts at promoting an increase in CFP's sub-loan maturities; this could be done by laying down a minimum average maturity of sub-loans made out of Bank finance alongside the present practice of specifying a minimum maturity as done under the first two Bank loans. 13. Funds made available under the loan for technical assistance financing were not used fully for that purpose, reflecting the little demand by CFP's clients for this kind of loans. Only 11 technical assistance sub-loans, totalling US$60,000, were made under the loan, another US$90,000 being allocated to research financing which the loan had provided for. The balance US$350,000 of the technical assistance component was transferred to the fixed asset lending component (see para. 3.21 of the PCR). As of February 1979, i.e., a year after the second Bank loan was declared effective, US$26,000 had been committed under the second loan technical assistance component out of a total amount of US$100,000. CFP's technical assistance activities are dis- cussed in Section III below. -6- II. CFP's Organization, Overall Operations and Procedures Organization 14. When the Bank extended its first loan to CFP, 55.4% of CFP's share equity was held by the Banc?,Popular, a Government-owned commercial bank, and one of its subsidiaries- , a very large part of the remaining shares being held by the Government. Banco Popular had helped set up CFP in 1967 and maintained thereafter close control over CFP's activities though its majority representation on the latter's Board and Loan Com- mittee. This had led CFP to assume a predominantly commercial banking approach in its lending operations with tight lending terms and excessive emphasis on the sub-borrower's collateral position, somewhat in contra- diction with the development objectives which had led to its creation. One condition of effectiveness of the Bank loan was that CFP's statutes be modified to ensure that Banco Popular would renounce its majority representation on CFP's Board and Loan Committee. CFP's statutes now bar any of the institution's shareholders from securing a majority position on its Board. In any case, following a number of share capital increases in which Banco Popular did not parti pate, its share in CFP's equity had, by December 1978, dropped to 33.9%- ; the control of Banco Popular on the operational and financial policies of CFP is now consider- ably reduced and the Government now holds directly the responsibility of guiding CFP's policies and operations. Since 1975 the composition of the Board is quite diversified and includes personalities from the Govern- ment, the industrial community as well as national technical assistance institutions. CFP's Board has adopted a developmental attitude in its discussions of the institution's policies and operations but, on the basis of its actions on the subject, it appears to have been passive in dealing with wider issues such as the elaboration of a comprehensive technical assistance program as a complement to CFP lending, reflecting the Government's lack of clear strategy in this field. 15. A second element in the Bank's assessment of CFP, at the time of making its first loan, was that the internal organization and control procedures had not fully kept pace with the institution's rapid growth since its establishment. To identify organizational weaknesses and ways to deal with them, CFP, as agreed with the Bank, appointed prior to loan effectiveness, a management consulting firm to carry out an organizational study (see para. 4.05 of PCR). This study, which consisted of two parts covering CFP's organizational structure and its internal system and 1/ Corporacion de Ferias y Exposiciones. 2/ Including Corporacion de Ferias y Exposiciones. - 7 - procedures was completed by end-1976; most of its recommendations were subsequently approved by CFP's management and Board members and have been implemented. The major measures.were the up-grading of regional offices in the overall structure of the company and the regrouping of departments dealing directly with sub-borrowers (i.e., the Credit Department and the Technical Assistance Department) under a common sub-manager (Operations Division). 16. The study also suggested the removing of the supervision function from the Technical Assistance Department and the establishment of a separate supervision department within the new Operations Division. However, a sharp drop in business in 1977 (partly on account of delays in processing the second Bank loan and its being declared effective) induced CFP's management to curtail severely its staff and to eliminate the newly-established Supervision Department, the supervision function being merged again with that of technical assistance, although specialized supervision officers remained in position in each regional office. CFP's availability of resources substantially improved in late 1978, but CFP's present management (in place since September 1978) has, in keeping with its policies of placing the institution on a more commercial basis, maintained a strict control on additional staff recruitment. 17. Anyhow, the need to separate the supervision and technical assistance functions within the structure of a SSI financing institution such as CFP appears open to debate. As regards the provision of tech- nical assistance, CFP has operated mostly as an intermediary between its sub-borrowers and external agencies, providing financing for specialized services but not acting itself as a prime source of services except in the case of some occasional group training of sub-borrowers. In this context, the rationale for proposing a separate supervision department was that the particular nature of CFP's technical assistance activities (which consisted mostly of lending activities) tended to push aside supervision to a secondary role. The creation of a separate supervision department, further induced by the need to separate portfolio control procedures from technical assistance considerations, was then seen as a possible instrumentality to foster the development of adequate supervision procedures within CFP. This rationale remains valid so long as technical assistance is viewed as a service to be provided on an occasional basis, as has been the case so far. A special department deals with portfolio control matters,while separately the Technical Assistance Department conducts simultaneously the two distinct functions of arranging technical assistance contracts for interested sub-borrowers, and coordinat- ing and monitoring the supervision work carried out by the regional offices. These two functions, supervision and technical assistance, could, however, assume a much higher degree of complementarity, if CFP took a more direct and systematic responsibility in the extension of technical assis- tance to its clients; in this event, the rationale of having a commom -8 - department dealing with both functions, supervision and technical assis- tance, would be enhanced, as the linking of both functions in one department could enable CFP to provide,technical assistance to its clients effectively as a problem-solving service (see also para. 27 below). 18. In line with the Colombian Government's decentralization policy, the regional orientation of CFP's organization has also been greatly strengthened. Besides the upgrading of regional offices suggested in the organization study, CFP's Board approved the establishment, in a number of cities across the country, of Regional Advisory Boards with substantial loan approval powers, giving CFP's regional offices a wider operational autonomy. Moreover, the number of regional offices increased in 1975 by five to 14, giving CFP's operations a broader regional coverage and, indeed, the geographical dispersion of CFP's most recent operations appears to be broader than that of sub-projects financed under the Bank's first loan. 19. CFP's overall operations, resources and profitability are discussed at some length in the PCR (see paras. 4.14 to 4.28). The volume of CFP's operations has increased substantially since the loan approval, total assets growing from Col$ 499 million (US$17 million equivalent) to Col$ 1,362 million (US$36 million) between 1974 and 1977, i.e., at a 40% average annual growth rate (equivalent to approximately a 12% rate in real terms). This is in excess of appraisal forecast but a much higher inflation rate than anticipated makes a direct comparison irrelevant. For factors largely beyond its control, CFP has not succeeded in widening its resource base as it had planned at the time of appraisal. The composition of its operations closely reflected that of its resource mix, long-term operations being restricted to the use of funds available under the two Bank loans and a US$5 million loan from USAID. Although Bank funds, so far intended mostly for fixed asset financing, have been committed and disbursed rapidly once made available to CFP, demand for working capital financing has proved even stronger, reflecting the needs of CFP's clients, particularly the smaller ones, and the low utilization of productive capacity prevailing in most of the SMI sector. In consequence, funds available under Banco de la Republica's Industrial Financing Fund (FFI), which constituted the main part of CFP's resources before the Bay loan, have been used almost exclusively to finance working capital- ; although the flow of FFI resources had at times been somewhat uneven in line with the Government's credit policies, the use of the FFI rediscounting line has over the years accounted for the larger share of CFP's term operations. 20. As a result of poor financial planning and indeed lack of effective cash flow projections, CFP has in the past suffered from 1/ An end-use which the Government has in any case been encouraging. -9- severe liquidity problems which were responsible for some delays in its use of the Bank loan. This also preveiYed its use of FFI resources to the extent it could have had otherwise- ,in that CFP did not have the necessary resources needed to complement FFI's rediscounting facilities (which could cover up to 65% of loans made to projects located in cities of 900,000 inhabitants and 80% of the others). During negotiations for the loan under review, the Bank had reached with CFP an under-standing on the need to strengthen the latter's liquidity position (Section 2.07 of the Project Agreement). A series of guidelines was provided to CFP to assist it in strengthening its financial and control procedures (referring particularly to the need for monthly cash flow position reports and annual cash flow projections); however, even though these were reportedly implemented, they were not translated into operational improvements until much later. 21. Spreads to financial intermediaries on FFI funds, being limited to a maximum of 4% (as opposed to 5.5% on Bank funds), fell progressively short of covering CFP's administrative costs as those increased with inflation and the expansion of operations to the less developed regions of the country. To remain profitable, CFP, with Government endorsement and the Bank's reluctant approval when the second loan was made, has since 1976 engaged in short-term export financing of large-scale industrial enterpris27 through the Government's Fondo de Promocion de Exportaciones (PROEXPO)- . These operations,which the Bank feared would divert CFP from its developmental orientation, accounted for 38% of CFP's disburse- ments during 1977 and appear to have taken up relatively little staff time, thus helping CFP to lower somewhat its administrative costs to about 7% of its average total assets. Under the second Bank loan, CFP agreed to limit its outstanding liabilities associated with short-term operations to the equivalent of two times its equity. During the OED mission, it was found that, to improve further its profitability, CFP had more recently undertaken to engage in short-term lending (mostly for import financing) on the basis of its own equity, obtaining a 26% 1/ As of end-1978,.CFP's cumulative use of the FFI rediscounting line amounted to Col$ 1.5 billion (approximately US$44 million equivalent), representing 20% of all credits approved by FFI (between 1969 and 1978). FF1 funds are provided upon demand by-any financial inter- mediary on the basis.of resourcee-availability, -assuming eligibility criteria are satisfied (see para. 4.14 of PCR). 2/ By end-year 1976 PROEXPO, an'agency of the Ministry of Economic Development, had also become a large shareholder of CFP with 33% of subscribed shares; it now owns a majority (52%) of CFP's equity. - 10 - return (in fact, 30% equivalent because of advance interest paymels), as opposed to 24% charged on sub-loans under the .second Bank loan- 22. For the reason mentioned above, CFP's profitability has remained largely below anticipated levels. For one thing, until 1977 administrative expenses did not decline as expected, averaging 7.2% of average total assets during 1977 as against a target of 5.8%. Since then, they have been further reduced, following a trimming of CFP's directly non-profit- able activities - including all research programs initiated under the loan (see para. 3.21 of PCR). Anyhow, profitability was not expected at appraisal to be high enough to compensate fully for erosion of net worth through inflation, and an agreement had been reached with the Government during negotiations that CFP would receive sufficient share capital contributions to maintain the value of its equity in real terms. Because of the priority given by the Government to the implementation of strict monetary policies in the face of soaring inflationary trends, this has not taken place even though CFP's share capital increased by 50% (from Col$ 200 million to Col$ 300 million) between end-1974 and end-1977, an increase which the Bank loan was reportedly instrumental in bringing about. CFP's new management is confident that, assuming a proportion of CFP's own funds are on-lent on a short-term basis and administrative expenses trimmed to the extent possible, CFP's lending operations can be made sufficiently profitable to maintain the institution's capital in real terms--albeit not sufficient to follow the real growth of the sector--and has decidedly made profitability one of its key objectives. A discussion of the profitability issue, however, is not complete without a review of CFP's activities which are not directly profitable, most of them related to the extension of technical assistance (paras. 28 to 35 below). 23. In its efforts at cutting down administrative expenses, CFP's latest management has endeavored to streamline and rationalize the institution's salary policy. CFP's staffing position has consistently been severely affected by the lack of competitiveness of its pay scale which was responsible for a very high staff turnover at the professional level. Before the cuts in personnel by the previous management for lack of business, staff turnover was estimated at 40% per year. Since then, CFP's new management has maintained a cost-conscious staffing policy, taking advantage of natural staff depletion to eliminate unnecessary positions; these affected mostly clerical jobs as also a number of redundant high-level positions without impairing the volume of operations. 