Colombia - Power Development Finance Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 13221 PERFORMANCE AUDIT REPORT COLOMBIA POWER DEVELOPMENT FINANCE PROJECT (LOAN 2401-CO) JUNE 27, 1994 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. EXCHANGE RATES Colombian Peso (Col$)/US$ Appraisal Estimate (2/1983) $1.00 = 92.53 Col$ Actual: Average 1983 = 78.85 Average 1984 = 100.82 Average 1985 = 142.31 Average 1986 = 194.26 Average 1987 = 242.61 Average 1988 = 299.26 Average 1989 = 382.57 Average 1990 = 502.26 ACRONYMS AND ABBREVIATIONS BR - Banco de la Republica (Central Bank) CHEC - Central Hidroelectrica de Caldas Limitada CORELCA - Corporacion Electrica de la Costa Atlantica CVC - Corporation Autonoma Regional del Valle del Rio auca DFC - Development Finance Company DNP - National Planning Department EEEB - Empresa de Energia Electrica de Bogota EMCALI - Empresas Municipales de Cali EPM - Empresas Publicas de Medellin FEN - Financiera Electrica Nacional Government - Government of Colombia ICEL - Instituto Colombiano de Energia Electrica ISA - Interconexion Electrica S.A. JNT - National Tariff Board PAR - Performance Audit Report PCR - Project Completion Report SAR - Staff Appraisal Report US$ - US Dollar FISCAL YEAR Government: January 1 - December 31 FCR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Office of Director-General Operations Evaluation June 27, 1994 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Performance Audit Report on Colombia Power Development Finance Project (Loan 2401-CO) Attached is the "Performance Audit Report on Colombia - Power Development Finance Project (Loan 2401-CO)" prepared by OED. The Audit covers the first loan made to the newly established development finance institution for the power sector, FEN, for US$170 million equivalent. The Bank also contributed US$30 million to one of the first B-loans made under the new co-financing facility. The audited project helped FEN close the financing gap of the sector over the years 1984-85. It also supported the strengthening of FEN, mainly regarding its efforts to open the domestic capital market to financing the medium and long-term needs of a power sector heavily investing in hydro dams. The results were disappointing, partly due to unrealistic assumptions at appraisal. By 1990 FEN's market share in the domestic capital market was still below 1% (PAR, para. 21). On the other hand, disbursements were rapid and project implementation was problem free. FEN was able to grow rapidly while being profitable. Larger borrowing to follow demand made FEN to default on the debt- equity covenant, but it should have been updated in due time (PAR, para. 14). The project is rated as satisfactory, its institutional development as substantial, and the sustainability of its benefits as likely. While not being the panacea to the huge financial problems plaguing an over-investing and over-borrowing power sector, FEN was a timely and appropriate response to channel all foreign currency funds through one entity. The government indicated its satisfaction about the benefits of the approach by broadening FEN's mandate to cater to the entire energy sector in 1989. Robert Picciotto by H. Eberhard K8pp Attachment 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.  PERFORMANCE AUDIT REPORT FOR OFFICIAL USE ONLY COLOMBIA POWER DEVELOPMENT FINANCE PROJECT (LOAN 2401-CO) TABLE OF CONTENTS Pa No. PREFACE .................................................. BASIC DATA SHEETS ..........................................u EVALUATION SUMMARY ........................................ iv I. BACKGROUND............................................. II. PROJECT OBJECTIVES AND IMPLEMENTATION ............... 1 Project Implementation ...................................... 3 Retroactive Financing ....................................... 3 Disbursements ............................................. 3 Project Costs .............................................. 3 Procurement .............................................. 4 IHl. INSTITUTIONAL BUILDING AND FINANCES ................... 4 Financial Covenants ........................................ 4 Maturity of Bank Loan, Grace Period and Exchange Risk ............ 5 Exchange Risk ......................................... 5 FEN's Domestic Resource Mobilization .......................... 6 Accounting and External Audit ................................ 6 Institutional Development .................................... 6 Environment .............................................. 7 Sustainability .............................................. 7 Outlook .................................................. 7 Directives ................................................ 7 IV. CONCLUSIONS AND LESSONS................................. 7 Project Rating ............................................. 7 FEN Performance ......................................... 7 Bank Performance ......................................... 8 This report was prepared by Jean-Franqois Landeau, Task Manager, who audited the Project in December 1993. Mrs. Maryvonne Mauprivez provided administrative assistance. 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.  