Rwanda - Fourth Banque Rwandaise de Developpement Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 12830 PROJECT COMPLETION REPORT RWANDA FOURTH RWANDESE DEVELOPMENT BANK (BRD) PROJECT (CREDIT 1650-RW) MARCH 9, 1994 Industry and Energy Operations Division South-Central and Indian Ocean Department Africa Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY AND EOUIVALENTS Currency unit = Rwandese Franc (RF) 1989 US$1 = RWF 80 1990 US$1 = RWF 82 1991 US$1 = RWF 125 1992 US$1 = RWF 125 (1st semester) 1992 US$1 = RWF 140 (2nd semester) FISCAL YEAR January 1 - December 31 WEIGHTS AND MEASURES Metric System ABBREVIATIONS AND ACRONYMS AGCD Agence Generale de Cooperation et Developpement BNR Banque Nationale du Rwanda BRD Banque Rwandaise de Developpement CER Caisse d'Epargne du Rwanda CFD Caisse Francaise de Developpement CHR Caisse Hypothecaire du Rwanda CSR Caisse Sociale du Rwanda DEG German Economic Development FMO Societe Neerlandaise pour le Developpement FSAC Financial Sector Adjustment Credit FSG Fonds Special de Garantie IMF International Monetary Fund KFW Kreditanstalt fur Wiederaufbau PFP Policy Framework Paper RF Rwandese Franc SAC Structural Adjustment Credit SAF Structural Adjustment Facility (IMF) SOE Statement of Expenditures UBP Union des Banques Populaires UNDP United Nations Development Program  FOR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A Office of Dlroctor-General Operations Evaluation March 9, 1994 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Rwanda - Fourth Rwandese Development Bank (BRD) Project (Credit 1650-RW) Attached is the Project Completion Report on Rwanda - Fourth Rwandese Development Bank (BRD) Project (Credit 1650-RW) prepared by the African Regional Office. Part II was prepared by the Borrower. The PCR is of satisfactory quality. It provides a clear and convincing account of the factors that undermined the sustainability of an otherwise well-managed, conservative and successful development finance institution. The economic decline of the late 1980s combined with the sudden liberalization of imports and the banking sector, impositions of non-viable loans by the Government on BRD, civil strife and armed conflict conspired to debilitate the financial health of this development bank. During 1992, however, the capital structure of BRD was strengthened through the enactment of a capital increase and a successful effort to collect arrears averted insolvency. Despite these actions, the financial situation and the competitiveness of the institution remain precarious unless BRD undergoes a thorough financial restructuring including a reformulation of its business strategy. The project outcome is unsatisfactory and it has had negligible institutional impact. Given the stalemate in the civil strife and its effect on the profitability of BRD, the sustainability of the project is rated as uncertain. The project is likely to form part of a cluster audit based on the result of the three similar projects implemented in Rwanda since the end of the 1970s. Attachment This document has a rcstricted distribution and may be used by recipients only in the performace of their official duties. It contents may not otherwise be disclosed without World Bank authorization.   FOR OFFICIAL USE ONLY PROJECT COMPLETION REPORT RWANDA FOURTH RWANDESE DEVELOPMENT BANK (BRD) PROJECT (Credit 1650-RW) TABLE OF CONTENTS Page No. PREFACE .............................................. i EVALUATION SUMMARY ................................... ii PART I: PROJECT REVIEW FROM THE WORLD BANK'S PERSPECTIVE ... I 1. Project Identity ................................... l 2. Background .......... ........................... 1 3. Project Objectives and Description ....................... 4 4. Adequacy of Project Preparation ........................ 5 5. Project Implementation .............................. 7 6. Project Results .................................... 9 7. Project Sustainability ................................ 11 8. Bank's Performance ................................ 12 9. Borrower's Performance ............................. 13 10. Lessons Learned .................................. 14 PART II: PROJECT REVIEW FROM THE BORROWER'S PERSPECTIVE .... 15 I. Introduction ...................................... 15 II. Assessment of the Project's Impact on the Country's Development ................................... 16 III. Comments on Part I of the PCR and responses to issues raised by the World Bank, following the Guidelines mentioned in the introduction . ............. .17 PART III: STATISTICAL INFORMATION . ......................... 22 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.   - i - PROJECT COMPLMTION REPORT RWANDA FOURTH RWANDESE DEVELOPMENT BANK (BRD) PROJECT (Cr. 1650-RW) PREFACE This Project Completion Report (PCR) covers the experience with Credit 1650-RW, the fourth of a series of lines of credit to finance enterprise development in Rwanda. Credit 1650-RW was approved by IDA in January 1986. The Credit was made to the Government; the financial intermediary channeling the proceeds to investment projects was the Banque Rwandaise de D3veloppement (BRD). This PCR was prepared by the Industry and Energy Division of the South-Central and Indian Ocean Department, Africa Region, and it is based, among others, on the Staff Appraisal Report, the Credit and Project Agreements, supervision reports, correspondence between the Bank and the Borrower, and internal Bank memoranda. Preparation of this PCR started during the Bank's final supervision mission of the project in October/November 1992. The PCR was sent to the Borrower for its comments which are incorporated as Part II.   - ii - PROJECT COMPLETION REPORT RWANDA FOURTH RWANDESE DEVELOPMENT BANK (BRD) PROJECT (Credit 1650-RW) EVALUATION SUMMARY Objectives 1. The project was envisaged as a continuation of previous successful efforts to support BRD as a viable and well managed development bank. The project provided a line of credit to finance enterprises which would contribute goods, services and employment to the economy of Rwanda. Almost half of the Credit funds were earmarked for financing small scale enterprises (SSEs) as a study conducted under the BRD III project (Cr. 1344-RW) had shown the importance of SSEs in the economic structure of the country and as a vehicle for promoting indigenous entrepreneurs. Policy measures to eliminate restrictions inhibiting SSEs' development, as a complement to the provision of financing, were also envisaged. As distortions in the foreign trade regime were in evidence at the time of appraisal, the project envisaged also strengthening BRD's capacity for economic evaluation of projects and a dialogue with the Government of Rwanda on the level and structure of import tariffs (paras. 3.1 to 3.4). Implementation Experience 2. Severe problems, that could hardly have been foreseen at the time of appraisal, seriously impaired the implementation of this project. Eirst, the economy of Rwanda which had experienced satisfactory growth in the 70's and early 80's stagnated and declined in the second half of the decade after the fall of coffee prices, the country's main export (para 2.7). The economic recession had an adverse impact on all business enterprises in Rwanda, including BRD's customers. Second, in 1987 and particularly in 1988, BRD increased lending operations and disbursements well above its own projections and financial means, falling into a serious financial squeeze. BRD, for the first time, had to resort to short term borrowing from the Central Bank and other financial institutions and private banks (para 5.1). Third, the introduction of liberalization measures by Government seriously affected BRD, initially in 1987 with the "despecialization" of the banking sector that eliminated, de facto, BRD's monopoly in term financing. In 1991, restrictions to imports were lifted, jeopardizing even more the financial viability of uneconomic import substitution enterprises, which constitute an important group of BRDs' customers (para 5.6). La, but not least, the civil strife and armed conflict which erupted in October 1990, seriously disrupted the country's import routes, logistics, tourism and economic activities in general (para 2.7).  - iii - 3. Even under severe economic distortions, BRD committed the totality of the Credit funds after the closing aate for commitments was extended from December 31, 1988 to June 30, 1990. SDR 4.9 million were committed for 11 large and medium size enterprises and SDR 2.5 million for 94 SSEs. All funds for large-medium projects have been disbursed but only SDR 1.4 million for SSEs. (para 5.3). 4. As it could be expected in a general environment so seriously degraded, project results have been well below expectations. Some of the large-medium size enterprises financed under the Credit have failed completely, those surviving face enormous difficulties to cope with adverse economic trends and with shortages and instability caused by the war. The major objective of supporting SSEs development through BRD produced mixed results. During 1987-89, BRD was able to process and approve a large number of SSEs projects but this action heavily taxed BRD's capability to properly handle its medium-large size customers portfolio. As civil strife ensued after October 1990, many SSE projects, particularly those engaged in agriculture and livestock, were seriously ravaged during the conflict. Several SSEs financed by this project did not start operations at all, for diverse reasons, and Credit funds committed for those projects had to be cancelled (para 5.5). Nevertheless, some 250 SSE projects remain active, representing 11% of BRD's portfolio. Under present circumstances this represents a good accomplishment for which BRD's management is to be commended. The Bank did not particularly focus its supervision work on the SSE program (para 6.4). 