1/ The wholesale price index for Colombia rose 27% in 1977 and 18% in 1978. For a discussion of interest rates under the loan, see para. 3.10 of the PCR. - 11 - Control Procedures and Quality of Portfolio 24. The quality of CFP's loan portfolio which was a maior source of concern for the Bank when the loan was made, worsened during 1976 but has subsequently improved following a progressive tightening of control procedures. By end-1978, loans affected by arrears represented 12.1% of total loan portfolio, down from 16.2% at the end of 1973 and 24.1% as of mid-1976. Reasons for CFP's difficult arrears situation in the past can be divided in three categories: (i) the uncertainty inherent in small-scale business as reflected in the difficulties experienced by sub-borrowers themselves, particularly in marketing their production; (ii) CFP's practice of lending on shorter-than-necessary terms; and (iii) its lax control and collection procedures. 25. Initially, portfolio control was carried out by the Credit and Portfolio Department, then also in charge of project appraisal and credit evaluation; consequently, adequate control systems did not get developed. In 1976 CFP was reorganized and a special department in charge of portfolio control was set up. Another contributing factor in improving the arrears situation was the establishment of a 32% penalty rate applicable to the overall balance of loans in arrears. An analysis of arrears by regional office reveals that some of the highest levels of arrears (as percentage of outstanding portfolio) are found in CFP's largest regional offices (i.e., Bogota, Medellin and Cali); this suggests that the business environment prevailing in the largest urban agglomerations, where personal knowledge is of lesser significance, tends to make portfolio control there more difficult, and points at a further tightening of control and collection procedures in these areas. Appraisal and Supervision Procedures 26. Although the overall quality of CFP's appraisalsimproved under the loan, progress achieved in this regard has been uneven, owing to the institution's high staff turnover which resulted from the payment of poor salaries and benefits and reduced the effectiveness of its efforts at training new analysts. Overall, the quality of appraisal work also varies widely among regional offices. In general, appraisal work has been rendered particularly difficult by the lack of accounting systems among a large majority of loan applicants. This, coupled with the uncertainties involved in the forecasting of SSI business, made it particularly strenuous for CFP to improve the sophistication of its appraisal techniques along the lines suggested by the Bank. Break-even point calculations and cash-flow projections were introduced only recently, and their impact on CFP's decision-making process (regarding the making of loans as well as their size and maturity) remains to be strengthened. - 12 - 27. Project supervision was almost non-existent until mid-1978 when the Bank insisted on a minimum program being carried out in relation to sub-projects financed under its loans. A program of visits has now been established under the Technical Assistance Department which monitors the supervision activities of the individual regional offices; reports are prepared following each visit, updating sub-project basic financial and economic data and providing a diagnosis of problems encountered. However, supervision is still considered a low-priority function within CFP, as indeed the sheer number of loans outstanding would make full- fledged, operationally effective supervision a much more time-consuming exercise than is considered worthwhile and profitable by CFP's current management. Such supervision appears, however, essential and could appropriatelybe linked to the extension of direct technical assistance by CFP, as suggested in the next section. III. Technical Assistance 28. Some combination of financial and non-financial assistance is generally needed to foster the development of small-scale enterprises. Accordingly, the Bank emphasized the importance that both types of assistance be effectively coordinated to maximize their combined effect and, in line with this general strategy, considered an improvement in the quality and frequency of technical assistance provided to CFP's borrowers to be a major objective of its loa 7 This position was later emphasized in the Bank's Sector Policy Paper- which left, however, the choice of the appropriate relationship between advisory services and credit administration to be decided on the basis of the particular institutional setting in existence in each country. 29. A variety of Colombian institutions is engaged in providing technical assistance to SMI. These include vocational training insti- tutions such as Servicio Nacional de Aprendizaje (SENA), technological institutions such as Instituto de Investigaciones Tecnologicas (IIT) or private consulting institutions such as Fundacion para el Fomento de la Investigacion (FICITEC). However, due to the sheer number of potential clients as well as the concentration of assistance agencies in the major industrial centers, the services provided by these institutions have remained short of meeting SMI's requirements. Moreover, when the loan was made, most of the available assistance programs were considered by the Bank as inadequate in that they did not provide sufficient direct advisory services tailored to the needs of individual enterprises, being focussed on providing general managerial advice and seldom extending to training in marketing, production techniques, quality control and product 1/ Employment and Development of Small Enterprises, February 1978. - 13 - planning. Moreover, technical assistance agencies have tended to con- centrate their direct assistance operations on medium-size enterprises which could better afford the relatively high cost of such services, estimated at appraisal at an average of US$3,000-4,000 equivalent per client. 30. Aside from organizing some group training of borrowers in accounting and general management, CFP's own technical assistance activities have consisted mostly of attempting to detect the major operational problems of loan applicants and to arrange for appropriate technical assistance by other institutions. In this context, the Bank's strategy (under its two loans to CFP) has been to make funds available for technical assistance credits to pay for advisory services provided to sub-borrowers by local independent consultants. The use of such funds under the first loan was small (US$60,000). and preliminary indications suggest that this will be so also under the second loan. The PCR refers to a grant in an amount of US$1 million, made by USAID in 1975 for tech- nical assistance purposes, as having largely contributed to the small use made of the Bank's technical assistance allocation. However, about 70% of the USAID grant amount went to cover CFP's own administrative expenses (generated by technical assistance activities) while part of the remaining 30% went to subsidizing training courses given by CFP or others. Thus, while the USAID loan did reduce use of Bank technical assistance funds, CFP's overall lending for direct technical assistance purposes has been, during the period covered by the loan, considerably smaller than expected. Moreover, sub-loans for technical assistance have been concentrated among the larger (medium-scale) of CFP's customers, further suggesting that the strategy adopted by the Bank was not effective in meeting the technical assistance needs of the SSI sector which represented its main target and that CFP was unable to interest SMSEs, to any appre- ciable extent, in borrowing money for direct technical assistance purposes. This experience appears to be similar to that in the few other cases where this approach was used and suggests that if technical assistance is to be provided, and used, it must be subsidized in one way or another. 31. The Bank suggested that CFP could have been more active in the adve pising of its technical assistance credits. While this is certainly true- , there is no doubt that the demand for such credit on the part of small-scale enterprises was, and is likely to remain, low, given the cost of direct assistance services provided by consultant agencies (technical assistance sub-loans made under the Bank loan were on average for US$5,500 equivalent); and, in this context, CFP's potential in overcoming rapidly 1/ It is noteworthy that, until recently, the Bank loans were not mentioned in CFP's Credit Manual as a possible source of financing for external technical assistance services. - 14 - the borrowers' reluctance in utilizing credit for individual services appears to be very limited. Another factor contributing to the low demand for credit for technical assistance has been the natural pre- ference expressed by mosti?f CFP's sub-borrowers for the free-of-charge services provided by SENA- , even though SENA's direct assistance activities are limited and not comprehensive. 32. CFP's successive managements have rarely demonstrated a strong commitment towards implementing a program of technical assistance to their sub-borrowers. A notable exception was the introduction in 1975, with the assistance of a UNIDO advisor, of general industrial development programs, comprising general managerial courses and seminars as well as training in accounting; these, however, have remained very limited in quantitative terms. Moreover, the position of CFP's present management is that, given its objective of bringing the institution's operations on a profitable footing, it could not consider direct technical assistance as part of CFP's functions. The Bank appears to support this policy, even though CFP's statutes explicitly mention direct technical assistance as one of its fields of activity. 33. Under the second loan, the Bank emphasized the need for increased coordination between the various technical assistance programs available to SSI. Accordingly, prior to the second loan becoming effective, and as agreed with the Bank, CFP signed a cooperation agreement with SENA; this agreement, however, was formulated in very vague terms and did not bring any substantial change in the collaboration between the two insti- tutions. The potential for such collaboration is in any case limited, as SENA's activities, which consist of vocational training for workers and general management training for entrepreneurs, are somewhat different from the kind of services needed by SSI, whether in production, financial management or marketing. 34. The experience of the first two loans clearly shows the inade- quacy of the Bank strategy in its dealing with the provision of technical assistance, despite the observed need for such assistance, particularly in the fields mentioned above, to improve SSI operations. It would be desirable, in view of this, to explore an alternative strategy or means of providing technical assistance, such as building up gradually in-house technical assistance capability within CFP or setting up a specialized institution to provide such services to SSI, in case the Bank considers a technical assistance program as an essential component in its SSI financing in Colombia. 1/ SENA is a public vocational training institution whose operational costs are covered by a flat-rate duty paid by all Colombian enterprises on their labor bill. - 15 - 35. Because of the difficulties experienced in the past by CFP in securing a steady flow of domestic resources (in the form of loans as well as share capital), CFP's management has been striving to conduct its operations in a way to minimize as much as possible the need for (direct or indirect) subsidized funds from Government. The introduction of direct technical assistance activities as part of CFP's business would obviously be counter to this basic philosophy and call for a separate accounting system for technical assistance expenditures, so as not to weaken the recent efforts at controlling CFP's strictly administrative expenses. As part of a loan agreement with USAID, the Government has, for the last year or so, been contributing to CFP's administrative expenses in the form of an annual allocation to its technical assistance (training and credit) activities. This allocation amounted in 1978 to Col$ 12 million (US$315,000 equivalent), i.e., approximately 15% of the institution's total administrative expenses during the year, a ratio higher than that of the Department of Technical Assistance staff (which is also coordinating supervision duties) to CFP's total staff. This indicates that, even though an expansion of CFP's technical assistance activities would generate some amount of financing difficulties for it, some kind of arrangements could be worked out through either the provision of subsidized funds from the Government or the earmarking of interest rate differentials on CFP lending for technical assistance activities; the working out of such an arrangement should indeed be the focus of the Bank's discussions with CFP and the Government regarding technical assistance issues. IV. Government Policies 36. Because of the larger labor content and wider regional dispersion of SMI in relation to industry as a whole, small- and medium-scale industrial enterprises play a vital role in Colombia's efforts to reduce unemployment and urban migration, decentralize industrial development and build up entrepreneurship. The Government is conscious of the developmental impact which SSI could have on the economy and has repeatedly given it an important position in the formulation of its development strategy, as stated in successive development plans. However, in the last few years it has had to give priority attention to the inflationary situation in the country; this led it to impose tight monetary policies which have most severely affected the smaller end of the industrial sector. For this reason, the general review of SSI policies,which the Bank was expected to initiate as a complement to its loan, in the hope of promoting Government policies more responsive to the needs of the SSI sector, has not taken place; rather, discussions between the Bank and the country's authorities (the Government and the Central Bank) have centered on CFP's resource position, the Bank insisting upon the Government keeping to its agreement to provide CFP with sufficient resources to maintain the real value of its equity. Through the making of Bank loans, as also through successive - 16 - increases in CFP's equity and access to the FFT discounting line, SSI's access to financial resources has markedly improved since 1975. However, owing largely to the monetary difficulties which have beset the Colombian economy for the last few years, it has not been possible, so far, to secure for CFP a steady, stable supply of long-term domestic resources, indicating that the Bank, in spite of its striving, had been only partially successful in its effort at enlarging the flow of resources to meet SSI's financial needs. 