PERFORMANCE AUDIT REPORT COLOMBIA POWER DEVELOPMENT FINANCE PROJECT (LOAN 2401-CO) PREFACE 1. This is a Performance Audit Report (PAR) on the first World Bank lending to the power sector of Colombia through the sector's development finance company, Financiera Electrica Nacional (FEN). This was also the fourth co-financing operation approved by the Bank. The project financing consisted of a "A" loan of US$170.0 million equivalent and of a contribution of US$28.2 million to a "B" loan of US$200.0 million equivalent. This dual assistance aimed at closing the financing gap in the sector's large on-going investment program for 1984-85. The A loan was approved on March 29, 1984. The loan was entirely used after a one-year extension of the closing date. Disbursements under the B loan were closed in July 1988. 2. The PAR is based on the Project Completion Report (PCR) prepared by the Latin America and the Caribbean regional department and issued on October 23, 1992, the Staff Appraisal Report, the loan documents, and a study of the project files. An OED mission visited Colombia in December 1993. The excellent cooperation and valuable assistance provided by FEN and by two power companies in Medellin in the preparation of this report is gratefully acknowledged. 3. The PCR provides a comprehensive account of the project experience. The PAR elaborates on selected aspects of project implementation, in particular the resource mobilization effort of FEN, its financing of the power sector, and its role since project completion. 4. Following standard OED procedures, copies of the draft PAR were sent to the Government and to the Borrower for comments. The Borrower and the Government had no comments.  ii PERFORMANCE AUDIT REPORT COLOMBIA POWER DEVELOPMENT FINANCE PROJECT (LOAN 2401-CO) BASIC DATA SHEET KEY PROJECT DATA Item Appraisal Actual Actual as % Expectations Estimates of Appraisal (US$ million) Estimates Total Project Cost 1,602.0 n.a. n.a. Loan Amount (A loan) 170.0 170.0 same Physical Components Completed 12/85 12/87 Economic Rate of Return (%) n.a. n.a. CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS (US$ million) As at December 31 FY84 FY85 FY86 (1) Appraisal Estimate 100.0 70.0 0 (2) Actual 80.5 78.5 11.0 (3) (2) as % of (1) 80.5 112.1 n.a. Date of Final Disbursement: 12/1/1987 PROJECT DATES Original Plan Revised Actual First Mention in Files 1/12/83 Appraisal no official appraisal Post-appraisal - 3/07/83 Negotiations - - 2/10/84 Board Approval - 3/29/84 Loan Agreement Date - 6/27/84 Effectiveness Date 5/84 - 9/27/84 Loan Completion - - 4/30/87 Closing Date 12/86 12/87 12/31/87 iii STAFF INPUTS (Staff-weeks) MISSION DATA No. of Month/Year Persons Man-weeks Preparation I 6/82 n.a. n.a. Preparation H 11/82 n.a. n.a. Preparation III 5/83 3 3 Preparation IV 6/83 3 3 Post-appraisal I 3/83 8 16.6 Post-appraisal II 4/83 1 1 Post-appraisal III 10/83 1 2 Supervision I 9/84 1 1 Supervision H 3/85 1 .5 Supervision m 10/85 2 3 Supervision IV 8/86 1 3 OTHER PROJECT DATA Borrower: Financiera Electrica Nacional Executing Agency: same Follow-up Project: Power Sector Project (Loan 2889-CO) iv PERFORMANCE AUDIT REPORT COLOMBIA POWER DEVELOPMENT FINANCE PROJECT (LOAN 2401-CO) EVALUATION SUMMARY Introduction 1. At the time of appraisal, in 1982, US$170 million, respectively) covered 23.1% Colombia was implementing an ambitious of the eligible financing gap for the period program of hydro power generation (PAR para. 4). investments. The magnitude of these capital outlays was beyond the resource mobilization 4. The Project was the fourth ability of the various utilities. It called for a cofinancing operation made by the Bank and specialized financial intermediary to both the first one in Latin America. More enhance the creditworthiness of the power importantly it was the first cofinancing sector and launch new financial instruments to operation soon after the break-out of the Debt increase the borrowing ability of the sector. Crisis in Latin America which underscores the This led to the establishment of Financiera strong guarantee relief offered by the Bank to Electrica Nacional (FEN) in January 1982 as the private banking community (PAR, para. 5). an independent financial institution owned jointly by the government and six major power 5. The loan amount was increased companies. Its mandate was broad but focused several times during project preparation. The initially on channeling more resources to the first and largest increase (US$75.0 million) was power sector (PAR, paras. I and 2). designed "to send a clearer signal to capital markets" that commercial bank financing was 2. The project had dual objectives. The an acceptable risk in Colombia. Eventually first was "to assist Colombia in its effort to the Bank's contribution both direct (A-loan) provide the electricity supply required to and indirect (B-loan) was almost equal support growth and employment" (SAR, para. (US$198.2 million) to that of the commercial 4.01). This was to be achieved by providing a banks (US$200.0 million) (PA, para. 7). relative large loan and enlisting the support of foreign commercial lenders through co- 6. Formal retroactive financing back financing. The second objective was to to January 1st, 1983 was earmarked in the support the development of FEN primarily Loan Agreement. It was limited to US$29.0 through the tapping of the domestic capital million and most (US$25.0 million) went to market (PAR, para. 