5. The policy dialogue on economic issues, like the exchange rate and trade regime, started under this project, was eventually pursued and measures were implemented under the Structural Adjustment program that the Government has begun to carry out since 1991 (para 2.7). Sustainability 6. The slowdown of BRD's operations in 1989-91, due to its financial crisis and the economic downturn, a capital increase enacted in 1992 and successful efforts in arrears collection, have allowed BRD to weather the storm. With a strong capital base, BRD has been able to make adequate provision for all failed or doubtful customers. However, BRD's sustainability in the medium term, with a much reduced performing portfolio, is only assured if drastic rationalization measures to cut administrative costs are carried out. These measures alone, however, would not guarantee its long term viability in the new, competitive, environment. The design of a new strategy to operate BRD, diversifying its services, is urgently required. Such a strategy is most likely to be successful if BRD is privatized or if it would merge with a private bank. While efforts to privatize or merge BRD need to wait for political normalization, BRD should not delay implementing its restructuring, thus preventing any risk to its financial stability (para 7.2). Lessons Learned and Action Taken 7. In intermediation projects, when Bank funds are loaned to private sector enterprises through local financial institutions, if serious distortions prevail in the economy (i.e. overvalued exchange rate, highly protective import barriers, subsidies), an enterprise benefiting from the project may obtain financial gains which would most likely not represent economic gains for the country.  - iv - The introduction in project evaluation of an economic rate of return calculation as a means for overcoming those distortions has not proven to be an adequate solution. That has been the experience in this type of project in many countries and it was repeated in the BRD IV project. Either the analysis becomes overly complex or changes in the pattern and level of distortions render the evaluation meaningless. In the absence of an effective program of economic reforms, therefore, support for intermediation projects should be delayed until distortions are corrected and market prices become a fair representation of the cost of production factors. 8. BRD IV was one of the last "DFC type" projects carried out by the Bank in Sub-Saharan Africa. In the mid-80s, as many countries in the Region engaged in a process of structural adjustment to reverse economic decline, new market conditions became adverse for DFCs as their main customers, import substitution industries, were seriously affected by liberalization of imports. As liberalization measures reached also the financial and banking sector, the single DFC approach to channel IDA funds for term lending to private entrepreneurs also became obsolete. Such objective is pursued now through lines of credits for term financing made available, usually through the Central Bank, to all eligible financial institutions. Survival of existing DFCs in the new environment is by no means assured. For that they are required not only to streamline and cut operating costs but also to diversify their services, a strategy more likely to succeed under privatization. 9. In the late 1980s, it became clear that BRD's problems could not be dealt with in isolation without addressing macroeconomic issues in general, as well as financial sector issues. To that effect, the Government embarked since 1991 on an economic reform program which was designed with the assistance of the Bank and the IMF. Furthermore, in 1991, the Bank carried out a Financial Sector Review which provided the basis to design a program of adjustment for the Financial Sector as a whole, including provisions to address BRD's more immediate financial problems (para 8.4). More recently, an IDA financed Private Sector Development Project, approved by the Board in September 1993, includes a Plan to rationalize BRD in order to cut operational costs, increase its efficiency and make it eligible for the Apex line of credit included in that project (para 8.4). Beyond these measures, a corporate strategy needs to be designed to address BRD's long term sustainability (para 6 above).   - I - PROJ.ECT COMPLETION REPORT RWANDA FOURTH RWANDESE DEVELOPMENT BANK (BRD) PROJECT (CREDIT 1650-RW) PART I: PROJECT REVIEW FROM WORLD BANK'S PERSPECTIVE 1. Project Identity Name Fourth Rwandese Development Bank (BRD) Project Credit Number 1650-RW RVP Unit Africa Region Country Rwanda Sector Industry and Finance 2. Background 2.1 Origin of the Project. The World Bank Group has been intimately linked with the Banque Rwandaise de Developpement (BRD). Created in 1967, BRD was the main source of investment financing in Rwanda until 1987. BRD is jointly owned by the Government and Public Agencies of Rwanda (56 percent) and bilateral donor agencies ' (33 percent), while local private investors hold the balance of 11 percent. 2.2 This project completion report covers the Fourth IDA Credit to BRD. Prior to this Credit, BRD had been the exclusive beneficiary of three IDA Credits since 1979; the fourth Credit (No. 1650-RW) of US$9 million was approved on January 14, 1986, and became effective October 31, 1986. The Credit is 100 percent committed, 75 percent disbursed, and will be formally closed on the original target date of December 31, 1993. Project Completion Reports (Nos.4921-RW and 8799-RW) dated February 10, 1984, and June 28, 1990, on the first IDA Credit (No.655-RW), and the Second and Third IDA Credits (Nos. 896-RW and 1344-RW) respectively, were circulated to the Board. The two PCRs concluded generally that the first two IDA Credits focused on institution building and economic development objectives, and achieved satisfactory results. However, it was also stressed that these operations failed to address issues affecting overall financial sector performance or its impact on the real economy. When the PCRs were issued Credit 1650-RW was well underway. In the last PCR, a broader sectoral approach through a financial sector adjustment operation and apex lines of credit were recommended for future financial intermediary operations. I CCCE (France), AGCD (Belgium), FMO (Holland) and DEG (Germany).  - 2 - 2.3 At the time of approval of the Fourth Credit in 1986, BRD had developed into a relatively well managed, financially sound and profitable development institution. BRD was the exclusive recipient of foreign exchange lines of credit which it channelled to long-term financing of medium and large manufacturing and agro-industrial projects in public and private sectors. Its overall economic impact was positive but limited to narrowly focused and urban based branches of activity, with limited potential for employment generation. The institution's project appraisal and lending procedures were sound and reliable, and decision-making procedures remained relatively free of Government interference. Its loan portfolio was performing well, as BRD enjoyed virtual monopoly on long-term credits which, unlike credit ceilings imposed on commercial banks, were not controlled by Banque Nationale du Rwanda (BNR), the central bank. This BNR policy enabled BRD to select judiciously its borrowers, relying mainly on technical and financial appraisal criteria. Moreover, BRD implemented a systematically conservative policy of high provisions against general and loan specific risks of losses, while reinvesting earnings to shore up its capital base. 2.4 Until the early 1980s, both BRD and commercial banks did not extend any significant credit facilities to the SSE and agricultural sectors, owing mainly to the high intermediation costs and financial risks involved, aggravated by Government-imposed low interest rates for SSEs. In 1982, after reviewing its own experience with SSE lending and following the conclusions of an SSE sector study financed by IDA under BRD III, BRD's Board approved an SSE program which included the creation of an SSE lending Division, the simplification of procedures for SSEs' applications and appraisals and the creation inside BRD of its own SSE Guarantee Fund. 2.5 Continued and sustained IDA financial, and especially institutional, support was maintained and deepened. BRD appraisal and lending policies required changes, inter alia, (i) to reorient lending towards more emphasis on SSEs and agriculture, (ii) to generalize economic rates of return screening of investment projects, (iii) to consolidate and further reinforce the real value of its equity base by curtailing dividend payments and protecting BRD against foreign exchange risks on international borrowing. Real interest rates were generally positive during the period under review, and BNR policies did not raise any particular problems. 2.6 T'he BRD IV project addressed some issues in the industrial and enterprise sector, including the need to review the import tariff structure, eliminate price controls, strengthen the Special Guarantee Fund created by the Government to promote SSEs, and revise the investment code to extend incentives in favor of SSEs. No specific conditions were, however, imposed by IDA, as the price control system was deemed to be more of an administrative ex-post registration than a constraint to manufacturers and importers. As to the tariff structure and policies, a US-AID funded study by the Ministry of Finance was already under way and IDA did not wish to interfere at that moment. It was agreed that IDA would offer its advice upon completion of that study. In the final analysis most of these issues were addressed in the context of IDA adjustment operations as reviewed below. 