37. Because it focussed most of its discussions on financial issues, the Bank devoted little attention to the lack of clear Government policies regarding the provision of technical assistance to SSI, which represented the second facet of its sectoral objective. The thinking of the Bank on this issue appears in any case to have remained very vague at the time of the approval of the first loan, whose experimental nature was then acknowledged. The Bank repeatedly emphasized the need for improved coordination between national technical assistance agencies, while, at the same time, pointing out the inability of these agencies to cope with SSI's assistance needs. The lack of prospects offered by this strategy has by now been amply demonstrated,as existing technical assistance agencies either provide in-depth assistance to medium-scale firms (as smaller companies are unable or unwilling to assume the cost of such services), or provide too general managerial-type training to have more than a marginal impact on the performance of SSEs. In view of this, discussions with the Government should now focus on alternative approaches to achieve the same end. V. Conclusions 38. The major benefits which could be expected from the development of the SMI sector (employment generation, geographical decentralization of industrial activity) and were seen to fit the general objectives pursued by the Government's overall economic development strategy, induced the Bank to initiate lending to this sector of the Colombian economy. The loan under review was the first Bank loan to SMI in Latin America and, as such, featured a number of experimental characteristics, including (i) a strict definition (by asset size) of the population target to which funds available under the loan were to be channelled and (ii) a special component designed to provide sub-borrowers with finance for technical assistance purposes in view of the acknowledged needs of SSEs for such services. 39. On the basis of available information, the economic benefits derived from investments financed under the loan turned out to be at par with expectations; in particular, the average (fixed asset) investment cost per job created was about US$5,000-6,000 and, although the con- centration of Bank-financed sub-projects in Colombia's main urban centers was higher than forecast, geographical dispersion increased later as CFP - 17 - provided greater autonomy to its regional branches. A comprehensive survey of sub-projects conducted by the Bank when the second loan was appraised emphasized the existence of a very strong correlation between fixed asset size and investment cost per job created, indicating that the Bank's insistence on allocating 75% of the loan amount to the small- scale portion of the SZU sector was appropriate and that the criteria specified by the Bank could be made even tighter,as shown by the rapid acceleration of the incremental capital-labor ratio for firms with total assets above Col$ 5 million. 40. A major issue arising out of an attempt at institutionalizing credit to SMI relates to the'financial viability of such DFCs, as arrears, larger than those experienced by ordinary DFCs because of uncertainties inherent in small-scale business, and higher administrative expenses are likely to keep profitability low. CFP's profitability has so far been very poor and its arrears have, until recently, remained on the high side. A progressive tightening of portfolio control procedures and the introduction of cost-conscious administrative policies have led to a relative improvement, although increased cost-consciousness has also brought the relinquishment of CFP's research activities as well as part of its technical assistance programs. The Bank has endeavored to discuss CFP's financial situation with the Government in terms of total availabi- lity of local funds and this appears to have been the right approach, even though the Bank's efforts in this regard have not been entirely successful, owing to the higher priority given by the Government to enforcing strict monetary policies. 41. CFP's financial situation represents only one facet of the more general issue of the type of functions the institution is to assume and the choice made by its management between the two extreme viewpoints of managing CFP as a profit-making financial intermediary and as a public service. In referring to CFP's exclusively commercial lending practices at the time of the appraisal of the loan under review, the Bank had demonstrated its awareness of the need to reach a balanced view between these two functions. The position of CFP's current management appears to be in favor of concentrating on activities directly yielding a financial return. It is significant that the continuation of low profitability, high administrative costs, comparatively high--even reduced-- arrears, have apparently pushed CFP to complement its SMI lending activities with short-term export financing activities in relation to large-scale enterprises through PROEXPO. This appears appropriate (particularly since staff-time requirements for export financing are not large), as such a policy improves CFP's status in the country and decreases its reliance on Government subsidies, and as long as these new activities are viewed as complementing, and not substituting for, CFP's developmental activities and procedures such as appraisal banking, supervision and technical assistance. Operations Evaluation Department August 30, 1979 - 18 - PROJECT COMPLETION REPORT SMALL-SCALE INDUSTRY PROJECT LOAN 1071-CO I. INTRODUCTION 1.01 Corporacion Financiera Popular (CFP) was founded in 1967 by Banco Popular, a government-owned commercial bank, in response to an earlier study by the Stanford Research Institute on the prospects and needs of Colombian small scale industry (SSI). CFP was created and expanded in recognition of the need to undertake more aggressive promotion of small industry, given the existing private development banks' (financieras) focus on medium and large firms and to offer a broader range of services deemed necessary for an effective SSI development effort. CFP's basic objectives as defined in its statutes are to (i) contribute to the development of small manufacturing, agroindustrial, and mining enterprises; (ii) encourage import substitution and export promotion projects; and (iii) stimulate investment in small-scale enterprise (SSE). 1.02 The Bank had supported Colombia in its industrial development through five "financiera" loans totalling US$162.5 million from 1966 through 1973. However, only 6% of these Bank funds through 1973 had gone to SSE clients with total assets below Col$10 million, and the average subloan amount had been US$400,000. Thus the Bank concurred with the government that a different mechanism was necessary for addressing the needs of SSI. CFP, the only institution dedicated entirely to supporting small enterprises, was the logical entity for developing into an effective SSI term financing institution. CFP's growth had been rapid through 1973, with over 3,500 loans approved and Col$417 million (US$16.7 million) in assets. Preparation of the project by the Bank began in June 1973, and appraisal took place in November 1973 and May 1974. This report covers the period beginning with the Bank's involvement with CFP through mid-1978, although certain activities that pertain mainly to the second loan 1/ are omitted. 1/ Loan 1451-CO for US$15 million, approved by the Bank in June, 1977. Due to CFP's objectives and experience with the Bank, the institution was selected to channel an additional US$4 million of term funds, 50% of which would be Bank funds, to very small firms in the 23 cities partici- pating in the Bank's First Urban Project in Colombia (Loan 1558-CO). As the target group was smaller than that generally served under the Bank's two SSI projects, the design of that component was rather similar to an existing CFP-administered credit line for lending mainly to artisans. While the Bank had generally discouraged CFP from undertaking small, fragmented operations that may not further the long-run development of the institution, the high priority of the overall project and CFP's capability for carrying out the component eventually led to CFP's inclusion in the project. - 19- II. OBJECTIVES AND EXPECTATIONS OF THE BANK 2.01 The first loan to CFP was considered to be experimental because it dealt with a new type of specialized institution and represented the Bank's first SSI project in Latin America. The clearly stated principal objectives of the project were to: (a) provide term financing on reasonable conditions to SSI; (b) strengthen CFP, particularly by improving its resource allocation ability, internal organization and procedures, and ensuring a sufficient resource base; (c) help improve technical assistance to CFP's borrowers and provide financing for it; and (d) initiate reviewing of SSI policies with the government. 2.02 The Bank loan of US$5.5 million was made to the Banco de la Republica (BR), Colombia's Central Bank, to be onlent to small industries through CFP. Under the US$5 million lending program, ultimate beneficiaries were given the choice of borrowing at 24% p.a. in pesos or at 13% p.a. in foreign exchange. Considering the greater risk and administrative expendi- tures of lending to SSI, CFP's spread was set at 5.5% for peso subloans and 5.25% for foreign exchange subloans, i.e., significantly higher than the 3% p.a. spread applicable to the Bank's "financiera" lending at that time. To ensure satisfactory terms for borrowers, subloans were subject to a four-year minimum. The Bank loan was made on a 12-year fixed schedule, including a 3-1/2 year grace period, rather than a composite amortization schedule, to allow CFP to build up term resources. Reflecting CFP's limited experience in fixed-asset financing and the small size of its projects, the limit for loans not requiring the Bank's prior approval.(the free limit) was a comparatively low Col$1.5 million (about US$43,000 equivalent). Under the lending component, preference was given to financing the smaller end of the SSI spectrum by (i) stipulating that ultimate beneficiaries could not have total assets in excess of Col$20 million at the time of subloan approval, and (ii) earmarking US$3.75 million (75% of the lending component) for firms with total assets below Col$10 million. The maximum cumulative assistance to any one firm was limited to Col$2.5 million for beneficiaries in the latter category and to Col$4.0 million for those with total assets between Col$10-20 million. Under the US$500,000 technical assistance component, subloans carried an interest rate of 15% p.a. in pesos and CFP's spread was limited to 3%. Most technical assistance subloans were expected to have a maturity of 10 years and were to include a grace period of 3 years. Part of the technical assistance funds were expected to be used by CFP for studies to improve its operations or its sector knowledge. 2.03 As part of loan preparation, the Bank identified a number of specific areas whose improvement would strengthen the overall impact of national SSI policies and programs. Most important of these was to upgrade CFP's term lending capability. Over 70% of CFP's previous lending had been for working capital, with most loan maturities being four years or less. The quality of - 20 - appraisal work varied widely as a result of high staff turnover and lack of standardized appraisal procedures and guidelines. Internal organization and control had not kept up with CFP's growth, and frequent management changes had been detrimental to the establishment of medium- and long-term objectives and to their implementation. Loan supervision and technical assistance were receiving little attention from CFP, although emphasis on this area increased in early 1974 with the assistance of a UNIDO expert. 2.04 In addition to the institutional development of CFP, the Bank loan was expected to have a sectoral impact by triggering (a) the establish- ment of a national council for small and medium scale industry (SMI) and (b) the carrying out of a comprehensive study by the government of the effects of industrial policy on SMI, as had been proposed in the Stanford Research Institute study of 1962. The objective was to help build up sufficient information on the SMI sector and its problems and to create an effective group for policy making and interagency coordination with which the Bank could maintain a dialogue on SMI development. Finally, the Bank made some rough quantitative estimates concerning the probable impact of the loan. However, given the rather limited Bank experience with SSI projects at that time, the speculative nature of these estimates was clearly recognized. 2.05 This rather comprehensive set of loan objectives was established in order to seek maximum results both in improving programs for SSI and gaining experience in undertaking SSI projects. The Bank and CFP agreed from the beginning on the importance of collecting extensive data on loan performance, and have both benefited from the extra experience derived from close monitoring. Also, given the experimental nature of the project, the Bank had agreed to have a "mid-term review" of the project prior to releasing the second US$2.5 million lending tranche of the loan. As the project was proceeding substantially as anticipated, the second tranche was released without any modification of the loan conditions. On the whole, most goals regarding CFP itself and the use of loan proceeds were achieved, while less progress was made on sectoral goals. III. UTILIZATION OF LOAN PROCEEDS Economic Environment 3.01 When the loan was made in 1975 some of the industrial sector's dynamism of the 1968-74 period was lost as the Colombian economy was hit by the world recession. The slackening in export demand was affecting all productive sectors of the economy, which were also suffering from the tight monetary policy introduced to stem inflation that had reached 27% in 1974. In 1975 industrial exports declined by 22% in value compared to 1974, and industrial production grew by only 3%. However, industrial production recovered with a 6.8% growth rate in 1976, led by buoyant domestic demand resulting from the boom in coffee prices that began in late 1975 (Annex 1). In 1977 the industrial sector grew by an estimated 6-7%, with manufatured exports growing only slowly in US dollar terms. - 21 - 3.02 Industrial investment declined by 16.5% in 1975, mainly as a result of the economic recession, a significant decline in corporate profits, and the tight credit policies. In 1976 investment recovered somewhat (by 7% over the previous year) as the economic outlook brightened, and investment appears to have grown by a further 7-10% in 1977. However, in real terms, the rate of investment in 1977 barely reached that of 1974. Because of the 1975-76 investment lull and continuing high demand, many industries, especially the larger ones, are now working close to full capacity. 3.03 There are no separate data available to determine the impact of these changes in world and national economic conditions on Colombian SMI. However, some subjective observations are possible based on discussions with CFP clients and a survey of its clients undertaken by CFP in 1977. 1/ Most importantly, rapidly rising costs created a need for additional financing, yet government anti-inflation measures contributed to keeping credit to SMI scarce. Although development loans to SMI 2/ increased by 45% from 1975 to 1976, with 30% of the loans financing fixed assets, demand outstripped supply and reportedly many good projects could not obtain financing. Also, many small industrialists apparently misinterpreted inflationary profits as real profits and thus failed to build up adequate reserves for future investments. Still, most small industries have proven quite resilient because of their operational flexibility and low fixed investment, and many demonstrated their ability to weather unfavorable conditions better than larger firms, as indi- cated below. 