3). the San Carlos and Mesitas hydro projects already financed by the Bank (Loans 1582-GO, 3. The project was conceived as a time- 1978 and 1725-CO, 1979; and 1628-GO, 1979, slice of FEN's loan disbursements over 1984- respectively). Retroactive financing highlights 85 to power utilities for eligible investment the gap funding nature of the project (PAR, subprojects. The two loans provided directly para. 9). and jointly by the Bank (US$200 million and V 7. The disbursements followed reasonably 11. Other key indicators monitored well the SAR projections. This is largely due during supervision were: FEN's gross spread to the DFC nature of the project as well as increase in net worth and increase in loans the pent-up demand for funds in the power outstanding to utilities. The margin has sector. Forty days after effectiveness, the decreased to below 2% which is on the low Bank had committed 95% of the A-loan and side for a specialized intermediary. On the 100% of the B-loan. The disbursement other hand, this trend contradicts some of the performance is noteworthy given that FEN critics that FEN's intermediation was more started operations only in 1983 (PAR, para.10) expensive (PAR, para. 15). 8. Given the DFC nature of the project, 12. The Bank Loan was granted directly there was no cost overruns. Their financing to FEN for 16 years and 7 months of maturity, remains the responsibility of the sub- including 4 years and 1 month of grace. FEN borrowers. To this extent it is a limitation of took the exchange risk, but it passed it on the DFC approach which may ignore entirely to its subborrowers. FEN subloans altogether the issue of cost overruns even if had maximum maturity of 15 years including 4 some are legitimate business uncertainties and years of grace in 1988. FEN's Board approved not due to mismanagement (e.g., geological the extension of the grace period to alleviate uncertainties) (PAR, para. 11). the debt burden of utilities. But the Bank correctly reminded FEN that it should have 9. In less than ten years, FEN has consulted prior to the decision (PAR, para. developed as an effective financial 16). intermediary of the power sector, able to mobilize funds abroad on a large scale, and to 13. The exchange risk due to the effect transformation of maturities to fit the currency-pooling system accumulated until medium to long-term needs of the sector, and February 28, 1994 was substantial. The fact yet be profitable and financially healthy that utilities have no revenues in foreign without government subsidies (PAR para. 13). currencies to hedge against the exchange risk has contributed to the negative perception 10. Several covenants were included in the attached to the cost of these funds (27.5% in Loan Agreement to protect FEN's financial 1993 for one utility) (PAR, para. 19). viability. FEN was required to maintain a liquidity buffer of at least two months of 14. The PCR is critical of FEN for disbursements and six months of debt having done little to open the domestic capital repayments and a debt-equity ratio no greater market to financial instruments dedicated to than 7:1. The later covenant was met until the power sector. The PCR does not give 1987, but the rapid growth of FEN's portfolio sufficient value to the SAR caveat that it obliged it to borrow beyond the original would take time to attract funds of adequate multiplier. In 1986 FEN proposed to change maturities to meet the power sector's needs. the covenant to 12:1, but the Bank responded Any realistic assessment would have concluded to the request only three years later and by that it would take substantial time and granting only a "temporary exception". Given pressure to change the behavior of a market the huge funding and refinancing requirements geared to short-term investments. The of the sector, it could have been expected that covenanted targets were not met but should the ceiling of 7:1 would become a constraint not have been covenanted in the first place. (PAR, para. 14). FEN is correctly disputing the PCR criticism as unfair given the context of the mid-1980s. Opening the Colombian financial market to medium- vi and long-term papers from the notoriously mitigated by better procedures and greater capital-hungry power sector was difficult. That experience. On the other hand, it is the FEN did it is an achievement in itself and nature of a financial intermediary of the type more important, in balance, than the relative of FEN to add a screen between the actively targets realized (PAR, paras. 