2.7 Economic Background. During the 1970 - 1985 period, Rwanda enjoyed a measure of political and economic stability, low inflation rates and sustained real growth averaging 4 percent per annum. However, the fourth line of Credit for BRD was approved and implemented during 1986-1990, when Rwanda's economy went through a period of declining and mixed performance and the Government was compelled to implement an economic adjustment program:  - 3 - * In 1986-1990 Rwanda was faced with precipitous declines in world market prices for coffee, Rwanda's principal foreign currency earner, and unfavorable climatic conditions which had an adverse impact on agricultural production. Real income per capita dropped by almost 16.5 percent from 1986 onwards. Initial Government policy responses to these adverse developments were inappropriate, and serious internal and external imbalances ensued, further worsening the declining per capita income through 1990. * In 1990-1992 a series of policy and institutional reform measures were conceived to correct these economic imbalances through a comprehensive economic reform program supported by: (i) an IDA financed Public Enterprise (PE) Reform Technical Assistance Credit (No. 2113-RW) in 1990; (ii) IDA's First Adjustment Credit (No.2271-RW) approved in June 1991; and (iii) a Financial Sector Adjustment Credit (FSAC) negotiated in October 1991. Although final approval of the FSAC by IDA was delayed by the civil war, and the lack of an agreed macroeconomic framework, progress has been made in implementing the reforms. The objectives were to achieve macroeconomic stability, improve competitiveness in enterprises and banks, rationalize monetary and credit policies, and stimulate economic recovery. The PE reform measures were to streamline and rationalize that sector, restructure and privatize or liquidate non viable enterprises, and curtail and eliminate their negative fiscal impact. This project was reoriented in June 1993 to give greater emphasis to privatization. * The adjustment program, by liberalizing the tariff structure and import and export regime and eliminating Quantitative restrictions (QRs) has had a substantial impact on highly protected manufacturing industries by subjecting them overnight to strong external competitive pressures at a time of civil strife and declining domestic demand. The financial situation of many of these enterprises, which are also long-term BRD borrowers, deteriorated significantly and led many of them to fall in arrears of payments with BRD and other banks since 1990. * Finally the disruptions and partial destruction of economic infrastructures by the civil war, particularly for international road transports in the North upon which Rwanda's economic livelihood depends, have had further detrimental effects on the local manufacturing industry. Lack of foreign exchange for imported inputs have compounded the problems leaving key industries operating below capacity, if at all. 2.8 The results of the adjustment program have been affected by the war and the consequent effect on economic activity, productive investments and government finances. The budget deficit widened to 15 percent of GDP in 1992 as a consequence of increased non-developmental expenditures stemming from the October 1990 breakout of civil strife and armed conflicts in the country. BRD has consequently been affected by these developments which have led to a decline in the performance of BRD loan portfolio and to an erosion of its equity base through increased loan loss provisions.  -4 - 3. Proiect Objectives and Description 3.1 The project's overall objective was to continue Bank's support to an effective financial intermediary - and the main instrument in Rwanda for developing the private sector - to help it continue providing term financing to well-conceived investment projects. More specifically the project objectives were to: (a) support BRD lending ability as the institution was moving more forcefully into the areas of SSE and agricultural lending; (b) improve BRD capability for economic evaluation of projects; (c) provide the needed foreign exchange resources for priority investments by private entrepreneurs; (d) help improve the policy framework for SSEs; and (e) assist the Government in its review of Rwanda's import tariffs system. 3.2 Under the project and for the first time, IDA specifically targeted the development of small scale enterprises (SSEs) and allocated US$4 million of the $9 million Credit to their development. Although BRD had been financing SSEs and had developed a small, highly selected SSEs loan portfolio, secured by its own guarantee fund backed by a KfW line of credit, it was concluded that BRD should further diversify its lending operations away from medium-large borrowers and spread its risk coverage over a wider sector spectrum. Moreover, SSEs were shown to account for an estimated 60 percent of industrial value-added, and 65 percent of employment generated (i.e., 26,000 workers), and it was concluded that they deserved greater support to help solve Rwanda's pressing employment problems. 3.3 To expand its SSE lending operations, BRD was implementing a two-phase program. In the first phase, which had been supported under the Third IDA project, emphasis was placed on processing a large backlog of SSE applications which were not investigated in the past and in adapting BRD loan processing procedures and organizational structure to facilitate SSE lending. Application forms, appraisal reports and loan contracts for SSE lending were thus considerably simplified and a specific Division in charge of SSE lending was created. Supervision of SSE projects was reinforced. BRD also increased its lending limit to 80 percent of total project costs for SSEs (as compared with 50-65 percent for medium-sized and larger enterprises), and established its own guarantee fund in 1982 (para 3.2 above). 3.4 Expanded BRD lending for SSEs thus became a major objective of the Fourth IDA Credit to BRD. Diversification of BRD lending was considered essential to pursue broad objectives of fairness and economic growth, providing increased opportunity to indigenous entrepreneurs. It was considered that more experience in the sector would lead to improved repayment performance through better criteria for selecting projects and promoters. Second, streamlined loan processing procedures for SSEs would hold down administrative costs. Third, an increase in BRD lending rates  would help improve its average spread and profitability. Fourth, access by BRD to a government loan guarantee scheme which would complement its own (para 3.2) would help minimize the financial impact of bad SSE loans. 4. Adequacy of Project Preparation. 4.1 A Bank mission to BRD in December of 1984 found that the Third IDA Credit (Cr. 1344-RW), which represented the bulk of BRD's resources, had already been 46% committed and full commitment was expected by the third quarter of 1985. The mission reported that the Government and BRD's management were eager to avoid a hiatus in BRD's resources and recommended a prompt appraisal of BRD IV, which took place in February of 1985. 4.2 As a repeat operation of an ongoing and increasingly successful DFC which had the monopoly for medium and long-term lending, no change to the status quo, with BRD as sole intermediary of IDA funds for private enterprise financing was envisaged. During preparation and appraisal, the main concerns were to improve the overall policy framework for SSEs and to ensure BRD's solvency. 4.3 An SSE study financed by IDA under BRD III recommended several actions to ease the constraints to SSEs' growth, namely: improvement of operation of the Special Guarantee Fund, simplification of administrative procedures for establishing new enterprises and elaboration of a special Investment Code for SSEs. These positive measures were adopted under the project and while in theory they could contribute somehow to SSEs' development, experience has shown that they can hardly be effective unless the overall policy framework is sound and is conducive to competition with equal opportunities for all economic operators. While the appraisal pointed out to distortions in the economy, like overvaluation of the exchange rate, high effective protection and import restriction, a solution to these problems was considered to be dealt with more appropriately under a Structural Adjustment Loan (SAL) operation which the Government had already requested. Unfortunately, the reform program was slow in being agreed and implemented even in the face of a serious economic decline since 1986 (para 2.7). 4.4 At appraisal, it was considered that although the overall quality of BRD's portfolio was good and adequate provisions had been set aside for probable losses from doubtful investments, BRD's collection performance had not improved adequately. Even though high by development bank standards, in 1984, only 89 percent of principal billed was collected. If the collection ratio would remain at that level, arrears were bound to increase at a time when BRD was reorienting its lending activities towards SSEs, a sector with potentially difficult collection problems. A major objective of the project was therefore to assist BRD in improving its arrears position. This objective was pursued with a covenant in the Project Agreement that BRD would ensure that the collection ratio, defined as the ratio of amounts collected to amounts fallen due in a given year, would reach 95% in 1985, 100% in 1986, 102% in 1987 and 105% in 1988 and thereafter, and that performance to achieve these ratios was to be closely monitored by the Bank. 4.5 Since the average cost of funds for BRD was expected to increase from 7 to 8 percent, the Bank proposed, and it was agreed, that BRD's lending rates would also increase to 12.5% for "productive" sectors and to 13.5% for services, thus ensuring an adequate spread of 5% to BRD. BRD notes in Part II of this Report that 7% is more adequate margin for a development bank.  - 6 - 4.6 As it was evident that local enterprises operated under a high level of effective protection, it was agreed under the project that an economic rate of return of no less than 10% was required for medium to large size subprojects to be financed under the line of Credit, thus avoiding the prospects of financing inefficient enterprises that would not stand competition under a more liberal trade regime. 