3.04 CFP's survey of clients offers interesting indications, although the sample clearly represents only the stronger firms that are able to borrow and thus is not fully representative of the SMI sector. During 1975-76, the value of production increased 34% for the more than 300 firms that responded, or about 8% in real terms between the two years, i.e. virtually equal to the estimated 7.7% growth for large industry. 3/ Most sectors performed reasonably well except leather products, where decreased supplies of suitable raw materials especially hampered small industrialists. The survey found a favorable outlook of the respondents towards the year 1977, with most anticipating substantial production increases and preparing investment plans superior to 1976. 3.05 Sales of the surveyed firms increased 10% in real terms in 1975-76, i.e. slightly more than production, reflecting some inventories reduction as some small industrialists were caught in the inflation spiral. 4/ Exports 1/ Encuesta Industrial, August 1977. 2/ From CFP, BR's Fondo Financiero Industrial (FFI), and the government's Caja Agraria. 3/ Estimate of FEDESARROLLO. 4/ Industrial inventories generally increase during inflation to avoid higher expected future costs and make profits on inflationary gain on the inventory. It appears that many small firms were unable to obtain sufficient credit to maintain inflated inventories. - 22 - increased 8% in real terms, and virtually all respondents (except leather) saw increasing export potential for 1977. Final product prices increased 17% and raw materials 23%, both less than national indices and indicative of the profit squeeze that small firms often suffer diring inflation. Employment increased 16% among the surveyed firms, far ahead of the estimated 3% national increase, with the average value of production per worker decreasing slightly in real peso terms. 1/ However, salary levels decreased in real terms, indicating difficul- ties which many workers have in maintaining the level of their salaries in real terms during accelerating inflation. Finally, over 50% of investment funds for 1976 were generated by the firm or its owners, while only 20% came from financial institutions. Commitments and Disbursements 3.06 The loan became effective on May 20, 1975, four months after it was signed. After an initial start-up delay of several months, subloan commitments built up rapidly and soon ran ahead of initial forecast. By May 1976, six months before the original terminal date for submission of subprojects, the US$5 million lending portion of the loan had been practically committed. The US$1.25 million allocation for medium enterprises was committed even sooner, by March 1976, due to the much smaller size of the allocation in relation to the absorptive capacity of the target group. Given the low demand for tech- nical assistance subloans and the continuing strong need for productive subloans, the Bank, at the request of CFP, reallocated US$0.35 million of the US$0.5 million technical assistance component to the lending component in May 1976. -However, in response to delays in the effectiveness of Loan 1451-CO, CFP maintained small uncommitted balances through 1977 for several subloan requests under review, and the terminal date for subproject submission was extended twice to December 31, 1977. Bank funds made a significant contri- bution to CFP's financing, representing about 14% of total CFP commitments between mid-1975 and year-end 1976. 3.07 Disbursements under the loan did not start until October 1975, i.e., four months after loan effectiveness, but disbursements were only less than three months behind schedule by December, 1976 (Annex 2). The small disburse- ment lag resulted mainly from (i) necessary adjustments in the Bank's and CFP's disbursement procedures; (ii) initial commitment delays; and (iii) small entrepreneurs' delays in meeting the legal and other conditions of subloan effectiveness. Also, at times the average lag of about five weeks between CFP's submission of disbursement requests and the Bank's disbursements caused some liquidity problems for CFP but neither BR nor the Monetary Board nor the Bank were prepared to allocate funds to a revolving fund to speed up disburse- ments. As of December 31, 1977, the original closing date for the loan, aggregate disbursements amounted to US$4.98 million, equivalent to 91% of the loan. As of December 31, 1978, the undisbursed balance amounted to approxi- mately US$69,000, and CFP expects to submit final disbursement requests in January 1979. Overall, the commitment and disbursement experience of the loan was quite good, especially considering the innovative nature of the project. 1/ But increasing in dollar terms to US$5,600 due to the inflation/ devaluation gap. - 23 - Characteristics of Subprojects Financed 3.08 Given the scarcity of local funds for long term fixed asset financing, CFP experienced little difficulty in identifying a large number of eligible projects of small manufacturing enterprises. A total of 305 subprojects were financed under the loan, as compared to an original target of 200. Of these subloans, 278 benefited enterprises with assets below Col$10 million and 27 with assets between Col$10-20 million; 20 subloans financed technical assistance. Only 21 subloans were above the Col$1.5 million "free limit" for Bank review, but they accounted for 25% of total commitments under the loan. A detailed breakdown of subloan approvals under the loan is presented in Annex 3. 3.09 Subloan size and maturities. The average size of subloans under the lending component was Col$0.7 million (about US$19,900 equivalent 1/) i.e., significantly lower than the US$25,000 forecast at the time of appraisal. About 66% of the subloans by amount were in the range of Col$0.3-Col$1.5 million and only 11.8% above Col$3 million. All recipients opted for subloans in pesos at a 24% p.a. interest rate. Only about 10% of the subloans had maturities greater than 6 years, 68% had maturities of 4-6 years, and 22% had periods of exactly four years, the minimum allowable under the loan. The relatively short average period of subloans reflects in part a tendency by CFP to set rather optimistic payback periods for projects, but also a reluctance by many entrepreneurs to indebt themselves for long periods. However, as recommended by the Bank, in 1976 CFP introduced a detailed flow of funds analysis for all subloans exceeding Col$350,000 to improve the adequacy of the maturities and grace periods granted in the light of the subprojects' cash-flow characteristics. 3.10 Interest rates. Average inflation in Colombia was unusually high during the loan commitment period, moving from 23.6% in 1975, the first year of the loan, to 20.0% in 1976 and 34.7% in 1977. 2/ This implies that term interest rates in Colombia turned negative during the period. The Bank has traditionally used the much larger financiera loans as the vehicle for addressing industrial interest rate policy, and was thus mainly concerned that the interest rate on the first SSI loan fit within the prevailing interest rate structure for industrial lending. Thus CFP agreed to charge a 24% p.a. interest rate under the lending component, i.e., slightly lower than the 25% rate under the Bank's financiera loans 3/ at that time, but above FFI's 21% average relending rate and CFP's lending with own resources at 22% p.a. 1/ An average exchange rate of US$1 = Col$35 for the commitment period of the loan is used. 2/ The respective December-December inflation figures are 17.9%, 25.9%, and 29.3%. These figures are exceptionally high compared to the historic rate of inflation in Colombia which averaged about 8% p.a. during 1950-1972. 3/ The effective interest rates are somewhat higher since both CFP and the financieras charge interest quarterly in advance. - 24 - The annualized inflation rate fell to an estimated 18% for 1978, so the interest rate on the Bank loan has been marginally positive during the disbursement period (using average inflation figures) and it now appears likely that the interest rate will result in the expected "acceptable positive cost of capital" over the effective life of subloans. 1/ 3.11 Subborrovers' characteristics. The ultimate beneficiaries' charac- teristics adhered closely to initial expectations. Two-thirds of the assisted concerns were each owned by less than five owners/shareholders, had total assets below Col$3 million (about US$85,000 equivalent) and employed less than 30 workers. About three-fourths of the firms had annual earnings exceeding 10% of total assets at the time of subloan approval. Similar to CFP's other lending programs, assisted enterprises represented a wide variety of indus- trial subsectors with food (19%), chemicals (11%), apparel and footwear (9%), metal working (7%), textiles (5%) and non-metallic mineral industries (6%) accounting for about 56% of approved subloans. Owing to the indivisibility of their managements, limited size of their market, the lack of suitable facilities and other difficulties to operate firms during the night hours, less than 20% of the assisted concerns were operating more than one shift per day. 3.12 Subproject characteristics. The average total cost per subproject amounted to about Col$1.6 million (US$45,700 equivalent) as compared to an original estimate of Col$1.2 million (US$34,300 equivalent). As a result of these larger than anticipated average subproject costs, the total cost of the project will reach US$13.2 million, i.e., 29% higher than expected. In line with these figures, the Bank's participation amounted to 41% of project cost (Table 1) compared to an original forecast of 54%. Reflecting its severe local currency resource constraints, CFP contributed only 8% of total subproject costs, 2/ i.e., substantially less than the 30% forecast, while subborrowers contributed 34% (16% forecast) and other local sources, including suppliers' credit and the extra-bank market, provided 17% (none forecast). Overall, institutional credit financed only 49% of subproject costs, compared with 84% projected at appraisal, since most of the medium scale firms were found to have easier access to institutional sources of finance, while the smaller firms had to rely more heavily on internally generated funds and share capital increases to complete their subproject financing. 1/ Inflation is projected at 16% for 1979, and 14% for 1980. 2/ CFP was required to contribute a minimum 10% of the subloan amount for each subproject. - 25 - Table 1: SOURCES AND USES OF FUNDS FOR SUBPROJECT FINANCING UNDER LOAN 1071-CO Uses Fixed Working Technical Sources Assets Capital Assistance Total in Col$ millions CFP Credits CFP Own Resources 0.1 4.2 4.3 Fondo Financiero Industrial 2.0 25.8 - 27.8 World Bank 208.7 - 3.4 212.1 USAID 0.4 2.9 - 3.3 Proexpo - 2.8 - 2.8 Ultimate Borrowers' Resources: Increase in Share Capital 17.9 18.4 0.1 36.4 Internally Generated Funds 87.0 50.5 3.0 140.5 Other Financial Institutions 13.1 22.3 0.3 35.7 Others 7.3 45.4 52.7 Total 336.5 172.3 6.8 515.6 3.13 As shown in Annex 3, most subprojects were for capacity expansion (90%) and primarily made use of locally manufactured inputs (90%). Only about 6% of subprojects were geared towards exporting. As expected, purchases of imported machinery and equipment (62%) substantially exceeded those from local manufacturers (381). Contrary to initial forecast, Y/ however, an unexpectedly large percentage (35Z) of the amount oneat vent for industrial building con- struction. Most recipients of these industrial construction subloans operated in inadequate industrial premises or had been ordered by municipal authorities to relocate to comply with zoning regulations and these subprojects were thus justifiable. However, since limited other sources existed for construction financing, such as the Central Mortgage Bank, the Bant urged CFP with some success to focus these operations on clients who had no access to alternative funds. 3.14 An analysis of a sample of 75 subprojects revealed the following average distribution of subproject costs. 1/ Only about one-fifth of the subloans were expected to involve financing of industrial building construction. - 26- Table 2: BREAKDOWN OF SUBPROJECT COSTS BY MAJOR EXPENSE CATEGORY Subproject Item Percentage of Total Subproject Cost Machinery & Equipment 46 Working Capital 28 Construction 22 Other (mainly land and technical assistance) 4 100 Based on these average subproject cost percentages and estimates of foreign exchange content of the above categories, the average foreign exchange content of subprojects financed under the loan has been calculated at about 41% as compared to 56.8% expected at the time of project appraisal. This significant variance has been due almost entirely to a much greater need for construction financing than originally anticipated. However, given the lower than expected Bank contribution to total project financing, the loan financed only the foreign exchange component of the project, as expected. 3.15 Employment generation. In order to gain a fuller understanding of the employment generation impact of the loan, the Bank undertook an extensive study 1/ as part of the appraisal of the second CFP loan (for details see Appendix 5 of Report No. 1512b-CO). According to the projections included in the subloan appraisals, the typical subproject created 7.65 jobs, at an average investment cost (including permanent working capital costs required as a component of the investment project) of US$6,900 (the accuracy of these projections is discussed in paras. 3.22-3.25), as compared with an Urban Poverty Program threshold of US$7,470 for Colombia. The cost per job for fixed asset investment only was US$5,100. These figures are roughly in line with national estimates 2/ for SSI in Colombia.. 3.16 Table 3 presents the average total investment cost per job generated by different sizes of firms. It shows that the cost rises rapidly with the increasing size of firms, with the investment cost per job for the smallest firms being one-third of that for larger firms. Notwithstanding the wide variance within each group, these figures provide ample evidence 3/ that 1/ The study encompassed 235 of the 305 subloans, using 10 data items collected during subproject appraisal and 9 computed variables. The Statistical Package for Social Sciences (SPSS) was used for processing. 2/ By the Departamento Nacional de Estadistica (DANE) and others. 