21 and 23). borrowing utilities and their traditional foreign financiers. The better performer among the 15. By the nature of being subloans Colombian utilities are indeed penalized by recovered from the borrowing utilities, the having to go through FEN for credits. This direct project benefits are likely to be Audit believes that, on balance, the sustained. At a second degree, as the subloans strengthening of FEN was beneficial to the are financing on-going hydro investments sector despite some increase in administrative which had been initially financed directly cumbersomeness (PAR para. 30). between the Bank and the concerned utilities, these benefits are also likely to be sustained 19. The Bank assistance was timely in (PA9 para. 25). FEN's role could be that it provided directly much needed funding expanded in the future (PAA para. 26). to continue with the massive program of capital investments undertaken in the late 16. The Audit rates the overall 1970s. It also provided a timely catalyst for performance of the project as satisfactory and commercial fundings (especially soon after the its institutional development as substantial. Debt Crisis) by contributing to one of the first The sustainability of the benefits of the various B-loan in its new cofinancing program. A subloans is likely (PAR, para. 28). contrario, it also illustrates the drawback of leaving large financing gaps unfunded in 17. This Audit does not agree with the investment programs of the magnitude PCR criticism that FEN did not fulfill implemented in Colombia. Funding for the expectations by failing to mobilize local entire program, including likely cost overruns, resources dedicated to the power sector. The should be identified ex ante, not be subjected original focus was on providing additional to reached efforts (PA, para. 31). funding to the sector without appearing to finance cost overruns. It was more important 20. The Audit gave the opportunity to for FEN to mobilize its counterpart funds in assess the Bank ability to implement DFC the amount needed given the urgency in the projects in non-industrial sectors. The sector than the composition of this funding. magnitude of capital investments on Colombia The institutional building component of justified setting up a specialized financial resources diversification, especially locally, was intermediary to cater to the sector's needs. added later on in the project preparation The Bank lacked realism in formulating the process. While legitimate, it did not have the targets for FEN to mobilize resources on the same degree of urgency as the financing gap domestic capital market given its preference funding. Furthermore, given both the untested for short-term financial instruments. This was nature of tapping the domestic market for partly due to the absence from the Bank team power sector financing and the lack of of a DFC specialist who would have brought experience of FEN a reasonable lag should some check-and-balance pointers to carry out have been expected in meeting the targets the project supervision. Without the banking (PAR, para. 29). viewpoint, the risk is high that insufficient attention will be given to some key 18. The complaints by some utilities about institutional and financial aspects of the the delays created by FEN screening of project (PA, para. 31). subloan withdrawals have merits and can be vii 21. While not being the panacea to the huge financial problems plaguing an over- investing and over-borrowing sector, FEN was a timely and appropriate response to channel all foreign currency funds through one entity. The government indicated its satisfaction about the benefits of the approach by broadening FEN's mandate to cater to the entire energy sector in 1989 (PAR, para. 33). PERFORMANCE AUDIT REPORT COLOMBIA POWER DEVELOPMENT FINANCE PROJECT (LOAN 2401-CO) I. BACKGROUND 1. At the time of appraisal in 1982, Colombia was focusing on hydro investments as the medium for power generation. These were capital-intensive and lumpy investments which required large funding. The Bank had direct lending relationships with several utilities until then. It was, however, felt that the financing needs called for a specialized financial intermediary to both enhance the creditworthiness of the power sector and launch new financial instruments to increase the borrowing ability of the sector. 2. Financiera Electrica Nacional (FEN) was established in January 1982 as an independent financial institution owned jointly by the government and six major power companies (ISA, EEEB, EPM, CVC, ICEL, CHEC and CORELCA). Its supervising ministry is the Ministry of Mines and Energy and financial regulation is provided by the central Bank, Banco de la Republica (BR). Its mandate was broad: - to restructure and coordinate power sector finances; - to grant medium and long-term loans; - to mobilize domestic savings; - to mobilize external resources; and - to provide brokerage services for its shareholders This provided FEN with a unique status: it was the only state enterprise in Colombia dedicated to finance its shareholders; it was the only rediscount development bank beside the central bank; and it was the only bank lending exclusively to a single sector. II. PROJECT OBJECTIVES AND IMPLEMENTATION 3. The project had dual objectives. The predominant objective was to fill the immediate financing gap of the power sector. As stated in the SAR, it was "to assist Colombia in its effort to provide the electricity supply required to support growth and employment" (SAR, para. 4.01). This was to be achieved by providing a relative large loan and enlisting the support of foreign commercial lenders through co-financing. The second objective was to support the development of FEN primarily through the tapping of the domestic capital market. 