4.7 The project did not address the legal and regulatory constraints faced by commercial banks and BRD in the legal recovery of delinquent loans, but instead focused on strengthening the Government controlled Special Guarantee Fund (SGF). In the Rwandese court system, loan recovery by legal means and, generally, the reinforcement of contractual obligations were systematically delayed and obstructed. These legal constraints prompted BRD, and commercial banks in particular, to impose stiff loan security requirements which most SSE borrowers could hardly satisfy. As a result, and in order to promote SSEs, IDA project preparation targeted the Government-run SGF to alleviate the guarantee problems faced by banks in financing SSEs. Issues of a guarantee fund managed by Government officials were not raised, nor were the operational and legal procedures of the fund properly and sufficiently addressed and analyzed. The SGF operational deficiencies had a larger effect on the commercial banks' behavior and their reluctance to lend to SSEs. BRD, on the other hand, had created in 1982 its own Guarantee Fund for SSEs (para 3.2), covering 80% of principal defaults. With SSEs risks adequately covered by its own Fund, BRD also justified its own view that no additional increase to the interest rate charged to SSEs was necessary for that purpose. 4.8 The need for sufficient interest margins to compensate banks and BRD for the higher administrative costs in SSE lending was not adequately addressed. Although IDA did insist on an increase in BRD's lending rates, it settled for less than required upon the Government's insistence. The only financial sector issue addressed was in the context of reviewing BRD's interest rate levels. No general approach to financial sector policies was undertaken. Likewise, the looming consequences on BRD's portfolio, when local manufacturing industries are subjected to external competitive pressures through import liberalization and tariff reduction, were neither addressed during project preparation. Institutional development and other financial issues faced by BRD were, on the other hand, well covered, and assumptions for projections reliably established. 4.9 With the forecast improvement of the collection ratio and the assurance of an adequate spread, it was considered that adequate measures had been taken to ensure BRD's profitability and stability. Moreover, BRD's management had performed quite well under the previous projects and its continuity gave all reasons to believe that BRD's prudent policies of the past would continue. BRD's Board composition, which included pro-active external donors, provided further confidence in the institution's maintaining its performance record. Finally, at the time of appraisal, while the country's economy was facing difficulties, foreign exchange reserves were still adequate, the Government had earned a good reputation among donors for its conservative but forward looking development policies. The Bank, unable to predict the steep economic decline that ensued in 1986 and thereafter, had reasons to believe that the good performance of previous BRD projects would be once more repeated.  -7- 5. Project Implementation 5.1 The evolution of BRD lending operations is shown in Annex 7a. Approvals of new loans remained at the predicted levels in 1985 and 1986, around FRW 600 million a year, but increased significantly in 1987 to FRW 900 million and reached an all-time, unusual high level of FRW 1,724 million during 1988 while BRD's operational forecast for that year was only of FRW 740 million. About a third of the FRW 1,724 million of approvals in 1988 represented not the financing of new projects but either payment, on behalf of BRD customers, of suppliers' credits guaranteed by BRD (FRW 194 million) or extension of ongoing projects (FRW 442 million). But even taking this into account, new lending by BRD in 1988 was close to FRW 1,100 million, an unusually high figure. This high level was due to strong pressure from the Government to use BRD lending to counter the effects of the economic recession. Facing adverse trends, the Government turned to BRD rather than launching a program of economic reforms that would have, in fact, improved the business climate and promoted investments. BRD's management was unable to maintain the institution's autonomy and went ahead supporting doubtful projects that it would most likely have rejected under normal circumstances. 5.2 Disbursements followed a similar pattern and, as Credit 1650-RW was temporarily suspended (see para 5.4), BRD exhausted its available resources to finance such high level of lending. To make up for the short fall, BRD borrowed short-term in the local market, and mainly from BNR, to lend long-term, thus creating a potentially serious problem (para 5.6). The consequence of these developments, aggravated by the economic recession and the civil strife, was a severe financial crisis. BRD had to contract its lending which decreased to FRW 570 millon in 1989, FRW 220 million in 1990 and an all-time low of FRW 147 million in 1991. 5.3 The 1988 peak in lending reflects the large number of loans approved by BRD in that year, 126, as compared with only 50 in 1987, an increase of 152%. Lending for SSEs, an objective pursued by the Bank and strongly reinforced that year by the Government, was responsible for the increase in terms of number of new operations, 106, although in amounts they represented only 17% of total lending. While this attention to SSEs corresponded largely to the objectives of support to indigenous small entrepreneurs and to diversification in terms of activities and location, it surely had a diminishing impact on BRD's capacity to handle its supervision and collection efforts in general and particularly of its large size customers at a time when firms in general were facing adverse economic conditions. As it could be expected in any lending institution, quality control and overall management of BRD deteriorated considerably as the volume of operations increased so dramatically in a relatively short period. The Bank went along in this behavior, approving financing for a large number of small and large subloans. A supervision mission which visited BRD in mid-1988 noted the large expansion of BRD's lending but did not find any reason to be alarmed. The only problem mentioned was the lack of funds for new commitments. The next Bank mission which visited BRD in mid-1989 could already find clear signs of deterioration of BRD's portfolio and profitability. By then, it was too late to arrest the serious negative impact on BRD's finances and general outlook. 5.4 Commitment of Credit 1650-RW started only in late 1987 and accelerated in the second half of 1988 after the Bank lifted a suspension of commitments which extended from February to July of that year (see para 5.6). The agreed closing date for commitments, end-December 1988, was extended first to end-1989 and ultimately to end June 1990, by which date the Credit was fully  - 8 - committed. The Credit financed 11 large and medium size enterprises for loans totalling SDR 4.9 million and 78 SSEs for a total of SDR 2.1 million. Annexes 9a, and 9b provide data and analysis of subprojects financed under Credit 1650-RW.2 As distortions in the economy became more pronounced during the implementation of BRD IV, the objective of strengthening BRD's capacity to perform economic evaluation of projects could hardly be achieved. 5.5 Credit disbursements were very slow in 1987 and 1988 but accelerated in 1989, 1990 and 1991. As of the date of this PCR, SDR 4.6 million have been disbursed for the large and medium size projects (around SDR 300,000 committed were cancelled at BRD's request). Only SDR 1,4 million have been disbursed for the SSE projects. BRD has been requested to examine all SSEs accounts and make a final request for disbursements. It appears that many approved SSE projects, particularly small farming and livestock projects were severely affected or were literally wiped out by the civil war. 5.6 Under BRD IV, BRD agreed to review the level and structure of rates at least once a year to ensure that they remain positive in real terms. Two interest rates were set: for "productive" sectors (12.5%) and for services (13.5%). In July 1987, as the rate of inflation fell below 5%, the Government, which had always wanted to use interest rates as instrument to encourage private investments, decreed a reduction of the maximum interest rates to 9% for priority sectors and 9-12% for other sectors. After consultation with IDA, BRD reduced interest rates for projects financed under Credit 1650-RW to 9% and the onlending rate to BRD was also reduced to 5%. However, BRD also reduced interest rates unilaterally on 13 subloans financed under BRD III (Cr.1344-RW) without IDA's consent, and as a result, IDA suspended new commitments under Credit 1650-RW. BRD's management was caught between its legal commitment to the Bank and the imminent risk of loosing some of its best customers to the newly created competition from other local banks (see next paragraph) who could refinance existing loans with lower rates. The suspension was lifted in July 1988, after the Government and BRD agreed with IDA on a number of steps that would ensure that the change in interest rates would not affect BRD's profitability. These included a requirement that BRD amend the credit agreements with clients benefitting from the reduced interest rates to enable BRD to increase interest rates if lending rates increase in the future. Since the suspension took place when the Credit resources were badly needed, BRD's management should probably have acted more forcefully to bring all parties to an agreement while withholding fresh loans until the matter was solved. 5.7 Along with the reduction of interest rates, the Government de facto did away with BRD's monopoly of term financing. Measures of "despecialization" allowed commercial banks to compete for that business, with the advantage of being able to be flexible in lending rates charged and to provide customers with services for foreign trade and short-term transactions, leaving BRD at a clear disadvantage. While the liberalization of the financial sector should be considered a positive measure, it was carried out without giving BRD the chance to prepare to face the new competition, a situation which prevails today. As the framework under which the BRD IV had been appraised had radically changed, the Bank should have also reconsidered the viability of this project in the new environment, and restructured it accordingly. 