3/ The Pearson correlation coefficient (zero order) for these two variables-- size of firm and cost per job--is 0.4570, significant at the 0.001 level. - 27 - within the small industry subsector in Colombia the smaller firms create more jobs per peso invested than the larger firms. Excluding all construc- tion subprojects (defined as subprojects with over 80% of investment costs spent on construction) from the analysis, the results do not change signifi- cantly. Table 3: BREAKDOWN OF JOB GENERATION COSTS BY ASSET SIZE OF FIRM Average Cost per Assets Number of Job Generated (Col$ 000's) Firms (Col$ 000's) (US$ 000's) Less than 500 19 130 3.7 500-1,000 29 139 4.0 1,000-2,000 50 175 5.0 2,000-3,000 29 227 6.5 3,000-4,000 19 231 6.6 4,000-5,000 12 336 9.6 5,000-10,000 23 378 10.8 10,000 + 19 410 11.7 US$1 = Col$35 3.17 There was also a strong correlation between the total investment cost and cost per job, and between the number of jobs generated and the cost for each job. However, it is likely that this is caused by the close relation- ship of these variables with the asset size of the firm. No correlation was found between various ratios such as percentage of investment cost spent on machinery, return on assets, and percentage increase in assets caused by the project and cost per job. Again, this may be due to the fact that asset size and cost per job is the dominant relationship, overshadowing the others in a statistical analysis. The data also indicate that the clothing and textiles, and wood and furniture industries have relatively low costs per job created, while the chemical industry has high costs per job. 1/ 3.18 Industrial decentralization. The regional distribution of Bank financing was not as wide as anticipated. Assisted enterprises were geographic- ally dispersed as follows: Cundinamarca including Bogota (33.4%), Antioquia 1/ A Pearson correlation coefficient could not be calculated since the industry variable is not continuous. A chi-square test gave a signif- icance of 0.0031 when industry and investment per job generated were calculated, indicating that cost per job does vary significantly by industry. - 28 - including Medellin (19.0%), Valle including Cali (12.5%) and the rest of the country (35.1%). The concentration of subloans in the Cundinamarca province was the result of various factors, including (a) the smaller size and lower business potential in the regional markets than in Colombia's main population center; and (b) the limited operational capacity of CFP's newly established regional offices during much of the commitment period. Overall, however, CFP's industrial decentralization efforts have been satisfactory, although some of the new regional offices may initially serve too small a market to achieve efficient operating levels. 3.19 Contribution to industrial output. Using the sales forecasts for individual subprojects at the time of subloan request, the project is estimated to increase the average annual sales of recipient firms by 89% from Col$4.4 million to Col$8.3 million. The projections would indicate a relatively low incremental capital/output ratio (ICOR) of about 1:1 for SSI in Colombia. CFP surveyed 76 firms with projects operating for at least one year to assess the accuracy of these projections (Annexes 10 and 11). While in the aggregate the sales projections for the firms (Col$650 million) virtually equal the actual figure (Col$656 million), the deviations in individual cases are substantial as only about one-fifth of the projections fell within 20% of the actual figures (paras. 3.21-3.23). 3.20 Income distribution. The profits of all subborrowers are expected to approach Col$300 million annually, almost half of which will be due to the project. Substantial additional income also accrues to workers, owners, and government through wages and taxes. Since no income distri- bution data were collected on the subprojects, due to the difficulty that would have been involved, some estimates can be made by using the Colombia Special Study of Financiera-Assisted Projects (Report No. 1037-CO). According to this study, unskilled workers receive 22% of the income from "small" projects; government, 76%; owners 37%; and there is a net loss to consumers and creditors of 35%. Thus the income distribution impact of the loan is heavily dependent upon the spending pattern of government, which, according to Urrutia, I/ has a positive impact on income distribution. If a set of distribution weights 2/ is used to adjust the net benefits accruing to the project, and account is taken of the industrial sector allocation of the loan, an income distribution index of 121 3/ is calculated as compared to 108 for the projects reviewed in the Special Study, i.e., the income distribution impact of the project favors the lower income groups. 3.21 Technical assistance financing. As of December 31, 1976, little use had been made of this financing by either CFP or small scale industrial 1/ Urrutia, "Revista del Banco de la Republica," June 1972. 2/ Government 1.2, Owners 0.8, Workers 1.4, and Others 1.0, as estimated by the Colombia Special Study. 3/ Indices above 100 are assumed to indicate improvement of the existing distribution, with the higher indices favoring lower income groups more heavily. - 29 - enterprises. As of that date, only 11 technical assistance subloans had been extended for an amount of Col$2.1 million (about US$60,000 equivalent) as compared to 100 technical assistance subloans totalling US$500,000 expected at the time of appraisal. As explained in para. 3.06, US$0.35 million of the US$0.5 million technical assistance component was transferred to the lending component, taking into consideration the lower than expected demand. CFP committed the remaining US$90,000 equivalent in technical assistance funds in early 1977, mainly to sponsor SMI research. While these funds financed some worthwhile studies, the new management 6irtailed all further research and thus may have virtually eliminated any long-term impacts of the Bank loan on building up CFP's research program. An important constraint to technical assistance financing of SMI has been the lack of awareness by small entrepre- neurs of the potential benefits of technical assistance and the scarcity of qualified consultants to provide in-depth individual technical assistance at reasonable prices. Although the project achieved little success in channeling technical assistance financing to SMI, it was helpful in developing the capa- bilities of CFP's technical assistance department, as shown by the experience gained under the second CFP loan. Results of Subloans 3.22 CFP undertook a survey of 76 subborrowers in order to assess the impact of subloans and judge the accuracy of projections made at the time of subproject appraisal. Annexes 10 and 11 contain data comparing projected and actual sales, profits, employment, and project cost of firms. On the whole, there appears to have been a tendency to overestimate sales and profits, as over 60% of the projections made did so. Nevertheless, actual sales were slightly higher than projected in aggregate volume, due to a few large sub- projects. Also, the main reason for the optimistic bias of many projections appears to have been unrealistic expectations about how quickly the projects can be implemented and new sales and profit levels achieved. 3.23 Project costs were underestimated in 75% of cases, mainly due to higher than expected inflation in Colombia, but also in part because CFP analysts were often reluctant to allow adequate contingencies. 1/ In aggregate, cost overruns were about 16%. It is interesting to note that projections in the large cities of Bogota and Medellin resulted as more conservative than in other cities, probably because they employ relatively more experienced staff and due to favorable local conditions which allowed the subprojects to be implemented faster. In over 60% of the subprojects employment increases were actually underestimated, and about 9% more employment was created than expected. The average investment cost per job, based on actual figures one year after subproject implementation, for the client firm as a whole was US$6,000, i.e., substantially below the Urban Poverty Program threshold for Colombia of US$7,470. On balance, the actual aggregate achievements of the Bank-financed subloans were roughly equal to those projected during subproject appraisals (paras. 3.12-3.19). 1/ The Bank has frequently pointed out the need for this in commenting on subprojects. - 30 - 3.24 Despite these overall encouraging results, it should be noted that case-by-case disparities in projected and actual performance are substantial in the sales and profits categories, where only one-fifth and one-eighth, respectively, of the projected figures fell within 20% of actual figures. CFP analysts make serious efforts to prepare accurate, detailed income statements and cash flow projections as part of subproject appraisals. However, many factors, such as competition in small local markets or the likely impact of business cycles on SSI, can significantly affect sales and profits projections. Overall, the projections seem a worthwhile tool to indicate the potential viability of the project and point up areas of high sensitivity, but substan- tial efforts aimed towards refining further the accuracy of estimates are likely not justified. 3.25 Data on subloans in arrears are still not representative because most subloans were still in or very near the grace period as of June 30, 1978. As of that date, 6 of the 305 subborrowers were bankrupt and out of business, and about 10 more were considered beyond recovery. In only 3 cases was CFP likely to suffer significant losses following liquidation of subborrowers. Overall, the arrears performance of Bank subborrowers is expected to be similar to that of other CFP clients (paras. 4.29-4.30). IV. INSTITUTION BUILDING Management and Staff 4.01 In early 1975, shortly after Board approval of Loan 1071-CO, CFP experienced administrative problems that led to a large-scale turnover involving management and key staff members at headquarters and in the regional offices. A new general manager was appointed who proved capable. He helped to create a dynamic and progressive attitude within CFP and strengthened the organization. Personnel nearly doubled to 354 between 1974 and 1976, and most senior posi- tions were held by capable and experienced individuals. Differences in opinion between the general manager and the Minister of Economic Development (the ex officio Chairman of CFP's Board) eventually led to the resignation of the general manager in August, 1977, dealing a blow to the institutional progress that had been made over the previous two years, especially since all 3 sub-managers and 5 of 14 local office managers also resigned. 4.02 The new management of CFP was faced with the difficult task of guiding an unfamiliar and complex institution with little help, as few expe- rienced senior staff had remained to assist them. Also, CFP continued to suffer from a high turnover of professional staff, partly as a result of unattractive salaries, with 57 resignations in 1977 and 38 in the first six months of 1978 compared to only 21 in 1976 (Annex 4). By mid-1978 most department heads, especially in operations, had been in their jobs less than one year, and about half less than six months. The new management, finding that the rapid growth of CFP had resulted in some overlapping of functions and lax administrative controls, has altered the institution's emphasis towards greater cost-consciousness and profitability, reducing the staff and CFP's non-income generating activities. GYP is now more closely administered and its - 31 - portfolio is well controlled. While the new management deserves credit for substantial achievements in these areas during the past year, some of CYP's former dynamism may have been lost in the process. I/ 4.03 Above all, there is one important management lesson to be learned from CFP's experience under the loan, i.e., the importance of stable, steady growth of operations and resources. The former manager expanded CFP rapidly to manage the first Bank loan, a proposed KfW credit, a USAID credit, and several proposed urban development projects with SSI components. However, the second Bank loan and several other credit lines were substantially delayed, and 1977 saw a pronounced slowdown in CFP's growth, especially in terms of 1976 expectations. Management reduced operations to the size justified by the then available resources. However, when in early 1978 the second Bank loan became effective, and other funds became available, CFP's organizational capacity was too limited to allocate these funds as rapidly as demand would permit. In conclusion, CFP has suffered substantially from this recurring "feast and famine" of resources (para. 4.14, Table 5), as would any organiza- tion subjected to external factors that imply frequent and dramatic changes in its modus operandi. 4.04 In connection with Loan 1071-GO, CFP diversified its board and lessened the commercial banking influence of Banco Popular. This move has generally benefited CFP except during a short period in 1977 when jurisdic- tional discussions between the Ministers of Finance and Economic Development 2/ slowed the flow of resources to CFP. The role of the board in directing the institution has been substantial and generally quite positive despite frequent changes in the de facto chairman, the Secretary General of the Ministry of Economic Development, due to non-CFP related matters. CFP has received substantial.guidance from many of its board members and benefited from their diversified skills and experience during a period of rapid growth. Organization 4.05 CFP's organization consists of a headquarters and 14 regional offices, of which the 5 newest were established in 1975. As required under Section 2.09 of the Project Agreement, in May 1975 CFP contracted the services of a management consulting firm, Asesoria Gerencial Ltda., to carry out an organization study to (i) make a critical analysis of its organizational arrangements and internal systems and procedures and (ii) help implement the needed changes that would enable the institution to maintain an efficient internal administration and to achieve its operational and institutional objectives. The study's recommendations were adopted by CFP's management and board in early 1976 and the process of implementation began shortly thereafter. At appraisal of the second loan to CFP (in October 1976) the Bank reviewed the organizational structure and was satisfied thatthis objective had been achieved. I/ CFP's general manager changed again in September 1978, following the accession of the new government. It is too early to predict the impact of this change. 2/ Banco Popular is under the Ministry of Finance while GYP is under the Ministry of Economic Development. Banco Popular has been by far CFP's most important source of unrestricted funds. - 32 - 4.06 The new organization was short lived, however, since the new manage- ment, as part of an effort to contain operating expenditures,downgraded several positions. The revised organizational structure consolidates responsibilities under 2 newly hired submanagers rather than 5, and de-emphasizes long-term planning and research activities. Still, the organizational revisions are appropriate to the changed emphasis (para. 4.02), although, combined with high management turnover, they have at times resulted in placing new responsibilities upon staff who were not fully prepared. The new management has also increased the role of'the Regional Advisory Boards (para. 4.12) and given substantial responsibility to regional office managers for loan processing and cost control. Systems and Procedures 4.07 Improvements in appraisal techniques, loan supervision, portfolio monitoring, internal auditing, information systems, and financial management were key objectives of the loan and received substantial attention during the Bank's loan supervision. Subloan appraisals improved substantially during 1976 through better documentation and the preparation of a detailed credit manual. However, appraisal quality had decreased by the time the second Bank loan was declared effective, due to large staff turnover in the operations area of CFP during the relatively long period between access to Bank loans. 