4. The project was conceived as a time-slice of FEN's loan disbursements to power utilities for eligible investment subprojects. Accordingly there were no physical component as such but subloans granted to creditworthy utilities for eligible capital expenditures projects. Two years were considered, 1984 and 1985, with total investments of US$1,133 million and US$916 million, respectively. Out of these US$2,049 million, 2 US$1,602 million were eligible for Bank co-financing, with the two loans provided directly and jointly by the Bank (US$200 million and US$170 million, respectively) covering 23.1%. 5. The project was the fourth cofinancing operation made by the Bank and the first one in Latin America. More importantly it was the first cofinancing operation soon after the break-out of the Debt Crisis in Latin America which underscores the strong guarantee relief offered by the Bank to the private banking community. 6. Given the financing gap funding nature of the project there was an effort at rapid processing. There was no formal appraisal but four preparation missions over one year (June 1982-June 1983) and three post- appraisal missions over six months (March-October 1983). It was further processed in a short time (five months from the last post-appraisal to Board approval). Loan effectiveness took another six months although there was only one condition to fulfill (PAR, para. 7). Eventually the whole process was slower than anticipated (27 months between first preparation mission and effectiveness) partly because the project scope was changed (PAR, para. 7). 7. The loan amount was increased several times during project preparation. Table 1 shows the rapid chronology. The sector's financing gap was estimated at US$920.0 million in 1983 and US$1,024.0 million in 1984, and that of FEN at US$180.0 million and US$219.0 million, respectively.' The first and largest increase (US$75.0 million) was designed "to send a clearer signal to capital markets"' that commercial bank financing was an acceptable risk in Colombia. Moreover, these changes aimed at financing a larger share of FEN's needs while staying within exposure limits. Accordingly the last request for increasing the Bank share in the B-loan from US$30.0 million to US$40.0 million was not accepted. Bank management felt that "there was a long-standing and possibly undue concentration on power in the Colombia lending program (48.6% over FY1978-1983)"'. Eventually the Bank limited its financing between the direct A-loan and the share in the B-loan to an amount (US$198.2 million) slightly lower than the commercial banks' participation in the B-loan (US$200.0 million). Table 1: Changes in the Loan Amount A Loan B Loan A + B Loans January 1983 25.0 80.0 105.0 March 1983 100.0 150.0 250.0 May 1983 90.0 150.0 240.0 August 1983 n.a. n.a. 200.0 August 1983 170.0 n.a. n.a. Final 170.0 200.0 370.0 1/ Issues Memorandum (January 12, 1983). 2/ Internal memorandum (March 1, 1983). 3/ Internal memorandum (August 15, 1983). 3 Project Implementation & Given the need to make Bank funds available rapidly to the power sector, including retroactive financing (PAR, para. 9), there was only one condition of effectiveness imposed. To protect FEN against the risk of term transformation (i.e., borrowing short-term and lending long- term), the Bank insisted that FEN be granted a short-term borrowing facility by the Central Bank alike commercial banks and other development finance institutions (DFIs). The framework existed (Resolution 72 of December 1, 1982 of the Monetary Board) but it took some time to accept FEN into it. In retrospect, making this a condition of Board presentation would have been preferable for it was a technical condition which had to be met without delay to optimize FEN's operating environment. After all this issue was deemed important enough by the Bank to make it an event of default if access to short-term borrowing from BR was changed (SAR, para. 6.01 (b)). Retroactive Financing 9. Formal retroactive financing was earmarked in the Loan Agreement. It went back to January 1st, 1983 soon after only the second preparation mission (partly because a short processing time was originally anticipated) but it was limited to US$29.0 million (Loan Agreement, Section 2.02 (d)). Most (US$25.0 million) was going to the San Carlos and Mesitas hydro projects (Loans 1582- CO, 1978 and 1725-CO, 1979; and 1628-CO, 1979, respectively). Retroactive financing represented 17.1% of the loan amount and this helps explain the good disbursement performance (PAR, para. 10). Most of all it highlights the gap funding nature of the project. Disbursements 10. The statistical introduction to this report shows that disbursements followed reasonably well the SAR projections. This is largely due to the DFC nature of the project. The pent-up demand for funds in the power sector is also a powerful explanation. Forty days after effectiveness, the Bank had committed 95% of the A-loan and 100% of the B-loans. Disbursements of the B-loan was even faster (100% of the Bank contribution by June 1986 compared to 83.7% for the A-loan). The disbursement performance is noteworthy given that FEN started operations only in 1983. Project Costs 11. Given the DFC nature of the project, there was no cost overruns to report. FEN has no automatic provision to finance cost overruns on the subprojects it approved. Cost overruns financing remains the responsibility of the sub-borrowers. To this extent it is a limitation of the DFC approach which may ignore altogether the issue of cost overruns even if some are legitimate business uncertainties and not due to mismanagement (e.g., geological uncertainties). 