2 A total of 90 projects for SDR 2.5 million were approved, but 12 did not materialize because the promoter withdrew before or after BRD loan signing.  5.8 Although BRD had generally avoided many of the problems that have plagued development banks in other countries, it now faces a number of such problems. As BRD doubled its lending in one year, and because a number of the projects approved were of questionable viability, BRD external donors (including IDA) declined to finance several of them. BRD resource position was not geared for such large volumes of lending as the institution was unable, on such short notice, to mobilize the required resources. This put additional strains on its available financial resources, prompting the institution to seek revolving and substantial short-term bridge funding from BNR, the central bank, to cover medium- and long-term lending commitments. In order to avoid the damaging consequences of further mismatch of lending and borrowing, and due to resource constraints the following years, BRD had to substantially curtail its lending activities. It was too late, however, to limit the expected damage both to the portfolio and financial situation of BRD. 5.9 Moreover, BRD loan portfolio is highly concentrated by sector and geographic location. Large industrial borrowers in the area of Kigali, the capital, and some seven enterprises account for over 50 percent, and 40 percent of loans outstanding, respectively. Although BRD policy statement limits individual financial risk exposure to a maximum 20 percent of its unimpaired equity, to 70% of total project costs in the case of SSEs, and to 60% and 50% for large and medium projects respectively, such concentration in few borrowers is detrimental in the long run. 6. Project Results 6.1 Annexes 7a, 7b, 7c, 7d and 7e detail BRD projected and actual financial operations during 1985-92. Actual financial performance was below expectations. The 1988 commitments, the economic downturn after 1986, aggravated in 1989, as well as the civil strife that started in 1990, could not be anticipated by the appraisal mission. Commitments and disbursements exceeded projections by 38 and 112 percent, respectively, in 1988. Provisions for risks of losses, net earnings and equity requirements were also underestimated but the lags were within acceptable and manageable bounds. The key aspects of its performance are highlighted below: (a) Portfolio performance: As import substituting industries account for 40 percent of the portfolio, BRD vulnerability to the slightest economic downturn is evident. Lending for a large number of risky SSE borrowers exacerbated the declining performance of the loan portfolio as collections plummeted sharply because of the worsening economic situation and the war. Overall loan collections in principal and interest fell to a low 58 percent in 1989, but improved thereafter, while remaining below the agreed minimum target of 100 percent (Annex 8b). The non-performing portfolio increased substantially as a percentage of total loan portfolio. Overall loan collection rates have dropped to dangerously low levels when contentious loans are taken into account.3 The loan portfolio, and collection problems faced by BRD, and to a lesser degree commercial banks, also find their roots in administrative and legal 3 BRD's loan classification system relegates contentious loans, which have fallen due 100%, to a separate category thus underestimating non-performing loans.  - 10 - constraints faced by all institutional lenders.4 (b) Profitability: The need to constitute additional loan loss provisions led BRD to declare steep operating losses of RwF 300 million (US$2.3 million) and RwF66.3 million ($510,000) in 1990 and 1991, respectively - the first losses in its history. In 1992 BRD seemed to have overcomed its difficulties with increased operating revenue (up 13 percent), lower cost of borrowing (down 47 percent) and a modest profit of RwF23 million, after RwF470 million in allocations to provisions compared to RwF370 million for 1991. BRD's auditors consider, however, that additional provisions of RWF 88 million should be made. It is likely that further operating losses will be incurred owing both to lower operational revenues, higher allocations to provisions for risks, and maintenance of the same staff and administrative levels of expenses. (c) Capital structure: Low collections and sharp increase in disbursements led to a serious liquidity squeeze during 1988-90. In 1991/92, BRD was forced to a capital increase of RwF364 million ($2.8 million) to offset the cumulated losses, maintain an adequate equity level, diversify activities and establish a commercial service. As a condition of the capital increase, BRD's Board introduced some restructuring and cost reducing measures, and mandated a slow-down of approvals and a focus on loan recovery activities. As a result, liquidity loan improved dramatically and BRD was able to fully reimburse BNR's short-term advances contracted in 1988. BRD capital base is relatively strong by regional standards, and its capital risk coverage ratio also relatively high. However, at current levels of loan collection these are inadequate. 6.2 BRD most pressing challenge stems from a traditional client base - 42 percent of which are import substitution industries -- that is threatened by the recession and the adjustment program. The financial sector reforms introduced progressively since 1987 to regulate banks and eliminate directed credit policies in Rwanda have also subjected BRD to greater competition from the commercial banks, for the first time, particularly in medium-term loans to "blue chip" companies. Consequently, BRD needs to undertake a radical review of its traditional vocation as a development bank, probably to diversify its lending activities, in order to provide new commercial banking services and short-term lending to its current customer base. 6.3 Performance under the Fourth Credit must be analyzed, taking into account that BRD management had little control over the many economic and political developments adversely affecting its lending operations and portfolio. IDA did not insist on ensuring that BRD retain a degree of freedom during the steep lending increase of 1988, although IDA and other donor agencies did use their leverage to refuse refinancing risky lending operations. Moreover, IDA did not anticipate the above problems and lay down a consistent and coherent strategy to face up to the looming challenges brought about, inter alia, by the adjustment process, including reforms of monetary and credit policies, tariff liberalization, etc. Action was taken only in the course of appraisal and negotiations 4 SSE loans are guaranteed under Guarantee Fund and BRD should recover its losses once legal problems are solved.  - I1 - of the Financial Sector Adjustment Credit (para 8.3) and first Apex line of credit channelled through BNR, the central bank, to the entire banking system (para 8.4). 6.4 The adverse effects of economic recession and civil strife do not allow a proper analysis of the evolution of SSEs in the economy of Rwanda and whether the project had a significant impact in their support and well-being. Despite the problems encountered by SSEs during recent years, 250 SSE projects financed by BRD remained active at the end of 1992, constituting 11% of BRD's portfolio. BRD's management deserves all the credit for this achievement as the Bank did not direct its supervision efforts towards SSEs improvement. 7. Project sustainability 7.1 BRD's management should be praised for its resilience during recent years, when the economy of Rwanda seriously deteriorated and life in general was severely disrupted by civil strife. Due to its efforts, in 1991 BRD's shareholders ratified their support to the institution by subscribing a capital increase of FRW 364 million. The capital increase and successful efforts in 1992 to collect FRW 209 million in arrears from its delinquent borrowers, have averted any fears of BRD's insolvency. On the contrary, the equity injection tied to the reduction of disbursements following a retrenchment of commitments in 1990 and 1991 and the lack of suitable new projects, have significantly improved BRD's short-term liquidity. 7.2 However, BRD's future is unclear: the changes operated in the economy after measures of liberalization were enacted, affected BRD's competitive position in the banking sector. In order to assist the Government and BRD's management to draw a plan for BRD's future, a detailed independent special audit of BRD's portfolio was undertaken in November 1992, by a team of three experienced, professional consultant bankers hired by IDA, with financing from a Japanese grant. The consultants concluded that, at the end of 1992, BRD's portfolio was underprovisioned by RWF 488 million. In order for BRD to be sustainable and profitable with a loan portfolio reduced by 40% after these additional provisions were taken into account, the consultants further recommended drastic rationalization measures, inter alia, a substantial reduction of personnel, closure of branches, streamlining of management; and an increase in the spread from 6.5 to 7%. The Government would also be asked to maintain, for at least another five years, the income tax exemption already granted to BRD. 7.3 The Government has agreed to the restructuring plan and its implementation would make BRD eligible to draw funds from the Apex Credit (para 8.4). BRD's management considers that under the conservative approach of this plan, BRD would possibly be able to maintain a share, albeit much reduced, of Rwanda's term financing market but its long-term prospects and sustainability would be questionable: facing stiff competition for the most profitable medium and large size projects, to which BRD's competitors can offer a wide range of services, BRD could be left with a portfolio constituted largely by SSEs with high risks and administrative costs. The feasibility of diversifying BRD's activities and undertaking the services of a commercial bank, as proposed by BRD's management, has yet to be studied. It is urgent that such study be carried out to allow BRD's shareholders to decide on the future strategy and action plan to be followed by the institution. The strategy study should consider, among other, options for BRD's full privatization and/or merger with a local or foreign commercial bank.  - 12 - 7.4 As the financial system of Rwanda moves further towards market mechanisms and away from direct controls, as appropriate regulatory framework will be necessary to identify troubled institutions at an early stage as well as to create an environment conducive to greater competition between institutions. A program of reforms of the regulatory and legal framework is underway under the proposed Financial Sector Adjustment Program (para 2.7). The BNR has already begun changing the regulations on provisioning, setting clear guidelines on when and how much to provision and when and how to deal with interest accrual on non-performing financial institutions, and will spell out the policy instruments that will enable BNR to conduct monetary policy, including the rediscount of long-term credit. The revised legislation will include provisions for BNR to have a clear mandate over the regulation and supervision of financial institutions as well as to set appropriate capital levels. Risk exposure limits will be set to limit a bank's exposure to a single party or group. New regulations will clarify the treatment of non-performing loans. In addition, external audits of all financial institutions will be required every year. 8. Bank's Performance 8.1 As a repeat project for a successful DFC, Bank appraisal of the project was carried out correctly. Appropriate measures were introduced to improve the overall framework for SSE's development in Rwanda which was a major project objective. Measures to ensure BRDs' financial performance and stability were also incorporated in the project design. If in a narrow sense the Bank's performance could be considered acceptable, the opportunity for raising important macroeconomic policy issues, like distortions of the exchange rate and in the import tariffs system, was not forcefully pursued by the Bank. In that respect, it should be mentioned that at the time of appraisal the economy of Rwanda had experienced some recovery. In 1984 the balance of payments registered a surplus, gross foreign exchange reserves were equivalent to 5 months of imports and the debt service burden was negligible. Inflation remained moderate, about 6%. It was estimated that GDP grew between 3.5% to 4% and the prospects for 1985 were favorable as the coffee harvest looked promising. It would be fair to assume that, in view of the positive economic outlook, the Bank did not feel it was urgent to condition the BRD IV project to the adoption of radical economic measures. 8.2 Supervision of the project was adequate in 1986, but no supervision mission visited BRD in 1987, the year of the Bank's reorganization. The first supervision after this hiatus, in mid- 1988, concentrated its effort in solving the problem that led to the suspension of commitments (para 5.6) and was unable to detect the dangers involved in BRD's accelerated lending expansion. In fact, in September 1988, the Bank gave its approval to the financing of several medium, large and SSE subprojects under Cr. 1650-RW. The Bank did voice its concern to BRD management and the Government in 1989, but by that time BRD's performance had started to deteriorate. Because of its concerns about serious distortions in the economy and the recession already in place, the Bank refused to consider a Fifth BRD project which BRD and the Government had requested. As a compromise, and in order to alleviate the critical resource situation of BRD, the Bank extended twice the deadline for commitments under the credit. 8.3 The Bank's refusal to consider a fifth line of credit for BRD did not imply that the objectives pursued in the four BRD projects were abandoned, but a clear realization that BRD's problems could not be dealt with in isolation, without addressing basic issues affecting the economy in general and the financial sector in particular. Following that strategy, the Bank, in cooperation  - 13 - with the IMF, assisted the Government in the preparation of a Structural Adjustment Program whose implementation started in 1991 (para 2.7). Furthermore, the Bank carried out a Financial Sector Review (Report N# 8934-RW) which laid the basis for an adjustment Program for the Financial Sector. In this Program measures were designed to strengthen BRD, inter alia, BRD's capital increase, rescheduling of BRD's debt to BNR and the preparation of a detailed five-year plan for BRD to diversify its activities and improve the quality of its portfolio. The plan has been partially implemented, allowing BRD to somehow consolidate its financial position. 8.4 More recently, an IDA financed Private Sector Development Project in Rwanda was approved in September 1993. The Project includes a US$7 million Apex line of credit that the Central Bank would onlend to qualified Participating Financial Intermediaries (PFI) to finance productive investments. As part of the Project the Government has agreed on a Plan of Action to restructure the operational and managerial organization of BRD to reduce costs, increase accountability and improve efficiency. The Plan also envisages measures concerning risk management and provisions, budgeting control and programming. Implementation of the Plan would allow BRD to qualify as PFI for the US$7 million Apex line of credit included in the project. 9. Borrowers' Performance 9.1 Borrowers' performance has been mixed. The Government and BRD's management strayed away briefly in 1987 and 1988 from the prudent, rather conservative practices that characterized BRD's behavior since its creation and had gained widespread support to BRD from bilateral and multilateral aid agencies, including the Bank. This deviation from past practices came about as pressure grew from Government to accelerate the promotion of nationals in business. The Government's attitude at that time is understandable as it wanted to count on BRD's financing to pursue more vigorously the promotion of indigenous enterprises and, at the same time, counteract the effects of the economic recession. BRD reluctantly, and under pressure, went along financing a large number of risky or unsound projects, large, medium and small, many of which it would have refused to support under normal circumstances. As the economic recession intensified in the ensuing years, the policy followed in 1987 and 1988 proved to have had serious negative effects on BRDs' finances and future prospects. 9.2 While measures adopted by the Government in 1987 to liberalize the financial sector constituted a positive step, no parallel action was considered to allow BRD to face the new competition by the private banks, better equipped with their services and flexibility to attract good customers away from BRD. 9.3 Since 1989, BRD's management performance has substantially improved. Dictated by the economic crisis and its own financial limitations, a prudent lending policy has been followed. At the same time, efforts in supervision and collections have been reinforced, with good results. Despite the disruption caused by the armed conflict, BRD has maintained sound administrative and personnel practices, which is very commendable. The challenge for the Government and the institution is now to face its structural problems and adopt the tough measures necessary for BRD's future survival and progress.  - 14 - 10. Lessons Learned and Action Taken 10.1. In intermediation projects, when Bank funds are loaned to private sector enterprises through local financial institutions, if serious distortions prevail in the economy (i.e. overvalued exchange rate, highly protective import barriers, subsidies), an enterprise benefiting from the project may obtain financial gains which would most likely not represent economic gains for the country. The introduction in project evaluation of an economic rate of return calculation as a means for overcoming those distortions has not proven to be an adequate solution. That has been the experience in this type of project in many countries and it was repeated in the BRD IV project. Either the analysis becomes overly complex or changes in the pattern and level of distortions render the evaluation meaningless. Therefore, in the absence of an effective program of economic reforms, support for intermediation projects should be delayed until distortions are corrected and market prices become a fair representation of the cost of production factors. 10.2. BRD IV was one of the last "DFC type" projects carried out by the Bank in Sub- Saharan Africa. The suggestions made by BRD in Part II, para. 3.7 (c) on ways to improve BRD World Bank cooperation have been taken into account in the follow-up operation, i.e. Private Sector Credit. In the mid-80s, as many countries in the Region engaged in a process of structural adjustment to reverse economic decline, new market conditions became adverse for DFCs as their main customers, import substitution industries, were seriously affected by liberalization of imports. As liberalization measures reached also the financial and banking sector, the single DFC approach to channel IDA funds for term lending to private entrepreneurs also became obsolete. Such objective is pursued now through lines of credits for term financing made available, usually through the Central Bank, to all eligible financial institutions. Survival of existing DFCs in the new environment is by no means assured. For that, they are required not only to streamline and cut operating costs but also to diversify their services, a strategy more likely to succeed under privatization. 