1/ However, the credit manual facilitates greatly the training of new analysts, and in mid-1978 CFP was taking adequate steps to elevate the quality of appraisals to former levels. 4.08 In the past, CFP relied heavily upon university students to supervise loans. While this was adequate for collecting information on clients, most students were not able to provide guidance to clients or identify problem areas. However, the students gain experience from their supervision visits and a number of them are employed by CFP upon graduation. CFP's new management has substantially upgraded the scope and thoroughness required of supervision visits and in mid-1978 issued guidelines greatly increasing the emphasis on supervision. 4.09 Under the new management portfolio monitoring has shown remarkable improvements. Detailed information on loans in arrears is produced quarterly and available within a few weeks of the end of the period. Also, CFP now produces monthly individual status reports for each loan in arrears. CFP's control and administrative procedures for subprojects financed under the loan have proved effective. Adequate records provide ready information on the status of individual subloans and the overall position under the loan. CFP maintains supporting documentation for Bank disbursements in its own files instead of submitting them to the Bank, and reviews of these files by the Bank's Controller's Department have found that expenditures are adequately documented. 1/ Most staff who were trained in project appraisal under Loan 1071-CO were no longer with CFP when Loan 1451-CO became available for commitment in early 1978. - 33 - 4.10 The quality and scope of statistical information on CFP operations have improved remarkably over the past several years, partly due to the impact of Bank reporting requirements under the loan. With encouragement by the Bank, CFP began developing an automated information system in 1976. While as much of the work as possible was being done by internal staff, the Bank offered to finance the services of external consultants from the technical assistance component of the second loan. The 1977 management change slowed progress in this area, as the cost-conscious new management has given diminished priority to the system's development. Although CFP has signed a two-year contract for advisory services in creating the system as required under Loan 1451-GO, it is proceeding on a somewhat smaller scale than originally envisioned. 4.11 CFP took several steps in 1975-76 to upgrade its internal control procedures. CFP now prepares cash-flow forecasts monthly and three-year financial and operational projections annually. The internal auditors ensure adequate financial control procedures, and the external auditors have found these procedures satisfactory. Also, CFP has recently instituted strict monthly budget ceilings in its various regional offices, which have been effective in assisting cost control. CFP's external auditors through 1976 performed adequately, but were changed by the new management. The 1977 audit report, prepared by an audit firm new to CFP but with experience in auditing Bank-financed projects, is acceptable overall. 4.12 Controls at the regional office level. Decentralization through increasing the regional offices' authority and responsibility occurred pro- gressively. This process was most evident in relation to loan approval and portfolio control. Although an Advisory Board for each of the regions was operational by the time of Bank loan approval, only in 1978 were the boards' terms of reference significantly expanded and their composition standardized. Currently consisting of four prominent members of the respective local community, one of whom is the local manager of Banco Popular and another the representative of ACOPI, the national SSI association, these boards' functions include the critical examination of analysts' appraisals and, jointly with the regional manager, the decision upon loan action up to a certain limit in respect of the particular regional office. The managers interviewed held favorable views on the practical worth of these boards in facilitating the quality of loan approval operations at the regional level. 4.13 The previously ad hoc and sporadic loan supervision work became systematized in 1976 under control of the headquarters' supervision depart- ment. Activities took the form of 'visits,' frequently of no more than two hours' duration, by CFP full-time supervision staff, or by credit analysis in the smaller regional offices. Reportedly for reasons of decentralization, the headquarters' supervision department was eliminated in early 1977; however, this move was not accompanied by effective procedural arrangements and follow-up action to foster supervision activities in the local offices. Consequently, supervision activities generally fell off in both quantity and quality. Headquarters' decision to decrease the use of final year university students for supervision, coupled with the high incidence of staff resignations, obliged many regional offices to deploy their more experienced credit analysts to help out with such supervision work as was performed. For the 1975-78 - 34 - period as a whole, supervision was not fully satisfactory, although CFP's concern with satisfying requirements of Loan 1071-CO tended to imply better supervision of Bank-financed subloans. In May 1978 the management issued guidelines requiring supervision of 90-100% of loans, but many regional offices feel this target is too ambitious. Resources 4.14 During 1975-76 CFP relied primarily on its own resources and its traditional creditors--FFI and Banco Popular--to support its working capital lending programs. Facing a lack of government capitalization and erosion of its assets through inflation, FFI was compelled to adopt stricter eligibility criteria 1/ in 1975, thus limiting access for many SSIs to its rediscounting facilities. Reflecting this trend, CFP's borrowings from FFI rose somewhat from Col$138 million in 1974 to Col$165 million in 1976 but fell from 27.6% to 16.3% of CFP's total resources. 2/ Increases in equity, on the other hand, played an important role due to the government's commitment under the first loan to maintain CFP's share capital in real terms. As of December 31, 1976, equity amounted to 31% of total resources as compared to only 24% three years before. Due to shortages of term resources, in 1976 CFP began engaging in short-term export financing of large enterprises with resources of the Fondo de Promocion de Exportaciones (PROEXPO), with these operations amounting to 12.7% of CFP's total resources by the end of the year. 3/ 4.15 During 1977 CFP's total borrowings increased by over 50%, with borrowings related to foreign lines of credit nearly doubling, mainly as a result of subloan disbursements committed in 1975 and 1976 under Bank and USAID loans. Despite great uncertainties attached to increasing the number of CFP's sources of borrowings, the Bank had hoped for some diversification at the time of appraisal. However, borrowings from FFI rose 65% between 1974-77 as the government expanded FFI's resource base, and there was even an increasing 1/ In Medellin, Cali and Bogota firms with total assets above Col$10 million had to export 50% of their production to be eligible for FFI's financing. Cancellation of at least 50% of any prior obligation with FFI and expansion of equity at least equal to FFI's original financing were made prerequisites for renewed access to the Fund and FFI's fixed asset financing was limited to the lower of 20% of the ultimate beneficiaries' net worth or Col$2 million. 2/ Banking regulations limited Banco Popular's maximum financial assistance to CFP to 10% of Banco Popular's equity, i.e., about Col$100 million as of year-end 1976. Nevertheless, FFI and Banco Popular accounted for 45% and 19%, respectively, of CFP's resource inflows between year-end 1973 and 1976. 3/ Mainly because of their great importance for CFP's profitability, the Bank accepted CFP's PROEXPO-type operations during negotiation of the second CFP loan provided they did not exceed 1.5 times equity and profits generated were used for technical assistance funding or studies. - 35 - reliance on the traditional sources of credit. The only specific objective established by the Bank under the loan was for the government to contribute sufficient capital to maintain CFP's equity in real terms, which was achieved through 1976 although with some delayed contributions. 1/ Although in 1977 the government failed to provide CFP with the full equity contribution that the Guarantee Agreement of Loan 1451-CO required, these funds were made available in early 1978. Operations 4.16 Antecedents. CFP's overall operations had grown rapidly during 1970-75 as shown in Table 4 below: Table 4: CFP CREDIT OPERATIONS, 1970-75 Credits in millions of current Col$ Year Approved Disbursed 1970 89 67 1973 274 210 1974 331 286 1975 467 385 Of the nine regional offices operating in 1974, Bogota, Medellin, and Cali accounted for 63% and 51% of all loans by amount and number, respectively. Operations to 1975 consisted mainly of rediscounting SSI loans with FF1. 4.17 Lending operations. Table 5 below and Annex 5 give the disburse- ments of the principal lines of credit utilized prior to and during the loan period: 1/ The government did not accept this formula under the second loan and specific contribution amounts were established for 1977 onwards. - 36 - Table 5: LOAN DISBURSEMENTS (in millions of current Col$) 1st. half Source 1974 1975 1976 1977 1978 FF1 j 166 155 166 283 165 PROEXPO 18 35 185 329 44 World Bank /b - 20 121 55 27 USAID /c - - 81 169 17 Others 63 108 47 30 25 Total 247 318 600 866 278 /a Includes CFP's counterpart contribution, about 28% of total. lb Includes CFP's counterpart contribution, about 10% of total. Ic Includes CFP's counterpart contribution, about 50% of total. Three main observations can be made from the above figures: (i) following relatively modest disbursements in 1974-75, sharp increases were achieved in 1976 and 1977, with the first six months of 1978 showing a decline to below the 1976 level; (ii) local lines of credit have continued to represent the bulk of CFP lending operations; and (iii) CFP operations are subjected to substantial variations as a result of wide fluctuations in availability of resources but have exceeded overall the appraisal forecasts (Annex 5). Except for 1976 the bulk of CFP term lending operations was made with FFI funds. The substantial decreases in all except FFI lending in the first half of 1978 largely reflect the delay in effectiveness of the second CFP loan. 4.18 In line with its statutory objectives, CFP has focussed its lending primarily on the lower end of the SSI spectrum. Over the period 1975-77 almost 80% of CFP's loans went to enterprises with total assets below Col$2 million, with an average loan amount of about Col$300,000. Assisted firms continue to represent a wide range of industrial subsectors, with apparel and footwear, metal working, non-metallic minerals, food and chemical industries accounting for about 62% of aggregate approvals. CFP's loan periods have been increasing significantly, as 14% of loans in 1975, 28% in 1976, and 39% in 1977 were for more than three years. Additionally, CFP has been able to double its lending outside the three major departments (Cundinamarca, Antioquia, and Valle). However, as much as 85% of CFP financing went for working capital in 1977, the highest percentage in its history, reflecting its resource mix (para. 4.15). 4.19 Other financing. Equity participations still remain marginal within CYP's overall operations. Return on equity investments has been poor, and several companies are in serious trouble. Over the past several years CFP has not invested in manufacturing companies, but only in several industrial estates. - 37 - 4.20 Technical assistance. Prior to 1975 CFP's technical assistance activities were scanty, confined to group training of borrowers in bookkeeping, financial control and administration practices, and were mainly directed to the Bogota region, with some attention to Medellin and Cali. Supported by a US$1 million technical assistance loan from USAID, implementation of a comprehensive and integrated technical assistance program began in January 1976. The USAID allocation for technical assistance turned out to be ample and this was a major cause of CFP not using the bulk of the Bank's technical assistance allocation (para. 3.20). The basic philosophy of the program was that CFP's technical assistance staff should be the means of detecting the precise needs of borrowers and then of channelling and monitoring indicated services from external sources; thus CFP itself was not envisioned as the prime source of services. 4.21 Despite the program's newness and the high staff turnover (17 tech- nical assistance staff members resigned during the 18 months to June 1978), the three main components of the assistance program (training/motivation courses, industrial in-plant consultation work, information services) provided both effective complementary support to CFP's lending operations and helped to advance the productivity and efficiency of SSI, although on a rather limited scale. The subsidy element of the USAID loan also enabled some services to be subsidized, e.g., participants in training courses contributed no more than 50% of actual cost, and preliminary diagnoses on industrial problems were free of charge. While the time required to build up and gain clients' ready acceptance of technical assistance inputs was significantly underestimated, this observation does not detract from the pioneering nature and overall positive contribution of the program and CFP's learning experience during the period. In many cases, CFP found that public organizations such as SENA often do not have the experience to provide adequate advice to solve the problems of individual firms, but that good private consultants seem expensive to small entrepreneurs. Still, experience has shown that the extra cost of private consultants is often justified in terms of quality of services, and CFP is slowly convincing clients of this through demonstration projects in various cities. As a result, demand for private consultants for technical assistance picked up substantially in 1977 and 1978. The technical assistance program's principal accomplishments are summarized in Table 6 below: Table 6: TECHNICAL ASSISTANCE ACTIVITIES Numbers 1st. half Type of Activity 1976 1977 1978 Seminars, courses, etc. 64 100 38 Borrowers receiving indust. consult. services 158 180 88 Information services: Monthly issues of "Impulso" /a 6,000 7,000 8,000 Question and answer service n.a. 592 800 /a A technical bulletin/newsletter for SMI. - 38 - Financial Position and Results 4.22 Financial structure. Projected and actual financial statements and ratios from 1974 to the first half of 1978 are shown in Annexes 6, 7 and 8. Despite its management changes and the tightening of its local currency position, CFP has substantially achieved the resource growth forecast under the loan. Borrowings in domestic currency have even exceeded expectations, mainly due to the unforeseen use of PROEXPO funds for short-term export credits, while foreign borrowings have been lower than expected, mainly because of the delays in effectiveness of a bilateral loan. 4.23 The portfolio size has developed as projected, despite liquidity levels that have increased substantially recently as a result of sudden resource availability and limited processing capacity (para. 4.03). Total assets stood at about US$38 million equivalent in June 1978, as compared to US$18 million equivalent at year-end 1974. Overall, a real growth in pesos of about 10-15% p.a. occurred, although the growth was sporadic and occurred mainly in 1976 and early 1977. 