4/ Internal memorandum (January 25, 1984). 5/ Telex to FEN (November 7, 1984). 4 Procurement 12. Procurement was not an issue during project implementation. Given the type of capital investment involved, it was decided that International Competitive Bidding (ICB) would apply to all procurement including the Bank share of the B-loan. As far as the commercial share of the B-loan was concerned, the respective lender's procurement norm could be followed. Since the project financed on-going investments, ICB prevailed. III. INSTITUTION BUILDING AND FINANCES 13. In less than ten years, FEN has developed as an effective financial intermediary of the power sector, able to mobilize funds abroad on a large scale, and to effect transformation of maturities to fit the medium to long-term needs of the sector, and yet be profitable and financially healthy without government subsidies. Financial Covenants 14. Several covenants were included in the Loan Agreement to protect FEN's financial viability. FEN was required to maintain a liquidity buffer of at least two months of disbursements and six months of debt repayments and a debt-equity ceiling ratio no greater than 7:1 (SAR, para. 6.01 (i)). The later covenant was met until 1987, but the rapid growth of FEN's portfolio (3.5 times SAR projections by 1990) obliged it to borrow beyond the original covenant. In September 1986 FEN proposed to change the covenant to 12:1, but the Bank responded to the request only in November 1989 and yet by granting only a "temporary exception" to the debt-equity ceiling of 7:1. This was not an adequate reaction to a predictable pattern. Given the huge funding and refinancing requirements of the sector, it could have been expected that the ceiling of 7:1 would become a constraint. Although relatively high for a recent DFC, 12:1 was an acceptable ratio in view of FEN's policy to lend only to utilities which signed a "contract programme" to straighten their finances. A timely agreement on this new ratio was warranted by the overall quality of FEN's portfolio, and it would have shown a proper understanding and support of the sector's new financial intermediary. 15. Other key indicators monitored during supervision' were: FEN's gross spread to ensure a reasonable profitability out of the intermediation, increase in net worth to build an equity base enabling further borrowing, and increase in loans outstanding to utilities to assess lending exposure to the power sector. The margin has decreased substantially from the early days (14.3% in 1983) to below 2% which is on the low side for a specialized intermediary. On the other hand, this trend contradicts some of the critics that FEN's intermediation costed too much. 6/ Supervision Report (April 18, 1985). 5 Bank Loan's Maturity, Grace Period and Exchange Risk 16. The Bank Loan was granted directly to FEN for 16 years and 7 months of maturity, including 4 years and 1 month of grace. FEN took the exchange risk, but it passed it on entirely to its subborrowers (PAR, paras. 17 and 18). The subloans had maximum maturity of 15 years including 4 years of grace. FEN Board approved the extension of the grace period to alleviate the debt burden of utilities in 1988. But the Bank correctly reminded FEN that it should have consulted prior to the decision. The onlending interest rate was a blend of the rates on the A and B-loans plus 0.50 percentage point spread. 17. Average Loan Life. From FEN's viewpoint, the availability over time of borrowed funds can be measured by an indicator called the Average Loan Life (ALL). It is defined as: the ratio of (a) the sum, until maturity, of the loan balances (in US dollars) outstanding at the end of each year over (b) the loan amount net of cancellations. This indicator measures the number of years during which the entire loan proceeds stay effectively at the borrower's disposal assuming up-front disbursement. The faster the disbursements and the slower the repayments, the longer the availability of loan funds. In the case of Loan 2401-CO, it had an expected average loan life of 10.1 years at the time of appraisal, a high figure due to the anticipated rapid disbursements. The slightly slower than expected disbursements reduced it by less than 3 months to 9.8 years. The depreciation of the Colombian peso affected most the ALL. It was cut by 3 years assuming (optimistically) that the accumulated exchange risk would disappear, and by another 22 months to a low 5.3 years assuming that the exchange would neither worsen (stili optimistically) nor disappear. Although this did not affect FEN directly as it passed on the exchange risk, it reduced its effective term transformation by 48% from the viewpoint of its borrowers, thus making the domestic resource mobilization a comparatively attractive proposition (PAR, para. 21). Exchange Risk 18. The exchange risk due to the currency-pooling system accumulated until February 28, 1994 was substantial. It amounted to US$92.1 million or 128.1% of the outstanding loan amount. To a large extent it was predictable given the rapid pace of inflation in the 1980s. The exchange rate forecasts made by the Bank were, however, considerably incorrect because inflation forecasts were not realistic. Table 2 