10.3. In the late 1980s, it became clear that BRD's problems could not be dealt with in isolation without addressing macroeconomic issues in general, as well as financial sector issues. To that effect, the Government embarked since 1991 on an economic reform program which was designed with the assistance of the Bank and the IMF. Furthermore, in 1991, the Bank carried out a Financial Sector Review which provided the basis to design a program of adjustment for the Financial Sector, including provisions to address BRD's more immediate financial problems (para 8.3). More recently, an IDA financed Private Sector Development Project, approved by the Board in September 1993, includes a plan to rationalize BRD in order to cut operational costs, increase its efficiency and make it eligible for the Apex line of credit included in that project (para 8.4). Beyond these measures, a corporate strategy needs to be designed to address BRD's long-term sustainability. The feasibility of BRD's privatization should be seriously considered.  - 15 - PART IT: PROJECT REVIEW FROM BORROWER'S PERSPECTIVE I. INTRODUCTION The World Bank's partnership with BRD is becoming something of a tradition, the latter now having received four IDA lines of credit. The purpose of this fourth line of credit, which is the subject of this project completion report, was to finance enterprises in the productive and service sectors. The credit package totaled SDR 8.5 million, SDR 4.7 million of which was for large and medium size enterprises and SDR 3.8 million for small-scale enterprises (SSEs). Since BRD has already received the project review from the World Bank's perspective and is in full agreement with the analysis made, we shall attempt to avoid repetition when addressing an aspect not already touched on by the Bank, namely the assessment of the project's impact on the country's development, and in our response to the issues raised by the Bank. We shall follow the guidelines provided to us in point 2, which are as follows: (a) Confirm the adequacy and accuracy of the factual information contained in Part III of the PCR. (b) Comment on the analysis presented in Part I of the PCR. (c) Evaluate the Bank's performance during the evolution and implementation of the project, with special emphasis on lessons learned that can be relevant for the future. (d) Evaluate the Borrower's own performance during the evolution and implementation of the project, with special emphasis on lessons learned that may be relevant for the future. (e) Assess the effectiveness of the relationship between the Bank and the Borrower during the evolution and implementation of the project, with special emphasis on lessons learned that may be relevant for the future. (f) Evaluate the performance of cofinanciers, and assess the effectiveness of their relationship with the Borrower during the evolution and implementation of the project. It must be remembered, when considering the comment and evaluations that follow, particularly with respect to (b) above, that BRD agrees with the various points in the analysis contained in Part I and does not wish to restate them for fear of getting into a tedious repetition of what has already been described by the World Bank.  - 16 - II. ASSESSMENT OF THE PROJECT'S IMPACT ON THE COUNTRY'S DEVELOPMENT 2.1 Status of projects in general 2.2 Current status of interventions financed under the line of credit The IDA IV line of credit has enabled BRD to finance 89 enterprises in all: 7 LSEs, 8 MSEs and 74 SSEs. Of all these projects, 30 are going well, 10 are in slight difficulty, 9 are in serious difficulty but might recover, 18 are failing and 22 have disappeared, 9 for reasons attributable to the war, and 13 as a result of poor management and market problems. 2.3 Value added Among the projects financed under the IDA IV line of credit and for which financial statements are available, value added generated in 1992 totaled RF 445,656,581 for a production of RF 1,168,730,430, i.e. 38%. It should be pointed out that most of the small-scale projects do not yet have reliable accounting systems. 2.4 Contribution to the production of goods and services to meet the needs of the population - Agri-foodstuffs production Thanks to the use of modern production methods (chemical fertilizer, plant protection products, concentrated feed, drugs, artificial insemination, machinery, sophisticated equipment and tools), agri-foodstuffs production has increased for such items as potatoes, milk, cheese, eggs, meat, tea, pineapples, bananas, vegetables, etc. - Industrial and crafts production Thanks to this same line of credit, industrial and crafts production has increased. Mention should be made of the contributions of industrial enterprises such as SORWATHE, ECOMIRWA, SODEPARAL, SOBOLIRWA, ZEM INDUSTRIES, PACAREM, GATETE Polycarpe and PROMETAL, and SOBAGEB, Claude Kampayana, COMECOMU, Theobald Mungwarakarama, Josias Rubare, Frangois Xavier Gasana, Joseph Bavakure, Joseph Rugigana, Jacques Rusirare, Elie Kanuma and Emmanuel Basimbizi in the crafts sector. - Production of services The IDA IV line of credit has also made it possible to increase the production of services. Noteworthy contributions include hotel projects, boarding facilities for secondary school students, medical offices, rural pharmacies and gas stations.  - 17 - 2.5 Creation of jobs The impact of the line of credit is also measured by the number of jobs created. Enterprises for which data are available for 1992 have a total payroll of 1,230, i.e. 94 per enterprise on average, with a total wage bill of RF 97,588,137 or RF 7,506,780 per enterprise. 2.6 Savings of foreign exchange Certain enterprises financed under the IDA IV line of credit have contributed to foreign exchange savings, by fully or partially substituting imports (ECOMIRWA, SQBOLIRWA, ZEM INDUSTRIES, PACAREM, PROMETAL, GATETE, SODEPARAL) or, better still, by bringing foreign exchange into the country (SORWATHE). 2.7 Taxation Given that most of the enterprises financed are still for tax exemption under the investment code, it is not necessary to quantify this aspect. It should be recognized, however, that the Treasury does collect revenues from such levies as the payroll tax [taxe professionnelle sur les salaires du personnel des entreprises], import duty on raw materials, turnover tax and business tax [patente]. 2.8 Special features of certain projects 2.9 Contribution to exports and to foreign exchange returns Among the enterprises financed under this line, only SORWATHE is exporting officially. The other enterprises whose products cross the frontier operate within the framework of border trade, which does not involve the transfer of foreign exchange. 2.10 Import of raw materials and the outflows of foreign exchange Certain enterprises financed under this same line of credit have made it possible to limit foreign exchange outflows to the cost of raw materials in place of the cost of the finished products. This is so in the case of ECOMIRWA, ZEM INDUSTRIES, PACAREM, GATETE Polycarpe and PROMETAL. III. COMMENTS ON PART I OF THE PCR AND RESPONSES TO ISSUES RAISED BY THE WORLD BANK. FOLLOWING THE GUIDELINES MENTIONED IN THE INTRODUCTION 3.1 Confirm the adequacy and accuracy of the factual information contained in Part III of the PCR The statistical information contained in Part III is sufficient and accurate, with the exception of the following:  - 18 - a. Annex 6 The capital structure at end 1992 is as follows: (RF millions) (as percentage) Rwandese Public Sector: 827.75 55 Government of Rwanda 714.26 48.1 OCIR-CAFE 50.92 3.5 OCIR-THE 13.70 0.9 CAISSE SOCIALE DU RWANDA 48.87 3.3 Rwandese Private Sector: 155.66 i105 Commercial banks 60.00 4.0 Other private investors 95.66 6.5 Foreign Institutions: 500.99 33.7 CCCE/CFD 135.71 9.1 DEG 133.68 9.0 FMO 96.00 6.5 Belgian Aid Agency 126.67 8.5 Bank of Tokyo 8,93 0.6 TOTAL 1,484.40 100 b. Annex 9a The number of SSEs financed should be 78 rather than 74. 3.2 Comments on the analysis presented in Part I of the PCR 3.3 Evaluation summary With reference to paragraph 6, "Viability" [sicl on page iii, BRD agrees with the World Bank that the new operating strategy for BRD would be most likely to succeed if it were to be privatized and if Government influence were reduced, but it is hardly necessary to remind you that BRD is governed by a law specifying the distribution of its capital stock. Privatization would therefore depend on the amendment of its governing legislation and action to redistribute its capital stock. BRD considers that the report should stress the rationalization effort, with projections for the future, which will include plans for diversification and redeployment toward short-term services (Commercial Services), in particular.  - 19 - 3.4 Part n: Project review from World Bank's perspective With reference to paragraph 4.5, it should be pointed out that the estimated 5% spread is well below the margin generally allowed development banks by the banking profession. The minimum spread would be set at 7%. With reference to paragraph 4.7, it should be mentioned that BRD's Guarantee Fund for SSEs covers 80% of the risks (principal + interest) and not 100%. With reference to paragraph 5.4, the number of SSEs financed is 78 rather than 94, the loans totaling SDR 2.1 million instead of SDR 2.5 million. We take this opportunity to mention that the financial statement dated September 30, 1993 gives the number of SSEs financed as 90, while BRD's records show only 78 SSEs. For the sake of uniformity, the following should be taken into account: SubDroject No. Comment SSE-30 Canceled on October 10, 1991 SSE-37 Promoter withdrew SSE-52 Canceled on November 17, 1989 SSE-63 About to be canceled, if not already, because the contract was never signed by the promoter SSE-66 About to be canceled, if not already, because the project has not been carried out SSE-102 Promoter withdrew SSE-106 and These projects are not shown in the "draft report" at the end of the SSE-107 fourth IDA line of credit but appear in your monthly statement of projects financed under IDA IV at September 30, 1993. These projects should therefore be removed from this monthly statement SSE-108 Promoter withdrew SSE-109 Idem SSE-117 Canceled on October 10, 1991 SSE-133 Promoter withdrew With reference to point 5.9, the figure of 70% of total project costs holds true only in the case of SSEs, the normal limit on individual financial risk exposure in the case of large-medium projects being 60% and 50%, respectively. With reference to point 6.1(a), it is true that the worsening economic situation and the war brought about a decline in recoveries and in loan portfolio performance, but given the fact that most SSEs are covered by the FSG, BRD hopes to be able to recover its funds as soon as the legal dispute with the FSG is settled.  - 20 - With reference to point 6.1(b), BRD did not consider it necessary to make additional provisions of RF 88 million at December 31, 1992, in view of the special situation of certain debts and the guarantees furnished. There are plans, however, to increase the provisions in 1993, based on a case-by-case assessment following the pertinent directives issued by the Board of Directors. Adopting this strategy, provisions of around RF 300 million will be made at the end of 1993 (an additional RE 110 million) with a projected profit of RF 62.8 million. With reference to point 6.1(c), while it is true that the sixth capital increase was aimed at maintaining an adequate equity level, it must be kept in mind that it was also aimed at diversifying activities and at establishing a commercial service. With reference to footnote 2 on page 10, the classification of loans as contentious does, of course, mean that they have fallen due 100%, but such loans still generate interest until a judgment is handed down. With reference to point 7.2, additional provisions of RF 488 million would seem impossible to justify given the fact that, thanks to the recovery effort undertaken to collect debts regarded as uncollectible by the consultants, BRD currently has a liquidity of RF 1,115 million, which will probably reach RF 1.3 million by end 1993. 3.5 Bank's performance during the evolution and implementation of the project. with special emphasis on lessons learned that may be relevant for the future BRD recognizes the World Bank's efforts to bring the project to a successful conclusion; it also recognizes that it was hard to predict the country's economic future and agrees with the comment made in Part I of the PCR. It appreciated the facility offered by the opening of a special account at BNR and the speed with which disbursements were made, in keeping with the terms of the agreement. BRD would like to see some relaxation in the 90-day rule requirement for the submission of invoices for payment. 3.6 Borrower's own performance during the evolution and implementation of the proiect. with special emphasis on lessons learned that may be relevant for the future BRD agrees with the Bank's comments and conclusions contained in Part I of the PCR. It wishes to stress that at times when finances have been tight and the economic situation has been tough, it has managed to overcome its difficulties, improve its financial situation enormously, and recover its profitability despite the combination of the devastating consequences of the economic downturn and the war of October 1990. 3.7 Effectiveness of the relationship between the Bank and the Borrower during the evolution and implementation of the proaect. with special emphasis on lessons learned that may be relevant for the future The relationship between the Bank and the Borrower was marked by a good climate of cooperation conducive to the effective overcoming of any dispute or problem that might arise during  - 21 - project implementation. Several meetings were held to exchange ideas. Nevertheless, there were two instances when a strain was placed on this effective cooperation: (a) The incident of the lending rate and suspension of the fourth IDA line of credit Following the Rwandese Government's decision, BRD had no choice but to lower the interest rates on certain loans financed out of the third IDA line of credit. BRD was bound to comply with this decision, a regrettable situation, since the suspension of the IDA IV line of credit should have been the outcome of direct contact between the Government and the World Bank. BRD was merely the victim of the decisions, which were beyond its control and resulted in a financial crisis. (b) Conseguence of "despecialization" of the banking sector BRD is discovering that the way to handle the "despecialization" of the banking sector decreed in 1987 is to diversify its activities and to establish a commercial service. Unfortunately, some differences of opinion still remain concerning the conditions for establishing a commercial service of this kind and these have so far made it impossible to carry out this project. Speaking in general terms, there are three suggestions that might be made in order to improve BRD-World Bank cooperation in the future: 1. BRD should negotiate a loan agreement on more concessional terms and scrupulously comply with every single provision it contains. 2. The World Bank should, within the limits of the sectors previously approved for intervention, allow more freedom of choice in terms of the selection of projects to be financed. 3. The World Bank should simplify disbursement procedures, with particular attention to the requirement to remit huge numbers of photocopies of various supporting documents. Since the Bank is represented in Rwanda and conducts a fair number of missions in our country, the existence at any time of a given project and the submission of proof of disbursement by BRD should be sufficient. 3.8 Performance of cofinanciers and assessment of the effectiveness of their relationship with the Borrower during the evolution and implementation of the project The project covered by Credit 1650-RW did not have any cofinanciers.  -22- Annex I Page 1 of 1 PART mI: STATISTICAL INFORMATION FOURTH DEVELOPMENT FINANCE PROJECT (Credit 1650-RW) Related Bank Credits Credit Title Purpose Year of Status Approval 655-RW Long-term finance for productive investments; 1976 96.9X disbursed and strengthening of SRO's appraisal cLosed. capabilities. 896-RW Funding for foreign exchange cost of 1979 86.21 disbursed and productive investments in Rwanda; staff closed. training and upgrading of appraisal capacity for BRD; technicaL assistance to Government for establishment of an auditing capacity in Rwanda. 1344-RU Fundfng for foreign exchange cost of 1983 831 disbursed and closed. productive investments in Rwanda; technical assistance for development of small scale antreprises, including pilot SSE financing;  -23- Annex 2 RWANI)A FOURTH DEVELOPMENT FINANCE PROJEC (Cr. 1650RW) Project Processing and Implementation Timetable Item Planned Revised Actual Preappraisal mission 12/84 Appraisal Mission 02/85 02/85 Credit Negotiations 09/85 10185 10/85 Board Approval 11/85 01/86 01/86 Credit Signature 02/86 01/86 01/86 Credit Effectiveness 03/86 10/86 10/86 Credit Closing 12/93 12/93 12/93 Credit Completion 06/94 12/93 TBD  -24- Annex 3 Page 1 of 1 FOURTH DEVELOPMENT FINANCE PROJECT (Credit 1650-RW) Loan Disbursements ($ million) Fiscal Year Estimated Actual Actual as % Actual as % Cumulative Cumulative of Estimated of IDA Standard 1987 1.2 0 0.00% 0.00% 1988 2 0.3 15.00% 12.82% 1989 3.6 4.04 112.22% 89.78% 1990 5.6 6.5 116.07% 103.17% 1991 7.2 7.8 108.33% 100.78% 1992 8.4 8.22 97.86% 97.16% 1993 9* 8.34* ERR 0.00% * Note: As all IDA credits are nominated in SDR's, Credit 1650-RW total amount was SDR 8.5 million. As of end December 1993, SDR 6,338 million have been disbursed. The undisbursed amount is SD 2,162 million.  -25- Annex 4 FOURTH DEVELOPMENT FINANCE PROJECT (Credits 1650-RW) Status of Compliance with Major Covenants Covenant Status of Comgliance 1. BRD shall maintain (i) real positive interest In compliance. rates and (ii) a sufficient interest spread to preserve its financial soundness. 2. BRD shall carry out the Project in accordance In compliance with its Statutes and Statement of Policy. 3. BRD shall furnish to the Association all such In compliance. information as the Association shall reasonably request concerning the administration, operations and financial condition of BRD. 4. BRD shall furnish to the Association audited In compliance. financial statements not later than six months after the end of the fiscal year. 5. BRD's ratio of collection of loan repayments in principal and interest to amounts falling due and billed to borrowers, will progressively reach and be maintained above 100 percent, specifically by reaching at least 95 percent in 1985, 100 percent in 1986, 102 percent in 1987, and 105 percent in 1988 and thereafter. 6. The total amount of BRD's, and all of its In compliance Subsidiaries'consolidated debt shall not exceed three times BRD's and all of its Subsidiaries consolidated capital and surplus. 7. BRD shall take steps satisfactory to the In compliance. Association as shall be necessary to protect itself against risk of loss resulting from changes in the rates of exchange between the various currencies (including Rwandese Francs) used in its borrowing and lending operations.  -26- Annex 5 RWANDA FOURTH DEVELOPMENT FINANCE PROJECT (Cr. 1650-RW) Use of Bank Resources - Mission Data Stage of Date No. of Days in Specialization Performance Project Cycle mo/yr Persons Field Represented Rating Status Throgh A&raisal Pre-appraisal 12/84 2 14 op. officer economist Appraisal 05/85 2 Supervision Supervision 10/86 2 4 Op. officer 1 Fin. analyst Supervision 05/88 1 5 Industrial op. officer 2 Supervision 07/89 3 12 Sr. op. officer/op. officer 2 economist Supervision 12/89 2 10 Sr. op. officers 2 Supervision 08/90 2 16 Sr. op. officers 3 Supervision 04/91 2 10 Sr. op. officer/consultant 2 Supervision 10/92 1 15 Sr. op. officer 2  -27- Annex 6 Page 1 of 1 FOURTH DEVELOPMENT FINANCE PROJECT (Credit 1650-RU) BRD: Evolution of Share Capital Structure (in percentages) End 1984 end-1992* Total Share Capital (million RwF) Shareholder: Rwandese Public Sector: 617.00 (55X) 827.75 (55.8) Govermrent of Rwanda 714.26 (48.1) OCIR-CAFE 50.92 (3.5) OCIR-THE 13.70 (0.9) Caisse Sociale du Rwanda 48.87 (3.) Rwandese Private Sector: 139.00 (12.5X) 155.66 (10.5) Commercial banks 60.00 (4.0) Other private investors 95.66 (6.5) Foreign Institutions: 363.40 (32.5X) 500.99 (33.7) CCCE/CFD 135.71 (9.1) DEG 133.68 (9.0) FMO 96.00 (6.5) Belgian Aid Agency 126.67 (8.5) Bank of Tokyo 8.93 (0.6) TOTAL 1,120.00 100 1,484.40 100  Annex 7a BANQUE RWANDAISE DE DEVELOPPEMENT (BRD) PROJECTED AND ACTUAL LENDING OPERATIONS (RwF million) December 31 1985 1986 1987 1988 1989 1990 1991 1992 Proj. Actual Proj Actual Mrj. Actual Pr