4.24 From 1974 to 1977, borrowings increased sharply from Col$265.5 million to Col$988.9 million and equity expanded from Col$200 million to Col$300 million, with the total debt to equity ratio rising from 1.3:1 in 1974 to 3.3:1 in mid- 1978. CFP's overall financial structure has thus remained rather conservative and well within the 4:1 debt ceiling defined under the loan. 1/ 4.25 As CFP had experienced some difficulty in maintaining an adequate liquidity position due to lagging debt collections and rigid borrowing terms, an understanding was reached with CFP at loan negotiations on measures to maintain a satisfactory liquidity position. Since then, its current ratio has remained satisfactory between 1.6:1 and 1.7:1, as compared to 1.05:1 in 1973. 4.26 Profitability. Due to a rather stable interest rate structure, revenues as a percentage of average total assets amounted to 17.9% in 1977, as compared to 17.1% in 1974. However, as a result of inflation and CFP's reorganization and rapid expansion of its branch network in 1975,and contrary to initial expectations, administrative expenditures--particularly salaries and other personnel expenses--have shown an increasing trend. This trend is apparent in the ratio of administrative expenses to average total assets, which increased from 6.6% in 1974 to 7.8% in 1976, instead of decreasing slowly to 6.1% by 1976 as had been anticipated. However, subsequent tight cost controls lowered the ratio to 7.2% in 1977. As agreed with the Bank under the second loan, CFP is gradually reducing its current level of expendi- tures until it reaches 5.8% 2/ in 1981. As of mid-1978, administrative 1/ Given the quality of CFP's financial management and the expected improve- ment of its local currency borrowing, the debt/equity limitation was increased to 5.5:1 at the time of the second loan. 2/ While somewhat high compared to most "traditional" development banks associated with the Bank, such level of administrative expenses is not excessive considering the nature of CFP's activities involving a large number of small-scale enterprises and the maintenance of 14 regional offices. - 39 - expenses as a percentage of average total assets amounted to 6.2% (on an annual basis), reflecting CFP's recent efforts to contain administrative costs. 4.27 Based on CFP's experience so far, the cost-to-asset ratio close to 6% may be the minimum achievable without hampering CFP's effectiveness and given its current scope of operations. A governmental institution supporting SSI development such as CFP must engage in certain development activities. While these additional expenses cannot be separated so that a minimum administ- rative cost can be estimated accurately, CFP estimates that operating cost levels of 5% for operations in large cities and 7% in small cities might be a realistic minimum. 4.28 Although net profits reached a record high of Col$8.4 million (3.8% of equity) in 1975, they decreased to Col$2.2 million in 1976, and to Col$5.3 million in 1977, leading to a marginal 0.8% and 1.6% return on average equity, respectively. CFP's low profitability resulted partly from the fact that (i) the spreads it enjoyed on most of its borrowings, parti- cularly from FFI, had become too thin to cover administrative expenditures associated with operations involving the low end of the SSI spectrum or beneficiaries located in less developed regions, and (ii) high inflation causes costs to rise faster than income for a bank that makes term loans at fixed interest rates. However, the recent emphasis on cost controls has helped to increase profitability, with profits for the first six months of 1978 slightly exceeding profits for all of 1977. Quality of Portfolio 4.29 At the time of the appraisal and negotiation of the loan, the quality of CFP's loan portfolio was one of the major concerns of the Bank as it had been affected by a steady rise in arrears over the past years, particularly in 1974-75. As of year-end 1975, total principal in arrears had reached a high of Col$51.1 million, as compared to Col$24.0 million in 1973. On the same dates, portfolio affected by arrears amounted to 20.2% and 16.2%, respectively (Annex 9). Besides a rapid operational growth and the higher risks inherent in lending to SSI, this increase in arrears resulted mainly from (i) the slow- down of the Colombian economy in 1974 and early 1975; (ii) CFP's lax collection procedures; and (iii) frequent failure of CFP to obtain adequate loan guarantees. 4.30 Since mid-1976 CFP has taken corrective measures to improve its arrears position and quality of portfolio. Apart from the establishment of a properly staffed portfolio supervision department, a firm and concerted program of action was initiated to tighten portfolio control and reduce arrears to satisfactory levels. As a result of continued efforts by CFP, the quality of its portfolio has been improving steadily, and by year-end 1977 portfolio in arrears had fallen to 12.8%. The major factors in decreas- ing arrears were: (i) improved appraisal techniques and better follow-up on clients in arrears; (ii) more stringent insistence upon real guarantees for - 40 - loans; 1/ (iii) high penalty interest rates; and (iv) the use of a private collection agency, with fees of the agency charged to the delinquent account. Good appraisals and effective supervision are the best means of reducing arrears, but these require additional expenses and in some cases more experi- enced staff than CFP salaries can attract. V. SECTORAL IMPACT 5.01 In connection with the loan, the Bank staff suggested that in order to remedy Colombia's lack of a coordinated policy framework to promote SMI, a high level advisory council on SMI be established by the government to (i) examine past and propose new policies which affect SMI development, (ii) help ensure a sufficient and timely flow of resources to SMI, and (iii) take initial steps conducive to country-wide promotion and coordination of technical assistance programs. Additionally, it was also recommended that the government carry out a comprehensive evaluation of the effects of indus- trial policies on SMI and of its potential to.assist in accomplishing govern- ment priority objectives. However, while the government is aware of SMI's economic and social importance in creating employment opportunities and furthering regional development, it has not yet sufficiently coordinated its industrial policies which continue to pay little attention to the special needs of the small scale industrial sector. 5.02 The lack of progress in this field stems from several factors including that efforts to control inflation over the past several years have taken priority over many other issues as the government has .tried to restrict credit. Also, the Ministry of Economic Development, formally responsible for developing an SMI policy, may have preferred not to transfer its function to an inter-ministerial committee or other agency. The main sectoral impact achieved under the loan was the upward adjustment of interest rates for SMI, in particular in rediscounting terms of the FFI, making them more attractive to financial intermediaries and bringing them in line with comparable sources of industrial finance. Interest rates for SMI lending were raised to a maximum of 24% and FFI's spreads increased to a maximum of 4%. Also, when pursuing the creation of a national SMI policy council became unrealistic, CFP itself undertook the initiative to improve coordination with other institutions dealing with SMI. In late 1976 and 1977 CFP signed cooperation agreements with five other public and private agencies. 5.03 CFP has been successful in achieving a sectoral impact at the regional level, mainly as a result of the creation of the Regional Advisory 1/ Except in certain cases for very small firms or where personal guarantees are very strong, CFP limits the loan to 40-50% of the assessed value of machinery and 70% of the assessed value of real property offered as guarantees. Still, used machinery is often difficult to resell and the realizable cash value of the guarantee is substantially less than the appraised value of the property. - 41 - Boards (para. 4.12). In many cases these boards have become top-level com- mittees for promoting institutional coordination and designing integrated programs to benefit the smallest firms. Because the boards have substantial impact in directing the operations of CFP, their plans and programs are backed by CFP's implementation capability. The boards have also become a forum for airing the needs and views of small businessmen and channelling these to the national government level. VI. CONCLUSIONS 6.01 Overall, CFP's progress in institutional development since 1974 has been substantial. An important factor in this development was the dynamic leadership provided by management, particularly until August 1977. The management team that took over in 1977 also achieved substantial results, mostly in terms of consolidating operations and establishing internal control systems. The Bank played an important institution building role both in connection with the appraisal of the two loans and through close supervision. The Bank loan supported 305 subprojects through an average subloan amount of US$19,900 for subprojects averaging US$45,700. Most of the subborrowers were small with two-thirds having total assets of less than US$85,000 prior to receiving the subloans. The projected investment cost per job was US$6,900, slightly lower than the US$7,470 Urban Poverty Program threshold for Colombia. The interest rates on Bank subloans have been marginally positive over the disbursement period, and it is projected that they will be significantly positive over the amortization period of the subloans. Overall, this first Bank loan specifically designed to benefit SSI in a Latin American country was successful in reaching the intended target group to finance increases in production and employment. 6.02 The project's initial assumption that a public institution such as CFP could support SMI development effectively at a modest cost to the government has proved correct. In terms of providing financing to the sector, CFP applied criteria similar to a private institution except that the greater margins under the Bank loan (about 2-3 points higher than for the financiera loans) and the lower profitability expected of a public institution allowed CFP to make much smaller loans and undertake greater supervision and technical assistance activities per dollar lent that are needed in lending to SMI. CFP's loan appraisals, while less detailed in recognition of the small loans, improved substantially during the loan period and have become an effective means of assessing subproject viability. Additionally, CFP achieved an important role at the regional level in promoting SMI development (para. 5.03) that went substantially beyond mere subproject financing. 6.03 CFP's profitability has remained low during the loan commitment and disbursement period, but this does not necessarily represent a loss to the economy. High staff turnover has been a major cause of low efficiency, but this cost to CFP may have been a net gain for the economy as most staff went on to use their training in positions in private industry or banks in Colombia. CFP has undertaken certain activities such as technical assistance and research - 42 - that are not cost recovering but are justified on economic and social develop- ment grounds. These activities would normally be funded from the national budget, but in many cases CFP has used its operating income to finance them, thus lowering accounting profitability. While CFP lends mainly to SSI, most clients are experienced entrepreneurs with proven ability in the businesses, and CFP has had only limited success developing new entrepreneurs. Lending to new entrepreneurs is difficult, time consuming and risky, and an institution with a wide clientele such as CFP cannot likely be effective in fostering new entrepreneurship on a substantial scale. 1/ Finally, CFP has maintained offices in certain areas where the level of operations initially could not cover office expenses, but the offices are playing an important role in support of national policies of industrial decentralization. 6.04 On the other hand, CFP's continuing great dependency on the govern- ment for resources, the availability of which proved quite uneven, and frequent changes of management and short term objectives had adverse impacts. However, numerous public institutions in Colombia, including those with which the Bank is associated, have suffered from these problems in recent years, and CFP has been relatively successful in minimizing their effects. Experience has shown that there is a need for stable predictable growth of an institution like CFP, and this should certainly be an objective for the future. Also, little progress was made in achieving sectoral objectives, including the establishment of a policy framework or institutional coordination to provide greater support to SMI development. 6.05 In spite of the factors mentioned above that have impeded CFP's development, it has proved itself to be an effective SMI development institu- tion and worthy of continued support by the Bank. While Bank supervision reports have frequently referred to insufficient attention by the national government towards CFP and the SMI sector, CFP's internal policies and pro- cedures have generally been found satisfactory. In retrospect, the reason for CFP's success seems to lie in the sound banking approach it has used in lending. CFP provides substantial project promotion and loan preparation assistance, and focusses operations almost exclusively on small firms, but is rather strict in lending for sound projects, obtaining adequate guarantees, and pressing for loan collection. The Bank has helped CFP to maintain a good balance between lending for development, with the inherent risks, and maintain- ing a sound portfolio and keeping expenses at reasonable levels. 