shows the discrepancy between projections and actual exchange rates when inflation is projected only linearly. Table 2: Assumptions for Exchange Rates (Pesos/US$) FY84 FY85 FY86 FY87 FY88 FY89 FY90 FY91 Proj. Inflation (%) 20 20 20 18 18 18 18 18 Proj. Rates 113.6 139.7 157.9 173.7 188.5 207.4 231.3 257.9 Actual Rates 100.8 142.3 194.3 242.6 299.2 382.6 502.3 633.1 19. Although very high, the exchange risk on the Bank loan does not affect FEN financially because the entire risk is passed on to its borrowers. The fact that they have no revenues in foreign 6 currencies to hedge against the exchange risk has contributed to the negative perception attached to these funds. Utilities correctly complained about the high cost of borrowing foreign funds. EPM of Medellin, for example, computed that over twelve months ending October 1993 the cost in pesos of its foreign borrowings for the Energy Department had been 24.06%. The subloans from FEN was more expensive 27.46% for the comparable period. 20. B-Loan. There were two B-loans, one in US dollars (#B006) for US$25.0 million, and one in other currencies (#B007) for US$3.2 million, both approved by the Board and signed in June 1984. Both had the same grace period (4 years) and the same maturity (10 years). While the first B-loan was fully disbursed, US$430,000 of the second one remained undisbursed. FEN's Domestic Resource Mobilization 21. One of the expectations for FEN was to open the domestic capital market to financial instruments dedicated to the power sector. The PCR is very critical of FEN in that regard (PCR, paras. 5.13-18). It acknowledges, but does not give sufficient value to, the SAR caveat that it would take time to attract funds of adequate maturities for the power sector's needs. Any realistic assessment would have concluded that it would take substantial time and pressure to change the behavior of a market geared to short-term investments. As a result the covenanted targets were not met and should not have been covenanted in the first place. FEN is correctly disputing the PCR criticism as unfair given the context of the mid-1980s. Indeed the SAR was more balanced in its expectation: "Additionally, the loan would further enhance FEN's image in the domestic capital market and, thus, help facilitate its access to future savings." (SAR, para. 4.02). This characterization did not imply any rushing to meet demanding quantitative targets. Nevertheless it is what was designed into the project. This Audit finds more a problem of internal consistency with the SAR (e.g., realistic in the qualitative objectives; exacting quantitative covenants) than a problem of failure by FEN. Opening the Colombian financial market to medium- and long-term papers from the notoriously capital-hungry power sector was difficult. That FEN did it is an achievement in itself and more important, in balance, than the relative targets realized. Indeed, by 1986, FEN's market share in the domestic capital market was below 1%, not a level conducive to dramatic breakthroughs. Accounting and External Audit 22. The external audit of FEN's accounts has been carried out by one of the major auditing firms in Colombia. The annual audit reports were not qualified. Institutional Development 23. Although predominantly a financing gap funding project, it envisioned a substantial evolution of FEN as a development bank. Established the year the project was appraised, FEN had to set procedures to screen subloan applications and train new staff to the new development banking business. Priority was given to the borrowing and onlending activities, and the coordination of the power sector's finances was left initially with ISA. The PCR was critical of FEN's performance as a DFC, but overlooked the necessary lag for a new institution to go from a mainly funding stage to the more involved stage as sector monitoring and coordinating APEX which requires experience and credibility. FEN reached that stage about 1990. This was after project completion, after its financing mandate had been broadened to the entire energy sector in 1989, and after the government consolidated the debts of some power companies. In other words, the desired outcome occurred but 7 not during project implementation as expected. This Audit confirmed that FEN is drawing praises for its role as overseer of the sector's finances. Environment 24. Environment was not directly a prominent issue in this project. Indirectly, however, each subproject financed under the loan had been previously covered by extensive SARs and their environmental impact assessed according to the criteria used in the early 1980s. Sustainability 25. By the nature of being subloans recovered from the borrowing utilities, the direct project benefits are likely to be sustained. At a second degree, the subloans financing on-going hydro investments which had been initially financed directly between the Bank and the concerned utilities, these benefits are also likely to be sustained. The limited rescheduling of subloans was not of a nature to jeopardize either sustainability. Outlook 26. Some maturity has gone into the power sector since the investment excesses of the 1980s. FEN is partly to credit for this evolution. A key tool has been the "program contracts" to reach and maintain creditworthiness signed with utilities as a condition for granting them loans. This has introduced a standardization in the way utilities are being appraised. Additional roles that FEN could consider in the future are the refinancing of debts whenever a utility's cash flows cannot cover both debt-servicing and self-financing of investments, and the hedging of the foreign exchange risk now that FEN deals with exporting entities in the energy sector whose foreign currencies receipts could be swapped for local receivables or debts. More financial engineering is one route for FEN to become more relevant as an APEX for the power sector. Directives 27. The Audit finds that all relevant Operations Manual Statements, Operations Policy Notes and Bank guidelines have been observed and complied with. IV. CONCLUSIONS AND LESSONS Project Rating 28. The Audit rates the overall performance of the project as satisfactory and its institutional development as substantial. The sustainability of the benefits of the various subloans is likely. FEN Performance 29. The PCR implies that FEN did not fulfill expectations by failing to mobilize local resources dedicated to the power sector. This Audit does not agree with this criticism. The project files reveal 8 that the original focus was on providing additional funding to the sector without appearing to finance cost overruns. This explains why the loan amount was increased several times (PA, para. 7) and why the project justified tapping the then innovative cofinancing window (PA, para. 5). In addition it was more important for FEN to mobilize its counterpart funds in the amount needed given the urgency in the sector than the composition of this funding. The institutional building component of resources diversification, especially locally, was added later on in the project preparation process. While perfectly legitimate, it did not have the same degree of urgency as the financing gap funding. Furthermore, given both the untested nature of tapping the domestic market for power sector financing and the lack of experience of FEN a reasonable lag should have been expected in meeting the targets. 30. The complaints by some utilities about the delays created by FEN screening of subloan withdrawals have merits and can be mitigated by better procedures and greater experience. On the other hand, it is the nature of a financial intermediary of the type of FEN to add a screen between the actively borrowing utilities and their traditional foreign financiers (mainly the Inter-American Development Bank and the World Bank). The better performer among the Colombian utilities are indeed penalized by having to go through FEN for credits. The financial woes of the sector (documented in a recent OED study)' justified the streamlining of the sector's finances through a specialized intermediary. This Audit believes that, on balance, the strengthening of FEN was beneficial to the sector despite some increase in administrative cumbersomeness. Bank Performance 31. The Bank assistance was timely in that it provided directly much needed funding to continue with the massive program of capital investments undertaken in the late 1970s. It also provided a timely catalyst for commercial fundings (especially soon after the Debt Crisis) by contributing to one of the first B-loan in its new cofinancing program. This prevented any disruption in the funding of key programs geared to provide electricity to a rapidly growing economy. A contrario, it also illustrates the drawback of leaving large financing gaps unfunded in investment programs of the magnitude implemented in Colombia. Funding for the entire program including likely cost overruns should be identified ex ante, not being subjected to scrambling efforts. 32. This project provides also the opportunity to assess the Bank ability to implement DFC projects in non-industrial sectors. The magnitude of capital investments on Colombia justified setting up a specialized financial intermediary to cater to the sector's needs. The Bank lacked realism in formulating the targets for FEN to mobilize resources on the domestic capital market given its preference for short-term financial instruments. This was partly due to the absence from the Bank team of a DFC specialist who would have brought some check-and-balance pointers to carry out the project supervision. Without the banking viewpoint, the risk is high that insufficient attention will be given to some key institutional and financial aspects of the project. 33. While not the panacea sometimes implied to the huge financial problems plaguing an over- investing and over-borrowing sector, the setting-up of FEN at the Bank's urging was a timely response to channel all foreign-currency funds which Colombia could mobilize through one entity. The respective utilities (with perhaps one exception) could not have mobilized the foreign funds in the 7/ Colombia: The Power Sector and the World Bank, 1970-1987 (June 28, 1990). 9 amounts required given their excessive indebtedness and the over-exposure fears felt by financial institutions after the outbreak of the Debt Crisis. This role as first-and-last resort financier of the power sector (such as negotiating a share of the several annual "jumbo loans" mobilized by Colombia on the international financial markets) cannot be stressed enough.