6.06 Much of the information contained in this completion report was available during the appraisal of the second SSI project (Loan 1451-CO--see Annex 4, Report No. 1512b-CO), and was taken into account in designing that project. The favorable experience of the first project and CFP lending projec- tions justified a substantially larger second loan of US$15 million. The 1/ An SMI survey in Colombia being conducted by the Bank's Development Economics Department suggests that most new enterprises need little initial investment and that institutional credit is frequently not necessary to start a firm. Therefore, CFP's limited ability to serve new enterprises should not pose a significant constraint on SMI develop- ment. - 43 - subproject and subborrower characteristics were considered satisfactory, so few changes were made in onlending terms; however, the slow utilization of technical assistance funds led to an initial allocation of US$100,000 as com- pared with US$500,000 for the first project. In order to promote more indus- trial decentralization and help CFP's profitability, margins were increased from 5.5% to 7% on subloans outside the Bogota, Medellin, and Cali areas. CFP's internal systems and procedures were generally adequate, but operations had grown to such a level that the Bank allocated funds to finance the computer- izing of CFP's information systems. The second loan also required that the government diversify CFP's sources of funds from 1977-80, but through 1978 little had been accomplished towards achieving this goal. The Bank continued supporting the idea of a national SMI council and a revision of SMI policies, but elections and the change of government during the second half of last year have not been conducive to any significant achievements. LCPI2 January 29, 1979 -44 - ANNEX 1 COLOMBIA PROJECT COMPLETION REPORT CORPORACION FINANCIERA POPULAR Loan 1071-CO Manufacturing Performance Indicators % growth p.a. 1960 1968 1974 1975 1976 60/68 68/74 74/76 Manufacturing Value Added (1970 Col$ billions) 11.7 18.1 29.7 30.8 32.7 5.5 8.6 4.9 1/ Manufacturing Investment (Index 1970=100) 44.0 59.3 137.0 1.14.0 122.0 4.1 14.7 -6.1 Manufacturing Exports (US$ millions) 0.3 39.8 388.5 303.3 392.1 84.2 46.2 0.4 GDP at Factor Cost (1970 Col$ billions) 71.9 105.8 156.7 164.4 171.2 4.9 6.8 4.5 Total Exports 382.0 508.0 1415.6 1442.6 1773.7 (US$ millions) Manufacturing Value Added as % of GDP 16.3 17.1 19.0 18.7 19.1 Manufacturing Exports/Total Merchandise Exports 0.1 7.8 27.4 21.0 22.1 1/ This index of industrial investment is based on surveys of 800 industrial companies carried out by FEDESARROLLO. It reflects mainly the level of investment in medium and large companies rather than in small ones. - 45 - AINNEX 2 COLOMBIA PROJECT COMPLETION REPORT CORPORACION FINANCIERA POPULAR Loan 1071-CO Projected and Actual Cumulative Disbursements - Loan 1071-CO (in US$ 000's) Percentage of ,ercenta?e of IBRD Fiscal Year and Quarter Forecast Total Loan Actual Total Lean 1974/75 June 30, 1975 500 9 - 1975/76 September 30, 1975 1,000 18 - - December 31, 1975 1,500 27 616 11 March 31, 1976 2,100 38 903 16 June 30. 1976 2,800 51 1,600 29 1976!77 September 30, 1976 3,400 62 2,478 45 December 31, 1976 1/ 4,000 73 3,285 60 March 31, 1977 4,500 82 4,107 75 Jure 30, 1977 4,900 89 4,239 1977178 September 30, 1977 5,300 96 4,711 rr December 31, 1977 2/ 5,500 100 4,984 91 March 31, 1978 - - 5,074 ,une 30, 1978 - - 3,74 1978/79 3/ September 30, 1978 - 5,300 9 December 31, 1978 - 5,500 1/ Original terminal date for submission of subprojects. 2/ Origi-Al closing date. 3! Froiect,id. (The closing date has bean postponed for a second time November 30, 1978.) -46 - ANNEX 3 Page 1 COLOMBIA PROJECT COMPLETION REPORT CORPORACION FINANCIERA POPULAR Loan 1071-CO Analysis of Subloan Approvals (in Col$ 000's) By Size of Approvals No. % Amount % Up to 200.0 56 18.4 7,509 3.5 200.1 to 300.0 33 10.8 8,984 4.2 300.1 to 500.0 84 27.5 .33,459 15.9 500.1 to 700.0 36 11.8 22,155 10.4 700.1 to 1,000.0 42 13.8 37,475 17.7 1,000.1 to 1,500.00 34 11.1 46,648 22.0 1,500.1 to 2,000.0 8 2.6 16,000 7.5 2,000.1 to 3,000.0 6 2.0 14,900 7.0 Above 3,000.0 6 2.0 25,000 11.8 TOTAL 305 100.0 212,128 100.0 By Duration Up to 4 years 122 40.0 44,882 21.2 4 to 5 years 103 33.8 100,830 47.6 5 to 6 years 48 15.7 43,114 20.3 6 to 7 years 25 8.2 18,944 8.9 7 to 8 years 5 1.6 2,755 1.3 8 to 9 years 1 0.3 303 0.1 Above 9 years 1 0.3 1,300 0.6 TOTAL 305 100.0 212,128 100.0 By Size of Enterprise (total assets) Up to 200.0 1 .3 100 - 200.1 to 600.0 31 10.2 7,815 3.7 600.1 to 1,000.0 30 9.8 10,323 4.9 1,000.1 to 2,000.0 76 24.9 33,193 15.6 2,000 to 3,000.0 51 16.7 31,968 15.1 3,000.1 to 4,000.0 33 10.8 23,083 10.9 4,000.1 to 5,000.0 22 7.2 21,514 10.1 5,000.1 to 10,000.0 34 11.1 31,177 14.7 More than 10,000.0 27 8.8 52,955 25.0 TOTAL 305 100.0 212,128 100.0 4 - ANNEX 3 Page 2 By nature of enterprise No. % Amount % New enterprise 12 3.9 10,852 5.1 Existing enterprise 293 96.1 201,276 94.9 TOTAL 305 100.0 212,128 100.0 By Location Cundinamarca 102 33.4 81,512 38.3 Antioquia 58 19.0 34,056 16.1 Valle 38 12.5 20,526 9.7 Atlantico 5 1.6 4,400 2.1 Santander 16 5.2 12,340 5.8 Tolima 10 3.3 4,360 2.1 Risaralda 18 5.9 12,214 5.8 Others 58 19.0 42,710 20.1 TOTAL 305 100.0 212,128 100.0 By End Use Machinery & Equipment 186 61.0 134,170 63.2 Construction 99 32.5 74,565 35.2 Technical Assistance 20 6.5 3,393 1.6 TOTAL 305 100.0 212,128 100.0 By Origin of Goods Purchased Imported 145 47.5 79,688 37.6 Locally Manufactured 160 52.5 132,440 62.4 TOTAL 35 100.0 212,128 100.0 By Type of Industrial Activity Apparel and Footwear 44 14.4 18,797 8.9 Metal Products 25 8.2 15,416 7.3 Non-metallic Minerals 17 5.6 12,840 6.1 Food Products 37 12.1 40,160 18.9 Furniture 7 2.3 3,710 1.7 Chemicals 21 6.9 23,686 11.2 Textiles 18 5.9 11,370 5.3 Diverse Manufacturing 4 1.3 1,846 0.9 Basic Metal Industries 6 2.0 5,641 2.7 Leather Products 12 3.9 8,092 3.8 Others 114 37.4 70,S70 33.2 TOTAL 305 100.0 212,128 100.0 Amount approved per Job Created 139.6 Average Approvals 695.8 COLOMBIA PROJECT COMPLETION REPORT CORPORACION FINANCIERA POPULAR Loan 1071-Co Evolution of the Corporacion Financier. Ponular'a Stgff at Headquarters and in the Regional Offices during 1976 through June 1978 Staff Evolution During the Period Status Transfer] Status at Beg. of Year Entry Resignation Promotion at End of Year Professional Other Professional Other Professional Other Professional Professional Other 1976 Headquarters 50 85 16 37 2 20 - - 64 102 Regional Offices 57 69 51 52 19 22 - - 89 99 Total 107 154 67 89 21 42 - - 153 201 1977 Headquarters 64 102 9 11 21 10 2 1 55 103 Regional Offices 89 99 22 18 36 9 (2) 1 74 108 Total 153 201 31 29 57 19 0 2 129 211; 1978 (Jan.-June) Headquarters 55 103 6 4 17 9 3 2 49 98 Regional Offices 74 108 15 11 21 5 (3) 4 69 114 Total 129 211 21 15 38 14 0 6 118 212 COLOMBIA PROJECT COMPLETION REPORT CORPORACION FINANCIERA POPULAR Loan 1071-CO ProJected and Actual Loan Approvals and Dibursemente for the Period 1974 - June 1978 (in Col$ millions) 1974 1975 1976 1977 197R 6/78 Fiscal Year Forecast Actual horecast Actual Forecast Actual Forecast Actual Forecast cu 1u (throug1 juze Approvals Loans 397,0 331.0 520.0 466.5 648.0 808.0 780.0 971.4 898.0 401.8 Equity 3.6 1.0 6.4 3.5 10.0 - 15.0 - 15.0 - TOTAL 400.6 332.0 526.4 470.0 658.0 808.0 795.0 971.4 913.0 401.8 ... . . .. Disbursements Loans 331.0 286.0 434.0 385.2 540.0 600.4 650.0 866.4 749.0 277.6 Equity 3.6 1.0 6.4 3.5 10.0 - 15.0 - 15.0 - TOTAL 334.6 287.0 440.4 388.7 550.0 600.4 665.0 866.4 764.0 277.6 z COLOMBIA PxujECT WunfTLO REPORT CORPORACTON FINANCIERA POPULAR Loan 1071-CO Projected and Actual Balance Sheets as of December 31, for the Period 1974-78 (in Col$ 000's) 1974 1975 1976 1977 1978 6178 7orecast Actual Forecast Actual Forecast Actual Forecast Actual Forecast Actual ASSETS Cash and Banks 17,547872 1241 2 202 291 60,734 33,310 20042 Portfolio: Loans 507,924 466,461 688,924 591,935 899,924 855,078 1,169,924 1,125,305 1,449,924 1,190,450 Equity Investments and Others- 8,500 9,2 _14900 _10,41 290 _5,514 39.900 97,852 54,900 185,490 Total Portfolio 516,424 475,666 703,824 602,345 924,824 880,592 1,209,824 1,123,157 1,504,824 1,375,940 Less: Provisions for: Losses on Doubtful Loans (18,619) (15,231) (24,119) (22,941) (31,919) (30,795) (40,719) (41,925) (53,119) ( 45,339) Losses on Doubtful Equity (1,097) (597) (2,097) (597) (3,597) ( 597) (5,297) ( 597) ( 6,997) ( 597) Investments - - - Taxes - (32110) - (8019) - - Net Portfolio 496,708 456,728 677,608 570,788 899,308 849,200 1,163,808 1,180,635 1,444,708 1,330,004 Fixed Assets (net) 4,524 3,670 5,924 4,678 7,124 13,679 9,124 44,083 10,624 44,615 Other Assets 31,037 29.997 44,427 38 666 49,127 125,255 73,127 76,346 89,127 73,826 TOTAL ASSETS 549,816 .499,382 749,467 626,544 981,718 1,008,755 1,275,218 1.361,798 IA577 769 1.468 487 .. . a l===== = - - - - = = -... ...b. .....A _ LIABILITIES AND NETWORTH Borrowings in Domestic Currency 318,416 265,505 384,067 359,108 480,718 499,797 619,318 708,933 753.969 787,833 Borrowings in Foreign Currency - - _ke,500 20,047 181,300 1422125 267.300 27,917 354,800 286,484 Total Borrowings 318,416 265,505 . 452,567 379,155 662,018 641,922 886,618 988.850 1 108,769 1,074,317 Other Payables 18,000 20,345 21,600 24,662 29,500 38,65Z. 652 38.500 4 3.* 49,600 50,069 Total Liabilities 336,46 285,850 167 403,817 691,518 683574 925,118 1.029A82 t,158,369 1,124,386 2/ 3/ Paid-in Capital 200,000 199,998 250,000 200,000 250,000 300,000 290,000 300,001- 330,000 310,848 Profit of the year 5,000 1,044 11,900 8,352 14,900 2,205 19,900 5,330 29,300 5,590 Retained Earnings 12490 124 1 2 2 40,200 626*98 , 100 27,663 Total Networth 213,400 213,532 275,300 2 9 328,181 350100 332,316 " 419,400 344,101 TOTAL LIABILITIES AND NETWORTH 54 6 49 32 74467 626 544 9878 0 221a755 i4 & 2A 281577769 -I4685487 1/ Tncludes short-term deposits. 2/ Tncliides ColS 65.8 million subscribed but not vet paid in. 3/ Includes Col$ 15.8 million subscribed but not yet paid in. COLOMBIA PROJECT COMPLETION REPORT CORPORACION FINANCIERA POPULAR Loan 10)1-cu Projected and Actual Income St t- ecember 31, for the Period 1974-78 (in Col$ 000's) 1974 1975 1976 1977 1978 6/78 Forecast Actual Forecast Actual Forecast Actual Forecast Actual Forecast Actual INCOME Income from Loans 72,840 74,419 114,114 104,467 163,646 140,719 230,700 202,415 301,200 123,355 Other Income 1,900 3,949 2,000 1,707 3000 6,277 4,200 10_047 5,400 10,164 Total Income 74,740 78,368 116,114 106,174 166,646 146,996 234,900 212,462 306,600 133,519 EXPENSES Financial Expenses 34,540 37,588 52,298 41,954 84,964 67,413 131,631 107,749 177,590 75,392 Administrative Expenses 29,100 30,463 40,816 42,069 52,952 64,813 _5,869 85,490 78,110 43,852 Total Expenses Before Provisions n and Depreciation 63,640 68,051 93,114 84,023 137,916 132,226 197,500 193.239 255,700 119i244 Provisions for Doubtful Loans 5,000 5,719 9,500 7.710 11,800 9,702 14,800 11,467 18,400 5,666 Provisions for Equity Investments 500 - 1,000 - 1,500 - 1,700 - 1,700 - Depreciation 600 444 600 583 800 819 1.000 1,923 1,500 1.154 Total Non-Cash Expenses 6,100 6.163 11,100 8,293 14,100 10,521 17,500 13,390 21,600 6.820 Total Expenses 69.74 7 104,214 92,316 152.016 142,747 215,000 206,629 277,300 125,794 Taxes - 3,110 - 5,506 - 2,042 - 503 - 2,135 Net Profit 5,000 1,044 11,900 8,352 14,630 2,207 19,900 5,330 29,300 5j590 x -a COLOMBIA PROJECT COMPLETION REPORT CORPORACION FINANCIERA POPULAR Loan 1071-CO Projected and Actual Financial Ratios for the Period 1974-78 1974 1975 1976 1977 1978 6/78 Forecast Actual Forecast Actual Forecast Actual Forecast Actjual Ferecast Actual Total Income as % of Average Total Assets 15.4 17.1 17.9 18.9 19.2 18.0 20.8 17.9 21.4 9.4 Financial Expenses as Z of Average Total Assets 7.1 8.2 8.1 7.5 9.8 8.2 11.7 9.1 12.4 5.3 Gross Profit as % of Average Total Assets 8.3 8.9 9.8 11.4 9.4 9.8 9.1 8.8 9.0 4.1 Administrative Expenses as % of Average Total Assets 6.0 6.6 6.3 7.4 6.1 7.8 5.8 7.2 5.4 3.1 Profit Before Provisions 6 Depreciation as % of Average Total Assets 2.3 2.3 3.5 4.0 3.3 1.9 3.3 1.6 3.6 1.0 Provisions & Depreciation as Z of Average Total Assets 1.3 1.3 1.7 1.5 1.6 1.3 1.5 1.1 1.5 0.5 Net Income Before Taxes as % of Average Total Assets 1.0 1.0 1.8 2.5 1.7 0.6 1.8 0.5 2.1 0.5 Set Income Before Taxes as % of Average Equity 3.2 2.7 4.9 6.3 5.2 1.5 6.2 1.8 7.6 2.3 Set Income Before Provisions & Depreciation as % of Average Equity 7.1 6.6 9.4 10.0 10.3 5.3 11.7 5.8 13.2 4.2 Net Income After Taxes as % of Average Equity 3.2 0.7 4.9 3.8 5.2 0.8 6.2 1.6 7.6 1.6 Net Income After Taxes as % of Average Total Assets 1.0 0.2 1.8 1.5 1.7 0.3 1.8 0.4 2.0 0.4 Interest & Commission as % of Average Portfolio 17.8 17.6 19.1 19.5 20.6 19.0 22.3 20.2 23.1 9.5 Financial Expenses as X of Average Borrowings 12.4 13.1 13.6 13.0 15.2 13.2 17.0 13.2 17.8 7.3 Cross Spread 5.4 4.5 - 5.5 6.5 5.4 5.8 2.6 7.0 5.3 2.2 Total Debt/Equity 1.6 1.3 1.8 1.8 2.4 2.1 2.6 3.1 3.7 3.3 o Percentage Increase in Total Assets 31.2 19.8 36.3 25.5 31.0 61.1 29.9 35.0 23.7 7.8 Provisions as % of Loan Portfolio 2.6 3.3 3.7 3.9 3.5 3.6 3.5 3.7 3.5 3.8 MameC CMUTLW Ef01T (CCoN F15lew A 9?I Len 1071-0D Asan of Arr.ar •EolutIon beteen ectr 31. 1973 and Jue 30. 1978 bc~mber 31. 1973 December 31. 1971 December 31. 1976 Dcember 31. 197 Jun 30. 1978 Azing of Arreare Totl1 Portfolo Affected _ Totaj lortfolio Afiectöd % Total PortliOAffc % Tota r Afecte Ljnagl fri Attfe i Cliente Aun i clienta i y el1e21t Amount 1 ; LzlntA p1 f 3 1-30 days ) ) 6.4 25.8 4.4 4.3 20.5 2.4 3.5 26.0 2.4 3.9 35.4 3.0 31-60 " 11.5) 21.7) 3.9 2.7 12.6 1.1 1.6 10.2 1.2 1.6 13.8 1.2 1.1 10.3 0.9 61-90 " ) ) 1.6 4.6 0.8 1.1 1.3 0.4 1.0 9.4 0.9 1.0 9.0 0.8 91-l80 3.7 3,8 1.6 2.7 10.9 1.9 1.7 13.4 1.6 1.7 9.0 0.4 1.4 11.3 1.7 181-350 " 3.5 8 5 2.3 .L6 _3.7 4.0 ..i.1 IL.! LI JJ 19l -L.2 .._.j .2a il,.! Subtotal 18.7 36. _j 100 a. 77. IL nj a wL l 10:0 68.4 id ALs ..UJ. w Li oer 360 days 8.2 23.5 6.4 11,5 407 j-LI Lä m.1 I 13:3 69.9 i IL> 1L1 I us las Total au -mu lu U,.4 121.7 j., uj Iw au aj ju Principal La Ar~a 24.0 6.> .51.1 C.5 59.9 1.1 69.9 6.4 76.3 6. 1/ Arers fr primejma ånly, 1J ås percenage of total portfolto COLOMBIA PROJECT COMPLETION REPORT CQ,PORACION FINANCIERA POPULAR Comparison of Projected and Actual Performance of Subborrowers Total Sales Total Profits Total Assets Employment Project Cost (Col$ millions) (Col$ millions) (Col$ millions) (Col$ millions) No. of Cities Firms Projected Actual Projected Actual Before After Projected Actual Projected Actual Call 3 34.4 22.7 3.0 2.3 20.1 24.1 294 296 11.8 8.6 Armenia 6 65.6 48.3 9.0 9.8 42.3 55.3 235 253 30.5 36.8 Cucuta 4 7.7 8.2 1.5 2.1 5.8 13.0 163 79 6.8 10.4 Tunja 9 38.3 26.0 2.7 1.3 22.5 43.0 250 206 18.3;2/ 21.3 2/ Pereira 15 145.3 91.5 21.0 9.1 70.3 113.6 604 721 15.1 1/ 23.0 1/ Bucaramanga 7 25.3 23.4 4.2 3.0 17.3 36.4 157 164 10.1 2/ 12.7 2/ Ibague 7 32.9 23.9 5.7 3.6 19.4 27.5 121 106 14.0 15.1 Neiva 1 1.0 .6 .2 .1 1.0 1.1 9 6 1.2 1.1 Medellin 17 192.4 300.1 22.4 1/ 24.0 1/ 101.9 307.4 727 934 44.9 3/ 40.5 3/ Bogota 7 107.2 111.4 8.4 10.8 28.3 78.3 241 288 16.8 28.1 TOTAL 76 650.1 656.1 78.1 66.1 328.9 699.7 2,801 3,053 169.5 197.6 1/ Excludes non-responses for four firms. 2/ Excludes non-response for one firm. 3/ Excludes non-response for thirteen firms. - 55 - ANNEX 11 COLOMBIA PROJECT COMPLETION REPORT CORPORACION FINANCIERA POPULAR Loan 1071-CO Comparison of Accuracy of Projections of CFP Subproject Appraisals (Number of Cases) Sales Profits Employment Project Costs Projected Lower than Actual 29 28 38 42 Projected Higher than Actual 47 44 30 14 Projected Equal to Actual - - 8 1. TOTAL 76 72J Projected 20% Below Actual 22 18 17 26 Projected 20% Above Actual 39 45 16 11 Projected Within 20% of Actual 15 9 43 20 TOTAL 1/ 76 72 76 57 1/ Totals not all equal due to non-responses.